Moderate antiangiogenic activity by local, transgenic ... - CiteSeerX

Uncorrected Version. Published on June 22, 2006 as DOI:10.1189/jlb.1105644

Moderate antiangiogenic activity by local, transgenic expression of endostatin in Rip1Tag2 transgenic mice Birgit Schaffhauser,* Tanja Veikkola,†,1 Karin Strittmatter,* Helena Antoniadis,* Kari Alitalo,† and Gerhard Christofori*,2 *Institute of Biochemistry and Genetics, Department of Clinical-Biological Sciences, University of Basel, Switzerland; and †Molecular/Cancer Biology Laboratory and Ludwig Institute for Cancer Research, Biomedicum Helsinki, University of Helsinki, Finland

Abstract: Many previous reports have demonstrated that systemic administration of endostatin (ES), a proteolytic cleavage product of collagen type XVIII and an endogenous angiogenesis inhibitor, represses tumor angiogenesis in different preclinical tumor models with varying efficacy. For example, systemic delivery of recombinant ES to rat insulin promoter 1 (Rip1)T-antigen 2 (Tag2)transgenic mice, a mouse model of pancreatic ␤-cell carcinogenesis, has repressed tumor angiogenesis efficiently and with it, tumor growth. Here, we report that the transgenic expression of ES in Rip1ES-transgenic mice only interferes moderately with tumor growth in Rip1Tag2;Rip1ES double-transgenic mice. Tumor incidence is not reduced by the local expression of ES, and tumor outgrowth and progression to tumor malignancy are only retarded slightly. A significant effect of local ES expression on tumor angiogenesis is only apparent during the early stages of tumor development, where less angiogenic hyperplastic lesions are observed. Although efficiently produced and secreted by transgenic ␤ cells, locally expressed ES appears to be sequestered in the microenvironment, and its systemic levels are not increased. The results indicate that the antiangiogenic functions of ES critically depend on the mode of delivery and the site of expression: Although its systemic application represses tumor angiogenesis and tumor growth efficiently, locally expressed ES appears to be less effective, and hence, additional mechanisms of solubilization or activation of latent ES seem to be required. These results have important implications about the modes of delivery used in antiangiogenic, therapeutic strategies, which are based on the antiangiogenic activities of ES. J. Leukoc. Biol. 80: 000 – 000; 2006.

also demonstrated by the fact that blood vessel formation and remodeling play a critical role in various pathological situations, including rheumatoid arthritis, macular degeneration, psoriasis, and also tumor progression [1, 2]. It is notable that tumor progression depends on the induction of tumor angiogenesis, an event called the angiogenic switch, which occurs when the balance between pro- and antiangiogenic factors is shifted in favor of angiogenesis inducers [3]. Important proangiogenic factors include vascular endothelial growth factor (VEGF), angiopoietins, fibroblast growth factors (FGF), hepatocyte growth factor, platelet-derived growth factor B, and many others (reviewed in refs. [4 –7]). Conversely, two types of angiogenesis inhibitors can be distinguished: therapeutic compounds, including small chemical entities or therapeutic antibodies designed to repress the function of known angiogenic factors, their receptors, or other processes critically involved in angiogenesis, and naturally occurring antiangiogenic factors, including endogenous inhibitors of angiogenesis, which have been found to play important roles in the regulation of physiological and pathological angiogenesis, including thrombospondins, angiostatin, endostatin (ES), tumstatin, canstatin, or interferon-␣/␤ (IFN-␣/␤; reviewed in ref [8]). Antiangiogenesis therapy has already found its way to the clinics (reviewed in refs. [9, 10]). This concept is based mainly on the inhibition of proliferation and migration of tumor-endothelial cells, thereby preventing the formation of new blood vessels and thus, tumor outgrowth. As angiogenesis usually does not occur in healthy individuals, with the exception of the reproductive cycle in females and during wound healing, such therapeutic approaches are not expected to cause significant side-effects. Moreover, as tumor endothelial cells are nontransformed cells, the development of resistance to a particular antiangiogenic regimen is not expected. Out of the many different angiogenesis inhibitor types, ⬃60 inhibitors are currently in various clinical trials (for a selected listing, see ref. [11]). One of the factors that is currently under intense investigation is ES, a naturally occurring angiogenesis antagonist. It

Key Words: angiogenesis 䡠 cancer 䡠 inhibitors 䡠 tumorigenesis 1

INTRODUCTION Tight control of vasculogenesis and angiogenesis represents one important aspect in maintaining tissue homeostasis, as it is 0741-5400/06/0080-0001 © Society for Leukocyte Biology

Current address: Amgen Finland, Keilaranta 16, 6th Floor, P.O. Box 86, 02101 Espoo, Finland. 2 Correspondence: Institute of Biochemistry and Genetics, Department of Clinical-Biological Sciences, University of Basel, Mattenstrasse 28, CH-4058 Basel, Switzerland. E-mail: [email protected] Received November 10, 2005; revised April 16, 2006; accepted April 17, 2006; doi: 10.1189/jlb.1105644.

Journal of Leukocyte Biology Volume 80, September 2006

Copyright 2006 by The Society for Leukocyte Biology.

1

represents the C-terminal fragment of the basement membrane component collagen type XVIII, which is generated by proteolytic cleavage, for example, by elastase or cathepsin L [12–15]. Released ES can be found in vessel walls and in platelets and circulating in blood serum [16 –19]. Mice deficient in collagen XVIII develop normally with only marginally disturbed vascular morphogenesis [20]. A large number of reports have documented the involvement of ES in the regulation of physiological and pathological angiogenesis. Thereby, ES specifically affects tumor-related vasculature but spares normal, resting endothelium [8]. Moreover, ES has been shown to inhibit the growth of tumor xenografts without acquiring any therapy resistance by the endothelial cells and eventually leading to a complete remission of the tumors [21]. However, depending on the experimental systems used, the therapeutic efficacy of ES treatment is ranging from a complete block of tumor growth to the absence of any significant effect (for review, see ref. [22]). Several physiological functions of ES have been reported. ES can suppress vascular permeability, and elevated levels of ES are found in certain cancers and chronic inflammatory diseases [23, 24]. For therapeutic experiments, various methods, including direct systemic application of ES or viral- or liposome-mediated gene or protein delivery, were used in different experimental systems [25–27]. In clinical phase I trials, ES has not shown signs of toxicity, even when high doses (up to 700 mg/m2) were applied over several months [28, 29]. One remarkable and as-of-yet unexplained phenomenon is that ES has no major effect on angiogenesis during pregnancy or tissue-regenerative processes as well as wound healing, although it is able to modulate it [30, 31]. The precise molecular mechanisms underlying the antiangiogenic effects of ES are still rather elusive. Heparin/heparan sulfate binding might be critical for the antiangiogenic function of ES [32, 33]. In vitro, ES inhibits endothelial cell proliferation and migration and induces endothelial cell apoptosis, and it specifically inhibits tumor angiogenesis without affecting normal angiogenesis in vivo (reviewed in ref. [34]). On the level of the targeted endothelial cells, ES activates caspase 3 and represses Bcl-2 and Bcl-XL expression, thereby promoting endothelial cell apoptosis [26, 35]. ES is also able to repress VEGF/VEGF receptor (VEGFR)-mediated signal transduction by interfering with the downstream extracellular signal-regulated kinase 1/2 effector-signaling pathway [24, 26, 36]. Other studies suggest that ES inhibits cell-matrix adhesion and signaling through its binding to the ␣5␤1 integrin [37, 38]. Moreover, ES is thought to exert its antiangiogenic activity via binding to glypicans, cell surface glycosylphosphatidylinositolanchored, heparan-sulfate proteoglycans with as-of-yet poorly defined functions in the angiogenic process [39]. More examples of the pleiotropic functions of ES include the inhibition of metalloproteinase activities and the repression of Wnt signaling by blocking T cell factor-mediated c-Myc and cyclin-D1 gene expression [40 – 43]. Finally, gene expression profiling and protein phosphorylation status experiments have revealed that ES treatment of human dermal microvascular endothelial cells modifies a large cluster of genes, which are involved in endothelial growth and development [44]. In particular, the expression of proangiogenic factors, such as VEGF-A, inhibi2


tors of differentiation, hypoxia-inducible factor-1␣, and signal transduction and transcription factors, seemed to be downregulated by ES treatment, whereas antiangiogenic factors, such as thrombospondin-1 and kininogen, are up-regulated. Thus, ES might serve as a modifier to shift the balance from endothelial cell proliferation and vessel formation to endothelial cell apoptosis [14, 34, 45, 46]. Here, we have used the rat insulin promoter 1 (Rip1)Tantigen 2 (Tag2)-transgenic mouse model of pancreatic ␤-cell carcinogenesis to study the effects of local expression of ES during tumor progression. In Rip1Tag2-transgenic mice, simian virus 40 large Tag, under the control of the Rip1, is expressed specifically in pancreatic ␤ cells, resulting in multistage insulinoma development [47]. Beginning at 5–7 weeks of age, in a still-premalignant stage of tumorigenesis, an angiogenic switch activates the quiescent vasculature and promotes angiogenesis in hyperproliferative lesions. This progression to active angiogenesis is characterized by endothelial cell proliferation, dilated blood vessels, and microhemorrhages [48]. Among a number of proangiogenic factors expressed in the angiogenic islets, VEGF-A and FGF are critically involved in the angiogenic switch during tumor progression in these mice. Conditional ablation of VEGF-A expression or adenoviral expression of soluble FGF or VEGFRs has blocked tumor angiogenesis efficiently and with it, tumor outgrowth [49, 50]. The expression status of endogenous angiogenesis inhibitors in the Rip1Tag2 model has not been established. Yet, systemic treatment of Rip1Tag2 mice with a number of antiangiogenic compounds, including specific inhibitors of VEGFR signaling, ES, angiostatin, TNP470, and IFN-␣, has led to a significant repression of tumor angiogenesis and tumor outgrowth [51–53]. Here, we set out to determine how an early and continuous expression of ES in the islets of Langerhans affects the onset of the angiogenic switch as well as tumorigenesis in Rip1Tag2 mice. In particular, we were interested to assess whether the local expression of ES would have a different impact on tumorigenesis as compared with the systemic treatment with recombinant ES reported previously [52].

MATERIALS AND METHODS Transgenic mice Generation and phenotypic characterization of Rip1Tag2 mice has been described previously [47]. Rip1ES mice were generated by cloning a cDNA fragment encoding human ES under the control of the Rip1 fragment and injecting the transgene construct according to standard procedures, as described previously [47]. Genotypes of four founder mice were confirmed by Southern blot and polymerase chain reaction (PCR) analysis. Transgene expression was confirmed by reverse transcriptase-PCR and immunohistochemical and immunoblotting analyses, and one mouse line with highest levels of ES expression was used for further experimentation (Line C43). Rip1ES female mice were crossed with Rip1Tag2 male mice to generate double-transgenic Rip1Tag2;Rip1ES mice. All mouse lines were kept in a strict C57Bl/6 background. Mice were killed between 8 and 13 weeks after birth. Tumor incidence was determined using a dissecting microscope, and tumor volumes were calculated from the tumor diameter by assuming a spherical shape of the tumors.

Histology and immunohistochemistry Pancreata were fixed for 6 h to overnight at 4°C in 4% paraformaldehyde in phosphate buffer (pH 7.4) before processing and embedding in paraffin.

http://www.jleukbio.org

Sections (5 ␮) were cut for immunohistochemical analysis. For bromodeoxyuridine (BrdU) labeling, mice were injected intraperitoneally with 100 ␮g BrdU (Sigma Chemical Co., St. Louis, MO) per gram of body weight 2 h before sacrifice. Antibodies used for immunohistochemistry were hamster antimouse podoplanin (8.1.1, Developmental Studies Hybridoma Bank, Iowa University, Iowa City), biotinylated mouse anti-BrdU (Zymed, South San Francisco, CA), rat antimouse CD-31 (PharMingen, Franklin Lakes, NJ), and mouse antihuman ES (Abcam, Cambridge, MA). Apoptotic cells were visualized using the in situ cell death detection kit, peroxidase [POD; deoxyuridine triphosphate nick-end labeling (TUNEL); Roche, Basel, Switzerland]. All biotinylated, secondary antibodies were used at a 1:200 dilution, and positive staining was visualized using the ABC horseradish POD kit (Vector Laboratories, Burlingame, CA) and diaminobenzidine POD substrate (Sigma Chemical Co.) or angiotensin-converting enzyme (Vector Laboratories), according to the manufacturer’s recommendations. For the analysis of tissue morphology and tumor grading, slides were stained with hematoxylin and eosin (H&E). For the quantitation of blood vessel density, proliferation (BrdU incorporation), and apoptosis (TUNEL staining), the numbers of vascular profiles positive for CD-31 or the number of nuclei staining positive for BrdU or the TUNEL reaction were determined in a comparable number of fields per section at 400⫻ magnification, and the mean of positive cells ⫾ SD/field was calculated. Lectin perfusion experiments were performed as described [54].

Collagen gel assay Islets of Langerhans were isolated from single-transgenic Rip1Tag2 and Rip1ES, from double-transgenic Rip1Tag2;Rip1ES and from C57/B6J control mice at 7– 8 weeks of age as described previously [48, 55]. Human umbilical vein endothelial cells (Passage 3) were cultured in Medium 199 (Sigma Chemical Co.) supplemented with 20% fetal calf serum (FCS; Sigma Chemical Co.), 2 nM glutamine, 40 ␮g/ml bovine pituitary extract, and 80 units/ml heparin. Cells were trypsinized, resuspended in RPMI 1640 (Sigma Chemical Co.), 10% FCS, and 2 mM glutamine and cocultured with islets in a threedimensional (3D) collagen matrix as described [48]. After 48 –96 h, the response of endothelial cells to the cocultured biopsies was scored.

Enzyme-linked immunosorbent assay (ELISA) ES protein levels in the supernatant of isolated, cultured islets of Rip1ES mice and in serum samples of the various single-transgenic, double-transgenic, and control mice were determined using an antihuman ES ELISA kit (Accucyte, College Park, MD), according to the manufacturer’s instructions.

Statistical analysis The statistical significance was tested using the ␹2 test, including Yates correction, Fisher exact test, or Student’s t-test.

RESULTS Generation of Rip1ES-transgenic mice To investigate the potential effects of the angiogenesis inhibitor ES on the development and physiology of ␤ cells in the islets of Langerhans and during tumorigenesis in the Rip1Tag2transgenic mouse model of pancreatic ␤-cell carcinogenesis, transgenic mice were generated expressing human ES under the control of the Rip1ES. Although in islets of Langerhans of a control mouse, human ES was not detectable by immunohistochemical staining (data not shown), in transgenic islets of Rip1ES mice, human ES was found expressed at high levels in 80 –100% of ␤ cells in 100% of the islets in the pancreas (Fig. 1). Thereby, ES was mainly detectable in granular secretory structures in the cytoplasm of ␤ cells. The concentration of ES secreted by ␤ cells was determined by ELISA in the conditioned medium of islets isolated from Rip1ES or control mice.

Fig. 1. ES expression in islets of Rip1ES-transgenic mice. Histological analysis of H&E-stained 5 ␮m paraffin sections of pancreata of Rip1ES mice reveals that expression of human ES in the islets of Langerhans does not alter morphology of the islets (A, B). Immunohistochemical staining of islets in single-transgenic Rip1ES mice with an antihuman ES antibody shows that the transgene is expressed in the cytoplasm of 80 –100% of ␤ cells in all pancreatic islets (C, D). Arrows indicate islets of Langerhans positively staining for human ES. E, Exocrine pancreas; I, islet. Original scale bar, 100 ␮m.

Islets from control mice did not secrete a significant level of human ES into the culture medium, whereas cultured islets from Rip1ES mice exhibited an average level of 0.44 ng/ml per islet in the culture medium. This level of protein corresponds to a calculated, total secreted amount of 176 ng/ml ES for the whole pancreas, which contains ⬃400 islets. It is isurprising that quantification of ES levels in serum of control and Rip1ES mice by a specific ELISA for human ES did not reveal significant amounts of human ES in the serum of Rip1ES mice (C57Bl/6 control mouse: 1.17⫾0.45 ng/ml; Rip1ES: 0.99⫾0.5 ng/ml), indicating that ES secreted by ␤ cells is not freely diffusible and may be sequestered by the islet’s microenvironment. Rip1ES mice were healthy and fertile and did not exhibit any apparent defect in islet physiology, glucose metabolism, or any other parameter investigated. The expression of human ES also did not affect islet morphology or vasculature density, as determined by histopathological analysis (Fig. 1 and data not shown). Lectin perfusion experiments also did not reveal any apparent alteration in vessel morphology, density, or architecture (data not shown). As the transgenic Rip1ES mice were fully viable, we investigated the effects of ␤-cell-specific ES expression on islet carcinogenesis in Rip1Tag2 mice. For this purpose, Rip1ES Schaffhauser et al. Transgenic expression of endostatin

3

mice (founder line C43) were crossed with transgenic Rip1Tag2 mice to generate double-transgenic Rip1Tag2; Rip1ES mice. Double-transgenic Rip1Tag2;Rip1ES mice and control single-transgenic Rip1Tag2 littermates were killed at different time-points during tumorigenesis to investigate a potential effect of long-term ES expression during ␤-cell tumorigenesis. Transgenic ES expression in tumors was analyzed by immunohistochemical stainings of tumor sections of Rip1Tag2; Rip1ES mice and Rip1Tag2 mice. No expression of human ES could be detected in normal islets and in the different stages of tumor progression in Rip1Tag2 single-transgenic mice, also demonstrating the specificity of the immunohistochemical staining procedure (Fig. 2). In comparison with single-transgenic Rip1ES mice, which displayed high levels of ES expression in 100% of the islets, the expression of ES was markedly lower in ␤-cell tumors of Rip1Tag2;Rip1ES double-transgenic mice and seems to decline in large tumors (Fig. 2). It is notable that ES expression was lowest in carcinomas with a dedifferentiated phenotype or in anaplastic tumors. Immunohistochemical staining of insulin expression revealed insulin expression by ␤-tumor cells was not lost concomitantly with the diminished ES expression, indicating that the reduced ES expression was not a result of a loss of insulin-promoter activity (data not shown). These results thus suggested that a selection against ES expression occurred with tumor progression. Similar to Rip1ES single-transgenic mice, ES levels in the serum of Rip1Tag2;Rip1ES double-transgenic mice were not at all increased as compared with Rip1Tag2 mice (Rip1Tag2: 0.53⫾0.21 ng/ml; Rip1Tag2;Rip1ES: 0.38⫾0.25 ng/ml) and were below the detection limit of the ELISA assay for human ES (C57Bl/6 control mice: 1.17⫾0.45 ng/ml; see above). The absence of significant amounts of human ES in the serum of Rip1ES and Rip1Tag2;Rip1ES mice hence indicated that high local expression of ES by ␤-tumor cells did not result in a systemic availability of human ES during Rip1Tag2 tumor progression.

Transgenic ES moderately represses tumor outgrowth and progression Next, we investigated whether the local expression of ES had any effect on Rip1Tag2 tumor development. A moderate, yet

not significant reduction in tumor volumes in Rip1Tag2; Rip1ES mice versus control Rip1Tag2 mice was apparent when sacrificing the mice at 12–13 weeks of age, just before the mice would succumb to insulinoma-induced hypoglycemia (Table 1). Also, no significant difference in tumor number was apparent between single-transgenic Rip1Tag2 and doubletransgenic Rip1Tag2;Rip1ES mice (Table 1). To assess whether transgenic ES repressed tumor progression, tumors were staged according to their morphology into normal islets, hyperplastic islets, adenomas, and carcinomas of three grades with increasing malignancy [56]. These analyses revealed that the ratio was shifted in favor of benign stages in Rip1Tag2;Rip1ES mice as compared with control Rip1Tag2 animals (P⬍0.005; ␹2-test with Yates correction; Table 1). It is interesting that the presence of ES treatment did not prevent progression to anaplastic carcinomas (2.92% in Rip1Tag2 and 3.37% of all islets and tumors in Rip1Tag2;Rip1ES) or lymph node metastasis (21.43% of Rip1Tag2 mice and 30% of Rip1Tag2;Rip1ES mice analyzed). Tumor cell proliferation, as determined by counting tumor cells, which stained positive for incorporated BrdU, was unchanged between Rip1Tag2 and Rip1Tag2;Rip1ES mice (Table 2). Also, the average number of cells undergoing apoptosis in large tumors was unaltered by the presence of ES (Table 2). As the incidence of apoptosis within the different tumor stages and tumor sizes varied tremendously, the statistical evaluation of apoptosis was performed in a more detailed manner. As apoptotic variance was highest in hyperplastic islets, and angiogenesis first occurs in hyperplastic islets, we specifically analyzed these early stages of tumor development and observed two types of apoptotic levels: lesions with low apoptosis and lesions with high apoptosis. Although the lesions with low apoptosis did not differ in their apoptotic index between Rip1Tag2 and Rip1Tag2;Rip1ES mice (10.2⫾4.1 and 10.2⫾4.4, respectively), a trend to increased apoptosis was observed in islets and small tumors with high apoptotic rates in Rip1Tag2;Rip1ES mice as compared with Rip1Tag2 mice (37.9⫾21.7 and 28.1⫾8.1, respectively; Fig. 3). However, this observation was not statistically significant (evaluated by Student’s t-test).

Fig. 2. ES expression during tumor progression in Rip1Tag2;Rip1ES mice. Immunohistochemical analysis of tumors in 13-week-old Rip1Tag2; Rip1ES (A–C) and Rip1Tag2 (D–F) mice with an antihuman ES antibody reveals that expression of the transgene decreases with tumor progression. Although hyperplastic islets (A, B) of double-transgenic mice express high levels of ES, its expression is decreasing with tumor progression from adenoma (B) to carcinoma (A) stage. (C) A higher magnification of B with cytoplasmic ES staining at varying intensities. Islets and tumors of Rip1Tag2 mice do not express detectable amounts of human ES (D–F). Ca, Carcinoma; Ad, adenoma. Original scale bar, 100 ␮m.

4



TABLE 1.

Effect of Local Expression of Human Endostatin on Tumorigenesis in Rip1Tag2 Mice Tumor progression (%)a Adenoma

Carcinoma

Rip1Tag2

72.59

27.41

Rip1Tap2; Rip1ES

80.64

P ⬍ 0.025*

19.36

Tumor volume (mm3)b

Tumor incidencec

40.92 ⫾ 25.48 (Median 41.89) 29.75 ⫾ 31.73 (Median 15.71)

6.67 ⫾ 2.74 5.28 ⫾ 2.05

Tumors and islets (n ⫽ 270) of Rip1Tag2 transgenic mice (N ⫽ 10) and tumors and islets (n ⫽ 408) of Rip1Tag2; Rip1ES double-transgenic mice (N ⫽ 15) were characterized as adenoma (normal, hyperplastic, angiogenic islets, and adenoma) and invasive carcinoma stages as described [56]. b Average and standard deviation (Av ⫾ SD) of tumor volumes in mm3 per mouse of Rip1Tag2 (n ⫽ 100, N ⫽ 15) and Rip1Tag2; Rip1ES mice (n ⫽ 95, N ⫽ 18). z-test, P ⬍ 0.5. c Average and standard deviation (AV ⫾ SD) of macroscopically visible tumors (ⱖ1 mm) per mouse of Rip1Tag2 (N ⫽ 15) and Rip1Tag2; Rip1ES mice (N ⫽ 18). z-test, P ⬍ 0.5. * ␹2-test; P ⬍ 0.025. n, Number of tumors analyzed; N, number of mice analyzed. a

Blood vessel density and morphology are not affected by ES expression

Inhibition of the angiogenic switch in Rip1Tag2; Rip1ES mice

We next investigated whether the expression of human ES had any effect on hallmarks of active angiogenesis, such as blood vessel density and vessel morphology. Immunohistochemical stainings with antibodies against CD31 did not reveal a significant effect on blood vessel density, despite the early and continuous expression of the angiogenesis inhibitor ES during tumor progression (Tab1e 2). Moreover, immunohistochemical analysis with antibodies against podoplanin did not reveal any ES-dependent changes in lymphatic vessel density or organization. In both genotypes, lymphatics were found to be rarely associated with larger surface areas of the tumors, and only a few lymphatics were detectable within the tumor mass. Based on the observation that transgenic ES expression declined with tumor growth and tumor malignancy, it was possible that ES exerted a more profound effect during earlier phases of tumor development. To investigate this possibility, we analyzed tumor progression of 8-week-old mice, a stage when first, small, solid tumors formed. However, no significant differences in tumor progression, tumor cell proliferation, and blood vessel density could be observed between the two genotypes (data not shown).

Next, we assessed whether the early expression of ES had an impact on the onset of angiogenesis. The angiogenic switch can be visualized in Rip1Tag2 mice by coculturing isolated islets of Langerhans at different stages of tumorigenesis with primary endothelial cells in a 3D collagen gel matrix [48]. In the case of an angiogenic islet, which is actively inducing angiogenesis, endothelial cells respond by chemotactic migration, proliferation, and tube formation, whereas in the case of a nonangiogenic islet, endothelial cells do not respond. Islets were isolated from single-transgenic Rip1ES and Rip1Tag2 mice, from double-transgenic Rip1Tag2;Rip1ES mice and from C57Bl/6 nontransgenic littermates between 7 and 8 weeks of age, and their angiogenic potential was compared. This experiment re-

TABLE 2.

Effect of Local Expression of Human Endostatin on Tumorigenesis in Rip1Tag2 Mice

Rip1Tag2 Rip1Tag2; Rip1ES

Tumor proliferationa

Tumor apoptosisb

Tumor microvessel densityc

99.90 ⫾ 48.7 95.3 ⫾ 42.3

10.6 ⫾ 8.0 11.7 ⫾ 11.5

35.8 ⫾ 17.5 37.8 ⫾ 21.2

a Average and standard deviation (AV ⫾ SD) of BrdU incorporation into tumor cells of Rip1Tag2 (n ⫽ 80, N ⫽ 9) and Rip1Tag2; Rip1ES mice (n ⫽ 51, N ⫽ 14) per high-powered field. Student’s t-test, P ⬎ 0.5. b Average and standard deviation (Av ⫾ SD) of apoptotic cell numbers (TUNEL staining per high-powered field) in tumors of Rip1Tag2 (n ⫽ 144, N ⫽ 11) and Rip1Tag2; Rip1ES mice (n ⫽ 81, N ⫽ 10). z-test, P ⬎ 0.5. c Average and standard deviation (Av ⫾ SD) of blood vessels (CD31 staining per high-powered field) in tumors of Rip1Tag2 (n ⫽ 68, N ⫽ 9) and Rip1Tag2; Rip1ES mice (n ⫽ 70, N ⫽ 18). Student’s t-test, P ⬎ 0.5. n, Number of tumors analyzed; N, number of mice analyzed.

Fig. 3. Increased apoptosis in hyperplastic islets and in small adenomas of Rip1Tag2;Rip1ES mice. TUNEL staining of histological sections from Rip1Tag2;Rip1ES (A, B) and Rip1Tag2 (C, D) mice visualizes apoptotic cells in islets and small adenomas of Rip1Tag2;Rip1ES and Rip1Tag2 mice, as indicated. Note that hyperplastic islets and small tumors in double-transgenic mice display a slightly higher incidence of apoptosis than comparable lesions in single-transgenic, control littermates. However, this difference is statistically not significant (Student’s t-test). Original scale bar, 100 ␮m.

Schaffhauser et al. Transgenic expression of endostatin

5

vealed a significantly reduced angiogenic capability of islets from double-transgenic Rip1Tag2;Rip1ES mice as compared with single-transgenic Rip1Tag2 mice. Only 9.1% of cultured islets from Rip1Tag2;Rip1ES mice but 30.8% of islets from Rip1Tag2 mice were angiogenic in this in vitro assay (P⬍0.05 using ␹2-test with Yates correction), indicating that the local expression of ES by ␤-tumor cells interferes with the activity of proangiogenic factors released by angiogenic islets. As expected, islets from nontransgenic, control littermates or singletransgenic Rip1ES mice did not show any angiogenic activity [48]. Together, the results indicate that transgenic, local expression of human ES exerted a significant repression of the angiogenic switch during Rip1Tag2 tumorigenesis, i.e., during the onset of tumor angiogenesis. Yet, it exhibited only a moderate tumor-repressing effect, i.e., during the maintenance of tumor angiogenesis. Moreover, expression of transgenic ES appeared to be down-regulated during tumor progression, and secreted ES was not detectable in the circulation, potentially explaining the failure of locally expressed ES to act as an efficient angiogenesis inhibitor in this mouse model.

DISCUSSION The aim of our study was to assess whether the local tumoral expression of the angiogenesis inhibitor ES would exert a similar effect on tumor growth as the systemic application of high doses of ES. For these studies, we used the Rip1Tag2transgenic mouse model of ␤-cell carcinogenesis, which has been previously used to investigate the molecular mechanisms underlying the angiogenic switch and for the systemic treatment with angiogenesis inhibitors, including ES [48, 52]. Transgenic expression of ES in Rip1ES mice is already active during embryonic development before birth, yet no effects on islet development and physiology are apparent. At the time, the angiogenic switch occurs in Rip1Tag2 mice (between 5 and 8 weeks of age), and ES is expressed at high levels within islets and early, preneoplastic lesions of double-transgenic Rip1Tag2;Rip1ES mice. At this stage of tumor development, transgenic expression of ES in Rip1Tag2;Rip1ES doubletransgenic mice results in a significant repression of the onset of angiogenesis, as determined by an ex vivo collagen assay. Our experiments also reveal a trend to increased apoptosis in hyperplastic and angiogenic islets as well as small adenomas in ES-expressing tumors. Possibly, this enhanced apoptosis participates in the delayed tumor progression and the significantly reduced tumor volume in Rip1Tag2;Rip1ES double-transgenic mice. In a similar manner, systemic treatment with ES has been most efficient during early stages of tumorigenesis, whereas the effect of antiangiogenic therapy on late-stage tumors has been rather moderate, also in Rip1Tag2 tumorigenesis [52]. Yet, the local expression of ES apears to be much less efficient in repressing tumor growth as compared with the systemic delivery of recombinant protein. Although expressed at significant levels and detectable in the supernatant of cultured islets of Rip1ES mice, ES is not detectable in the serum of Rip1ES mice, indicating that a local sequestration may interfere with the antiangiogenic activities of ES. Hence, ad6


ditional mechanisms of release and activation of such latent ES may be required for its full activity. The lack of the antiangiogenic efficiency of ES may also be a result of the fact that the transgenic expression of ES declines with tumor progression, independent of insulin expression, suggesting that a counter-selection against ES-expressing tumor cells occurs and that outgrowing tumors escape the local activity of ES. Such counter-selection during Rip1Tag2 tumorigenesis is not unprecedented. Conditional inactivation of VEGF-A in Rip1Tag2 mice results in the outgrowth of tumors, which have escaped the genetic ablation of VEGF-A gene function or have started to use the activity of other angiogenic factors to support tumor outgrowth, such as members of the FGF family [51]. Previous experiments from our and other laboratories demonstrated that VEGF-A plays a critical role in the initial phase of tumor angiogenesis, whereas FGFs may be critical for the maintenance of angiogenesis at later stages of tumor progression [49 –51]. This notion, together with the observation that ES interferes with VEGF- but not FGF-induced angiogenesis [24, 36], is consistent with our results, demonstrating ES mediated antiangiogenic activity specifically during the early phases of tumorigenesis in Rip1Tag2 mice. Further supporting this notion, prevention trials with systemic delivery of purified ES to Rip1Tag2 mice at early stages of tumorigenesis resulted in an efficient reduction of angiogenesis [52]. Hence, treatment with ES seems to be most effective during the early phases of tumorigenesis, when tumor angiogenesis ensues. Other parameters of tumor growth, such as tumor cell proliferation and apoptosis or tumor blood vessel or lymphatic vessel density, are not affected significantly by ES expression in Rip1Tag2;Rip1ES mice. Systemic treatment of Rip1Tag2 mice with high doses of recombinant ES revealed comparable results with no or only a slight reduction in proliferation and blood vessel density [52]. In contrast to our results, those experiments have achieved effects reaching from 88% reduction in tumor size in an intervention trial to only moderately retarded tumor growth in an end-stage disease-regression trial. The different results are most likely a result of the experimental set-up, the localization and sequestration of ES in the tumor microenvironment, and thus, the freely available concentrations of bioactive ES. Moreover, the site of exposure of the tumor endothelium to the inhibitor could be of importance. Systemic treatment usually delivers the inhibitor via the circulation inside the tumors, where endothelial cells can be targeted easily. In contrast, local sequestration in the extracellular matrix (ECM) may result in local, high levels of ES in the interstitial space, where the accessibility to endothelial cells may be restricted. For instance, in normal and hyperplastic islets of Rip1Tag2 mice, VEGF-A is not accessible to the receptor, as it is sequestered in the ECM. Matrix metalloproteinase-9, up-regulated in angiogenic lesions, releases VEGF-A, which in turn can bind and activate its receptors on endothelial cells, thereby promoting angiogenesis [54]. Such mobilization and activation of pro- and antiangiogenic factors may be a key regulatory event for the regulation of angiogenesis. In this context, it is also interesting to note that ES, produced by the primary tumor environment, has been discovered as a systemic repressor of metastatic outgrowth, further http://www.jleukbio.org

supporting the notion that systemic, high levels in the circulation are critical for the antiangiogenic functions of ES. ES is among the most promising but also most controversially discussed antiangiogenic molecules currently tested in clinical trials (for reviews, see refs. [28, 57]). ES is characterized by a variety of effects and particularly, by its pleiotropic antiangiogenic potential [57]. Its antiangiogenic potential may also depend highly on the activation or differentiation state of the target endothelial cells and therefore, also on their surface protein composition. Yet, its use as an angiogenesis inhibitor seems promising and successful to a certain point, and based on its pleitropic effects, it may be an appropriate agent for combinatorial therapies. For example, the combination of SU5416, a VEGFR-2 tyrosine kinase inhibitor, and low-dose ES has reduced tumor growth in xenograft transplantation models more efficiently than monotherapy alone [58]. Hence, also endogenously occurring inhibitors, such as ES, are promising candidates for further consideration as angiogenesis inhibitors, not at last as a result of their low toxicity and low susceptibility for the development of resistance [9, 21]. Our results indicate that the time of treatment, the mode of delivery, the tumor microenvironment, and the local bioactive concentrations are important issues to consider when designing therapeutic approaches involving ES.

ACKNOWLEDGMENTS This work was supported by the Roche Research Foundation (B. S., G. C.) and the EU-FP6 framework program LYMPHANGIOGENOMICS LSHG-CT-2004-503573 (G. C., K. A.). The monoclonal antibody against mouse podoplanin was obtained from the Developmental Studies Hybridoma Bank under the auspices of the National Institute of Child Health and Human Development. We are grateful to D. Kerjaschki, H. Weich, F. Lehembre, I. Crnic, and T. Schomber for reagents, advice, and scientific input.

REFERENCES 1. Folkman, J. (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat. Med. 1, 27–31. 2. Carmeliet, P. (2005) Angiogenesis in life, disease and medicine. Nature 438, 932–936. 3. Hanahan, D., Christofori, G., Naik, P., Arbeit, J. (1996) Transgenic mouse models of tumor angiogenesis: the angiogenic switch, its molecular controls, and prospects for preclinical therapeutic models. Eur. J. Cancer 32A, 2386 –2393. 4. Tammela, T., Enholm, B., Alitalo, K., Paavonen, K. (2005) The biology of vascular endothelial growth factors. Cardiovasc. Res. 65, 550 –563. 5. Ferrara, N. (2002) VEGF and the quest for tumor angiogenesis factors. Nat. Rev. Cancer 2, 795– 803. 6. Ferrara, N., Gerber, H. P., LeCouter, J. (2003) The biology of VEGF and its receptors. Nat. Med. 9, 669 – 676. 7. Bergers, G., Benjamin, L. E. (2003) Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 3, 401– 410. 8. Folkman, J. (2004) Endogenous angiogenesis inhibitors. APMIS 112, 496 –507. 9. Ferrara, N., Kerbel, R. S. (2005) Angiogenesis as a therapeutic target. Nature 438, 967–974. 10. Ferrara, N., Hillan, K. J., Novotny, W. (2005) Bevacizumab (Avastin), a humanized anti-VEGF monoclonal antibody for cancer therapy. Biochem. Biophys. Res. Commun. 333, 328 –335.

11. Tandle, A., Blazer III, D. G., Libutti, S. K. (2004) Antiangiogenic gene therapy of cancer: recent developments. J. Transl. Med. 2, 22. 12. Ferreras, M., Felbor, U., Lenhard, T., Olsen, B. R., Delaisse, J. (2000) Generation and degradation of human endostatin proteins by various proteinases. FEBS Lett. 486, 247–251. 13. Wen, W., Moses, M. A., Wiederschain, D., Arbiser, J. L., Folkman, J. (1999) The generation of endostatin is mediated by elastase. Cancer Res. 59, 6052– 6056. 14. O’Reilly, M. S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W. S., Flynn, E., Birkhead, J. R., Olsen, B. R., Folkman, J. (1997) Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell 88, 277– 285. 15. Felbor, U., Dreier, L., Bryant, R. A., Ploegh, H. L., Olsen, B. R., Mothes, W. (2000) Secreted cathepsin L generates endostatin from collagen XVIII. EMBO J. 19, 1187–1194. 16. Feldman, A. L., Pak, H., Yang, J. C., Alexander Jr., H. R., Libutti, S. K. (2001) Serum endostatin levels are elevated in patients with soft tissue sarcoma. Cancer 91, 1525–1529. 17. Miosge, N., Sasaki, T., Timpl, R. (1999) Angiogenesis inhibitor endostatin is a distinct component of elastic fibers in vessel walls. FASEB J. 13, 1743–1750. 18. Ma, L., Elliott, S. N., Cirino, G., Buret, A., Ignarro, L. J., Wallace, J. L. (2001) Platelets modulate gastric ulcer healing: role of endostatin and vascular endothelial growth factor release. Proc. Natl. Acad. Sci. USA 98, 6470 – 6475. 19. Ma, L., Hollenberg, M. D., Wallace, J. L. (2001) Thrombin-induced platelet endostatin release is blocked by a proteinase activated receptor-4 (PAR4) antagonist. Br. J. Pharmacol. 134, 701–704. 20. Fukai, N., Eklund, L., Marneros, A. G., Oh, S. P., Keene, D. R., Tamarkin, L., Niemela, M., Ilves, M., Li, E., Pihlajaniemi, T., Olsen, B. R. (2002) Lack of collagen XVIII/endostatin results in eye abnormalities. EMBO J. 21, 1535–1544. 21. Boehm, T., Folkman, J., Browder, T., O’Reilly, M. S. (1997) Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance. Nature 390, 404 – 407. 22. Dixelius, J., Cross, M. J., Matsumoto, T., Claesson-Welsh, L. (2003) Endostatin action and intracellular signaling: ␤-catenin as a potential target? Cancer Lett. 196, 1–12. 23. Hebbar, M., Peyrat, J. P., Hornez, L., Hatron, P. Y., Hachulla, E., Devulder, B. (2000) Increased concentrations of the circulating angiogenesis inhibitor endostatin in patients with systemic sclerosis. Arthritis Rheum. 43, 889 – 893. 24. Hajitou, A., Grignet, C., Devy, L., Berndt, S., Blacher, S., Deroanne, C. F., Bajou, K., Fong, T., Chiang, Y., Foidart, J. M., Noel, A. (2002) The antitumoral effect of endostatin and angiostatin is associated with a down-regulation of vascular endothelial growth factor expression in tumor cells. FASEB J. 16, 1802–1804. 25. Chen, Q. R., Kumar, D., Stass, S. A., Mixson, A. J. (1999) Liposomes complexed to plasmids encoding angiostatin and endostatin inhibit breast cancer in nude mice. Cancer Res. 59, 3308 –3312. 26. Dhanabal, M., Ramchandran, R., Waterman, M. J., Lu, H., Knebelmann, B., Segal, M., Sukhatme, V. P. (1999) Endostatin induces endothelial cell apoptosis. J. Biol. Chem. 274, 11721–11726. 27. Sacco, M. G., Cato, E. M., Ceruti, R., Soldati, S., Indraccolo, S., Caniatti, M., Scanziani, E., Vezzoni, P. (2001) Systemic gene therapy with antiangiogenic factors inhibits spontaneous breast tumor growth and metastasis in MMTVneu transgenic mice. Gene Ther. 8, 67–70. 28. Eder Jr., J. P., Supko, J. G., Clark, J. W., Puchalski, T. A., GarciaCarbonero, R., Ryan, D. P., Shulman, L. N., Proper, J., Kirvan, M., Rattner, B., et al. (2002) Phase I clinical trial of recombinant human endostatin administered as a short intravenous infusion repeated daily. J. Clin. Oncol. 20, 3772–3784. 29. Herbst, R. S., Hess, K. R., Tran, H. T., Tseng, J. E., Mullani, N. A., Charnsangavej, C., Madden, T., Davis, D. W., McConkey, D. J., O’Reilly, M. S., Ellis, L. M., Pluda, J., Hong, W. K., Abbruzzese, J. L. (2002) Phase I study of recombinant human endostatin in patients with advanced solid tumors. J. Clin. Oncol. 20, 3792–3803. 30. Berger, A. C., Feldman, A. L., Gnant, M. F., Kruger, E. A., Sim, B. K., Hewitt, S., Figg, W. D., Alexander, H. R., Libutti, S. K. (2000) The angiogenesis inhibitor, endostatin, does not affect murine cutaneous wound healing. J. Surg. Res. 91, 26 –31. 31. Ma, L., del Soldato, P., Wallace, J. L. (2002) Divergent effects of new cyclooxygenase inhibitors on gastric ulcer healing: shifting the angiogenic balance. Proc. Natl. Acad. Sci. USA 99, 13243–13247. 32. Kreuger, J., Matsumoto, T., Vanwildemeersch, M., Sasaki, T., Timpl, R., Claesson-Welsh, L., Spillmann, D., Lindahl, U. (2002) Role of heparan sulfate domain organization in endostatin inhibition of endothelial cell function. EMBO J. 21, 6303– 6311.

Schaffhauser et al. Transgenic expression of endostatin

7

33. Sasaki, T., Larsson, H., Kreuger, J., Salmivirta, M., Claesson-Welsh, L., Lindahl, U., Hohenester, E., Timpl, R. (1999) Structural basis and potential role of heparin/heparan sulfate binding to the angiogenesis inhibitor endostatin. EMBO J. 18, 6240 – 6248. 34. Sim, B. K. (1998) Angiostatin and endostatin: endothelial cell-specific endogenous inhibitors of angiogenesis and tumor growth. Angiogenesis 2, 37– 48. 35. Salvesen, G. S., Dixit, V. M. (1997) Caspases: intracellular signaling by proteolysis. Cell 91, 443– 446. 36. Kim, Y. M., Hwang, S., Pyun, B. J., Kim, T. Y., Lee, S. T., Gho, Y. S., Kwon, Y. G. (2002) Endostatin blocks vascular endothelial growth factormediated signaling via direct interaction with KDR/Flk-1. J. Biol. Chem. 277, 27872–27879. 37. Sudhakar, A., Sugimoto, H., Yang, C., Lively, J., Zeisberg, M., Kalluri, R. (2003) Human tumstatin and human endostatin exhibit distinct antiangiogenic activities mediated by ␣ v ␤ 3 and ␣ 5 ␤ 1 integrins. Proc. Natl. Acad. Sci. USA 100, 4766 – 4771. 38. Wickstrom, S. A., Alitalo, K., Keski-Oja, J. (2002) Endostatin associates with integrin ␣5␤1 and caveolin-1, and activates Src via a tyrosyl phosphatase-dependent pathway in human endothelial cells. Cancer Res. 62, 5580 –5589. 39. Karumanchi, S. A., Jha, V., Ramchandran, R., Karihaloo, A., Tsiokas, L., Chan, B., Dhanabal, M., Hanai, J. I., Venkataraman, G., Shriver, Z., et al. (2001) Cell surface glypicans are low-affinity endostatin receptors. Mol. Cell 7, 811– 822. 40. Kim, Y. M., Jang, J. W., Lee, O. H., Yeon, J., Choi, E. Y., Kim, K. W., Lee, S. T., Kwon, Y. G. (2000) Endostatin inhibits endothelial and tumor cellular invasion by blocking the activation and catalytic activity of matrix metalloproteinase. Cancer Res. 60, 5410 –5413. 41. Hanai, J., Gloy, J., Karumanchi, S. A., Kale, S., Tang, J., Hu, G., Chan, B., Ramchandran, R., Jha, V., Sukhatme, V. P., Sokol, S. (2002) Endostatin is a potential inhibitor of Wnt signaling. J. Cell Biol. 158, 529 –539. 42. Hanai, J., Dhanabal, M., Karumanchi, S. A., Albanese, C., Waterman, M., Chan, B., Ramchandran, R., Pestell, R., Sukhatme, V. P. (2002) Endostatin causes G1 arrest of endothelial cells through inhibition of cyclin D1. J. Biol. Chem. 277, 16464 –16469. 43. Shichiri, M., Hirata, Y. (2001) Antiangiogenesis signals by endostatin. FASEB J. 15, 1044 –1053. 44. Abdollahi, A., Hahnfeldt, P., Maercker, C., Grone, H. J., Debus, J., Ansorge, W., Folkman, J., Hlatky, L., Huber, P. E. (2004) Endostatin’s antiangiogenic signaling network. Mol. Cell 13, 649 – 663. 45. Goodwin, A. M., D’Amore, P. A. (2002) Wnt signaling in the vasculature. Angiogenesis 5, 1–9.

8


46. Benezra, R., Rafii, S. (2004) Endostatin’s endpoints— deciphering the endostatin antiangiogenic pathway. Cancer Cell 5, 205–206. 47. Hanahan, D. (1985) Heritable formation of pancreatic ␤-cell tumors in transgenic mice expressing recombinant insulin/simian virus 40 oncogenes. Nature 315, 115–122. 48. Folkman, J., Watson, K., Ingber, D., Hanahan, D. (1989) Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature 339, 58 – 61. 49. Inoue, M., Hager, J. H., Ferrara, N., Gerber, H. P., Hanahan, D. (2002) VEGF-A has a critical, nonredundant role in angiogenic switching and pancreatic ␤ cell carcinogenesis. Cancer Cell 1, 193–202. 50. Compagni, A., Wilgenbus, P., Impagnatiello, M. A., Cotten, M., Christofori, G. (2000) Fibroblast growth factors are required for efficient tumor angiogenesis. Cancer Res. 60, 7163–7169. 51. Casanovas, O., Hicklin, D. J., Bergers, G., Hanahan, D. (2005) Drug resistance by evasion of antiangiogenic targeting of VEGF signaling in late-stage pancreatic islet tumors. Cancer Cell 8, 299 –309. 52. Bergers, G., Javaherian, K., Lo, K. M., Folkman, J., Hanahan, D. (1999) Effects of angiogenesis inhibitors on multistage carcinogenesis in mice. Science 284, 808 – 812. 53. Parangi, S., O’Reilly, M., Christofori, G., Holmgren, L., Grosfeld, J., Folkman, J., Hanahan, D. (1996) Antiangiogenic therapy of transgenic mice impairs de novo tumor growth. Proc. Natl. Acad. Sci. USA 93, 2002–2007. 54. Bergers, G., Brekken, R., McMahon, G., Vu, T. H., Itoh, T., Tamaki, K., Tanzawa, K., Thorpe, P., Itohara, S., Werb, Z., Hanahan, D. (2000) Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis. Nat. Cell Biol. 2, 737–744. 55. Christofori, G., Naik, P., Hanahan, D. (1995) Vascular endothelial growth factor and its receptors, flt-1 and flk-1, are expressed in normal pancreatic islets and throughout islet cell tumorigenesis. Mol. Endocrinol. 9, 1760 – 1770. 56. Crnic, I., Strittmatter, K., Cavallaro, U., Kopfstein, L., Jussila, L., Alitalo, K., Christofori, G. (2004) Loss of neural cell adhesion molecule induces tumor metastasis by up-regulating lymphangiogenesis. Cancer Res. 64, 8630 – 8638. 57. Ramchandran, R., Karumanchi, S. A., Hanai, J., Alper, S. L., Sukhatme, V. P. (2002) Cellular actions and signaling by endostatin. Crit. Rev. Eukaryot. Gene Expr. 12, 175–191. 58. Abdollahi, A., Lipson, K. E., Sckell, A., Zieher, H., Klenke, F., Poerschke, D., Roth, A., Han, X., Krix, M., Bischof, M., Hahnfeldt, P., Grone, H. J., Debus, J., Hlatky, L., Huber, P. E. (2003) Combined therapy with direct and indirect angiogenesis inhibition results in enhanced antiangiogenic and antitumor effects. Cancer Res. 63, 8890 – 8898.
