Differential effects of the vascular endothelial growth factor receptor inhibitor PTK787/ZK on tumor angiogenesis and tumor lymphangiogenesis

Transcription

1 55 Differential effects of the vascular endothelial growth factor receptor inhibitor PTK787/ZK on tumor angiogenesis and tumor lymphangiogenesis Tibor Schomber, 1 Adrian Zumsteg, 1 Karin Strittmatter, 1 Ivana Crnic, 1 Helena Antoniadis, 1 Amanda Littlewood-Evans, 2 Jeanette Wood, 2 and Gerhard Christofori 1 1 Institute of Biochemistry and Genetics, Department of Biomedicine, University of Basel; 2 Novartis Pharma AG, Basel, Switzerland Abstract Halting tumor growth by interfering with tumor-induced angiogenesis is an attractive therapeutic approach. Such treatments include humanized antibodies blocking the activity of vascular endothelial growth factor (VEGF)-A (bevacizumab), soluble VEGF receptor (VEGFR) constructs (VEGF-Trap), or small-molecule inhibitors of VEGFR signaling, including PTK787/ZK (PTK/ZK), sorafenib, and sunitinib. PTK/ZK has been shown previously to specifically block VEGF-induced phosphorylation of VEGFR-1, -2 and -3 and thereby to inhibit endothelial cell proliferation, differentiation, and tumor angiogenesis. We have investigated the effect of PTK/ZK on tumor angiogenesis and tumor lymphangiogenesis using the Rip1Tag2 transgenic mouse model of pancreatic B cell carcinogenesis. In Rip1Tag2 mice, tumor angiogenesis is predominantly mediated by VEGF-A, and as expected, PTK/ ZK efficiently impaired tumor blood vessel angiogenesis and tumor growth. Double-transgenic Rip1Tag2;Rip1- VEGF-C and Rip1Tag2;Rip1VEGF-D mice not only exhibit VEGF-A-dependent blood vessel angiogenesis but also tumor lymphangiogenesis induced by the transgenic expression of VEGF-C or -D. In these mouse models, Received 7/18/08; revised 10/16/08; accepted 10/27/08. Grant support: Novartis Pharma (T. Schomber and G. Christofori), NCCR Molecular Oncology of the Swiss National Science Foundation (G. Christofori), EU-FP6 Framework Programme Lymphangiogenomics LSHG-CT (G. Christofori), and Swiss National Science Foundation (I. Crnic and G. Christofori). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Note: Current address for T. Schomber: Bayer HealthCare AG, Wuppertal, Germany; Current address J. Wood: S*BIO Pte Ltd., Singapore Science Park II, Singapore. Requests for reprints: Gerhard Christofori, Institute of Biochemistry and Genetics, Department of Biomedicine, University of Basel, Mattenstrasse 28, CH-4058 Basel, Switzerland. Phone: ; Fax: gerhard.christofori@unibas.ch Copyright C 2009 American Association for Cancer doi: / mct PTK/ZK also repressed tumor blood vessel angiogenesis and tumor growth yet failed to affect tumor lymphangiogenesis and lymphogenic metastasis. Adenoviral delivery of soluble VEGFR-3 also did not prevent tumor lymphangiogenesis in these mice. In contrast, spontaneous tumor lymphangiogenesis, as observed by the stochastic expression of VEGF-C and -D in tumors of neural cell adhesion molecule-deficient Rip1Tag2 mice, was repressed by PTK/ ZK and soluble VEGFR-3. The results indicate that the time of onset and the levels of VEGF-C/D expression may be critical variables in efficiently repressing tumor lymphangiogenesis and that pathways other than VEGFR signaling may be involved in tumor lymphangiogenesis. [Mol Cancer Ther 2009;8(1):55 63] Introduction Blocking tumor-induced angiogenesis has emerged as an attractive therapeutic goal in cancer treatment and is increasingly exploited in the clinics (1, 2). Notably in preclinical models, several approaches have shown high efficacy in blocking tumor angiogenesis and thus attenuating tumor growth. The angiogenic growth factor vascular endothelial growth factor (VEGF)-A has been found highly expressed in the majority ofcancer types and to exert a key function during the so-called angiogenic switch (3). Growth factors of the VEGF family signal through their cognate receptor tyrosine kinases, VEGF receptors (VEGFR) 1, 2, and 3 (4). Many approaches to repress tumor angiogenesis target the VEGF pathway by either depleting the VEGF-A ligand with specific antibodies (bevacizumab) or using soluble VEGFR constructs (VEGF-Trap; ref. 3). Another way to efficiently perturb VEGF-A signaling is to block the kinase activity ofvegfr-1 and -2 by small-molecule inhibitors, such as sorafenib, sunitinib, or PTK787/ZK (PTK/ZK; refs. 5, 6). The receptor tyrosine kinase inhibitor PTK/ZK has been shown previously to inhibit signal transduction by VEGFR-1, -2, and -3 in vitro and to repress angiogenesis by inhibiting human VEGFR-1 and -2 phosphorylation with very high efficacy (IC 50, and Amol/L, respectively) and, to a lesser extent, murine VEGFR-2 and human VEGFR-3 (IC 50, 0.27 and 0.66 Amol/L, respectively; ref. 7). Tumor lymphangiogenesis has been repeatedly shown to correlate with lymph node metastasis and poor prognosis in different types of cancer (8). Thus, repression of tumor lymphangiogenesis by interfering with the activities of the lymphangiogenic factors VEGF-C and -D and their cognate receptor VEGFR-3 may be an attractive avenue to interfere with tumor progression and metastasis. Accordingly, many experimental mouse models ofphysiologic lymphangiogenesis and tumor lymphangiogenesis have been

2 56 PTK/ZK, Tumor Angiogenesis, and Lymphangiogenesis employed to test different approaches for repressing lymphangiogenesis. For example, small interfering RNAmediated depletion ofvegf-c expression or sequestration ofvegf-c and -D by soluble VEGFR-3 reduces tumor lymphangiogenesis and, to varying degrees, metastasis to lymph nodes and distant organs (9 12). Moreover, neutralizing antibodies against VEGFR-3 has lead to an efficient reduction in regenerative lymphangiogenesis, corneal lymphangiogenesis, and tumor lymphangiogenesis and metastasis (13 17). Finally, repression ofinflammatory lymphangiogenesis and tumor lymphangiogenesis is also efficiently repressed by the pharmacologic inhibitors of VEGFR activity ZK261991, E7080, and indolinone derivatives (18 20). Recently, PTK/ZK has also been reported to inhibit tumor lymphangiogenesis in a syngeneic ear transplantation model ofmelanoma (21). We have here investigated the potential effect of PTK/ZK treatment on tumor progression in the Rip1Tag2 transgenic mouse model ofstochastic pancreatic h cell carcinogenesis. In Rip1Tag2 transgenic mice, SV40 T antigen is expressed under the control ofthe rat insulin promoter in the h cells ofpancreatic islets oflangerhans (22). During tumor progression in Rip1Tag2 mice, early proliferating hyperplastic islets can be distinguished from hyperplastic islets that have undergone an angiogenic switch and started to induce angiogenesis (23). These neoplastic lesions subsequently grow out as benign adenoma and eventually progress to invasive carcinoma. Metastasis rarely develops, because these mice succumb to insulinoma-induced hypoglycemia between ages 12 and 14 weeks. The angiogenic switch during Rip1Tag2 tumor progression has been extensively studied. VEGF-A is well known to play a critical role in the onset oftumor angiogenesis, and as expected, interfering with VEGF-A function, by genetic ablation ofthe VEGF-A gene, by adenoviral expression of soluble VEGFRs, or by pharmacologic interference, has resulted in impaired tumor angiogenesis and, with it, in reduced tumor growth (24 27). Previously, we have reported that the transgenic expression ofvegf-c and VEGF-D in pancreatic h tumor cells ofdouble-transgenic Rip1Tag2;Rip1VEGF-C and Rip1- Tag2;Rip1VEGF-D mice results in the induction oftumor lymphangiogenesis and metastatic spread to lymph nodes (VEGF-C) and to lymph nodes and lungs (VEGF-D; refs. 28, 29). Moreover, ablation of the function of neural cell adhesion molecule (NCAM) during Rip1Tag2 tumor progression in Rip1Tag2;NCAM / mice induces the expression ofvegf-c and -D and, with it, up-regulated tumor-associated lymphangiogenesis and lymph node metastasis (30, 31). Here, we report that treatment ofthese mouse models oftumor-associated lymphangiogenesis and lymphogenic metastasis with the VEGFR inhibitor PTK/ZK effectively inhibits tumor blood vessel angiogenesis and tumor growth but does not attenuate tumor lymphangiogenesis and metastasis to lymph nodes or lung. Moreover, although soluble VEGFR-3 represses tumor lymphangiogenesis in NCAM-deficient mice (30), it fails to interfere with tumor lymphangiogenesis in VEGF-C transgenic Rip1Tag2 mice. Together, the results suggest that the timing and levels ofvegf-c and D expression are critical variables for interfering with VEGFR-mediated lymphangiogenesis and that active and ongoing tumor lymphangiogenesis may rely on additional signals other than the activation ofvegfr-3. Materials and Methods Mice The generation and phenotypic analysis ofrip1tag2, Rip1Tag2;Rip1VEGF-C, Rip1Tag2;Rip1VEGF-D, and Rip1Tag2;NCAM / transgenic mice has been described previously (22, 28, 29). All mice were strictly kept in a C57BL/6 genetic background. PTK/ZK was kindly provided by Novartis Pharma. Mice were treated with either 100 mg/kg bodyweight PTK/ZK dissolved in polyethylene glycol 300 (PEG300; Sigma) or PEG300 alone by daily oral administration for 3 weeks starting at age 9 weeks. Recombinant E1/E3-defective adenovirus expressing soluble VEGFR-3 (AdsFlt-4) or EGFP (AdEGFP) was used as described previously (30). For treatment ofrip1tag2;rip1- VEGF-C double-transgenic mice, virus particles were tail vein-injected once a week for 3 weeks. Mice were sacrificed and tumor growth and metastasis to distant organs were assessed. Tumor incidence per mouse was determined by counting all macroscopically detectable pancreatic tumors with a minimal diameter of1 mm. Tumor volumes were calculated by measuring the tumor diameter and assuming a spherical shape ofthe tumors. All experimental procedures involving mice were done according to the guidelines ofthe Swiss Federal Veterinary Office and the regulations of the Cantonal Veterinary Office of Basel Stadt. Histopathologic Analysis Pancreata from transgenic and control mice were isolated and fixed in 4% paraformaldehyde overnight, dehydrated, and embedded in paraffin. Freshly isolated tissue was embedded in OCT compound (Tissue-Tek) and snap frozen in liquid nitrogen. For analysis ofcell proliferation, mice were i.p. injected with 100 Ag bromodeoxyuridine (Sigma)/ g bodyweight 90 min before sacrifice. Histologic analysis was done on H&E-stained paraffin sections. Histologic staging and grading oftumors and the quantification of tumor cell clusters circulating in lymphatic vessels were done in a blindfold manner on H&E-stained sections. For the analysis oflung metastasis in Rip1Tag2;Rip1VEGF-D mice, the entire lungs were cut into 10 Am sections and stained with H&E for microscopic examination. Lymph node metastases were detected by positive insulin staining of regional lymph nodes in pancreatic paraffin sections. The following antibodies were used: rabbit anti-mouse LYVE-1 (Reliatech), rat anti-mouse CD31 (Pharmingen), rabbit anti-mouse NG2 (Chemicon), rat anti-mouse F4/80, biotinylated mouse anti-bromodeoxyuridine (Zymed), and rabbit anti-mouse caspase-3 (Calbiochem). The secondary antibodies used were either conjugated with Alexa Fluor 488 or 568 (Molecular Probes). Nuclei were counterstained with 4,6-diamidino-2-phenylindole.

3 Molecular Cancer Therapeutics 57 Figure 1. PTK/ZK treatment of Rip1Tag2 transgenic mice results in decreased tumor volumes. A, representative H&E staining of Rip1Tag2 pancreatic tumors treated with PEG300 vehicle or PTK/ZK. Tumors of mice treated for 3 wk with PTK/ZK show decreased tumor sizes and central areas of necrosis. Dashed line, tumor margins. Bar, 100 Am. B, quantification of tumor volumes in Rip1Tag2 (n = 14; **, P = ; Mann-Whitney test), Rip1Tag2;Rip1VEGF-C(n = 8; P = 0.007; Mann-Whitney test), and Rip1Tag2;Rip1- VEGF-D (n = 7; P = ; Mann-Whitney test) treated with PTK/ZK or PEG300 vehicle for 3 wk or as indicated for 1 wk (n = 8 for PEG300 vehicle and n =9 for PTK/ZK; P = 0.13; Mann-Whitney test). Note that VEGF-D expression in Rip1Tag2;Rip1VEGF-D mice results in a repression of tumor angiogenesis and thus in smaller tumor volumes (29). CD31 + intratumoral vessels and the numbers ofcells positively staining for NG2 (pericytes), cleaved caspase-3 (apoptosis), and F4/80 (macrophages) were determined using ImageJ software (NIH). 3 Pericyte coverage ofmicrovessels was determined by counting CD31 + /NG2 + vessels. All sections were analyzed with either an Axioskop 2 Plus light microscope using Axiovision 3.1. Software (Zeiss) or a Nikon Diaphot 300 immunofluorescence microscope (Nikon) using Openlab Software (Improvision). Phospho-VEGFR-3 ELISA For the quantification of phospho-vegfr-3 in tumorassociated lymphatics, Rip1Tag2;Rip1VEGF-C mice were treated with either 100 mg/kg PTK/ZK or vehicle control for 3 weeks starting at age 9 weeks. At week 12, the mice (2 mice each) were sacrificed and macroscopically detectable tumors (5 per mouse) were excised and homogenized in ice-cold cell lysis buffer [1% NP-40, 20 mmol/l Tris (ph 8.0), 137 mmol/l NaCl, 10% glycerol, 2 mmol/l EDTA, 1 mmol/l activated sodium orthovanadate, 10 g/ml aprotinin, and 10 g/ml leupeptin]. After removal of cell 3 debris by centrifugation, the protein contents of the cleared lysates were determined and normalized amounts of protein lysates were used for quantitative ELISA assays as recommended by the manufacturer (R&D Systems). AlamarBlue BTumor Cell Proliferation Assay h insulinoma cells were seeded into 96-well plates at 7,000 per well in 150 AL DMEM/10% FCS at day -2. PTK787 was added at day 0. Medium was changed at days -1, 0, and 2. Metabolic activity ofcells, representing proliferation, was measured by removing 75 AL medium and adding 10 AL AlamarBlue (AbD Serotec). Fluorescence was measured 5 h past-alamarblue addition at excitation wavelength of560 nm and emission wavelength of590 nm. Results PTK/ZK Impairs Tumor Growth and Tumor Progression We have investigated the effect of the small-molecule inhibitor PTK/ZK on VEGFR-driven tumor angiogenesis in vivo in the Rip1Tag2 mouse model of h cell carcinogenesis. During multistage tumorigenesis in Rip1Tag2 transgenic mice, VEGF-dependent angiogenesis is first detected in hyperplastic lesions of6-week-old mice. After

4 58 PTK/ZK, Tumor Angiogenesis, and Lymphangiogenesis Figure 2. PTK/ZK retards tumor progression. A, treatment of Rip1Tag2 transgenic mice with PTK/ZK for 3 wk results in significantly reduced numbers of macroscopically detectable tumors (n = 8; ***, P = ). B, tumor progression is decreased in PTK/ZK- versus vehicletreated mice. Percentage of benign hyperplastic islets/adenomas in relation to malignant carcinoma. C, treatment of Rip1Tag2 transgenic mice with PTK/ZK results in increased tumor cell apoptosis. Apoptotic cells in the tumors of PTK/ZK- or vehicle-treated mice were analyzed by immunofluorescent staining for cleaved caspase-3 (n = 5;*, P = 0.04; unpaired Student s t test). D, no significant difference in tumor cell proliferation was detectable in PTK/ZK versus PEG300 control-treated tumors by counting bromodeoxyuridine-positive cells per high-power field (n = 4). the onset oftumor angiogenesis, the neoplastic lesions progress to adenoma and carcinoma at age 12 weeks, when the mice are sacrificed to forestall death by hyperinsulinemia/hypoglycemia. In different studies, these mice were instrumental to show the success ofvegf blockade during tumor growth by pharmacologic and also by genetic means (24, 27, 32). We assessed the efficacy of VEGFR inhibition by PTK/ZK during tumor angiogenesis and lymphangiogenesis in the Rip1Tag2 mouse model. Rip1Tag2 singletransgenic mice, double-transgenic Rip1Tag2;Rip1VEGF-C and Rip1Tag2;Rip1VEGF-D, and NCAM-deficient Rip1Tag2;NCAM / mice were treated daily with PTK/ ZK (100 mg/kg body weight) for 3 weeks starting at age 9 weeks, a time point at which tumor angiogenesis and tumor growth are most active in single-transgenic Rip1Tag2 mice and tumor lymphangiogenesis and lymph node metastasis are occurring in the composite VEGF-C/D transgenic or Figure 3. PTK/ZK does not affect h tumor cell proliferation. Established h tumor cell lines were cultured with increasing concentrations of PTK/ZK. At a concentration of 500 nmol/l, potential VEGF-R signaling is completely inhibited, but no decrease in tumor cell proliferation is observed. Fluorescence signals were normalized to 100% at day 0. Bars, SD. n.s., not significant (pairwise unpaired Student s t test). NCAM-deficient Rip1Tag2 mice. Histologic examination of tumor sections from PTK/ZK-treated and control-treated mice revealed that tumors ofptk/zk-treated mice were smaller and showed large areas ofinternal hemorrhaging in comparison with controls (Fig. 1A). Analysis ofthe total tumor volume in PTK/ZK-treated versus vehicletreated mice revealed a significant reduction in tumor burden per Rip1Tag2 mouse, which was also observed in Rip1Tag2;Rip1VEGF-C and Rip1Tag2;Rip1VEGF-D doubletransgenic mice (Fig. 1B) and Rip1Tag2;NCAM / mice (data nor shown). Moreover, a 1-week short-term treatment oflate-stage 12-week-old Rip1Tag2 transgenic mice also diminished the total tumor volumes when compared with control-treated mice (Fig. 1B). Tumor incidence, as determined by the number of tumors with a diameter >1 mm, was also significantly reduced by the treatment with PTK/ZK (Fig. 2A). Furthermore, histopathologic staging (33) of h cell tumors in PTK/ZK-treated mice revealed that the ratio of epithelial, differentiated adenomas to invasive carcinomas was increased in comparison with control-treated animals, showing that PTK/ZK treatment retarded tumor progression in Rip1Tag2 mice (Fig. 2B). Treatment ofrip1tag2 mice with PTK/ZK resulted in a significant increase of apoptotic cells within tumors as visualized by immunofluorescence staining for cleaved caspase-3 (Fig. 2C). However, tumor cell proliferation was unaffected by PTK/ZK treatment as determined by bromodeoxyuridine incorporation assays (Fig. 2D). Together, these results show that PTK/ZK impairs tumor growth and tumor progression in Rip1Tag2 transgenic mice by affecting tumor cell survival but not tumor cell proliferation. To exclude a direct toxic effect of PTK/ZK on h tumor cells, cultured h tumor cell lines established from Rip1Tag2 transgenic mice were treated with PTK/ZK at increasing doses. Cell proliferation was not significantly affected at the doses

5 Molecular Cancer Therapeutics 59 Figure 4. PTK/ZK treatment reduces tumor angiogenesis in Rip1Tag2 mice. A, analysis of CD31 + vessels by immunohistochemistry (top) and immunofluorescence (bottom) reveals a strong reduction in intratumoral blood vessels and in the total amount of NG2 + pericytes in PTK/ZK- versus controltreated Rip1Tag2 transgenic mice. Bar, 100 Am. B, quantification of CD31 + intratumoral blood vessels, NG2 + pericytes, CD31 + / NG2 + vessels, and F4/80 + macrophages as indicated (n = 5; ***, P < , unpaired Student s t test). comparable with the in vivo therapeutic regimen, and only at the highest dose of10 Amol/L final concentration h tumor cell proliferation was diminished (Fig. 3). PTK/ZK Efficiently RepressesTumor Angiogenesis Because PTK/ZK is known to inhibit VEGFR-mediated signal transduction in endothelial cells, we analyzed tumor microvessel density in PTK/ZK-treated versus vehicletreated mice by determining the number ofcd31 + intratumoral blood vessels. As shown in Fig. 4, blood vessel densities in PTK/ZK-treated tumors were significantly reduced when compared with control-treated tumors. Moreover, together with the reduction in CD31 + intratumoral blood vessels, the overall number ofng2 + pericytes was significantly diminished in PTK/ZK-treated compared with control-treated tumors (Fig. 4A). This decline in the number ofng2 + pericytes was comparable with the observed decrease in CD31 + blood vessels. However, the number ofcd31 + vessels covered by NG2 + pericytes was increased in PTK/ZK-treated versus vehicletreated tumors (Fig. 4A and B). The numbers oftumorinfiltrating F4/80 + macrophages were also significantly decreased in tumors ofptk/zk treated tumors compared with controls (Fig. 4B), indicating a process ofvessel normalization as induced by most antiangiogenic treatment protocols (34). Because VEGFR-1 activity plays an important role in monocyte recruitment, such decrease may be a

6 60 PTK/ZK, Tumor Angiogenesis, and Lymphangiogenesis Figure 5. PTK/ZK treatment does not affect tumor lymphangiogenesis or metastasis. A, representative immunohistochemical staining of LYVE-1-positive lymphatic endothelial cells in the tumor periphery of Rip1Tag2;Rip1VEGF-D tumors treated with PTK/ZK or PEG300 vehicle. Brown, LYVE-1-positive lymphatics. B, quantification of circumferential lymphangiogenesis in Rip1Tag2;Rip1VEGF-C/D double-transgenic mice treated with PTK/ZK or vehicle control (n = 8 for Rip1Tag2;Rip1VEGF-Cand n = 7 for Rip1Tag2;Rip1- VEGF-D). C, phosphorylation of VEFGR-3 is diminished in tumors of PTK/ZK-treated Rip1Tag2;Rip1VEGF-C mice. Tumors were removed from PTK/ZK- and PEG300-treated Rip1Tag2;Rip1VEGF-Cmice, and the tyrosine phosphorylation status of VEGFR-3 in tumor lysates was analyzed by a specific ELISA assay as described in Materials and Methods. D, analysis of lymph node metastasis (filled columns) and lung metastasis (open columns) in Rip1Tag2;Rip1VEGF-D double-transgenic mice treated with PTK/ZK or PEG300 vehicle (n = 7). direct consequence ofa PTK/ZK-mediated repression of VEGFR-1 activity. Decreased numbers ofmacrophages may also contribute to the reduction in tumor angiogenesis and tumor growth (35). Together, these results are consistent with the notion that PTK/ZK represses Rip1Tag2 tumor growth by interfering with tumor angiogenesis. PTK/ZK Does Not AffectTumor Lymphangiogenesis Previously, PTK/ZK has been shown to inhibit the signaling activity ofall three VEGFR tyrosine kinases (7). We thus examined whether PTK/ZK could attenuate VEGF-C/D-driven tumor lymphangiogenesis in Rip1- Tag2;Rip1VEGF-C and Rip1Tag2;Rip1VEGF-D doubletransgenic mice. In these mice, the transgenic expression ofvegf-c or -D in h tumor cells ofrip1tag2 mice leads to enhanced tumor lymphangiogenesis and metastasis to lymph nodes (VEGF-C) and lymph nodes and lung (VEGF- D; refs. 28, 29). Double-transgenic Rip1Tag2;Rip1VEGF-C and Rip1Tag2;Rip1VEGF-D mice were treated daily with PTK/ZK for 3 weeks starting at age 9 weeks, a period during which tumors exhibited dramatic volume expansion and in which peritumoral lymphatics expanded with the increasing tumor volumes (28, 29). Analysis ofthe extent of peritumoral lymphangiogenesis by measuring the LYVE-1- positive circumferential lymphatics revealed no significant reduction in tumor-induced lymphangiogenesis in PTK/ ZK-treated versus vehicle-treated double-transgenic Rip1- Tag2;Rip1VEGF-C and Rip1Tag2;Rip1VEGF-D mice (Fig. 5A and B). However, PTK/ZK efficiently reduced tumor growth as shown by the decrease in total tumor volumes (Fig. 1) and provoked a significant reduction of VEGFR-3 phosphorylation as determined by an ELISA assay specific for the detection of the phosphorylated, activated form of VEGFR-3 (Fig. 5C). Consistent with this observation, daily treatment of5-week-old Rip1VEGF-C single-transgenic mice with PTK/ZK for 3 weeks did not result in reduced lymphatic vessel densities in the periphery ofislets oflangerhans (data not shown), although during this time frame lymphangiogenesis is active in these mice (28). Macrometastases and micrometastases to the lymph node or the lung are only detectable beyond age 12 weeks of Rip1Tag2;Rip1VEGF-D double-transgenic mice, just before the mice have to be sacrificed to avoid complications by hypoglycemia. Hence, a treatment period between 9 and 12 weeks covers a time window at which most ofthe metastatic spread oftumor cells via the lymphatics occurs (28, 29). To evaluate the effect of PTK/ZK treatment on metastasis to the lymph nodes and lungs ofrip1tag2;rip1- VEGF-D double-transgenic mice, the mice were treated for 3 weeks with either PTK/ZK or PEG300 control starting

7 Molecular Cancer Therapeutics 61 at age 9 weeks. As shown in Fig. 5D, the incidence oflymph node and lung metastasis was unaffected by the treatment. These results indicate that, in our experimental mouse models, tumor lymphangiogenesis and metastasis cannot be abrogated by PTK/ZK-mediated VEGFR-3 (and VEGFR-1 and -2) inhibition. To corroborate the inability ofptk/zk to repress VEGF- C-induced tumor lymphangiogenesis in Rip1Tag2;Rip1- VEGF-C double-transgenic mice, we employed adenoviral expression ofa soluble VEGFR-3 construct (sflt-4), where the three extracellular immunoglobulin domains ofvegfr- 3 have been fused to a mouse immunoglobulin Fc domain (30). This treatment regimen only caused a marginal, nonsignificant reduction of tumor lymphangiogenesis in Rip1Tag2;Rip1VEGF-C double-transgenic mice (Fig. 6A). In contrast, as reported previously, adenoviral expression ofsoluble VEGFR3 in Rip1Tag2;NCAM / mice significantly repressed tumor lymphangiogenesis induced by the up-regulation ofendogenous VEGF-C and -D expression in Rip1Tag2;NCAM / mice (30). In these mice, PTK/ZK treatment resulted in a moderate, yet nonsignificant repression oftumor-associated lymphangiogenesis. However, the presence oftumor cell clusters in lymphatic vessels was completely abrogated by the treatment with PTK/ZK (Fig. 6B). Together, the results show that tumor lymphangiogenesis, forced by the continuous transgenic expression of either VEGF-C or VEGF-D during Rip1Tag2 tumorigenesis in Rip1Tag2;Rip1VEGF-C and Rip1Tag2;Rip1VEGF-D double-transgenic mice, is hardly repressed by interfering VEGF-C/D-mediated VEGFR-3 signaling. In contrast, stochastic tumor lymphangiogenesis, as observed by the up-regulation ofendogenous VEGF-C and -D expression in NCAM-deficient Rip1Tag2 mice, can be repressed at moderate and significant levels by PTK and the adenoviral expression ofsoluble VEGFR-3, respectively. Discussion Together, our experiments with the various Rip1Tag2 mouse models oftumor angiogenesis and tumor lymphangiogenesis show that the receptor tyrosine kinase inhibitor PTK/ZK efficiently represses tumor growth by attenuating tumor-induced blood vessel angiogenesis. Even in a 1-week treatment of12-week-old Rip1Tag2 transgenic mice, notably at the final stages of tumor progression, a clear reduction in total tumor volumes was detectable, showing that also at later stages oftumor progression VEGF-A-driven angiogenesis is still a valuable target for restraining tumor growth. Because PTK/ZK inhibits all three VEGFR tyrosine kinases in vitro, VEGF-C/D-induced tumor lymphangiogenesis mediated by VEGFR-3 activation should also be efficiently repressed by PTK/ZK treatment. Yet, in Rip1Tag2;Rip1- VEGF-C and Rip1Tag2;Rip1VEGF-D double-transgenic Figure 6. A, quantification of circumferential lymphangiogenesis in Rip1Tag2;Rip1VEGF-Cdoubletransgenic mice treated with adenovirus expressing soluble VEGFR-3 (AdsFlt-4; 3 mice, 47 tumors) or EGFP (AdEGFP; 2 mice, 59 tumors). B, top, histologic sections from tumors of Rip1Tag2;NCAM / mice to visualize tumor cell clusters within lymphatic vessels by staining with antibodies against VEGFR-3 (Flt-4) or the lymphatic vessel marker LYVE-1 as indicated. Bottom, quantification of circumferential lymphangiogenesis and tumor cell clusters within lymphatic vessels in Rip1Tag2;NCAM / mice treated with PTK/ZK (6 mice, 149 tumors analyzed) versus vehicle control (6 mice, 139 tumors analyzed).

8 62 PTK/ZK, Tumor Angiogenesis, and Lymphangiogenesis mice, where tumor lymphangiogenesis is induced by the transgenic expression ofhigh levels ofvegf-c and -D, respectively, inhibition ofvegfr signaling by PTK/ZK has been unable to efficiently block lymphangiogenesis. In contrast, stochastic tumor lymphangiogenesis, as observed by the expression ofendogenous VEGF-C and -D in Rip1Tag2;NCAM / mice, was efficiently repressed by soluble VEGFR-3 (30) and moderately diminished by PTK/ ZK (shown here). Our observations contrast in part with those ofprevious publications, where various experimental models ofregenerative and inflammatory lymphangiogenesis and tumor lymphangiogenesis with subsequent lymph node metastasis have been employed to effectively inhibit lymphangiogenesis by the use ofsoluble VEGFR-3 immunoglobulin, small interfering RNA-mediated depletion ofvegf-c expression, neutralizing antibodies against VEGFR-3, and various pharmacologic inhibitors against VEGFR activities (see Introduction). These models interfere with lymphangiogenesis either directly at or shortly after the onset oflymphangiogenesis in regenerative, inflammatory, or tumorigenic conditions, and their experimental setting may thus differ from the models used here, where ongoing tumor lymphangiogenesis has been targeted. Notably, the results presented here also indicate that the time ofonset and the levels ofexpression ofthe lymphangiogenic factors VEGF-C and -D are critical variables in defining the efficacy of antilymphangiogenic therapy by either soluble receptor constructs or pharmacologic inhibitors. Moreover, our data raise the intriguing possibility that established and actively ongoing lymphangiogenesis, the maintenance oftumor lymphangiogenesis, may no longer be dependent on VEGFR-3 signaling, which is in line with a previous publication, where anti-vegfr-3 treatment in newborn mice is only effective within the first 2 weeks of age, and lymphangiogenesis becomes VEGF-C independent at later time points (36). We therefore speculate that other signaling pathways replace VEGFR signaling at later stages oftumor lymphangiogenesis. Such switch from one angiogenic growth factor system to another is not without precedence in tumor blood vessel angiogenesis: persistent inhibition ofvegf-a-mediated signaling has resulted in the emergence oftherapy resistance and a switch to fibroblast growth factors and platelet-derived growth factor-mediated angiogenic signaling pathways (27, 32). Preliminary experiments employing soluble fibroblast growth factor receptor constructs did not reveal a function for fibroblast growth factors in this process (data not shown). Therefore, attenuating pre-established and ongoing tumor lymphangiogenesis and metastasis may require other therapeutic approaches than those that target exclusively VEGFR-3 signaling. Disclosure of Potential Conflicts of Interest A. Littlewood-Evans and J. Wood: employees of Novartis Pharma AG. T. Schomber and G. Christofori: received grant support from Novartis Pharma. Acknowledgments We thank U. Schmieder and R. Jost for excellent technical support and Drs. I. Albrecht and M. Cabrita for critically reading the article. References 1. Kerbel R, Folkman J. Clinical translation of angiogenesis inhibitors. Nat Rev Cancer 2002;2: Carmeliet P. Angiogenesis in life, disease and medicine. Nature 2005; 438: Ferrara N, Hillan KJ, Gerber HP, Novotny W. Discovery and development of bevacizumab, an anti-vegf antibody for treating cancer. Nat Rev Drug Discov 2004;3: Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med 2003;9: Escudier B, Eisen T, Stadler WM et al. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med 2007;356: Thomas AL, Trarbach T, Bartel C, et al. A phase IB, open-label doseescalating study of the oral angiogenesis inhibitor PTK787/ZK (PTK/ZK), in combination with FOLFOX4 chemotherapy in patients with advanced colorectal cancer. Ann Oncol 2007;18: Wood JM, Bold G, Buchdunger E, et al. PTK787/ZK , a novel and potent inhibitor of vascular endothelial growth factor receptor tyrosine kinases, impairs vascular endothelial growth factor-induced responses and tumor growth after oral administration. Cancer Res 2000;60: Stacker SA, Achen MG, Jussila L, Baldwin ME, Alitalo K. Lymphangiogenesis and cancer metastasis. Nat Rev Cancer 2002;2: He Y, Rajantie I, Pajusola K, et al. Vascular endothelial cell growth factor receptor 3-mediated activation of lymphatic endothelium is crucial for tumor cell entry and spread via lymphatic vessels. Cancer Res 2005; 65: Chen Z, Varney ML, Backora MW, et al. Down-regulation of vascular endothelial cell growth factor-cexpression using small interfering RNA vectors in mammary tumors inhibits tumor lymphangiogenesis and spontaneous metastasis and enhances survival. Cancer Res 2005;65: Lin J, Lalani AS, Harding TC, et al. Inhibition of lymphogenous metastasis using adeno-associated virus-mediated gene transfer of a soluble VEGFR-3 decoy receptor. Cancer Res 2005;65: Wong SY, Haack H, Crowley D, Barry M, Bronson RT, Hynes RO. Tumor-secreted vascular endothelial growth factor-cis necessary for prostate cancer lymphangiogenesis, but lymphangiogenesis is unnecessary for lymph node metastasis. Cancer Res 2005;65: Pytowski B, Goldman J, Persaud K, et al. Complete and specific inhibition of adult lymphatic regeneration by a novel VEGFR-3 neutralizing antibody. J Natl Cancer Inst 2005;97: Roberts N, Kloos B, Cassella M, et al. Inhibition of VEGFR-3 activation with the antagonistic antibody more potently suppresses lymph node and distant metastases than inactivation of VEGFR-2. Cancer Res 2006;66: He Y, Kozaki K, Karpanen T, et al. Suppression of tumor lymphangiogenesis and lymph node metastasis by blocking vascular endothelial growth factor receptor 3 signaling. J Natl Cancer Inst 2002;94: Kubo H, Cao R, Brakenhielm E, Makinen T, Cao Y, Alitalo K. Blockade of vascular endothelial growth factor receptor-3 signaling inhibits fibroblast growth factor-2-induced lymphangiogenesis in mouse cornea. Proc Natl Acad Sci U S A 2002;99: Shimizu K, Kubo H, Yamaguchi K, et al. Suppression of VEGFR-3 signaling inhibits lymph node metastasis in gastric cancer. Cancer Sci 2004;95: Matsui J, Funahashi Y, Uenaka T, Watanabe T, Tsuruoka A, Asada M. Multi-kinase inhibitor E7080 suppresses lymph node and lung metastases of human mammary breast tumor MDA-MB-231 via inhibition of vascular endothelial growth factor-receptor (VEGF-R) 2 and VEGF-R3 kinase. Clin Cancer Res 2008;14: Hos D, Bock F, Dietrich T, et al. Inflammatory corneal (lymph)angiogenesis is blocked by VEGFR-tyrosine kinase inhibitor ZK , resulting in improved graft survival after corneal transplantation. Invest Ophthalmol Vis Sci 2008;49: Kirkin V, Mazitschek R, Krishnan J, et al. Characterization of indolinones which preferentially inhibit VEGF-C- and VEGF-D-induced activation of VEGFR-3 rather than VEGFR-2. Eur J Biochem 2001;268: Sini P, Samarzija I, Baffert F, et al. Inhibition of multiple vascular endothelial growth factor receptors (VEGFR) blocks lymph node metastases but inhibition of VEGFR-2 is sufficient to sensitize tumor cells to platinum-based chemotherapeutics. Cancer Res 2008;68:

9 Molecular Cancer Therapeutics Hanahan D. Heritable formation of pancreatic h-cell tumours in transgenic mice expressing recombinant insulin/simian virus 40 oncogenes. Nature 1985;315: Folkman J, Watson K, Ingber D, Hanahan D. Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature 1989; 339: Inoue M, Hager JH, Ferrara N, Gerber HP, Hanahan D. VEGF-A has a critical, nonredundant role in angiogenic switching and pancreatic h cell carcinogenesis. Cancer Cell 2002;1: Bergers G, Hanahan D. Combining antiangiogenic agents with metronomic chemotherapy enhances efficacy against late-stage pancreatic islet carcinomas in mice. Cold Spring Harbor Symp Quant Biol 2002; 67: Bergers G, Javaherian K, Lo KM, Folkman J, Hanahan D. Effects of angiogenesis inhibitors on multistage carcinogenesis in mice. Science 1999;284: Compagni A, Wilgenbus P, Impagnatiello MA, Cotten M, Christofori G. Fibroblast growth factors are required for efficient tumor angiogenesis. Cancer Res 2000;60: Mandriota SJ, Jussila L, Jeltsch M, et al. Vascular endothelial growth factor-c-mediated lymphangiogenesis promotes tumour metastasis. EMBO J 2001;20: Kopfstein L, Veikkola T, Djonov VG, et al. Distinct roles of vascular endothelial growth factor-d in lymphangiogenesis and metastasis. Am J Pathol 2007;170: Crnic I, Strittmatter K, Cavallaro U, et al. Loss of neural cell adhesion molecule induces tumor metastasis by up-regulating lymphangiogenesis. Cancer Res 2004;64: Perl AK, Dahl U, Wilgenbus P, Cremer H, Semb H, Christofori G. Reduced expression of neural cell adhesion molecule induces metastatic dissemination of pancreatic h tumor cells. Nat Med 1999;5: Casanovas O, Hicklin DJ, Bergers G, Hanahan D. Drug resistance by evasion of antiangiogenic targeting of VEGF signaling in late-stage pancreatic islet tumors. Cancer Cell 2005;8: Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G. A causal role for E-cadherin in the transition from adenoma to carcinoma. Nature 1998; 392: Jain RK. Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy. Science 2005;307: Condeelis J, Pollard JW. Macrophages: obligate partners for tumor cell migration, invasion, and metastasis. Cell 2006;124: Karpanen T, Wirzenius M, Makinen T, et al. Lymphangiogenic growth factor responsiveness is modulated by postnatal lymphatic vessel maturation. Am J Pathol 2006;169:

10 Differential effects of the vascular endothelial growth factor receptor inhibitor PTK787/ZK on tumor angiogenesis and tumor lymphangiogenesis Tibor Schomber, Adrian Zumsteg, Karin Strittmatter, et al. Mol Cancer Ther 2009;8: Updated version Access the most recent version of this article at: Cited articles Citing articles This article cites 36 articles, 16 of which you can access for free at: This article has been cited by 5 HighWire-hosted articles. Access the articles at: alerts Sign up to receive free -alerts related to this article or journal. Reprints and Subscriptions Permissions To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, use this link Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.