SAR131675, a Potent and Selective VEGFR-3–TK Inhibitor with Antilymphangiogenic, Antitumoral, and Antimetastatic Activities

Published OnlineFirst May 14, 2012; DOI: 10.1158/1535-7163.MCT-11-0866-T

Molecular Cancer Therapeutic Discovery Therapeutics

Antoine Alam1, Isabelle Blanc1, Genevieve Gueguen-Dorbes1, Olivier Duclos2, Jacques Bonnin1, Pauline Barron1, Marie-Claude Laplace1, Gaelle Morin1, Florence Gaujarengues1,Fred erique Dol1, Jean-Pascal Herault 1, Paul Schaeffer1, Pierre Savi1, and Francoise¸ Bono1

Abstract SAR131675 is a potent and selective VEGFR-3 inhibitor. It inhibited VEGFR-3 tyrosine kinase activity and

VEGFR-3 autophosphorylation in HEK cells with IC50 values of 20 and 45 nmol/L, respectively. SAR131675 dose dependently inhibited the proliferation of primary human lymphatic cells, induced by the VEGFR-3

ligands VEGFC and VEGFD, with an IC50 of about 20 nmol/L. SAR131675 was found to be highly selective for VEGFR-3 versus 107 receptors, enzymes, ion channels, and 65 kinases. However, it was moderately active on VEGFR-2 with a VEGFR-3/VEGFR-2 ratio of about 10. SAR131675 had no antiproliferative activity on a panel of 30 tumors and primary cells, further showing its high specificity and indicating that SAR131675 is not a cytotoxic or cytostatic agent. SAR131675 was very well tolerated in mice and showed a potent antitumoral effect in several orthotopic and syngenic models, including mammary 4T1 carcinoma and RIP1.Tag2 tumors. Interestingly, it significantly reduced lymph node invasion and lung metastasis, showing its antilymphangiogenic activity in vivo. Moreover, treatment of mice before resection of 4T1 primary tumors was sufficient to prevent metastasis. Tumor-associated macrophages (TAM) play an important role in tumor growth and metastasis. The expression of VEGFR-3 on TAMs has been recently described. F4/80 immunostaining clearly showed that SAR131675 significantly reduced TAM infiltration and aggregation in 4T1 tumors. Taken together, SAR131675 is the first highly specific VEGFR-3-TK inhibitor described to date, displaying significant antitumoral and antimetastatic activities in vivo through inhibition of lymphangiogenesis and TAM invasion. Mol Cancer Ther; 1–13. 2012 AACR.

Introduction infiltrating stromal cells (3). In experimental tumor mod- In most solid tumors, metastases in regional lymph els, VEGFC and VEGFD expression induces lymphangio- nodes occur frequently in the initial process of cancer genesis and correlates with lymphatic invasion and nodal dissemination and are a strong indicator of poor patient metastasis (4, 5). In addition, agents neutralizing VEGFC prognosis (1). A number of clinical and experimental and VEGFD or blocking VEGFR-3 signaling suppress data suggest that migration of tumor cells into the development of new lymphatic vessels and tumor metas- lymph nodes is greatly facilitated by tumor lymphan- tasis in experimental cancer models (6, 7). giogenesis (2). Although VEGFC and VEGFD are the sole ligands One key molecule mediating tumor lymphangiogenesis described to date as binding directly to VEGFR-3 (8, 9), is the VEGF receptor-3 (VEGFR-3), a tyrosine kinase blunting of both ligands at the same time does not result in the embryonic blood vasculature defects seen in VEGFR- receptor recognized and activated by VEGFC and / VEGFD, commonly expressed in malignant and tumor- 3 mice (10–12). This suggests that additional ligands exist or that ligand-independent signaling occurs. Indeed, whereas VEGFA binds only to VEGFR-2, we and others Authors' Affiliations: 1Sanofi Recherche et Developpement, Early to have shown that VEGFA may induce formation of Candidate DPU, Toulouse; and 2Sanofi Recherche et Developpement, Fibrosis TSU, Chilly Mazarin, France VEGFR-2/-3 heterodimers (13, 14). This suggests that when cells express both receptors, VEGFR-3 inhibitors Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/). may also block some VEGFA signaling. In adults, VEGFR-3 is found only on endothelial cells Corresponding Author: Francoise¸ Bono, Sanofi Recherche et Develop- pement, Early to Candidate DPU, 195, Route d'Espagne, 31036 of lymphatic vessels. However, in malignant tumors, Toulouse, France. Phone: 33-053-463-2517; Fax: 33-053-463-2286; VEGFR-3 expression is found on vascular endothelial cells fi E-mail: francoise.bono@sano .com and particularly on angiogenic sprouts. Genetic targeting doi: 10.1158/1535-7163.MCT-11-0866-T of VEGFR-3 or blocking of VEGFR-3 signaling with 2012 American Association for Cancer Research. monoclonal antibodies results in decreased sprouting and

www.aacrjournals.org OF1

Alam et al.

vascular density, suggesting that VEGFR-3 plays a crucial Cell lines role in tumor angiogenesis (15, 16). Human tumor cell lines were obtained from DSMZ or There is considerable evidence of the expression of the American Type Culture Collection (ATCC). No fur- VEGFR-3 in several primary tumors, including colorectal, ther authentication was carried out. Cells were cultured in ovarian, gastric, and others (17–19), and its upregulation the culture medium in a humidified atmosphere of 5% in breast cancer has been shown to precede tumor cell CO2 and 95% air at 37 C. Cell viability was determined by invasion (20). Interestingly, coexpression of VEGFR-3 trypan blue exclusion and always exceeded 95%. with VEGFC in human cancer cells was associated with Complete cell culture medium for human cell prolifer- increased metastasis and poorer survival, suggesting a ation was RPMI-1640 (Gibco Laboratories) containing possible autocrine loop between VEGFC and VEGFR-3 in 10% heat-inactivated FCS (fetal calf serum; Gibco Labo- cancer cells (21, 22). ratories), 2 mmol/L L-glutamine, 1 mmol/L sodium pyru- Expression of VEGFR-3 has also been reported on vate, 100 UI/mL penicillin, and 100 mg/mL streptomycin macrophages after brain ischemia, kidney transplanta- (Gibco Laboratories). tion, chronic airway inflammation, or corneal injury Adult human lymphatic microvascular endothelial (23–27) and also on tumor-associated macrophages cells (HLMVEC) were purchased from Cambrex, main- (TAM) (28, 29). TAMs are critical regulators of angiogen- tained in culture as described previously (35), and were esis and lymphangiogenesis; they express VEGFA and tested and authenticated in the laboratory for the expres- VEGFC and induce tip-cell formation and fusion (16, 30– sion of VEGFR-3 and Prox1 as lymphatic cell markers. 33). In addition, they are thought to directly contribute to Cells were used up to passage 6. 4T1 mouse mammary lymphatic vessel formation by transdifferentiating into adenocarcinoma cells were obtained from the ATCC lymphatic endothelial cells (25, 28, 33, 34). (CRL-2539) and were maintained in RPMI-1640 contain- Collectively, these data implicate VEGFR-3 in different ing 10% heat-inactivated FCS and 2 mmol/L L-glutamine. processes of tumor growth and metastasis, suggesting Authentication of the cell line was done by expression of that therapeutic targeting of VEGFR-3 might selectively osteopontin as described below and used up to passage 25. reduce tumor growth and metastasis. To obtain maximal beneficial effects without causing Tyrosine kinase assay severe side effects, VEGFR-3 selective compounds are Multiwell plates were precoated with a synthetic poly- required. Although several marketed multikinase inhibi- mer substrate poly-Glu-Tyr (polyGT 4:1). The reaction tors, including sunitinib and sorafinib, have been reported was carried out in the presence of kinase buffer (10:50 to present anti–VEGFR-3 activity, currently there is no mmol/L HEPES buffer, pH 7.4, 20 mmol/L MgCl2, specific VEGFR-3–TK inhibitor under evaluation. Here 0.1 mmol/L MnCl2, and 0.2 mmol/L Na3VO4) supple- we describe, for the first time, a small molecule with high mented with ATP and dimethyl sulfoxide (DMSO) for þ selectivity for VEGFR-3 and show that SAR131675 is able the positive control (C ) or SAR131675 (ranging from 3– to reduce lymphangiogenesis, angiogenesis, and TAM 1,000 nmol/L). ATP was used at 30 mmol/L for VEGFR-1 infiltration and consequently reduce tumor growth and and VEGFR-3 and at 15 mmol/L for VEGFR-2. The phos- metastasis. phorylated poly-GT was probed with a phosphotyrosine specific monoclonal antibody (mAb) conjugated to horse- Materials and Methods radish peroxidase (HRP; 1/30,000; clone PT66; Sigma) Reagents and developed in the dark with the HRP chromogenic Accustain, Drabkin, and the human recombinant hemo- substrate (OPD). The reaction was then stopped by the m globin were obtained from Sigma. The proteome array addition of 100 L 1.25 mol/L H2SO4, and absorbance Phospho-MAPK (mitogen-activated protein kinase), the was determined using an Envision spectrophotometer at DuoSet ELISA Phospho-ERK1/2, and the recombinant 492 nm. proteins bFGF, VEGFA, and VEGFD were purchased Percent inhibition (I%) was calculated using the follow- þ from R&D and VEGFC from ReliaTech, huVEGFR-1-TK ing formula: I% ¼ 100 ([SAR131675] C )/(C C ) from Upstate, and huVEGFR-3-TK from Cell Signalling: 100, in which [SAR131675] represents the value obtained huVEGFR-2-TK was produced internally. in the presence of the indicated concentration of SAR131675. Animals All animal treatment procedures described in this study Autophosphorylation in HEK cells and survival and were approved by the Animal Care and Use Committee of migration assays of HLMVCs sanofi. Female BALB/cByj mice and RIP1.Tag2 mice on Experiments were conducted as described previously the C57Bl/6J background were obtained from Charles (35). Twenty-four hours after transfection, the cells were River France. treated during 1 hour with orthovanadate (100 mmol/L), harvested, collected, and, after counting, distributed in 5- Treatment mL tubes in the presence of the indicated concentration of SAR131675 was dissolved on the day of use in a 0.6% SAR131675. After 30 minutes of incubation, the reaction methylcellulose/0.5% Tween 80 solution. was stopped by the addition of cold PBS supplemented by

OF2 Mol Cancer Ther; 2012 Molecular Cancer Therapeutics

SAR131675, a Potent and Selective VEGFR-3 Inhibitor

orthovanadate. Cells were then lysed with 150 mLof thetized with tricaine, and the number of intersegmental radioimmunoprecipitation assay (RIPA) buffer over 15 vessels was analyzed by fluorescence microscopy (2.5 minutes at 4C and then centrifuged for 10 minutes at magnification). 10,000 g. Supernatants were distributed in duplicate on 96-well plates precoated with the anti-Flag and left for 1 Angiogenesis and lymphangiogenesis induced in a hour at room temperature. After 3 washes, the anti–phos- sponge implant mouse model photyrosine conjugated to the HRP was added and incu- Sterile sponge disks (Cellspon; Interchim) impregnated bated for 1 hour at room temperature. Wells were then with 200 mg of FGF2 or PBS were subcutaneously intro- washed 3 times with TBS buffer containing 0.5% Tween 20 duced on the back of anaesthetized mice. FGF2 was and 2 mmol/L MgCl2. The reaction was stopped with 50 reinjected into the sponges the first 2 days. Daily oral m Lof2NH2SO4, and the signal was read using an Envision treatment with SAR131675 (30, 100, and 300 mg/kg/d) spectrophotometer at 485 and 530 nm. started the day of sponge implantation. Seven days later, the animals were euthanatized and the sponges were Survival assay removed, harvested, and lysed in RIPA buffer at 4C. HLMVECs were seeded in 96-well plates coated with After a centrifugation at 6,000 g, the supernatants were 0.3% gelatin (5 103 cells per well). Cells were incubated collected for further analysis. in RPMI 0.1% FCS with VEGFA (10 ng/mL) VEGFC (300 ng/mL), VEGFD (300 ng/mL), or FGF2 (10 ng/mL) in the RIP1.Tag2/transgenic mouse models absence or presence of SAR131675. Five days later, viable For the prevention study, treatment of mice (9 per cells were quantified with the cell Titer-glo luminescent group) started at 5 weeks of age for 5 weeks and then the cell viability assay (Promega) as previously described number of Langherans islets (red islets) was determined (35). after retrograde perfusion with collagenase solution through the common bile duct (36). For the intervention Migration assay study, mice were treated daily from 10 weeks of age for 16 The migration assay was carried out with the BD Bio- days and tumor volume was measured and calculated as Coat Angiogenesis System-Endothelial Cell Migration for 4T1 tumors below. The tumor burden was calculated Kit (BD Biosciences). HLMVECs (1 105 cells per well as the sum of individual tumor volumes for each mouse. in RPMI) were added in triplicate in the upper chambers. For the survival study, daily treatment (20 mice per group) The lower chambers were loaded with RPMI alone, RPMI started at 12 weeks of age and mice were monitored daily with 50 ng/mL VEGF-A or 100 ng/mL VEGF-C. After 24 to detect moribund mice. hours at 37C, cells were labeled with calcein AM (Molec- ular Probes, Inc.) according to the manufacturer’s instruc- 4T1 mammary carcinoma model tions. Fluorescence of cells that had migrated was mea- 4T1 cells (105) were implanted into mammary fat pads sured on a TECAN GENios Microplate reader (TECAN). of BALB/c mice (15 per group; ref. 37). Daily oral treatment with SAR131675 (30 and 100 mg/kg/d) started at Proteome array and Erk phosphorylation day 5. Tumors were measured 2 to 3 times weekly with HLMVECs were plated in 6-well tissue culture plates at calipers. The tumor volume (V) was calculated using the a density of 2 105 cells for 48 hours. Cells were then formula V ¼ 0.52 a2 b (a: smallest tumor diameter and serum deprived for 2 hours and stimulated or not for 10 b: largest tumor diameter). At day 21, the tumors, the minutes with VEGFC (500 ng/mL) in the presence or not lungs, and the axillary lymph nodes were removed. The of the indicated concentration of SAR131675, lysed and number of metastases at the surface of each lung was subjected to phospho-MAPK array or to phospho Erk counted. The tumors and the lymph nodes were lysed in ELISA according to manufacturer’s instructions. RIPA buffer at 4C. The osteopontin level in lymph nodes was quantified using an ELISA kit (Assay Design). Quantification of VEGFR-2 phosphorylation by ELISA 4T1 tumor excision model Quantification of VEGFR-2 phosphorylation by ELISA Implantation of cells and treatment was started as was carried out as described previously using porcine described above, but primary tumors were removed at aortic endothelial cells (PAEC) expressing VEGFR-2 (35). day 15 in all groups and the tumor weight was evaluated. The mice treated with SAR131675 were then divided into 2 Blood vessel formation in zebrafish groups (12 mice per group); in the first group, treatment Embryos from transgenic zebrafish (Danio rerio) that with SAR131675 was stopped whereas in the second, mice specifically express the fluorescent protein copGFP in the were treated up to day 26. At day 26, lung metastatic foci vascular system were used. The chorion of the 24 hpf were counted. (hours post fertilization) embryos were removed, and embryos were dispensed in a 96-well assay plate in Immunochemistry embryo water mixed with SAR131675 solubilized in Tissues were fixed with 10% accustain for 24 hours, DMSO. After 24 hours at 28.5C, the embryos were anes- dehydrated, and embedded in paraffin. Immunostaining

www.aacrjournals.org Mol Cancer Ther; 2012 OF3

Alam et al.

A O HN

NNNH HO 2

B C 100 100 Figure 1. Effect of SAR131675 on 80 80 VEGFR tyrosine kinases. A, structure of SAR131675. B, effect 60 60 of SAR131675 on kinase activity of 40 40 recombinant human VEGFR-3 measured in the presence of 30 % Inhibition % 20 Inhibition % 20 mmol/L of ATP. C, effect of SAR131675 on VEGFR-3 0 0 autophosphorylation after overexpression in HEK cells. These 100101 1,000 1 10 100 1,000 data are representative of several assays. D, effect of VEGFA on [SAR131675] nmol/L [SAR131675] nmol/L VEGFR-2 phosphorylation in PAEC stably expressing VEGFR-2. E, D E 120 effect of SAR131675 on VEGFA- induced VEGFR-2 100 phosphorylation. 80

Control VEGF-A 60

Anti-VEGFR-2 200 kDa 40 % Inhibition 20 Anti-PY 200 kDa 0

101 100 1,000 10,000

[SAR131675] nmol/L

was done in sections incubated with anti–mouse CD31 Ab regression analysis, 2-way ANOVA with repeated mea- (1/50), LYVE1 (1/1,000; Santa Cruz) or F4-80 (1/50; eBios- sures was carried out. ciences, BD PharMingen) following incubation with the Vectastain ABC Kit (Vector laboratories) appropriate to Results the species of primary antibody, and antigens were devel- In vitro 0 effect of SAR131675 on VEGFR-3 tyrosine oped with 3,3 -diaminobenzidine peroxidase substrate kinase activity and counterstained with hematoxylin. Images were cap- SAR131675 is the result of high-throughput screening tured with a camera (SONY) mounted on a Nikon and chemical optimization (Fig. 1A). Its ability to inhibit microscope. the kinase activity of VEGFR-3 was first determined by ELISA using recombinant human (rh)-VEGFR-3, with Statistical analysis ATP and polyGT as substrates. SAR131675 dose depen- IC50 values were obtained using internal software dently inhibited rh-VEGFR-3–TK activity with an IC50 of Biost@t-SPEED v2.0 with the 4-parameter logistic model. 23 7 nmol/L (Fig. 1B, n ¼ 4). Under the same conditions, For the in vivo studies, results are given as mean SEM. SU11248 (sunitinib) inhibited VEGFR-3–TK activity with Differences between groups were examined for statistical an IC50 of 10 nmol/L (a representative set of data with significance using 2-wayANOVA following Dunnett test sunitinib is shown in Supplementary Fig. S1). SAR131675 or confidence interval calculation as specified in the is an ATP-competitive inhibitor; it inhibited VEGR-3–TK results part. Statistical analysis of data from the survival activity with a Ki of about 12 nmol/L, similar to its IC50 study was conducted using a log-rank test. For tumor value (Supplementary Fig. S1).

OF4 Mol Cancer Ther; 2012 Molecular Cancer Therapeutics

SAR131675, a Potent and Selective VEGFR-3 Inhibitor

Table 1. Effect of SAR131675 on human VEGFR-1 and VEGFR-2 using recombinant enzymes or after overexpression in HEK cells

VEGFR-1 VEGFR-2 rhVEGFR-1 HEK-VEGFR-1 rhVEGFR-2 HEK-VEGFR-2 SAR131675 I% ¼ 31 at 3 mmol/L I% ¼ 48 at 1 mmol/L 235 nmol/L 280 nmol/L

NOTE: The IC50 geometric means are represented.

The effect of SAR131675 on VEGFR-3 autophosphory- erate activity toward VEGFR-2 and with no cytostatic or lation was evaluated after overexpression in HEK cells. cytotoxic effects on tumor or primary cells. SAR131675 was seen to be cell permeable and inhibited VEGFR-3 autophosphorylation in a dose-dependent man- In vitro effect of SAR131675 on lymphatic cell ner with an IC50 ranging from 30 to 50 nmol/L (Fig. 1C). In survival and migration this assay, sunitinib inhibited VEGFR-3 autophosphory- We evaluated the effect of SAR131675 on the in vitro lation with an IC50 ranging from 10 to 30 nmol/L (not survival of primary human lymphatic cells induced by shown). SAR131675 was also evaluated on 2 different specific VEGFR-3 ligands, VEGFC and VEGFD, and by variants of VEGFR-3 and also on murine flt4. The results unrelated growth factors, VEGFA and FGF-2. As indicated that SAR131675 has a similar inhibitory effect on shown in Fig. 2A, SAR131675 potently inhibited lym- both human variants and murine VEGFR-3 (Supplemen- phatic cell survival induced by VEGFC and VEGFD tary Fig. S1). (IC50 of 14 and 17 nmol/L, respectively). It inhibited The selectivity of SAR131675 toward VEGFR-1 and VEGFA-induced survival with an IC50 of 664 nmol/L, VEGFR-2 was evaluated using rh-VEGFR-1 and rh- confirming its moderate activity on VEGFR-2 and had VEGFR-2 ELISA assays as for VEGFR-3. SAR131675 inhib- no effect on FGF2-induced proliferation. In the same > m ¼ ited VEGFR-1–TK activity with an IC50 3 mol/L (I% experiment, sunitinib inhibited VEGFA-, VEGFA-C-, m 31 at 3 mol/L) and VEGFR-2–TK activity with an IC50 of and VEGFA-D–induced survival with an IC50 of about 235 nmol/L (n ¼ 2; Table 1). 5 nmol/L and, consistent with reported data (38), it had In the same assay, sunitinib inhibited VEGFR-1 and no effect on FGF2-induced survival (Supplementary VEGFR-2 with an IC50 of 64 and 14 nmol/L, respectively Fig. S1). (not shown). In the autophosphorylation assay in HEK We then evaluated the effect of SAR131675 on cells, SAR131675 inhibited VEGFR-1 autophosphoryla- HLMVEC migration induced by VEGFA or VEGFC in m tion with an IC50 of about 1 mol/L and VEGFR-2 with an Boyden chambers. SAR131675 potently inhibited VEGFC- < IC50 of about 280 nmol/L (Table 1). These results con- induced migration with an IC50 30 nmol/L (Fig. 2B) and firmed that SAR131675 moderately inhibits VEGFR-2 and VEGFA-induced migration with an IC50 of about 100 has very little effect on VEGFR-1, showing a good selec- nmol/L (Fig. 2C). tivity for VEGFR-3, which is not the case for sunitinib. Using the Phospho-MAP array in HLMVECs, Erk was To further confirm the activity of SAR131675 on identified as the major phosphorylated kinase upon VEGFR-2, we used PAECs stably expressing human VEGFC stimulation (Fig. 2D). SAR131675, significantly VEGFR-2. SAR131675 inhibited VEGFA-induced and dose dependently inhibited VEGFC-induced Erk VEGFR-2 phosphorylation in a dose-dependent manner, phosphorylation, with an IC50 of about 30 nmol/L (Fig. with an IC50 of 239 nmol/L, confirming the results 2E), confirming that SAR131675 is a potent inhibitor of obtained in HEK293T cells after VEGFR-2 overexpression VEGFR-3 signaling. (Fig. 1D and E). The selectivity of SAR131675 was further confirmed as Effect of SAR131675 on angiogenesis and it was inactive on a panel of 65 kinases, on a panel of 107 lymphangiogenesis in nontumoral models in vivo nonkinase enzymes and receptors, and on 21 ion channels The in vivo activity of SAR131675 was first investigated (Supplementary Tables S1, S2A and B, respectively). We in embryonic angiogenesis using the zebrafish model ruled out the eventuality of any nonspecific cytotoxic or (Danio rerio) expressing GFP under the control of a vas- cytostatic effect of SAR131675 by investigating its effect on cular specific promoter. Treatment with SAR131675 the proliferation of tumor cell lines. SAR131675 did not started when blood flow began, 24 hours postfertilization. significantly inhibit the proliferation of any of the cell lines Embryos treated with 3 mmol/L SAR131675 for 24 hours studied at concentrations up to 10 mmol/L (Supplemen- showed fewer intersegmental vessels in the tail compared tary Table S3), thus confirming the specificity of with control embryos and at the dose of 10 mmol/L SAR131675. intersegmental vessels largely failed to form (Fig. 3A). Together, these results showed that SAR131675 is a Thus, in this model, SAR131675 efficiently impaired potent and selective inhibitor of VEGFR-3–TK with mod- embryonic vasculogenesis.

www.aacrjournals.org Mol Cancer Ther; 2012 OF5

Alam et al.

A B 2,500

120 2,000 VEGF-D 1,500 100 *** VEGF-C *** 1,000 *** 80 VEGF-A 500

FGF-2 Luminescence (AU) 60 0 NS 300100300 40 C VEGF-C + SAR131675 (nmol/L)

% inhibition % 2,500 20 2,000 *** 0 1,500 *** *** –20 1,000 0 10 100 1,000 500 Luminescence (AU) [SAR131675] nmol/L 0 0NS 300100 1,000 D E VEGF-A + SAR131675 (nmol/L) + 0.8 NS

0.6 ** ** + + 0.4 *** VEGFC + O.D.(450 nm) 0.2 *** *** Erk + Akt 0.0 + NS VEGF-C 300100301031 500 ng/mL 10 mn VEGFC + SAR131675 (nmol/L)

Figure 2. Effect of SAR131675 on human primary lymphatic cells. A, effect on lymphatic cell survival. Cells were plated in medium containing 0.1% FCS. Cells were then left unstimulated or stimulated with VEGFA (10 ng/mL), VEGFC (300 ng/mL), VEGFD (300 ng/mL), and bFGF (10 ng/mL) in the presence of increasing SAR131675 concentrations. Data are expressed as percentage of inhibition and plotted on SPEED software. Results are representative of 3 independent experiments. B and C, migration assay of human microvascular endothelial cells (HMVEC). Cells were seeded in triplicate in the upper wells of a fluoroblok chamber. RPMI alone or containing VEGF-A 50 ng/mL or VEGF-C 100 ng/mL was added in the bottom wells in the presence of increasing doses of SAR131675 (30–300 nmol/L for VEGFC and 100–1,000 nmol/L for VEGFA). The amount of migrating cells was quantified using Calcein labeling. Data are shown as mean SEM. D, effect of VEGFC on MAP kinases. HMVECs were left unstimulated or stimulated for 10 min with VEGFC (500 ng/mL) and lysed. þ, positive controls; the spots corresponding to Erk and Akt are indicated. Cell lysates were than incubated with array membranes. E, effect of SAR131675 on Erk phosphorylation in lymphatic cells. Cells were stimulated as described in B in the presence of increasing concentration of SAR131675. Erk phosphorylation was quantified by ELISA. Results are expressed as mean SEM; significant differences were assessed by ANOVA followed by Dunnett test versus VEGF-C. , P < 0.05; , P < 0.01; , P < 0.001.

Using a corneal model, a high concentration of FGF2 has Seven days later, SAR1316765 at 100 mg/kg/d had signi- been reported to induce VEGFR-3–dependent lymphan- ficantly reduced the levels of VEGFR-3 (Fig. 3H) and giogenesis (39). We therefore developed a similar hemoglobin content (Fig. 3I) by about 50%, (P < 0.001). approach by subcutaneous implantation of a sterile Treatment with 300 mg/kg/d gave levels of VEGFR-3 and sponge disk impregnated with FGF2 in mice (Fig. 3B). In hemoglobin that were similar to those of the control this model, FGF2 induces angiogenesis that is measured group. Thus, SAR131675 efficiently abrogates lymphan- by CD31 immunostaining and hemoglobin content (Fig. giogenesis and angiogenesis induced in vivo by FGF2. 3D), as well as lymphangiogenesis that is measured by Furthermore, at 100 mg/kg SAR131675 did not show LYVE-1 immunostaining and by VEGFR-3 quantification any effect on arterial pressure whereas at 300 mg/kg, a using ELISA (Fig. 3E). FGF2 induced a 3- to 5-fold increase minor and transient increase of blood pressure was of hemoglobin and VEGFR-3, thus confirming that in observed (Supplementary Fig. S2). Because the inhibition these conditions, FGF2 induces angiogenesis and lym- of VEGFR-2 signaling is known to increase blood pressure phangiogenesis (Fig. 3I). and given our in vitro results, we concluded that Treatment of mice with SAR131675 (30, 100, or 300 SAR131675 at a dose of 300 mg/kg is able to inhibit both mg/kg/d) started on the day of sponge implantation. VEGFR-2 and VEGFR-3 signaling in vivo. In consequence

OF6 Mol Cancer Ther; 2012 Molecular Cancer Therapeutics

SAR131675, a Potent and Selective VEGFR-3 Inhibitor

Figure 3. In vivo effect of SAR131675 on embryonic and adult angiogenesis and lymphangiogenesis. A, lateral view of zebrafish embryos that specifically Control SAR131675 3 µmol/L SAR131675 10 µmol/L express the fluorescent protein copGFP in the vascular system. Controls show the usual pattern of intersegmental blood vessels B D F G that failed to form after treatment with 3 and 10 mmol/L of SAR131675 (arrows). B and C, induction of angiogenesis and lymphangiogenesis by FGF2 (200 ng) in a sponge implanted in the back of C E mice. Angiogenesis and lymphangiogenesis were evaluated at day 7 by immunochemistry staining with CD31 (D) and anti- LYVE-1 antibodies (E), respectively. Visual effect of vehicle (F) and SAR131675 (100 mg/kg/d; G) on H I FGF2-induced angiogenesis and 3.5 lymphangiogenesis. Dose effect of 300 3.0 SAR131675 on VEGFR-3 levels (H) x 4 250 x 2.9 and on hemoglobin content (I) in the ** 2.5 ** 200 * *** sponge. Results are expressed as 2.0 *** 150 mean SEM, by ANOVA followed by *** 1.5 *** Dunnett test versus FGF-2– 100 stimulated group. , P < 0.01; 1.0 VEGFR3 (pg/mL) 50

, P < 0.001. Hemoglobin (mg/mL) 0.5 0 0 Vehicle 0 30 100 300 Vehicle 0 30 100 300

FGF-2 + SAR131675 (mg/kg/d) FGF-2 + SAR131675 (mg/kg/d)

100 mg/kg was the highest dosage used in subsequent in control mice. In the survival study, in which treatment vivo studies. started at week 12 and continued until the death of the mice, mice under vehicle treatment started to die at 12 Effect of SAR131675 on carcinogenesis in the weeks with a median survival age of 13 weeks. Daily RIP1.Tag2 mice model treatment with SAR131675 significantly extended the The RIP1.Tag2 transgenic mouse is a well-characterized median survival age of the mice to 15 weeks (Fig. 4D, multistep carcinogenesis model, which arises from tar- P < 0.01). Taken together, these results showed that geted oncogene expression in the insulin-producing beta SAR131675 is able to prevent the angiogenic switch and cells (40). We investigated the efficacy of SAR131675 on slow tumor growth. the angiogenic switch (prevention study), on asymptom- atic small tumors (intervention study), and finally on Effect of SAR131675 on growth and metastases of advanced near-end-stage cancers (survival study) as illus- 4T1 mammary carcinoma tumors in mice trated in Fig. 4A. 4T1 cells were implanted orthotopically into mammary In the prevention study, 5 weeks treatment with fat pads in mice who were then orally treated with vehicle SAR131675 was well tolerated and the number of angio- or SAR131675 (30 and 100 mg/kg/d) from days 5 to 21 genic islets in the pancreas of SAR131675-treated mice after implantation. was significantly decreased (42%, P < 0.001) compared SAR131675 was well tolerated as no loss of body weight with the vehicle-treated group (Fig. 4B). In the interven- was observed in the 2 treated groups. Treatment with tion study, daily oral administration of SAR131675 from SAR131675 significantly reduced the tumor volume (24% week 10 to week 12.5 caused a significant decrease in P < 0.05 and 50% P < 0.001 at 30 and 100 mg/kg/d, tumor burden (62%, P < 0.05, Fig. 4C). In these 2 studies a respectively; Fig. 5A). significant decrease of CD31-positive vessels was Development of 4T1 tumors was associated with impor- observed in the SAR131675-treated mice compared with tant vascularization and peritumoral lymphangiogenesis

www.aacrjournals.org Mol Cancer Ther; 2012 OF7

Alam et al.

† Figure 4. Effect of SAR131675 on spontaneous multistage tumor Wk 5 Wk 10 Wk 12 model (RIP1.Tag2; A) studies designed to target discrete stages Prevention Intervention Survival of tumoral growth in the RIP1.Tag2 transgenic mice leading to death of BCthe mice (†). In the prevention study (B), mice were treated with vehicle 18 or with SAR131675 (100 mg/kg/d), 16 from week 5 to week 10, and the 35 SAR131675 @ 10 Wk SAR131675 @12 Wk ) number of angiogenic islets (the 3 14 30 transition from prevascular 12 hyperplasia to highly vascularized 25 10 and progressively outgrowing 20 *** tumors) were measured at 10 8 * weeks of age. For the intervention 15 6 study (C), mice were treated with 10 4 vehicle or with SAR131675 Tumor burden (mm (100 mg/kg/d) starting at week 10 Nb of angiogenic islets angiogenic Nb of 5 2 and the total tumor volume was 0 0 calculated at week 12 by Control SAR131675 Control SAR131675 measuring each tumor volume with calipers. Error bars represent the D SEM of data from 9 mice per group. 100 , P < 0.05; , P < 0.001. Student t test versus control group. D, monitoring of RIP1.Tag2 mice 75 survival following daily treatment with vehicle or with SAR131675 50 SAR131675 (100 mg/kg) starting from week 12 (n ¼ 20 mice per group, P = 0.01 Control P ¼ 0.01, log-rank test).

% mice survival 25

0 10 12.5 15 17.5 Weeks

as measured by CD31 and anti-LYVE1 immunostaining (P < 0.05) at 30 and 100 mg/kg/d, respectively (Fig. 5D). (Supplementary Fig. S3A and C, respectively). Interest- Although sunitinib strongly reduced tumor growth (82%, ingly, as illustrated in Supplementary Fig. S3B and D, Supplementary Fig. S4A), it is interesting to note that in SAR131675 seems to reduce both parameters. Moreover, this model, sunitinib at 50 mg/kg/d had no effect on the SAR131675 at 30 and 100 mg/kg/d significantly reduced number of lung metastases (Supplementary Fig. S4A). VEGFR-3 levels in the tumors by 39% (P ¼ 0.05) and 51% Altogether, these results showed that SAR131675 is a (P < 0.01), respectively (Fig. 5B), thus confirming its potent potent antitumoral agent, acting through antiangiogenic antilymphangiogenic properties. and antilymphangiogenic effects. Moreover, it reduces Because human breast cancer leads to high levels of the migration of cancer cells into lymph nodes and lungs. osteopontin, we quantified osteopontin levels in 4T1 cells in vitro and showed a good correlation with cell number Effect of preoperative treatment with SAR131675 on (R2 ¼ 0.998, supplementary Fig. S3E and F). Interestingly, distant metastasis the osteopontin content dramatically increased in sentinel Recent reports suggest that antiangiogenic compounds lymph nodes, indicating an infiltration by metastatic 4T1 may promote tumor invasion after resection of the pri- cells (Fig. 5C). SAR131675, at 100 mg/kg, significantly mary tumor (41). We evaluated the effect of SAR131675 on reduced the osteopontin content in lymph nodes (56%, P < distant metastasis in the 4T1 model after resection of the 0.01), indicating that SAR131675 has a potent effect on primary tumor (Fig. 6A). Treatment was started on day 5- lymph node invasion by 4T1 cells (Fig. 5C). post cell injection, and the primary tumors were excised At day 21, the number of macroscopic lung metastases by surgery at day 15. By this time, the antitumoral effect of was reduced in the treated groups by 17% (ns) and 28% SAR131675 was already significant (Fig. 6B and C).

OF8 Mol Cancer Ther; 2012 Molecular Cancer Therapeutics

SAR131675, a Potent and Selective VEGFR-3 Inhibitor

A B 1,800 Vehicle 1,600 1,600 SAR131675 30 mg/kg/d ) 3 1,400 SAR131675 100 mg/kg/d Figure 5. Antitumoral effect of 1,200 SAR131675 treatment on a murine 1,200 P = 0.05 mammary carcinoma model. A, *** 1,000 ** ** tumor volume monitoring in Balb/c 800 ** 800 mice in control or mice treated with * *** 600 *** SAR131675 at 30 and 100 mg/kg/d. *** 400 *** VEGFR-3 (pg/mL) 400

B, murine VEGFR3 levels were Tumor volume (mm ** quantiﬁed by ELISA in the tumor * 200 lysates of each treated group at the 0 0 end of the experiment (day 21). C, 50 10 15 20 Vehicle SAR 131675 SAR 131675 quantiﬁcation of osteopontin levels Days 30 mg/kg/d 100 mg/kg/d by ELISA in the lymph nodes after C D vehicle or SAR131675 treatment. 400 35 Results are expressed as mean ns 350 SEM. , P < 0.05; , P < 0.01; 30 , P < 0.001. ANOVA followed by * 300 25 Dunnett test versus vehicle group. In 250 ** A, 2-way ANOVA analysis with 20 repeated measures was carried out. 200 D, number of metastases at the 15 150 surface of the lungs for each treated 10 group. 100 5

Osteopontin level Osteopontin level (pg/mL) 50 Metastasis number in lungs 0 0 Negative Vehicle SAR131675 Vehicle SAR131675 SAR131675 lymph nodes 100 mg/kg/d 30 mg/kg/d 100 mg/kg/d

SAR131675 induced an important decrease in the number model may also arise from a reduction in TAM of lung metastases (Fig. 6D), whether the treatment was infiltration. pursued after primary tumor resection or not. This indi- Together, these results from 2 models corroborate the cated that SAR131675 blocks early events of tumor escape hypothesis that tumor reduction by SAR131675 is associ- and metastasis and suggests that treatment before tumor ated with a reduction in TAM infiltration. resection could be sufficient to prevent tumor metastasis. As before, sunitinib was more efficient in reducing tumor Discussion growth, but showed no significant effect on the number of It is well established that VEGFR-3 is a key player in the lung metastases, whatever the treatment schedule (Sup- control of lymphangiogenesis and also angiogenesis. In plementary Fig. S4B). addition, an increasing number of studies have shown its upregulation in tumor cells and also on TAMs, suggesting Effect of SAR131675 on macrophage infiltration that selectively targeting VEGFR-3 might be an interesting TAMs play an important role in tumor promotion and therapeutic approach to reduce tumor growth and metas- metastasis (42). We analyzed macrophage infiltration tasis. Identification of such selective inhibitors could also (with F4/80 immunostaining), using sections from 4T1 have the advantage of providing maximal therapeutical mammary tumors and from the RIP1.Tag2 pancreas. effect without causing severe side effects. Here we In 4T1 tumors, macrophages were mainly located at the describe, for the first time, a small molecule with highly periphery of the tumor at day 12 postimplantation, where- restricted selectivity for VEGFR-3. SAR131675 is a potent as they infiltrated the tumor and formed clusters in more and selective VEGFR-3 inhibitor, with a moderate activity advanced tumors (day 21, Fig. 7A). Treatment with on VEGFR-2. It is able to block angiogenesis, lymphan- SAR131675 decreased macrophage infiltration and clus- giogenesis, and TAM infiltration and in consequence, it tering as illustrated in Fig. 7A and consistently reduced reduces tumor growth and metastasis. F4/80 levels in tumors as measured by an ELISA assay It has been reported that targeting VEGFR-3 with spe- (Fig. 7B). cific inhibitors may block new lymphatic growth exclu- A strong decrease of F4/80 staining in the RIP1.Tag2 sively, without any effect on the survival or function of model was also observed in the angiogenic islets and the existing lymphatic vessels in adult mice (43, 44). In agree- tumors from SAR131675-treated mice (supplementary ment with these reports, SAR131675 is able to block the Fig. S3G). These results indicated that reduction of formation of new lymphatic vessels in a sponge model as tumor growth induced by SAR131675 in the RIP1.Tag2 well during tumor-induced lymphangiogenesis, without

www.aacrjournals.org Mol Cancer Ther; 2012 OF9

Alam et al.

A Day 0: 4T1 cells implantation Day 5: Start of Day 15: p.o. treatment Excision Day 26 Lung SAR131675 (20 d treatment) Metastasis Figure 6. Antimetastatic effect of number SAR131675 after 4T1 tumor SAR131675 (10 d treatment) resection. A, experiment schedule: SAR131675 treatment at 100 mg/kg/d started 5 days after B C 4T1 cell implantation in mammary 0.8 500 fat pads; at day 15 primary tumors were removed in all groups and the 400 0.6 tumor weight was evaluated (B). In *** the group noted 10 days, the 300 ** SAR131675 treatment stopped at 0.4 day 15 and mice were studied 200 untreated until day 26. In the group noted 20 days, the treatment was 0.2 100 continued until the end of the experiment (day 26). C, murine Tumor weight at day 15 (g) weightat day Tumor VEGFR3 levels were quantiﬁed by 0

0.0 15 (pg/mL) at day VEGFR3 level ELISA in tumors lysates at day 15. Vehicle SAR131675 Vehicle SAR131675 D, the number of metastases was counted at the lung surface at day DE120 26. E, illustration of the number of lung metastases after Bouin ﬁxation. Results are expressed as 90 mean SEM, signiﬁcant differences between groups are assessed by ANOVA followed 60 ** ** by Dunnett test. , P < 0.01; , P < 0.001. 30

0 Lung metastasis Lung number at d26 Vehicle 10 d 20 d Vehicle SAR131675 (10 days treatment group) SAR131675

any sign of lymphedema. Moreover, treatment of mice Although angiogenesis inhibitors show antitumor with SAR131675 for about 3 months at 100 mg/kg/d did effects in several mouse models, they have been reported not result in any mortality. to concomitantly elicit increase of lymphatic and distant Vascular complications have emerged as relevant metastasis. Indeed, an anti–VEGFR-2 neutralizing anti- toxicities associated with angiogenesis inhibitors. Bev- body reduces tumor vasculature and volume and also acizumab, a mAb targeting VEGFA, has been linked to increases incidence of lymph node and liver metastasis in hemorrhage, arterial, and venous thrombosis, as well as RIP1.Tag2 mice (47, 48). In this same model, SAR131675 hypertension (45). However, at 100 mg/kg, SAR131675 reduced tumor volume in both prevention and interven- did not induce any significant hypertension in rats, tion studies without any prominent invasive fronts in the suggesting that inhibition of the VEGFR-3 pathway is treated animals. Increased metastasis was also observed not responsible for this adverse event. This also shows in mice receiving short-term therapy with sunitinib (41). that at 100 mg/kg, any effect of SAR131675 on VEGFR-2 In the 4T1 model, SAR131675, but not sunitinib, signifi- activity is not sufficient to induce hypertension in vivo. cantly reduced lymph node invasion and distant metas- This minimal activity on VEGFR-2 in vivo was also tasis. In addition, short-term treatment before tumor confirmed in zebrafish embryo studies. Indeed, in resection with SAR131675 was sufficient to reduce lung agreement with results obtained with a morpholino metastasis, but this was not the case with sunitinib. antisense nucleotide of VEGFC or through overexpres- In addition to its antimetastatic activity, SAR131675 sion of a soluble form of VEGFR-3 (46), treatment with significantly reduced the tumor volume in several exper- SAR131675 affected only intersegmental vessel devel- imental tumor models, including colon, mammary and opment but did not give risetosideeffectssuchasthe prostate, and in RIP1.Tag2 mice. These data are consistent defects in artery development seen in the VEGFR-2 KO with those reported previously that showed that blocking zebrafish (11). the VEGFR-3 pathway may block primary tumor growth

OF10 Mol Cancer Ther; 2012 Molecular Cancer Therapeutics

SAR131675, a Potent and Selective VEGFR-3 Inhibitor

A B F4/80 IHC on 4T1 mammary carcinoma model Day 12 Day 21 30 Control Control SAR131675 Figure 7. Effect of SAR131675 on macrophage infiltration at different 25 stages of tumor development. A, immunohistochemical staining of F4- 80 in 4T1 mammary tumors at days 20 12 and 21 in control and SAR131675- *** 40x 40x treated mice ( 40; 400 40x 15 magnification). B, quantification of F4-80 levels by ELISA dosage in the

4T1 mammary tumor lysates at day Arbitratry unit/ tumor 10 21 in vehicle and SAR131675- treated mice (100 mg/kg/d). 5 400x 400x 400x

Vehicle SAR131675 100 mg/kg/d

(49). Three major mechanisms could explain the role of SAR131675 strongly reduced TAM infiltration in the VEGFR-3 in primary tumor growth: an autocrine effect on 4T1 and the RIP1.Tag2 models. This reduction was asso- tumor cells, pathologic angiogenesis, and TAMs. ciated with reduced tumor growth, reduced metastasis, An increasing number of studies have shown the and extended median survival of RIP1.Tag2 mice. The expression of VEGFR-3 on tumor cells in human cancer effect of SAR131675 on macrophage differentiation and patients (5, 17–19, 21–22). However, this phenomenon has recruitment and on antitumoral immune response is never been reported in mouse tumor models. In agree- under evaluation. ment with this, VEGFR-3 immunostaining of RIP1-Tag2 and 4T1 tumors confirmed that VEGFR-3 is not induced Disclosure of Potential Conflicts of Interest on these tumor cells in vivo (not shown). P. Schaeffer has ownership interest (including patents) in Sanofi stock. Although VEGFR-3 is not expressed on adult blood vessels (3), high expression of VEGFR-3 has been reported Authors' Contributions Conception and design: A. Alam, I. Blanc, G. Gueguen-Dorbes, O. Duclos, in angiogenic sprouts, and blocking VEGFR-3 signaling P. Schaeffer, F. Bono with monoclonal antibodies results in decreased sprout- Development of methodology: A. Alam, I. Blanc, G. Gueguen-Dorbes, ing and vascular density in mouse angiogenesis models J. Bonnin, P. Barron, M.-C. Laplace, G. Morin, J.-P. Herault, P. Schaeffer, F. Bono (16). Thus, the reduction of tumor growth by SAR131675 Acquisition of data (provided animals, acquired and managed patients, could be due to inhibition of angiogenesis. This would be provided facilities, etc.): A. Alam, I. Blanc, G. Gueguen-Dorbes, J. Bonnin, consistent with the reduction of angiogenesis in the P. Barron, M.-C. Laplace, G. Morin Analysis and interpretation of data (e.g., statistical analysis, biostatis- sponge model, which expresses high levels of VEGFR-3 tics, computational analysis): A. Alam, I. Blanc, G. Gueguen-Dorbes, and also with the decrease in vascular density in the 4T1 O. Duclos, P. Barron, G. Morin, P. Schaeffer, F. Bono Writing, review, and/or revision of the manuscript: A. Alam, I. Blanc, and the RIP1.Tag2 models. Thus SAR131675 may block G. Gueguen-Dorbes, F. Dol, J.-P. Herault, P. Schaeffer, F. Bono angiogenesis through different nonexclusive mechan- Administrative, technical, or material support (i.e., reporting or orga- isms: by inhibiting VEGFR-3 signaling in endothelial cells nizing data, constructing databases): A. Alam, J. Bonnin, P. Barron, F. Gaujarengues or by blocking VEGFA- or VEGFC-induced VEGFR-2/ Study supervision: A. Alam, I. Blanc, G. Gueguen-Dorbes, O. Duclos, VEGFR-3 heterodimers. It is worth noting that the mod- F. Dol, P. Schaeffer, P. Savi, F. Bono erate VEGFR-2 activity of SAR131675 may also contribute to its activity on tumor angiogenesis. Acknowledgments Finally, the antiangiogenic effect may result from the The authors thank Carles Callol (Biobide) for the experiment carried out on zebrafish embryos; Sylvie Sordello (TSU Infectious Diseases, Sanofi effect of SAR131675 on TAM infiltration. TAMs promote Toulouse) for the experiment on tumor resection; and Natalio Vita (E2C, angiogenesis by releasing proangiogenic factors such as Sanofi, Toulouse) for his critical review of this manuscript and Fiona Ducrey (E2C, Sanofi, Chilly) for English revision. VEGFA and VEGFC and thereby induce tip-cell formation The costs of publication of this article were defrayed in part by the and fusion (16, 29, 32). It is well established that macro- payment of page charges. This article must therefore be hereby marked phages and also TAMs express VEGFR-3 (28–30). More- advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. over, VEGFR-3 inhibition decreased dendritic cell recruitment to the spleen and resulted in prolonged rat cardiac Received November 4, 2011; revised March 23, 2012; accepted April 19, allograft survival (50). Here we have shown that 2012; published OnlineFirst May 14, 2012.

www.aacrjournals.org Mol Cancer Ther; 2012 OF11

Alam et al.

References 1. Sleeman JP, Thiele W. Tumor metastasis and the lymphatic vascula- VEGF-C in human colorectal adenocarcinoma. Anticancer Res ture. Int J Cancer 2009;125:2747–56. 2002;22:1463–6. 2. Achen MG, Stacker SA. Molecular control of lymphatic metastasis. 22. Kodama M, Kitadai Y, Tanaka M, Kuwai T, Tanaka S, Oue N, et al. Ann N Y Acad Sci 2008;1131:225–3. Vascular endothelial growth factor C stimulates progression of human 3. Tammela T, Alitalo K. Lymphangiogenesis: Molecular mechanisms gastric cancer via both autocrine and paracrine mechanisms. Clin and future promise. Cell 2010;140:460–76. Cancer Res 2008;14:7205–14. 4. Gombos Z, Xu X, Chu CS, Zhang PJ, Acs G. Peritumoral lymphatic 23. Hamrah P, Chen L, Zhang Q, Dana MR. Novel expression of vascular vessel density and vascular endothelial growth factor C expression in endothelial growth factor (VEGFR-3) and VEGF-C on corneal Dendritic early-stage squamous cell carcinoma of the uterine cervix. Clin Cancer cells. Am J Pathol 2003;163:57–68. Res 2005;11:8364–71. 24. Cursiefen C, Chen L, Borges LP, Jackson D, Cao J, Radziejewski C, 5. Shida A, Fujioka S, Kobayashi K, Ishibashi Y, Nimura H, Mitsumori N, et al.. VEGF-A stimulates lymphangiogenesis and hemangiogenesis in et al. Expression of vascular endothelial growth factor (VEGF)-C and -D inflammatory neovascularization via macrophage recruitment. J Clin in gastric carcinoma. Int J Clin Oncol 2006;11:38–43. Invest 2004;113:1040–50. 6. Roberts N, Kloos B, Cassella M, Podgrabinska S, Persaud K, Wu Y, 25. Maruyama K, Ii M, Cursiefen C, Jackson DG, Keino H. Inflammation- et al. Inhibition of VEGFR-3 activation with the antagonistic antibody induced lymphangiogenesis in the cornea arises from CD11b positive more potently suppresses lymph node and distant metastases than macrophages. J Clin Invest 2005;115:2363–72. inactivation of VEGFR-2. Cancer Res 2006;66:2650–7. 26. Shin YJ, Choi JS, Choi JY, Hou Y, Cha J-H, Chun MH, et al. Induction of 7. Norrmen C, Tuomas T, Petrova TV, Kari A. Biological basis of thera- vascular endothelial growth factor receptor-3 mRNA in glial cells peutic lymphangiogenesis. Circulation 2011;123:1335–51. following focal cerebral ischemia in rats. J Neuroimmunol 2010;229: 8. Achen MG, Jeltsch M, Kukk E, Makinen€ T, Vitali A, Wilks AF, et al. 81–90. Vascular endothelial growth factor D (VEGF-D) is a ligand for the 27. Baluk P, Tammela T, Ator E, Lyubynska N, Achen MG, Hicklin DJ, et al. tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Pathogenesis of persistent lymphatic vessel hyperplasia in chronic Proc Natl Acad Sci U S A 1998;95:548–53. airway inflammation. J Clin Invest 2005;115:247–57. 9. Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, et al. A 28. Schoppmann SF, Birner P, Stockl J, Kalt R, Ullrich R, Caucig C, et al. novel vascular endothelial growth factor, VEGF-C, is a ligand for the Tumor-associated macrophages express lymphatic endothelial Flt4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. EMBO J growth factors and are related to peritumoral lymphangiogenesis. Am 1996;15:1751. J Pathol 2002;161:947–56. 10. Dumont DJ, Jussila L, Taipale J, Lymboussaki A, Mustonen T, Pajusola 29. Yang H, Kim C, Kim MJ, Schwendener RA, Alitalo K, Heston W, et al. K, et al. Cardiovascular failure in mouse embryos deficient in VEGF Soluble vascular endothelial growth factor receptor-3 suppresses receptor-3. Science 1998;282:946–9. lymphangiogenesis and lymphatic metastasis in bladder cancer. Mol 11. Covassin LD, Villefranc JA, Kacergis MC, Weinstein BM, Lawson ND. Cancer 2011;10:36. Distinct genetic interactions between multiple Vegf receptors are 30. Kerjaschki D. The crucial role of macrophages in lymphangiogenesis. J required for development of different blood vessel types in zebrafish. Clin Invest 2005;115:2316–9. Proc Natl Acad Sci U S A 2006;103:6554–9. 31. Schoppmann SF, Fenzl A, Nagy K, Unger S, Bayer G, Geleff S, et al. 12. Haiko P, Makinen T, Keskitalo S, Taipale J, Karkkainen MJ, Baldwin VEGF-C expressing tumor-associated macrophages in lymph node ME, et al. Deletion of vascular endothelial growth factor C (VEGF-C) positive breast cancer: impact on lymphangiogenesis and survival. and VEGF-D is not equivalent to VEGF receptor 3 deletion in mouse Surgery 2006;139:839–46. embryos. Mol Cell Biol 2008;15:4843–50. 32. Fantin A, Vieira JM, Gestri G, Denti L, Schwarz Q, Prykhozhij S, et al. 13. Alam A, Herault JP, Barron P, Favier B, Fons P, Delesque-Touchard N, Tissue macrophages act as cellular chaperones for vascular anasto- et al. Heterodimerization with vascular endothelial growth factor recep- mosis downstream of VEGF-mediated endothelial tip cell induction. tor-2 (VEGFR-2) is necessary for VEGFR-3 activity. Biochem Biophys Blood 2010;116:829–40. Res Commun 2004;324:909–15. 33. Yan A, Avraham T, Zampell JC, Aschen SZ, Mehrara BJ. Mechanisms 14. Nilsson I, Bahram F, Li X, Gualandi L, Koch S, Jarvius M, et al. VEGF of lymphatic regeneration after tissue transfer. PLoS One 2011;6: receptor 2/-3 heterodimers detected in situ by proximity ligation on e17201. angiogenic sprouts. EMBO J 2010;29:1377–88. 34. Attout T, Hoerauf A, Den ec e G, Debrah AY, Marfo-Debrekyei Y. 15. Kubo H, Fujiwara T, Jussila L, Hashi H, Ogawa M, Shimizu K, et al. Lymphatic vascularisation and involvement of Lyve-1þ macrophages Involvement of vascular endothelial growth factor receptor-3 in main- in the human onchocerca nodule. PLoS One 2009;4:e8234. tenance of integrity of endothelial cell lining during tumor angiogen- 35. Favier B, Alam A, Barron P, Bonnin J, Laboudie P, Fons P, et al. esis. Blood 2000;96:546–53. Neuropilin-2 interacts with VEGFR-2 and VEGFR-3 and promotes 16. Tammela T, Zarkada G, Wallgard E, Murtomaki€ A, Suchting S, Wirze- human endothelial cell survival and migration. Blood 2006;108:1243–50. nius M, et al. Blocking VEGFR-3 suppresses angiogenic sprouting and 36. Bergers G, Javaherian K, Lo KM, Folkman J, Hanahan D. Effect of vascular network formation. Nature 2008;454:656–60. angiogenesis inhibitors on multistage carcinogenesis in mice. Science 17. Su JL, Chen PS, Chien MH, Chen PB, Chen YH, Lai CC. Further 1999;284:808–12. evidence for expression and function of the vegf-c/vegfr-3 axis in 37. Pulaski BA, Ostrand-Rosenberg S. Reduction of established sponta- cancer cells. Cancer Cell 2008;13:557. neous mammary carcinoma metastases following immunotherapy 18. Donnem T, Al-Shibli K, Al-Saad S, Delghandi MP, Busund L-T, with major histocompatibility complex class II and B7.1 cell-based Bremnes RM. VEGF-A and VEGFR-3 correlate with nodal status in tumor vaccines. Cancer Res 1998;58:1486–93. operable non-small cell lung cancer: Inverse correlation between 38. Patyna S, Arrigoni C, Terron A, Kim TW, Heward JK, Vonderfecht SL, expression in tumor and stromal cells. Lung Cancer 2009;63:277–83. et al. Nonclinical safety evaluation of sunitinib: a potent inhibitor of 19. Garces CA, Kurenova EV, Golubovskaya VM, Cance WG. Vascular VEGF, PDGF, KIT, FLT3, and RET receptors. Toxicol Pathol 2008;36: endothelial growth factor receptor-3 and focal adhesion kinase bind 905–16. and suppress apoptosis in breast cancer cells. Cancer Res 2006; 39. Chang L K, Garcia-Cardena G, Farnebo F, Fannon M, Chen E J, 66:1446–54. Butterfield C, et al. A dose-dependent response of FGF-2 for lym- 20. Kurenova EV, Hunt DL, He D, Fu AD, Massoll NA, Golubovskaya C, phangiogenesis. Proc Natl Acad Sci 2004;101:11658–63. et al. Vascular endothelial growth factor receptor-3 promotes breast 40. Hanahan D. Heritable formation of pancreatic beta-cell tumours in cancer cell proliferation, motility and survival in vitro and tumor for- transgenic mice expressing recombinant insulin/simian virus 40 onco- mation in vivo. Cell Cycle 2009;8:2266–80. genes. Nature 1985;315:115–22. 21. Witte D, Thomas A, Ali N, Carlson N, Younes M. Expression of the 41. Ebos JM, Lee CR, Cruz-Munoz W, Bjarnason GA, Christensen JG, vascular endothelial growth factor receptor-3 (VEGFR-3) and its ligand Kerbel RS. Accelerated metastasis after short-term treatment with

OF12 Mol Cancer Ther; 2012 Molecular Cancer Therapeutics

SAR131675, a Potent and Selective VEGFR-3 Inhibitor

a potent inhibitor of tumor angiogenesis. Cancer Cell 2009;15: 47. Casanovas O, Hicklin DJ, Bergers G, Hanahan D. Drug resistance by 232–9. evasion of antiangiogenic targeting of VEGF signaling in late-stage 42. Solinas G, Marchesi F, Garlanda C, Mantovani A, Allavena P. Inflam- pancreatic islet tumors. Cancer Cell 2005;8:299–309. mation-mediated promotion of invasion and metastasis. Cancer 48. Paez-Ribes M, Allen E, Hudock J, Takeda T, Okuyama H, Vinals~ F, et al. Metastasis Rev 2010 29:243–8. Antiangiogenic therapy elicits malignant progression of tumors to 43. Pytowski B, Goldman J, Persaud K, Wu Y, Witte L, Hicklin DJ, et al. increased local invasion and distant metastasis. Cancer Cell 2009; Complete and specific inhibition of adult lymphatic regeneration by a 15:220–31. novel VEGFR-3 neutralizing antibody. J Natl Cancer Inst 2005;97:14–21. 49. Laakkonen P, Waltari M, Holopainen T, Takahashi T, Pytowski B, 44. Karpanen T, Wirzenius M, kinen TM, Veikkola T, Haisma HJ, Achen Steiner P, et al. Vascular endothelial growth factor receptor 3 is MG, et al. Lymphangiogenic growth factor responsiveness is modu- involved in tumor angiogenesis and growth. Cancer Res 2007;67: lated by postnatal lymphatic vessel maturation. Am J Pathol 593–99. 2006;169:708–18. 50. Nykanen€ AI, Sandelin H, Krebs R, Keranen€ MA, Tuuminen R, Karp€ anen€ 45. Kamba T, McDonald DM. Mechanisms of adverse effects of anti-VEGF T, et al. Targeting lymphatic vessel activation and CCL21 production therapy for cancer. Br J Cancer 2007;96:1788–95. by vascular endothelial growth factor receptor-3 inhibition has novel 46. Ober X. VEGFc is required for vascular development and endoderm immunomodulatory and antiarteriosclerotic effects in cardiac allo- morphogenesis in zebrafish. EMBO Rep 2004;5:78–84. grafts. Circulation 2010;121:1413–22.

www.aacrjournals.org Mol Cancer Ther; 2012 OF13

SAR131675, a Potent and Selective VEGFR-3−TK Inhibitor with Antilymphangiogenic, Antitumoral, and Antimetastatic Activities

Antoine Alam, Isabelle Blanc, Geneviève Gueguen-Dorbes, et al.

Mol Cancer Ther Published OnlineFirst May 14, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-11-0866-T

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2012/08/09/1535-7163.MCT-11-0866-T.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/early/2012/07/13/1535-7163.MCT-11-0866-T. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.