Egfl7 Promotes Tumor Escape from Immunity by Repressing Endothelial Cell Activation

Published OnlineFirst October 28, 2011; DOI: 10.1158/0008-5472.CAN-11-1301

Cancer Molecular and Cellular Pathobiology Research

Egﬂ7 Promotes Tumor Escape from Immunity by Repressing Endothelial Cell Activation

Suzanne Delfortrie1,2,3,4,Sebastien Pinte1,2,3,4, Virginie Mattot1,2,3,4, Chantal Samson1,2,3,4,Gaelle€ Villain1,2,3,4, Bertrand Caetano1,2,3,4,Geraldine Lauridant-Philippin1,2,3,4,5, Marie-Christine Baranzelli5, Jacques Bonneterre5, Francois¸ Trottein2,3,6,7, Christelle Faveeuw2,3,6,7, and Fabrice Soncin1,2,3,4

Abstract Downregulating the leukocyte adhesion molecules expressed by endothelial cells that line tumor blood vessels can limit the entry of immune effector cells into the tumor mass, thereby contributing to tumoral immune escape. Egfl7 (also known as VE-statin) is a secreted protein specifically expressed by endothelial cells in normal tissues and by cancer cells in various human tumors. High levels of Egfl7 correlate with higher tumor grade and poorer prognosis. Here we show that expression of Egfl7 in breast and lung carcinoma cells accelerates tumor growth and metastasis in immunocompetent mice but not in immunodeficient mice. Tumors expressing Egfl7 were infiltrated relatively poorly by immune cells and were characterized by reduced levels of immunostimulatory cytokines [IFN-g, interleukin-12 (IL-12)] and fewer endothelial adhesion molecules [intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1)]. In vitro studies revealed that Egfl7 inhibited the expression of leukocyte adhesion molecules by endothelial cells, preventing lymphocyte adhesion. In contrast, Egfl7 did not exert any effects on immune cell activation. Human breast cancer lesions expressing high levels of Egfl7 also expressed less ICAM-1 and VCAM-1 in their blood vessels, also indicating an inverse correlation between expression levels of Egfl7 and IFN-g. Thus, Egfl7 expression in tumors promotes tumor progression by reducing the expression of endothelial molecules that mediate immune cell infiltration. Our findings highlight a novel mechanism through which tumors escape immune control. Cancer Res; 71(23); 7176–86. 2011 AACR.

Introduction destroy tumor cells, circulating immune cells must first infil- trate the tumor mass. The tumor blood vessel endothelium, Blood vessels are essential for the growth of most solid although imperfectly tight, actively protects tumor cells from tumors and for metastasis. Within a tumor, new blood vessels, the immune system through its barrier function. The trans- formed mostly by endothelial cells, sustain the nutriment and migration of leukocytes through the endothelium is a normal oxygen supplies of the developing mass (1). The tumor endo- process, which, during inflammation, results in the infiltration thelium forms an imperfect, tortuous, and leaky barrier (2). of circulating cells from the blood stream into tissues. Such Metastatic cells enter the blood circulation through this endo- extravasation depends on a complex set of interactions thelium (intravasation) and spread across the organism as an between leukocytes and endothelial membrane receptors, alternative route to lymphatic metastasis (3–5). On the other which, on the endothelial side, involves E- and P-selectin, hand, infiltration of circulating immune cells, such as natural intercellular adhesion molecule 1 (ICAM-1), vascular cell adhe- þ killer (NK) cells and CD8 cytotoxic T lymphocytes, into the sion molecule 1 (VCAM-1), and platelet/endothelial cell adhe- tumor mass plays a part in the control of tumor growth sion molecule (PECAM)/CD31 as the main actors (10). Tumor through IFN-g and cytotoxic-based mechanisms (6–9). To escape from immunity may be achieved by preventing infiltration of effector immune cells (11) through the downregulation of these endothelial adhesion molecules (12–15). fl Authors' Affiliations: 1Centre National de la Recherche Scientifique, We originally characterized Eg 7 (VE-statin) as a gene UMR8161; 2Universite Lille-Nord de France; 3Institut Pasteur de Lille; specifically expressed by blood vessel endothelial cells in 4Equipe Labellisee la Ligue; 5Centre Oscar Lambret; 6Institut National de normal organs during development and in the adult (16, 17). 7 la Sante et de la Recherche Medicale, U1019; and CNRS, UMR8204, Lille, fi fl France In human cancer, this speci city of expression is lost as Eg 7 has been detected in tumor cells themselves, in addition to Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). endothelial cells. Its expression levels correlate with a higher tumor grade in glioma (18) and colon cancer patients (19), and S. Delfortrie and S. Pinte contributed equally to this work. with a poorer prognosis and higher metastatic score in hepa- Corresponding Author: Fabrice Soncin, Institut de Biologie de Lille, 1 rue tocarcinoma patients (20). Although these observations sug- du Pr. Calmette, 59021 Lille Cedex, France. E-mail: [email protected] gested a role for Egfl7 in cancer progression, its direct role in doi: 10.1158/0008-5472.CAN-11-1301 tumor development has not been studied yet. Here, we show 2011 American Association for Cancer Research. that Egfl7 is an endogenous regulator of endothelial cell

7176 Cancer Res; 71(23) December 1, 2011

Egﬂ7 Promotes Tumor Escape from Immunity

activation, which, when expressed by tumor cells, protects the probe, HPI). Mice were sacrificed after 30 minutes, after which tumor from the host immunity. tumors were collected, embedded in paraffin, and hypoxy- probe detected by immunohistochemistry. Material and Methods Immunohistochemistry Cells Tumors were either fixed in 4% paraformaldehyde and Mouse mammary carcinoma 4T1 [American Type Culture embedded in paraffin, and sectioned (7 mm) or frozen in OCT Collection (ATCC) CRL-2539], lung adenocarcinoma LLC1 compound, sectioned at 10 mm, and postfixed with 1% (ATCC CRL-1642), and Jurkat (ATCC TIB-152) cells were paraformaldehyde (5 minutes). Immunostaining was carried obtained from ATCC and were not further tested or authen- out using antibodies as listed in Supplementary Table S1. For ticated. 4T1 and Jurkat cells were cultured in RPMI, 10% FBS, necrosis analysis, sections were stained with hematoxylin 100 U/mL penicillin, and 100 mg/mL streptomycin. LLC1 cells and eosin, and necrotic areas were identified as unstained were cultured in Dulbecco's Modified Eagle's Medium regions. Apoptotic cells were visualized using the Terminal (DMEM), 10% FBS, 100 U/mL penicillin, and 100 mg/mL Transferase Recombinant Kit (Roche). Proliferating cells streptomycin. Human primary umbilical vein endothelial cells were detected by staining with a Ki-67 antibody (Roche). (HUVEC, Lonza) were cultured in endothelial cell growth For optical microscopy, sections were counterstained with medium (EGM-2) and used between passages 1 and 5. Condi- hematoxylin. For immunofluorescence microscopy, cell tioned medium was produced by incubating 4T1-Ctrl or 4T1- nuclei were labeled by incubating in 40, 6-diamidino-2- Egfl7 cells (2 104/cm2) in EBM-2 (Lonza), 0.2% BSA for 24 phenylindole (DAPI) solution (1 mg/mL, Sigma) for 1 minute hours. Medium was filtered (0.22 mm) before use. All cells were at room temperature. Slides were analyzed using an Axio- fi cultured in a humidi ed 95% air/5% CO2 incubator at 37 C. plan2 or an AxioImagerZ1 microscope (Zeiss) and AxioVs V4.8.2.0 software or a confocal LSM710 microscope (Zeiss) Cloning using the ZEN2008 software (MICPaL). Angiogenic hot spots The mouse Egfl7 cDNA was cloned in frame with a were analyzed by image analysis of CD31-stained sections C-terminal influenza hemagglutinin (HA)-coding sequence in using ImageJ v1.42q software (22). the pMSCV plasmid (Clontech), allowing the production of retrovirus after transfection in HEKGP cells. 4T1 and LLC1 Tumor endothelial cell isolation cells were infected with control or Egfl7-coding viruses, and Freshly collected tumors (n ¼ 6/group) were minced and whole-cell populations were selected for puromycin resistance dissociated in DMEM, 1 mg/mL collagenase-I, 10 mg/mL (4 mg/mL) for 7 days. DNase-I for 30 minutes at 37C, and filtered (90 mm and 40 mm mesh). Cells from each tumor were treated separately. Cells Transfection were incubated in red blood cell lysis buffer (Sigma) for 5 HUVECs were plated in 2-cm2 well plates (25,000 cells/cm2) minutes at room temperature. Washed cells were incubated in and transfected the next day with 10 nmol/L siRNA (Dharma- DMEM, 0.2% FBS containing a rat anti-mouse CD16/CD32 (Fc- con) in Primefect siRNA reagent (Lonza) mixed with EGM-2. block, BD-Pharmingen, 553141, 1 mg/106 cells) for 1 hour at 4C, After 24 hours, EGM-2 was added and cells were cultured for 24 then with anti-rat IgG–coated magnetic beads (Dynabeads, or 48 hours. Invitrogen) which had been incubated with a rat anti-mouse CD45 antibody (BD-Pharmingen, 550539) for 20 minutes at Tumor models 4C. CD45 cells were collected and incubated with magnetic Eight-week-old female Balb/c, C57Bl/6, and severe com- beads preincubated with a rat anti-mouse CD31/PECAM bined immunodeficient (SCID)-beige mice were purchased antibody (BD-Pharmingen, 553370) for 20 minutes at 4C. þ from Charles River (7 mice per experimental group in each CD45 /CD31 and CD45 /CD31 cells were separated and experiment unless otherwise stated). 4T1 and LLC1 cells lysed in TRIzol. Enrichment was evaluated by measuring the (5 105/50 mL) were injected either in the mammary fat pad expression levels of CD31. (Balb/c, SCID-beige) or s.c. (C57Bl/6). Tumors were measured using an electronic caliper and volume calculated as Quantitative reverse transcriptase-PCR volume ¼ width width length p/6 (21). Mice were Cells or tissues were homogenized in TRIzol (Life Technol- housed according to European legislation; all protocols were ogies). Total RNA was extracted and reverse transcribed approved by the local ethics committee (#CEEA02/2009). using a high capacity cDNA reverse transcription kit (Life Technologies). Quantitative reverse transcriptase PCR Tumor vessel perfusion (RT-PCR) was carried out using TaqMan gene expression Blood vessel leakage was assessed by injecting 100 mL of PBS, assays, reagents, and conditions (Life Technologies). 0.5 mg/mL Lycopersicon esculentum lectin-FITC (Vector Lab- oratories) and 2.5 mg/mL Dextran (70 kD)–Texas Red (Molec- Western blotting ular Probes) in the tail vein. Mice were euthanized after 10 Proteins were extracted in RIPA buffer, analyzed by 12% minutes, and tumors were collected and processed for cryo- SDS-PAGE, and blotted onto Immobilon-P. Egfl7-HA, cleaved section. Hypoxia was estimated after peritoneal injection of caspase-3, and actin were detected using specific antibodies 0.15 mol/L NaCl and 60 mg/kg pimonidazole-HCl (Hypoxy- from Covance (HA.11 Clone 16B12, 1:1,500), Cell signaling

www.aacrjournals.org Cancer Res; 71(23) December 1, 2011 7177

Delfortrie et al.

(9664S, 1:1,000), and Santa Cruz Biotechnology (sc-1615, Immune cell activation 1:1,000), respectively. Chemiluminescence was measured To generate dendritic cells, bone marrow–derived cells were using the Luminescent Image System (LAS3000, Fujifilm). cultured in Iscove's Modified Dulbecco's Medium supplemented with 10% FBS and 1% of supernatant from granulocyte macro- Adhesion assay phage colony-stimulating factor (GM-CSF)–expressing J558- For T-lymphocyte adhesion assays, Jurkat cells (1 106) GM-CSF-cells for 14 days. Cells (105 cells per well) were stim- were incubated with DiI (2 mmol/L, Molecular Probes) for 10 ulatedor not with1 mg/mL lipopolysaccharide (LPS) for 24 hours minutes at 37C and allowed to adhere (105 cells/cm2) onto a in the presence or absence of 150 ng/mL mouse recombinant monolayer of confluent HUVECs for 20 minutes at 22C. (r)Egfl7 (23). Interleukin (IL)-6 and IL-12p40 were quantified in Fluorescent Jurkat cells were counted under a UV microscope. the supernatants by ELISA (BD-Biosciences). For NK cell

Figure 1. Egfl7 promotes tumor growth and metastasis. A, 4T1-Ctrl and 4T1-Egfl7 cells were injected into the mammary fat pads of Balb/c mice and tumors were dissected after 27 days. Tumor extracts (75 mg) were analyzed by Western blotting using anti-HA and anti- actin antibodies to visualize the expression of exogenous Egfl7 in reference to actin in 2 representative samples of each condition. B, tumor sections were immunostained using a specific anti-Egfl7 antibody (brown) to visualize the higher levels of expression in 4T1-Egfl7 tumors; bar, 100 mm. C, 4T1-Ctrl (*)and 4T1-Egfl7(&) cells were implanted into the mammary fat pads of Balb/c mice at day 0 and growing tumors measured over time; inset, average final weights of 4T1-Ctrl and 4T1- Egfl7 tumors. D, left, lungs of mice bearing 4T1-Ctrl and 4T1-Egfl7 primary tumors were dissected and photographed; arrows, macrometastases; right, average numbers of lung surface macrometastases counted in each animal group. E, LLC-Ctrl (*)or LLC-Egfl7(&) were injected in the skin of C57Bl/6 mice and tumors measured at regular times after injection; inset, average weights of LLC-Ctrl and LLC-Egfl7 tumors after 25 days. Data are representative of a set of 4 and 3 experiments carried out in similar conditions using 4T1 and LLC1 models, respectively. , P < 0.05.

7178 Cancer Res; 71(23) December 1, 2011 Cancer Research

Egﬂ7 Promotes Tumor Escape from Immunity

function, mouse liver mononuclear cells were labeled with APC- Human tumor samples conjugated anti-CD5 and PE-conjugated anti-NK1.1 monoclonal Human tumors were collected, processed, and stored at the þ antibodies. CD5 /NK1.1 cells were sorted using a FACSAria anatomopathology laboratory of the Centre Oscar Lambret (BD-Biosciences), cultured in RPMI, 5% FCS (104 cells per well) (Lille, France). Tissues were fixed in 4% paraformaldehyde and stimulated with 10 ng/mL IL-12 and 10 ng/mL IL-18 in the (Prolabo) and embedded in paraffin. Sections (4 mm) were presence or absence of 150 ng/mL rEgfl7 for 48 hours. IFN-g stained using a polyclonal anti-hEgfl7 antibody (1:30, R&D, production was analyzed in supernatants by ELISA (eBios- AF3638) in a Discovery automat [Ventana (G. Lauridant-Phi- ciences). T cells were purified from spleens by negative selection lippin and F. Soncin; personal communication)]. Other immu- using a Dynal T-cell isolation kit (Life Technologies). T cells nostainings were carried out using specific antibodies against (105 cells per well) were cultured in RPMI, 5% FCS and stim- ICAM-1 (1:50, Abcam Ab53013) and VCAM-1 (1:500, Abcam ulated or not with plate-bound anti-CD3 (5 mg/mL) and soluble Ab98954). Slides were analyzed by 2 independent observers anti-CD28 (1 mg/mL), in the presence or absence of 150 ng/mL and compared with corresponding hematoxylin and phloxin rEgfl7 for 48 hours. IFN-g production was analyzed in super- safran-stained slides for identification of the tumor subregions. natants by ELISA. Proliferation was assessed 96 hours later, Tumors were classified from 0þ (no Egfl7 staining) to 3þ using AlamarBlue (Life technologies). (intense Egfl7 staining).

Figure 2. Increased necrosis, hypoxia, angiogenesis, and permeability in 4T1-Egfl7 tumors. A, left, 4T1-Ctrl and 4T1-Egfl7 tumor sections were stained with hematoxylin and eosin and necrosis identified as unstained areas; bars, 1 mm; right, necrotic areas were quantified as a percentage of total tumor area in the 4T1 and the LLC1 models. B, 4T1-Ctrl and 4T1-Egfl7 tumors were analyzed for hypoxia in 15 nonoverlapping sections of 4T1- Ctrl and 4T1-Egfl7 tumors. C, left, sections of 4T1-Ctrl and 4T1-Egfl7 tumors immunostained for CD31/ PECAM (brown) to visualize the tumor endothelium; bar, 100 mm; þ right, quantification of CD31 vessels in angiogenic hot spots in the 4T1 and LLC1 tumors; the area fraction represents the percentage of stained area in 20 nonoverlapping 2.4-mm2 fields. D, mice bearing 4T1-Ctrl or 4T1-Egfl7 tumors were injected with L. esculentum-fluorescein isothiocyanate (FITC) lectin and Dextran-Texas Red. Tumors were dissected and analyzed by fluorescent microscopy. Red staining outside FITC-stained vessels indicates blood leakage; bar, 100 mm. Histogram on the right indicates the mean percentage of leaky vessels counted in 25 independent 0.15-mm2 fields of 4T1-Ctrl and 4T1-Egfl7 tumors. , P < 0.05.

www.aacrjournals.org Cancer Res; 71(23) December 1, 2011 7179

Delfortrie et al.

Figure 3. Egfl7 prevents the infiltration of immune cells into 4T1- Egfl7 tumors. A, hematoxylin and eosin staining of 27-day tumors shows numerous small nucleated cells accumulated in the lumens of 4T1-Egfl7 tumor blood vessels (right, arrows), such cells are rare in 4T1-Ctrl tumor blood vessels (left, arrows); bar, 100 mm. B, CD3e immunostaining of tumor sections (brown) shows a large infiltration of T lymphocytes in 4T1-Ctrl tumors (left, arrows pointing to some). Most þ of the CD3e cells remained in the vessel lumens of 4T1-Egfl7 tumors (right, arrows); bar, 100 mm; right, average numbers of infiltrated þ CD3e cells (outside of blood vessels) counted in 30 independent 0.15-mm2 fields. C, average numbers of cells positive for the indicated molecule counted within the tumor masses in 25 independent 0.15-mm2 fields. Data are representative of 2 experiments carried out in similar conditions. , P < 0.01; , P < 0.001.

Statistics mammary fat pad of syngeneic Balb/c mice, 4T1-Egfl7 cells In each experiment, data points have been collected at least formed tumors that showed a marked accumulation of Egfl7 in triplicate. The Mann–Whitney and Wilcoxon tests were used (Fig. 1A and B). These tumors grew much faster in volume than to compare the mean values between groups (, P < 0.05; , P < 4T1-Ctrl tumors and, accordingly, the mean final weight of 0.01; , P < 0.001). Error bars in graphs represent the calcu- 4T1-Egfl7 tumors was twice as high as that of 4T1-Ctrl tumors lated SE. (Fig. 1C). Mice bearing 4T1-Egfl7 tumors also had 3.6 times higher frequency of lung metastasis when compared with 4T1- Results Ctrl tumors (Fig. 1D). As in vivo experiments can vary with the cell lines and animal models used, we carried out similar Egfl7 promotes tumor growth and metastasis experiments using the LLC1 lung carcinoma cell line implanted To study the potential role of Egfl7 in tumor development, in syngeneic C57BL/6 mice. LLC-Egfl7 cells also produced mouse 4T1 breast cancer cells were infected with a retrovirus tumors that grew faster than LLC-Ctrl cells (Fig. 1E). The encoding the full length, HA-tagged mouse Egfl7 (4T1-Egfl7) or effects of Egfl7 on tumor growth and metastasis in vivo were with a control virus (4T1-Ctrl). When implanted into the not due to intrinsic modifications of the tumor cell properties

7180 Cancer Res; 71(23) December 1, 2011 Cancer Research

Egﬂ7 Promotes Tumor Escape from Immunity

as expression of Egfl7 had no effects on 4T1 and LLC1 prolif- The effects of Egfl7 on tumor growth depend on the eration, migration, or clone formation in anchorage-indepen- immune system dent conditions in vitro (Supplementary Fig. S1). To directly assess the importance of the host immune system In vivo, 4T1-Egfl7 tumors displayed larger necrotic areas on the effects of Egfl7, 4T1-Ctrl, and 4T1-Egfl7, cells were than did 4T1-Ctrl tumors (Fig. 2A) and a similar tendency was injected in the mammary gland of immunosuppressed SCID- observed in LLC1 tumors. Hypoxia was higher in 4T1-Egfl7 beige mice, which lack functional T lymphocytes, B lympho- tumors than in controls (Fig. 2B), whereas the apoptosis, cytes, and NK cells. As expected, 4T1-Ctrl tumors grew much cleaved caspase-3 levels, and Ki-67 proliferation indexes were faster and induced a higher rate of metastasis in SCID-beige comparable between tumors (Supplementary Fig. S2). Micro- mice when compared with Balb/c immunocompetent mice vessel density in angiogenic hot spots was slightly increased in (Fig. 4), thus confirming the repressing effects of immune cells 4T1-Egfl7 tumors when compared with controls but was not on tumor development. On the other hand, expression of Egfl7 significantly different between LLC1 tumors (Fig. 2C). Inter- in tumor cells failed to accelerate tumor growth and metastasis estingly, 4T1-Egfl7 tumor blood vessels were much more in SCID-beige mice so that no differences between the growth permeable than were 4T1-Ctrl vessels (Fig. 2D), and although rates of 4T1-Ctrl and 4T1-Egfl7 tumors could be seen anymore. excessive edema was not observed, blood lakes were more Thus, the effects of Egfl7 on tumor growth and metastasis frequent in LLC-Egfl7 tumors than in LLC-Ctrl tumors (2.7% depend on an active host immune system. 0.93% and 1.1% 0.4% of the tumor area, respectively). Egfl7 represses leukocyte adhesion molecules in tumor Egfl7-expressing tumors are less infiltrated by immune endothelial cells cells We next addressed the potential effects of Egfl7 on the Interestingly, a detailed observation revealed the presence of functions of immune cells. rEgfl7, which was active as an numerous small round cells in the lumen of 4T1-Egfl7 tumor inhibitor of platelet-derived growth factor (PDGF)-BB– blood vessels, whereas these cells were less frequently observed induced smooth muscle migration [Supplementary Fig. S5A, within the tumor tissue itself. In contrast, 4T1-Ctrl tumors had (refs. 16, 23)], had no effect on the LPS-induced release of IL-6 veryfewcellsaccumulatedintheirbloodvesselsandthe (Fig. 5A) and IL-12p40 (Supplementary Fig. S5B) by dendritic tumor tissue appeared more granular than that in 4T1-Egfl7 cells. Similarly, rEgfl7 did not affect IFN-g production by NK tumors (Fig. 3A). As these cells resembled circulating or cells stimulated with IL-12 and IL-18 and failed to modulate the infiltrated immune cells, staining for CD3e, a global marker of T lymphocytes, was conducted. The 4T1-Egfl7tumor þ tissue contained 69% fewer CD3e cells than did the 4T1- þ Ctrl tumors, with most of the CD3e cells remaining in the þ lumens of the blood vessels (Fig. 3B). Staining for CD4 and þ CD8 cells showed that both T-cell populations were affected, with a 42% and 45% decrease in the respective number of cells infiltrated in 4T1-Egfl7 tumors when compared with 4T1-Ctrl (Fig. 3C; Supplementary Fig. S3). These effects were not restricted to the T-cell lineage as the numbers of þ þ infiltrated B-lymphocytes (CD19 ), macrophages (CD68 ), þ þ NK cells (NKp46 ), and dendritic cells (CD11c )were decreased by more than half in 4T1-Egfl7 tumors when compared with 4T1-Ctrl tumors (Fig. 3C; Supplementary Fig. S3). These findings were correlated with large differences in the transcript levels of immunostimulating cytokines IFNg and IL-12b,and,toalesserextent,IL-1a and IL- 1b, the levels of which were strongly reduced in 4T1-Egfl7 tumors when compared with controls (Supplementary Table S2). Similar results were obtained using the LLC-C57BL/6 model; LLC-Egfl7 tumors expressed much less CD3E and IFN-g transcripts than LLC-Ctrl tumors (Supplementary Table S3). On the other hand, the spleens of Balb/c mice fl carrying 4T1-Ctrl or 4T1-Egfl7 tumors showed no significant Figure 4. Eg 7 has no effects in the absence of a functional immune system. A, 4T1-Ctrl (*) and 4T1-Egfl7(&) cells were injected into the differences in the relative numbers of T lymphocytes þ þ þ mammary fat pads of immunosuppressed SCID-beige mice and the (CD3e ,TCR ) and B lymphocytes (CD19 ), or of NK cells developing tumors were measured overtime. Overlays of growth curves þ (NKp46 ), suggesting that the immune depletion was not of 4T1-Ctrl (dashed) and 4T1-Egfl7 tumors (dotted) implanted in Balb/c systemic in mice carrying 4T1-Egfl7 tumors but was locally mice in similar conditions are plotted at the same scale for comparison; inset, average final weights of 4T1-Ctrl and 4T1-Egfl7 tumors. B, left, restricted to the tumor tissue (Supplementary Fig. S4). lungs were dissected and photographed; arrows, macrometastases; Together, these results show that 4T1-Egfl7 tumors formed a right, average numbers of macrometastases counted at the surface of local immunodeficient environment. lungs of each animal group.

www.aacrjournals.org Cancer Res; 71(23) December 1, 2011 7181

Delfortrie et al.

Figure 5. Egfl7 alters the tumor endothelium characteristics. A, dendritic cells (1 105, left) were cultured for 24 hours in the presence of LPS and rEgfl7 where indicated (þ) and IL-6 was quantified in the supernatants. NK cells (5 104, middle) were incubated for 48 hours with IL-12 and IL-18 and with rEgfl7 where indicated (þ) and IFN-g was quantified in the supernatants. Purified splenic T lymphocytes (5 105, right) were stimulated with plate-bound anti-CD3 and soluble anti-CD28 in the presence of rEgfl7 where indicated (þ) for 48 hours and production of IFN-g in the supernatants was quantified. B, 4T1-Ctrl and 4T1-Egfl7 tumors were sectioned and immunostained for ICAM-1 (left, green) or VCAM-1 (right, red) and counterstained with DAPI (blue); bars, 100 mm. C, relative quantities (RQ) of ICAM-1 and VCAM-1 transcripts measured by quantitative RT-PCR in endothelial cells isolated from 4T1- Egfl7 and 4T1-Ctrl tumors. For each sample, levels were normalized to the actin expression levels. Data are expressed at 2ðÞDDCt , taking 4T1- Ctrl values as reference. D, expression levels of the indicated genes in endothelial cells isolated from 4T1-Ctrl (white bars) and 4T1- Egfl7 (black bars) tumors as assessed by quantitative RT-PCR as above. Data are representative of 2 experiments carried out in similar conditions.

anti-CD3/anti-CD28–stimulated production of IFN-g by a likely explanation for the observed deficit of immune cells T lymphocytes (Fig. 5A) as well as their proliferation (Supple- and cytokines within 4T1-Egfl7 tumors. In addition, endothe- mentary Fig. S5B). lial genes involved in promoting vessel integrity, maturation, We thus investigated whether Egfl7 could alter the recruit- and quiescence, such as Tie-2 and PAI-1, were repressed by ment of immune cells within the tumors. Expression of cell more than 80% in endothelial cells of 4T1-Egfl7 tumors, adhesion molecules such as E- and P-selectins, ICAM-1, VCAM- whereas expression of uPA was increased in these cells (Fig. 1, and CD31/PECAM, by endothelial cells plays a crucial role in 5D). Expression of Dll4 was strongly repressed, in agreement leukocyte rolling and adhesion before transendothelial migra- with the observed increased vascular density and hypoxia, but tion (10). We detected high levels of expression of ICAM-1 and in apparent contradiction with the enhanced tumor develop- VCAM-1 in 4T1-Ctrl tumor blood vessels, whereas expression ment (24). Of note, flt-1 expression was strongly increased in was weak in 4T1-Egfl7 tumors (Fig. 5B). To confirm this 4T1-Egfl7 tumor endothelial cells, and the expression levels of þ observation, we isolated CD45 /CD31 endothelial cells from CD31/PECAM, P-selectin, VE-cadherin, eNOS, and of the integ- tumors. In agreement with the immunostainings, the expres- rin subunits av, a3, b1, and b3 were not modified (not shown). sion levels of ICAM-1 and VCAM-1 in endothelial cells purified Together, these results suggest that the decreased frequency from 4T1-Egfl7 tumors were reduced by 57% and 70%, respec- of immune cells within 4T1-Egfl7 tumors is due to the inhi- tively, when compared with levels in controls (Fig. 5C). Endo- bition of their recruitment rather than to the repression of their thelial cells isolated from 4T1-Egfl7 tumors also expressed functions, and corresponds to phenotypic modifications of the much lower levels of E-selectin (Fig. 5D). These results provide endothelium.

7182 Cancer Res; 71(23) December 1, 2011 Cancer Research

Egﬂ7 Promotes Tumor Escape from Immunity

Egfl7 directly regulates leukocyte adhesion on affected (Fig. 6C). Furthermore, repressing either ICAM-1 or endothelial cells VCAM-1 (Supplementary Fig. S6B and S6C) in endothelial cells The possibility that Egfl7 directly affects the expression of treated with a siRNA targeting Egfl7 reduced the effects of this leukocyte adhesion molecules by endothelial cells was studied latter siRNA on T-cell adhesion (Fig. 6D), suggesting that the in vitro using the Jurkat T-cell lymphoma model of immune cell repressing effects of Egfl7 depends directly on the repression of adhesion on human primary HUVECs, a model known to ICAM-1 and VCAM-1. depend on ICAM-1 and VCAM-1 (25). Treatment of HUVECs with medium conditioned by 4T1-Egfl7 cells reduced the Expression of ICAM-1, VCAM-1, and IFN-g is repressed in number of T-lymphocytes adhering onto the endothelial human tumors expressing Egfl7 monolayer by half when compared with cells incubated with To validate our observations in human cancers, expres- a medium conditioned by 4T1-Ctrl cells (Fig. 6A). Repressing sion of ICAM-1, VCAM-1, and IFN-g was analyzed in a the endogenous egfl7 gene in endothelial cells using RNA series of human breast carcinomas which were selected on interference (Supplementary Fig. S6A) doubled the number of the basis of their expression levels of Egfl7 in tumor cells T-lymphocytes adhering to HUVECs when compared with (G. Lauridant-Philippin and F. Soncin; personal communi- control (Fig. 6B). This correlated with a large increase in cation). Within the same lesion, there was a 23% and 13% expression of E-selectin, VCAM-1, and ICAM-1 transcripts, decrease in the numbers of blood vessels expressing ICAM-1 whereas expression of P-selectin and CD31/PECAM was not and VCAM-1, respectively, when these vessels were in close

Figure 6. Egfl7 prevents the adhesion of T lymphocytes on endothelial cells. A, left, DiI-labeled Jurkat T lymphocytes were seeded onto a monolayer of confluent HUVECs that had been treated for 24 hours with a medium conditioned by 4T1-Ctrl (CM-Ctrl) or 4T1-Egfl7 (CM-Egfl7) cells; bar, 100 mm; right, values represent the average numbers of adhering Jurkat cells counted in 15 independent 2.4-mm2 fields. B, left, DiI-labeled Jurkat T lymphocytes were seeded onto a monolayer of confluent HUVECs that had been transfected with a control siRNA (si-Ctrl) or a siRNA targeting Egfl7 (si-Egfl7); bar, 100 mm; right, values represent the average numbers of adhering Jurkat cells counted (as in A). These experiments are representative of 3 experiments carried out in similar conditions. C, expression levels of E-selectin (E-sel), P-selectin (P-sel), VCAM-1, ICAM-1, and CD31/PECAM measured by quantitative RT-PCR in HUVECs transfected with a control siRNA (si-Ctrl) or a siRNA targeting Egfl7 (si-Egfl7). Levels are expressed at 2ðÞDDCt , taking si-Ctrl values as reference. D, DiI-labeled Jurkat T lymphocytes were seeded onto a monolayer of confluent HUVECs that had been previously transfected with a control siRNA (si-Ctrl) or siRNAs targeting Egfl7 (si-Egfl7), ICAM-1 (si-ICAM-1), or VCAM-1 (si-VCAM-1) and quantified; values represent the average numbers of adhering Jurkat cells counted in 9 independent 2.4-mm2 fields. , P < 0.05. RQ, relative quantities.

www.aacrjournals.org Cancer Res; 71(23) December 1, 2011 7183

Delfortrie et al.

The few previous studies that have described the expression of Egfl7 in human cancers all suggested that Egfl7could promote tumor growth and metastasis (18–20), but no experimental study had addressed the direct role of Egfl7 in tumor development. Here, the results obtained using 2 independent tumor models and mouse genetic backgrounds validate this initial hypothesis made on clinical observations. Furthermore, overexpressing Egfl7inexperimental tumors allowed us to understand its functions in more detail. Egfl7 is not an oncogene as it does not confer per se intrinsic proliferative or invasive properties to lung or breast tumor cells in vitro. In vivo,Egfl7showsnoeffectson tumorgrowthwhencomparedwithgrowthcontrols,inthe absence of a functional immune system. Our data indicate that the effects of Egfl7 on tumor growth and metastasis are rather indirect: Egfl7 promotes tumor escape from immunity, which, in turn, promotes tumor progression. Interest- ingly, Egfl7 has no effect on the immune cells themselves. Indeed, it does not directly activate dendritic cells, NK cells, or T lymphocytes and does not affect their activation status upon stimulation. The main effect of Egfl7istorepressthe tumor endothelium activation so that immune cells remain sequestered in the blood circulation, thus preventing their infiltration within the tumor mass. This explains why all immune cells analyzed were underrepresented in tumors expressing Egfl7, regardless of the cell type. Our in vitro adhesion assays indicate that this diminished recruitment of Figure 7. High expression of Egfl7 correlates with low VCAM-1, ICAM-1, fl and IFN-g in human tumors. A, expression of Egfl7 (top, brown) and immune cells is directly mediated by Eg 7throughthe ICAM-1 (bottom, brown) was analyzed by immunohistochemistry in a downregulation of endothelial adhesion molecules. The þ þ human breast ductal invasive carcinoma (noted SBR II, RE ,RP , and downregulation of endothelial adhesion molecules in HER2 0). The tumor tissue, which expresses large amounts of Egfl7 within – tumors (12 15) and in human cancers (12, 14, 26) had tumor cells (top left, ), shows low levels of ICAM-1 in adjacent blood already been observed. Egfl7 favors tumor escape from vessels (bottom left, arrows). The peritumoral region, which expresses low levels of Egfl7, mostly in blood vessels (top right, ), shows strong immunity by downregulating the expression of endothelial expression of ICAM-1 in blood vessel endothelium (bottom right, arrows). adhesion molecules through mechanisms that are still elu- B, expression of ICAM-1 and VCAM-1 in blood vessels was quantified as sive. Of note, Egfl7 was recently reported to downregulate 2 the mean percentage of positive vessels in 25 independent 0.15-mm the NF-kB pathway in human coronary artery endothelial fields in tumors expressing low (0þ, n ¼ 5) and high (3þ, n ¼ 5) levels of Egfl7. , P < 0.05. C, human breast carcinomas (n ¼ 12 per group) were cellsafteranischemia/reoxygenationtreatment(27).Based analyzed for expression of Egfl7 (black bars) and IFN-g (gray bars) by on this observation and our results, it is thus possible that a quantitative RT-PCR. For each sample, data were normalized to the b2- direct repression of the NF-kBpathwaybyEgfl7inendo- microglobulin expression levels. Data are expressed at 2ðÞDDCt , taking thelial cells contributes to the repression of ICAM-1 and, the lowest average value as reference for each gene. RQ, relative possibly, that of VCAM-1 and E-selectin. quantities. We also observed that expression of Egfl7 in tumors increases blood vessel permeability and decreases expression of fl proximity to tumor cells expressing high levels of Eg 7, Tie-2 when compared with that of controls, suggesting that fl when compared with areas where expression of Eg 7was the endothelium integrity is altered in the presence of Egfl7. low (peritumoral; Fig. 7A and B). Furthermore, the levels of Because Egfl7 was shown to inhibit PDGF-BB–induced smooth expression of IFN-g were inversely correlated with the levels muscle cell migration (16), it is likely that its expression fl of expression of eg 7 measured in a series of human breast prevents the recruitment of perivascular cells to newly formed tumor samples (Fig. 7C). tumor blood vessels, thus decreasing vascular tightness. Such a lack of vessel integrity is commonly observed in tumors (2) and Discussion is proposed to favor tumor spreading through metastasis (4, 28). Interestingly, overexpression of angiopoietin-2 in We show here that Egfl7 is a natural repressor of endothelial tumors also increases vessel permeability but correlates with cell activation. It inhibits the expression of endothelial adhe- a higher rate of immune cell infiltration within the primary sion molecules and consequently reduces the adhesion of tumors, and with a reduced tumor growth rate (29). Together, lymphocytes onto the endothelium. When placed in a tumor these observations suggest that, depending on the conditions, context, these effects result in an increased escape from the disruption of the endothelium might lead both to a reduced immunity and a more rapid tumor growth. primary tumor and to an increase in metastasis, which could

7184 Cancer Res; 71(23) December 1, 2011 Cancer Research

Egﬂ7 Promotes Tumor Escape from Immunity

explain the observed dual anti–primary tumor/prometastatic Disclosure of Potential Conflicts of Interest effects of antiangiogenic therapies (30, 31). F. Soncin is Directeur de Recherche of the Institut National de la Santeetdela In human breast tumor cells, we observed that expression of fl fl Recherche Medicale. No potential con icts of interest were disclosed by other Eg 7 corresponds to a local decrease in ICAM-1 and VCAM-1 authors. expression in adjacent blood vessels, whereas more distant vessels are not affected. This suggests that Egfl7 has local Acknowledgments effects on blood vessels in human tumors that are similar to The authors thank Drs. G. Pages and I. Wolowczuk for reagents; R. Shapiro those observed in experimental tumors in mice. The inverse and R. Pierce for correcting the manuscript; and T. Chassat for help with animals. correlation between the expression levels of Egfl7 in human tumors and those of IFN-g further validates our hypothesis and Grant Support suggests that Egfl7 produces an immunodeficient environment within human breast cancer tissues. The study was supported by Ligue Nationale contre le Cancer (Equipe Tumor escape from immunity is undoubtedly an interesting Labellisee La Ligue, F. Soncin), Institut National du Cancer (F. Soncin N2008-053, F. Trottein R08046EE/RPT08003EEA), and Association pour la process to consider for the design of therapeutic tools aimed at Recherche sur le Cancer (F. Soncin). S. Pinte was supported by a postdoctoral preventing cancer progression and metastasis. Egfl7 therefore grant from Institut National du Cancer, and S. Delfortrie was supported by a represents a new target for interfering with this process. The CNRS-Region Nord-Pas de Calais PhD fellowship. The costs of publication of this article were defrayed in part by the fact that Egfl7 maintains the normal endothelium in a non- payment of page charges. This article must therefore be hereby marked activated state, however, implies that one should investigate advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this this path with caution, as interfering with the properties of fact. fl Eg 7 for treating cancer may produce adverse side effects in Received April 19, 2011; revised September 22, 2011; accepted September 28, normal organs. 2011; published OnlineFirst October 28, 2011.

References 1. Chung AS, Lee J, Ferrara N. Targeting the tumour vasculature: insights 15. Dirkx AE, Oude Egbrink MG, Kuijpers MJ, van der Niet ST, Heijnen VV, from physiological angiogenesis. Nat Rev Cancer 2010;10:505–14. Bouma-ter Steege JC, et al. Tumor angiogenesis modulates leuko- 2. Jain RK. Normalization of tumor vasculature: an emerging concept in cyte-vessel wall interactions in vivo by reducing endothelial adhesion antiangiogenic therapy. Science 2005;307:58–62. molecule expression. Cancer Res 2003;63:2322–9. 3. Hashizume H, Baluk P, Morikawa S, McLean JW, Thurston G, Roberge 16. Soncin F, Mattot V, Lionneton F, Spruyt N, Lepretre F, Begue A, et al. S, et al. Openings between defective endothelial cells explain tumor VE-statin, an endothelial repressor of smooth muscle cell migration. vessel leakiness. Am J Pathol 2000;156:1363–80. EMBO J 2003;22:5700–11. 4. Mazzone M, Dettori D, Leite de Oliveira R, Loges S, Schmidt T, Jonckx 17. Lelievre E, Hinek A, Lupu F, Buquet C, Soncin F, Mattot V. VE-statin/ B, et al. Heterozygous deficiency of PHD2 restores tumor oxygenation egfl7 regulates vascular elastogenesis by interacting with lysyl oxi- and inhibits metastasis via endothelial normalization. Cell 2009;136: dases. EMBO J 2008;27:1658–70. 839–51. 18. Huang CH, Li XJ, Zhou YZ, Luo Y, Li C, Yuan XR. Expression and 5. Alitalo K, Carmeliet P. Molecular mechanisms of lymphangiogenesis in clinical significance of EGFL7 in malignant glioma. J Cancer Res Clin health and disease. Cancer Cell 2002;1:219–27. Oncol 2010;136:1737–43. 6. Koebel CM, Vermi W, Swann JB, Zerafa N, Rodig SJ, Old LJ, et al. 19. Diaz R, Silva J, Garcia JM, Lorenzo Y, Garcia V, Pena C, et al. Adaptive immunity maintains occult cancer in an equilibrium state. Deregulated expression of miR-106a predicts survival in human colon Nature 2007;450:903–7. cancer patients. Genes Chromosomes Cancer 2008;47:794–802. 7. Castermans K, Griffioen AW. Tumor blood vessels, a difficult hurdle for 20. Wu F, Yang LY, Li YF, Ou DP, Chen DP, Fan C. Novel role for epidermal infiltrating leukocytes. Biochim Biophys Acta 2007;1776:160–74. growth factor-like domain 7 in metastasis of human hepatocellular 8. Herberman RB, Nunn ME, Holden HT, Lavrin DH. Natural cytotoxic carcinoma. Hepatology 2009;50:1839–50. reactivity of mouse lymphoid cells against syngeneic and allogeneic 21. Tomayko MM, Reynolds CP. Determination of subcutaneous tumor tumors. II. Characterization of effector cells. IntJ Cancer 1975;16:230–9. size in athymic (nude) mice. Cancer Chemother Pharmacol 1989;24: 9. Shrikant P, Mescher MF. Control of syngeneic tumor growth by 148–54. þ activation of CD8 T cells: efficacy is limited by migration away from 22. Abramoff MD, Magalhaes PJ, Ram SJ. Image processing with ImageJ. the site and induction of nonresponsiveness. J Immunol 1999;162: Biophotonics Int 2004;11:36–42. 2858–66. 23. Caetano B, Drobecq H, Soncin F. Expression and purification of 10. Muller WA. Mechanisms of transendothelial migration of leukocytes. recombinant vascular endothelial-statin. Protein Expr Purif 2006;46: Circ Res 2009;105:223–30. 136–42. 11. Wu NZ, Klitzman B, Dodge R, Dewhirst MW. Diminished leukocyte- 24. Kuhnert F, Kirshner JR, Thurston G. Dll4-Notch signaling as a thera- endothelium interaction in tumor microvessels. Cancer Res 1992;52: peutic target in tumor angiogenesis. Vasc Cell 2011;3:20. 4265–8. 25. Chan JR, Hyduk SJ, Cybulsky MI. Alpha 4 beta 1 integrin/VCAM-1 12. Piali L, Fichtel A, Terpe HJ, Imhof BA, Gisler RH. Endothelial vascular interaction activates alpha L beta 2 integrin-mediated adhesion to cell adhesion molecule 1 expression is suppressed by melanoma and ICAM-1 in human T cells. J Immunol 2000;164:746–53. carcinoma. J Exp Med 1995;181:811–6. 26. Shioi K, Komiya A, Hattori K, Huang Y, Sano F, Murakami T, et al. 13. Kuzu I, Bicknell R, Fletcher CD, Gatter KC. Expression of adhesion Vascular cell adhesion molecule 1 predicts cancer-free survival in clear molecules on the endothelium of normal tissue vessels and vascular cell renal carcinoma patients. Clin Cancer Res 2006;12:7339–46. tumors. Lab Invest 1993;69:322–8. 27. Badiwala MV, Tumiati LC, Joseph JM, Sheshgiri R, Ross HJ, Delgado 14. Griffioen AW, Damen CA, Martinotti S, Blijham GH, Groenewegen G. DH, et al. Epidermal growth factor-like domain 7 suppresses intercel- Endothelial intercellular adhesion molecule-1 expression is sup- lular adhesion molecule 1 expression in response to hypoxia/reoxy- pressed in human malignancies: the role of angiogenic factors. Cancer genation injury in human coronary artery endothelial cells. Circulation Res 1996;56:1111–7. 2010;122:S156–61.

www.aacrjournals.org Cancer Res; 71(23) December 1, 2011 7185

Delfortrie et al.

28. Rolny C, Mazzone M, Tugues S, Laoui D, Johansson I, Coulon C, et al. 30. Paez-Ribes M, Allen E, Hudock J, Takeda T, Okuyama H, Vinals F, et al. HRG inhibits tumor growth and metastasis by inducing macrophage Antiangiogenic therapy elicits malignant progression of tumors to polarization and vessel normalization through downregulation of PlGF. increased local invasion and distant metastasis. Cancer Cell Cancer Cell 2011;19:31–44. 2009;15:220–31. 29. Fagiani E, Lorentz P, Kopfstein L, Christofori G. Angiopoietin-1 and -2 31. Ebos JM, Lee CR, Cruz-Munoz W, Bjarnason GA, Christensen JG, exert antagonistic functions in tumor angiogenesis, yet both induce Kerbel RS. Accelerated metastasis after short-term treatment with a lymphangiogenesis. Cancer Res 2011;71:5717–27. potent inhibitor of tumor angiogenesis. Cancer Cell 2009;15:232–9.

7186 Cancer Res; 71(23) December 1, 2011 Cancer Research

Egfl7 Promotes Tumor Escape from Immunity by Repressing Endothelial Cell Activation

Suzanne Delfortrie, Sébastien Pinte, Virginie Mattot, et al.

Cancer Res 2011;71:7176-7186. Published OnlineFirst October 28, 2011.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-11-1301

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2021/03/20/0008-5472.CAN-11-1301.DC1

Cited articles This article cites 31 articles, 12 of which you can access for free at: http://cancerres.aacrjournals.org/content/71/23/7176.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/71/23/7176.full#related-urls

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 Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/71/23/7176. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.