Published July 6, 2016, doi:10.4049/jimmunol.1600052 The Journal of Immunology

A Critical Role for CD200R Signaling in Limiting the Growth and Metastasis of CD200+ Melanoma

Jin-Qing Liu,* Fatemeh Talebian,* Lisha Wu,*,† Zhihao Liu,*,† Ming-Song Li,† Laichu Wu,‡ Jianmin Zhu,x Joseph Markowitz,* William E. Carson, III,* Sujit Basu,* and Xue-Feng Bai*,x

CD200 is a cell surface glycoprotein that functions through engaging CD200R on cells of the myeloid lineage and inhibits their functions. Expression of CD200 was implicated in a variety of human cancer cells, including melanoma cells; however, its roles in tumor growth and immunity are not clearly understood. In this study, we used CD200R-deficient mice and the B16 tumor model to evaluate this issue. We found that CD200R-deficient mice exhibited accelerated growth of CD200+, but not CD2002, B16 tumors. Strikingly, CD200R- deficient mice receiving CD200+ B16 cells i.v. exhibited massive tumor growth in multiple organs, including liver, lung, kidney, and peritoneal cavity, whereas the growth of the same tumors in wild-type mice was limited. CD200+ tumors grown in CD200R-deficient mice contained higher numbers of CD11b+Ly6C+ myeloid cells, exhibited increased expression of VEGF and HIF1a genes with increased angiogenesis, and showed significantly reduced infiltration of CD4+ and CD8+ T cells, presumably as the result of reduced expression of T cell chemokines, such as CXCL9 and CXCL16. The liver from CD200R-deficient mice, under metastatic growth of CD200+ tumors, contained significantly increased numbers of CD11b+Gr12 myeloid cells and Foxp3+ regulatory T cells and reduced numbers of NK cells. Liver T cells also had a reduced capacity to produce IFN-g or TNF-a. Taken together, we revealed a critical role for CD200R signaling in limiting the growth and metastasis of CD200+ tumors. Thus, targeting CD200R signaling may potentially interfere with the metastatic growth of CD200+ tumors, like melanoma. The Journal of Immunology, 2016, 197: 000–000.

umor-associated inflammation and immune responses are lymphoid cells, including B lymphocytes and activated T cells key components in the tumor microenvironment (TME) that (14). CD200R, the cognate ligand for CD200, is also an IgSF T regulate tumor growth, progression, and metastasis (1). protein (15). The expression pattern of mouse and human CD200R Tumor-associated myeloid cells (TAMCs) are a group of cells that is similar, with strong expression in macrophages, neutrophils, and play key roles in inducing tumor angiogenesis (2, 3), activating mast cells (16). Triggering CD200R suppresses myeloid cell ac- invasion/metastasis (4, 5), and regulating tumor-specific T cell re- tivity in vitro, and engagement of CD200R by CD200 inhibits sponses (6). Thus, identification and characterization of key path- their activation (17). Unlike most of the IgSF receptors, CD200R ways that can regulate TAMCs are of critical importance for lacks ITIM domains (18); however, its 67-aa cytoplasmic tail developing cancer immunotherapy. In this regard, our recent mouse contains three tyrosine residues, and the third tyrosine residue is studies (7, 8) suggest that CD200–CD200R interaction may be im- located within an NPXY motif, which is phosphorylated upon portant for the growth and metastasis of tumors, such as melanoma. ligation of CD200R (19). This leads to the recruitment and phos- CD200 (also known as OX-2) is a member of the Ig superfamily phorylation of Dok-2 and Dok-1, which then bind to RasGAP (IgSF) of proteins. It contains two extracellular Ig domains and a and SHIP (19–21). In macrophages and mast cells, this cascade small intracellular domain with no known signaling motif (9). was shown to inhibit the phosphorylation of ERK, P38, and JNK CD200 is expressed in a variety of normal tissues (10–13) and (20). CD200R signaling in macrophage appears to limit autoim- mune inflammation in animal models of multiple sclerosis and ar- *Department of Pathology and Comprehensive Cancer Center, The Ohio State thritis (22) and lung injury caused by viral infection (23), because † University, Columbus, OH 43210; Department of Gastroenterology, Guangdong CD200-deficient mice were found to have a significantly increased Provincial Key Laboratory of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou 510515, China; ‡Davis Medical Research Center, Columbus, disease severity due to hyperactivation of macrophages. CD200R- OH 43210; and xPediatric Translational Medicine Institute, Shanghai Children’s Med- deficient mice were also shown to be more susceptible to arthritis, ical Center, Shanghai Jiaotong University School of Medicine, Shanghai 200127, China presumably due to enhanced functions of macrophages (24). These ORCIDs: 0000-0001-5681-8459 (Z.L.); 0000-0003-2490-5464 (J.M.); 0000-0002- findings indicate that CD200–CD200R interactions are primarily in- 8272-0807 (S.B.). volved in limiting the cellular functions of myeloid lineages of cells. Received for publication January 11, 2016. Accepted for publication June 12, 2016. Expression of CD200 is found in multiple types of cancer (25–27), This work was supported in part by a grant from the National Cancer Institute (R01 CA138427) and a Pelotonia Idea Award from The Ohio State University Compre- including melanoma (28). It is generally considered that expression hensive Cancer Center. of CD200 on cancer cells has a protumor effect (26–31). However, this Address correspondence and reprint requests to Dr. Xue-Feng Bai, Department of conclusion was based primarily on experiments using allogenic models. Pathology and Comprehensive Cancer Center, The Ohio State University Medical Our previous studies using B16 melanoma and J558 plasmacytoma Center, 129 Hamilton Hall, 1645 Neil Avenue, Columbus, OH 43210. E-mail address: [email protected] mouse syngeneic models suggest that tumor-expressed CD200 may Abbreviations used in this article: DC, dendritic cell; IgSF, Ig superfamily; IHC, inhibit tumor growth and metastasis via shaping the TME (7, 8). In immunohistochemistry; OSU, The Ohio State University; MDSC, myeloid-derived this study, we further evaluated the role of CD200R signaling in tu- suppressor cell; TAM, tumor-associated macrophage; TAMC, tumor-associated my- mor growth and metastasis using CD200R-deficient mice. We eloid cell; TME, tumor microenvironment; Treg, regulatory T cell; WT, wild-type. found that CD200R-deficient mice exhibited accelerated growth of + 2 Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 CD200 , but not CD200 , B16 tumors. Strikingly, CD200R-deficient

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1600052 2 CD200R SIGNALING IN TUMOR GROWTH AND IMMUNITY

FIGURE 1. IHC analyses of CD200 expression in hu- man melanoma samples. Melanoma sample with overex- pressed CD200 in tumor cells (left panel) or whose tumor cells do not express CD200 (right panel) (original mag- nification 3200). Frozen sections were used for IHC staining. Anti-human CD200 mAb (OX-104) was pur- chased from eBioscience.

mice receiving CD200+ B16 tumor cells i.v. exhibited massive tumor CD200R gene was targeted in the embryonic stem cells of 129/Sv-C57BL6 growth in multiple organs, such as liver, lung, kidney, and peritoneal mice, and the CD200R-mutated mice were bred into the C57BL/6 strain cavity, whereas the growth of the same tumor in wild-type (WT) for 12 generations. All mice were maintained and cared for in The Ohio State University (OSU) laboratory animal facilities, which are fully mice was limited. Thus, we revealed a critical role for CD200R accredited by OSU Institutional Animal Care and Use Committee. signaling in limiting the growth and metastasis of CD200+ tumors. Generation of CD200+ and CD2002 B16 cells Materials and Methods B16.F10 melanoma cells were transfected with a pcDNA3 (Invitrogen) Mice expression vector containing the full length of the mouse CD200 cDNA or a control pcDNA3 expression vector to generate CD200+ or CD2002 B16 C57BL/6 mice were purchased from The Jackson Laboratory. CD200R2/2 cells, as previously described (7, 8). The resulting hygromycin-resistant mice were generated via a contract with Taconic Farms. Exon 1 of the cells were selected for CD200 expression using flow cytometry. The

FIGURE 2. Melanoma cell expression of CD200 inhibits tumor growth and alters the TME. (A) B16.F10 melanoma cells were transfected with a CD200 expression plasmid or a control plasmid to generate CD200+ (B16-CD200) or CD2002 (B16-Ctrl) cells. Flow cytometry was used to analyze CD200 expression on B16-Ctrl (dashed line) or B16-CD200 (solid line) cells. (B) A total of 1 3 105 of B16-Ctrl or B16-CD200 cells was injected s.c. into each C57BL/6 mouse, and tumor growth was observed over time. (C) Flow cytometry analysis of tumor-infiltrating leukocytes was performed on disassociated tumor cells. (D) T cell subsets and activation status were analyzed by flow cytometry. Five mice were included in each group; data represent three ex- periments with similar results. *p , 0.05, **p , 0.01, ***p , 0.001, Student t test. The Journal of Immunology 3 generated CD200+ (B16-CD200) or CD2002 (B16-Ctrl) cells were main- TC-39 (reverse); mTNF-a:59-ATGAGAAGTTCCCAAATGGC-39 (forward) tained in RPMI 1640 medium (Life Technologies) supplemented with 5% and 59-CTCCACTTGGTGGTTTGCTA-39 (reverse); mTSP-1: 59-TGGA- FBS and 1% penicillin/streptomycin. GATGGAATCCTCAATG-39 (forward) and 59-CAGCTGGTCTGGATTG- TGTT-39 (reverse); and mVEGF: 59-AGAGAGCAACATCACCATGC-39 Establishment of s.c. and metastatic tumors (forward) and 59-GGTCTGCATTCACATCTGCT-39 (reverse). The HPRT To establish s.c. tumors in C57BL/6 and CD200R2/2 C57BL6 mice, 1 3 gene was simultaneously amplified as endogenous control. The primers were 5 9 9 9 10 B16-CD200 or B16-Ctrl cells/mouse were used for the s.c. injection. 5 -AGCCTAAGATGAGCGCAAGT-3 (forward) and 5 -TTACTAGGCAGA- 9 Development of tumors was monitored, and tumors were measured for TGGCCACA-3 (reverse). Each sample was assayed in triplicate, and the length (a) and width (b) every 3 d using a digital caliper. Tumor volumes experiments were repeated twice. The relative gene expression was calcu- 2 lated by plotting the Ct (cycle number), and the average relative expression were calculated as ab /2 (32). To establish metastatic tumors, each mouse 2DDCt 5 for each group was determined using the comparative method (2 ). was injected with 1 3 10 B16-CD200 or B16-Ctrl cells via the tail vein. Mice were monitored for up to 3–4 wk. At the end of the experiments, Abs and flow cytometry mice were sacrificed, and different organs were collected and weighed or underwent various analyses. For CD200 and CD200R staining, PE-labeled anti-CD200 (clone OX-90) and FITC-labeled anti-CD200R (OX-110) Abs (Serotec) were used. Real-time PCR FITC-, PE-, allophycocyanin- or PerCP-labeled Abs to CD4, CD8a, g a Quantitative real-time PCR was performed using an ABI 7900-HT sequence CD11b, Gr1, Ly6G, Ly6C, DX5, IFN- , TNF- , and Foxp3, as well as system (PE Applied Biosystems) with the QuantiTect SYBR Green PCR kit isotype-matched control Abs, were purchased from BD Biosciences. Cells (QIAGEN), in accordance with the manufacturers’ instructions. PCR was done were incubated with Abs in 0.1 M PBS (pH 7.4), supplemented with 1% using previously determined conditions (33). The following primers were FCS and 0.1% sodium azide, on ice for 30 min. Cells were then washed three times and fixed in 1% paraformaldehyde, followed by flow cytometry used for amplifying specific genes: mArg1: 59-ACAACCAGCTCTGGGAA- analysis. The intracellular cytokine staining procedure was the same as we TCT-39 (forward) and 59-TGTACACGATGTCTTTGGCA-39 (reverse); mCCL2: 59-GGCTGGAGAGCTACAAGAGG-39 (forward) and 59-ATGTC- described (7). Data were analyzed using FlowJo software (TreeStar, TGGACCCATTCCTTC-39 (reverse); mCCL22: 59-AACCTTCTTGCTCC- Ashland, OR). TCTGGA-39 (forward) and 59-CTTTGTGGTCCCATATGCTG-39 (reverse); Immunohistochemistry and immunofluorescence mCXCL9: 59-GAACTCAGCTCTGCCATGAA-39 (forward) and 59-GCA- TCGTGCATTCCTTATCA-39 (reverse); mCXCL10: 59-GCTGCAACTG- Immunohistochemistry (IHC) was used for the staining of frozen tissue CATCCATATC-39 (forward) and 59-TTTCATCGTGGCAATGATCT-39 sections of human melanoma biopsy samples. Immunostaining were per- (reverse); mCXCL16: 59-CCAGTGGGTCCGTGAACTA-39 (forward) and 59- formed on cancer tissues using 4 mg/ml of mouse anti-human CD200 mAb GGTACTGGCTTGAGGCAAAT-39 (reverse); mHIF1a:59-TCAAGTTG- (OX-104; eBioscience) for 120 min. Ab binding was detected using anti- GAACTGGTGGAA-39 (forward) and 59-TATAGGGAGCCAGCATCTCC-39 mouse peroxidase-conjugated EnVision reagent and diaminobenzidine as (reverse); mMMP9: 59-TAAGGACGGCAAATTTGGTT-39 (forward) and 59- chromogen. Sections were also counterstained with Mayer hematoxylin. CTTTAGTGGTGCAGGCAGAG-39 (reverse); mMIF: 59-CCACCATGCC- Immunofluorescence staining was performed on mouse tumors. Briefly, TATGTTCATC-39 (forward) and 59-GTGCACTGCGATGTACTGTG-39 established mouse tumors were harvested and frozen in Tissue-Tek OCT (reverse); mIL-1b:59-CACTACAGGCTCCGAGATGA-39 (forward) and medium (Sakura Finetek), and 10-mm-thick slices were prepared. Tissue 59-TTTGTCGTTGCTTGGTTCTC-39 (reverse); mIL4.F: 59-TGTACCAGG- sections were fixed in ice-cold acetone for 30 s and stained with the corre- AGCCATATCCA-3 9 (forward) and mIL4.R: 59-TTCTTCGTTGCTGTGAG- sponding fluorescent Abs overnight at 4˚C. The Abs used for fluorescence GAC-39 (reverse); mIL-6: 59-ACTTCACAAGTCGGAGGCTT-39 (forward) staining of tumor sections were FITC–anti-CD31 and FITC–anti-VEGF. and 59-TCTGCAAGTGCATCATCGT-39 (reverse); mNOS2: 59-ACCTT- After washing with PBS, slides were mounted with DAPI-containing GTTCAGCTACGCCTT-39 (forward) and 59-CATTCCCAAATGTGCTTG- VECTASHIELD mounting medium (Vector Laboratories) and were examined

FIGURE 3. Generation of CD200R2/2 mice. (A) Targeting strategy for the generation of CD200R2/2 mice. (B) CD200R expression on myeloid cells and T cells from different organs were analyzed by flow cytometry. Spleen CD4+ and CD8+ T cells were analyzed for the expression of CD200R. 4 CD200R SIGNALING IN TUMOR GROWTH AND IMMUNITY and photographed on an inverted three-color fluorescence microscope expression of CD200 significantly inhibited tumor growth in system (Nikon Ti-U). Images were analyzed and quantified using ImageJ C57BL/6 mice. We analyzed the cellular components of CD200+ software (National Institutes of Health). and CD2002 tumors and found that CD200+ tumors contained sig- Statistical analysis nificantly higher numbers of CD4+ and CD8+ T cells compared with CD2002 tumors (Fig. 2C). T cell subset analysis revealed that The Student t test was used to compare tumor size and number differences + + between two groups. For comparison of mice survival, Kaplan–Meier tumor-infiltrating CD4 T cells are mainly Foxp3 regulatory T cells survival analysis and the log-rank test were used (version 10.0; SPSS, (Tregs) and IFN-g–producing Th1 cells, whereas the proportion of Chicago, IL). A p value ,0.05 was considered significant. Tregs and Th1 cells did not show a significant difference between + 2 Ethics statement CD200 and CD200 tumors (Fig. 2D). In contrast, we found that the proportion of CD8+IFNg+ T cells was significantly increased in This study and experimental protocols were approved by OSU Institutional CD200+ tumors (Fig. 2D). For the myeloid compartment, we ob- Animal Care and Use Committee (permit number 2008A0093-R2). served significantly decreased numbers of CD11b+ cells, in partic- ular CD11b+Ly6C+ cells (Fig. 2C). Thus, expression of CD200 in Results melanoma cells significantly altered the cellular components of tu- Expression of CD200 in melanoma cells alters the TME mors and enhanced tumor-specific T cell responses.

Recent studies revealed that CD200 is frequently expressed on + human cancer cells, such as melanoma cells (28). IHC staining of A critical role for CD200R signaling in limiting CD200 frozen sections of melanoma biopsy tissues from patients revealed melanoma growth and metastasis 2 that melanoma cells extensively express CD200 or are CD200 To determine whether the inhibition of tumor growth or metastasis (Fig. 1). It was suggested that overexpression of CD200 on mela- of melanoma-expressed CD200 was through CD200R, we gen- noma cells inhibits T cell responses to melanoma, thereby promot- erated CD200R2/2 mice. As depicted in Fig. 3A, exon 1 of the ing melanoma tumor growth and metastasis (28); however, we (8) CD200R gene was targeted in the embryonic stem cells of 129/Sv- showed that expression of CD200 in B16 melanoma cells inhibits C57BL6 mice, and the CD200R-mutated mice were bred with the tumor growth and metastasis in syngenic immune-competent the C57BL/6 strain for 12 generations. By comparison of WT and mice. We hypothesized that tumor-expressed CD200 interacts with CD200R2/2 mice, we found that CD200R was differentially expressed CD200R on tumor associated myeloid cells (TAMCs) and inhibits in CD11b+ myeloid cells from different organs (Fig. 3B); however, their function, thereby shaping the TME, which can directly or in- both CD4+ and CD8+ T cells in the peripheral lymphoid organs directly affect T cell responses in tumors. To determine whether this lacked CD200R expression (Fig. 3B). To test whether lack of is the case in the melanoma TME, we injected B16-CD200 or B16- CD200R signaling affects tumor growth and progression, we in- Ctrl cells (Fig. 2A) into C57BL/6 mice s.c. As shown in Fig. 2B, jected B16-CD200 melanoma cells s.c. into CD200R2/2 mice and

FIGURE 4. CD200R-deficient mice had accelerated in situ and metastatic growth of B16-CD200 tumors. A total of 1 3 105 B16-CD200 cells (A)or B16-Ctrl cells (B) was injected s.c. into each CD200R2/2 mouse and their WT breeding littermates, and tumor growth was observed. (C) B16-Ctrl and B16- CD200 tumors grew similarly in CD200R2/2 mice. A total of 1 3 105 B16-CD200 cells or B16-Ctrl cells was injected s.c. into each CD200R2/2 mouse to observe tumor growth. (D) A total of 1 3 105 B16-CD200 cells was injected i.v. into each CD200R2/2 mouse and control WT littermates. Mice were sacrificed 19 d after melanoma cell injection. Representative mice and melanoma metastasis to the lungs, livers, and kidneys are shown. The weights of lungs, livers, and kidneys were quantified. Data shown represent three (A–C) and five (D) experiments with similar results. Five mice per group were used for the experiments shown in (A)–(C). *p , 0.05, **p , 0.01, Student t test. The Journal of Immunology 5 their WT littermates. Accelerated tumor growth was observed in in CD200R2 tumors. CD200R deficiency does not significantly affect CD200R2/2 mice (Fig. 4A); however, CD2002 B16 tumors grew the expression of thrombospondin 1, IL-1b, and other M1/M2 genes, 2 2 similarly in WT and CD200R / mice (Fig. 4B), and B16-CD200 such as MMP9, CCL2, IL-4, IL-6, Arg1, and NOS2; however, the and B16-Ctrl tumors also grew similarly in these mice (Fig. 4C). expression of TNF-a gene was significantly reduced in CD200R2 + Thus, the growth difference in CD200 tumors between WT and tumors (Fig. 5B). Immunofluorescence staining of tumor sections 2/2 CD200R mice is CD200R dependent. To test whether CD200R revealed that CD200R-deficient B16-CD200 tumors contained a deficiency affects melanoma metastasis, B16-CD200 melanoma higher density of CD31+ blood vessels (Fig. 5C). Thus, CD200R 2/2 cells were injected i.v. into CD200R and control mice. Nineteen deficiency leads to the increased accumulation of myeloid cell 2/2 days after melanoma cell injection, CD200R mice exhibited populations and tumor angiogenesis. increased numbers of tumor foci in the abdominal cavity and or- gans, such as livers and lungs (Fig. 4D), such that increased weights CD200R deficiency leads to decreased T cell responses in + of livers and lungs from CD200R2/2 mice were observed. In some CD200 tumors cases, extensive tumor formation in the kidneys was also observed Because we found that CD200+ B16 melanoma grew much faster (Fig. 4D). Thus, CD200R signaling appears to play a critical role in in CD200R2/2 mice, we further investigated whether the enhanced + limiting CD200 melanoma tumor growth and metastasis. tumor growth of B16-CD200 tumors in these mice was due to reduced 5 Increased expansion of myeloid cells and angiogenesis in T cell responses in the tumors. B16-CD200 cells (1 3 10 /mouse) 2/2 CD200R-deficient tumors were injected s.c. into cohorts (n = 5) of WT or CD200R mice. Established tumors were disassociated and analyzed by flow cytom- Because CD200R is primarily expressed in the myeloid lineage of etry. As shown in Fig. 6A, we found that B16-CD200 tumors from cells, we hypothesized that lack of CD200R signaling in the TME 2/2 + + CD200R mice contained significantly reduced numbers of CD4 would affect the growth of CD200 tumors by affecting myeloid + cells. We first examined whether the myeloid populations were dif- and CD8 T cells compared with tumors from WT mice. However, ferent between B16-CD200 tumors grown in WT and CD200R2/2 CD200R deficiency does not appear to affect the production of cy- + + mice.AsshowninFig.5A,B16-CD200tumorsgrowninCD200R2/2 tokines, such as IFN-g and TNF-a,byCD8 (Fig. 6B, 6D) and CD4 + mice contained significantly increased numbers of myeloid cells, (Fig. 6C, 6D) T cells. The proportion of Foxp3 Tregs was also 2/2 which was dominated by the expansion of CD11b+Gr1+ cells, fol- similar in B16-CD200 tumors grown in WT and CD200R mice lowedbyCD11b+Ly6C+ and CD11b+Ly6C2 cells. CD11b2Gr1+ (Fig. 6B, 6D). Gene-expression analysis revealed that CD200R- neutrophils were also increased. Analysis revealed that the expression deficient tumors expressed significantly lower levels of CXCL9 and of VEGF, HIF1a, and migration inhibitory factor genes was elevated CXCL16 and elevated levels of CCL22 (Fig. 6E). Thus, CD200R

FIGURE 5. Increased tumor angiogenesis in CD200R-deficient tumors. B16-CD200 cells (1 3 105/mouse) were injected s.c. into cohorts (n =5)ofWT or CD200R2/2 mice. (A) Flow cytometry analyses of myeloid cells from B16-CD200 tumors grown in C57BL6 or CD200R2/2 mice. Representative flow cytometry analyses of myeloid cells in tumors and summary of their different subsets are shown. Established tumors were also analyzed for the expression of a number of genes by quantitative real-time PCR (B) and histological immunofluorescence staining for CD31 (C, left panels). Scale bars, 100 mm. (C) CD31+ cells were quantified from images captured under a fluorescence microscope and quantified using ImageJ software (right panel). Five tumors from each group were analyzed. *p , 0.05, **p , 0.01, Student t test. 6 CD200R SIGNALING IN TUMOR GROWTH AND IMMUNITY signaling appears to primarily affect the accumulation of CD4+ and significantly reduced NK cells compared with livers of WT mice, CD8+ T cells in the TME. whereas IFN-g productivity was not significantly different between Because CD200R deficiency had a particularly significant im- NK cells in livers of WT and CD200R-deficient mice (Fig. 7F). pact on B16-CD200 tumor formation in the liver, we tested whether the T cell responses in the liver were different between WT and Discussion 2 2 CD200R / mice during tumor growth. B16-CD200 cells were In this study, using the CD200R-deficient mouse model, we further 2 2 injected i.v. into cohorts (n = 5) of WT and CD200R / mice. revealed that CD200R signaling plays a critical role in limiting the Mice were sacrificed 19 d after tumor cell injection, and liver growth of CD200+, but not CD2002, melanoma tumors. CD200R leukocytes were isolated, followed by flow cytometry analyses. deficiency resulted in an altered TME, leading to increased tu- 2 2 We found that livers of CD200R / mice contained significantly mor angiogenesis and reduced T cell infiltration and/or effector 2 elevated numbers of CD11b+Gr1 myeloid cells (Fig. 7A), slightly functions. reduced total CD4+ T cells (Fig. 7B), and significantly increased Expression of CD200 was implicated in a variety of human Foxp3+ Tregs (Fig. 7C). CD4+IFNg+ (Fig. 7C) and CD8+TNFa+ cancer cells, including melanoma cells (28), and was thought to 2 2 (Fig. 7D) cells were also reduced in the livers of CD200R / mice play protumor effects via the direct inhibition of tumor-reactive bearing B16-CD200 tumors compared with WT mice bearing B16- T cells (25, 28, 29, 31). However, the fact that CD200R is pri- CD200 tumors. Thus, CD200R deficiency leads to increased Treg marily expressed in TAMCs, but not primarily on T cells in the expansion and reduced T cell responses in the liver during the TME (7), suggests that tumor-expressed CD200 acts mainly via growth of B16-CD200 tumors. Because NK cells were implicated interaction with myeloid cells in the TME. In syngenic mouse in melanoma cell liver colonization (34–36), we also examined the studies, it was shown in some slow-growth tumor models that the numbers of DX5+ NK cells in the livers of CD200R2/2 and WT CD200 signal, derived from tumor cells or host cells, inhibited mice bearing B16-CD200 tumors. As shown in Fig. 7E, we found antitumor immune responses (37–40). However, the results from that livers of CD200R2/2 mice bearing B16-CD200 tumors had these studies are more mixed, and none of these studies used the

FIGURE 6. Flow cytometry analyses of tumor-infiltrating lymphocytes from s.c. grown B16-CD200 tumors in CD200R2/2 and WT mice. B16-CD200 cells (1 3 105/mouse) were injected s.c. into cohorts (n = 5) of WT and CD200R2/2 mice. Established tumors ∼ 1 cm in diameter were disassociated, followed by flow cytometry analyses. (A) Disassociated tumors were stained for CD45, CD4, and CD8, followed by flow cytometry analysis. Data shown were gated on CD45+ cells, and a summary of data of five tumors from each group is shown. Intracellular staining and flow cytometry were used to analyze CD8+ (B) and CD4+ (C) T cells for the expression of IFN-g, TNF-a, and Foxp3. (D) Data shown are plotted from five tumors in each group. (E) Expression of chemokine genes in B16-CD200 tumors from WT and CD200R2/2 mice were quantified by quantitative PCR. Five tumors from each group were analyzed. *p , 0.05, **p , 0.01, Student t test. The Journal of Immunology 7

FIGURE 7. Flow cytometry analyses of leukocytes from the livers of CD200-deficient and WT mice with metastatic B16-CD200 tumors. B16-CD200 cells (1 3 105/mouse) were injected i.v. into cohorts (n = 5) of WT and CD200R2/2 mice. Mice were sacrificed 19 d after tumor cell injection, and liver leukocytes were isolated, followed by flow cytometry analyses. Leukocytes from livers were stained for CD45, CD11b, and Gr1 (A), CD45, CD4, and CD8 (B), or DX5 (E), followed by flow cytometry analysis. Data shown were gated on CD45+ cells, and summary of data of five samples from each group is shown. Intracellular staining and flow cytometry were used to analyze liver CD4+ T cells (C), CD8+ T cells (D), and DX5+ NK cells (F) for the expression of IFN-g, TNF-a, or Foxp3. Data shown are a summary of five samples from each group, representing two experiments with similar results. *p , 0.05, **p , 0.01, Student t test. more aggressive mouse melanoma model. Thus, the discrepancy CD200R2/2 mice, B16-CD200 tumor formation and metastasis between previously published results and this study suggests that were greatly accelerated (Fig. 4). Our study also revealed in- the role of CD200 in tumor immunity may differ in different tu- creased expression of VEGF, HIF1a, and migration inhibitory mor types (41) and may be associated with tumor aggressiveness factor genes (Fig. 5B) and CD31+ vasculatures (Fig. 5C) in and its ability to induce inflammation, as suggested by a recent CD200R-deficient B16-CD200 tumors. Thus, we hypothesize that study (42) using CD200R-deficient mice. lack of CD200R signaling in myeloid cells results in their accu- We demonstrated recently that TAMCsexpresshighlevelsof mulation and activation in the TME. This, in turn, leads to in- CD200R and are susceptible to CD200-mediated inhibition (7) in a creased production of an array of proangiogenic products (3, 43– CD200R-dependent manner (8). In this study, we found that CD200 45), such as VEGF, leading to increased tumor angiogenesis and on melanoma cells can significantly alter TAMC populations. In WT tumor growth/metastasis. Indeed, a recent study (46) revealed that mice, B16-CD200 tumors contained significantly lower proportions lack of CD200R signaling in bone marrow–derived macrophages of CD11b+Ly6C+ myeloid cells (Fig. 2). In s.c. grown B16-CD200 resulted in increased expression of VEGF, Arg1, and IL-1b in tumors, CD200R deficiency resulted in increased proportions of response to PGE2, which increased their ability to mediate an- Ly6C+ and Ly6C2 CD11b+ myeloid cells and CD11b2 neutrophils giogenesis. Moreover, this study also showed that laser-induced (Fig. 5A), whereas in the metastatic-grown B16-CD200 tumors, such choroidal neovascularization was enhanced in CD200R-deficient as those in the liver, significantly increased CD11b+Gr12 myeloid cells mice. In this study, the mechanism of CD200R-regulating VEGF were observed (Fig. 7A). Thus, it appears that expression of CD200 on expression remains unsolved. Because the expression of IL-1b melanoma cells can potentially affect all of the subpopulations of and TNF-a genes was reduced in CD200R-deficient tumors, it is TAMCs cells in the TME, depending on the locations of tumors. unlikely that these factors are responsible for the increased ex- TAMCs, such as Ly6C2 myeloid-derived suppressor cells pression of VEGF and HIF1a. Nevertheless, our study provides (MDSCs) and CD11b+Ly6C+ cells, play key roles in inducing the first evidence, to our knowledge, that CD200R signaling is tumor angiogenesis (2, 3). In this study, we demonstrated that, in involved in inhibiting angiogenesis in the TME. 8 CD200R SIGNALING IN TUMOR GROWTH AND IMMUNITY

Although CD200R expression was found primarily in macro- Disclosures phages and neutrophils, studies also revealed lower levels of The authors have no financial conflicts of interest. CD200R expression in dendritic cells (DCs) and some subset of T cells (16, 41, 47). CD200R signaling was shown to induce IDO expression in plasmacytoid DCs, initiating an immunosuppressive References 1. Mantovani, A., P. Allavena, A. Sica, and F. Balkwill. 2008. Cancer-related in- pathway of tryptophan (48). Thus, it is anticipated that T cell flammation. Nature 454: 436–444. responses and T cell effector functions are also affected by 2. Yang, L., L. M. DeBusk, K. Fukuda, B. Fingleton, B. Green-Jarvis, Y. Shyr, CD200R signaling. Indeed, in CD2002/2 mice, autoantigen- L. M. Matrisian, D. P. Carbone, and P. C. Lin. 2004. Expansion of myeloid immune suppressor Gr+CD11b+ cells in tumor-bearing host directly promotes specific T cell responses were normal (22, 24); however, CTL Cancer Cell 2 2 tumor angiogenesis. 6: 409–421. responses in CD200 / mice were elevated in influenza viral in- 3. Murdoch, C., M. Muthana, S. B. Coffelt, and C. E. Lewis. 2008. The role fection models (23, 49). CD200 signaling was also shown to be of myeloid cells in the promotion of tumour angiogenesis. Nat. Rev. Cancer 8: 618–631. required for the induction of T cell tolerance (40, 50). Thus, 4. Qian, B. Z., and J. W. Pollard. 2010. Macrophage diversity enhances tumor CD200R signaling in DCs and T cells appears to inhibit T cell re- progression and metastasis. Cell 141: 39–51. 5. Mantovani, A., T. Schioppa, C. Porta, P. Allavena, and A. Sica. 2006. Role of sponses. However, in this study, we found that B16-CD200 tumors tumor-associated macrophages in tumor progression and invasion. Cancer Me- + + in WT mice contained more CD4 and CD8 T cells compared with tastasis Rev. 25: 315–322. B16-Ctrl tumors (Fig. 2), whereas CD200R-deficient B16-CD200 6. Gabrilovich, D. I., and S. Nagaraj. 2009. Myeloid-derived suppressor cells as + regulators of the immune system. Nat. Rev. Immunol. 9: 162–174. tumors contained significantly reduced numbers of CD4 and 7. Wang, L., J. Q. Liu, F. Talebian, H. Y. El-Omrani, M. Khattabi, L. Yu, and + CD8 T cells compared with B16-CD200 tumors (Fig. 6A). Al- X. F. Bai. 2010. Tumor expression of CD200 inhibits IL-10 production by tumor- though the proportions of IFN-g–andTNF-a–producing CD4+ and associated myeloid cells and prevents tumor immune evasion of CTL therapy. + Eur. J. Immunol. 40: 2569–2579. CD8 T cells were similar (Fig. 6), the overall numbers of effector 8. Talebian, F., J. Q. Liu, Z. Liu, M. Khattabi, Y. He, R. Ganju, and X. F. Bai. 2012. T cells in tumors are reduced as a result of the global reduction in Melanoma cell expression of CD200 inhibits tumor formation and lung metas- effector T cells. Thus, it appears that tumor CD200-mediated tasis via inhibition of myeloid cell functions. PLoS One 7: e31442. + 9. Barclay, A. N., M. J. Clark, and G. W. McCaughan. 1986. Neuronal/lymphoid CD200R signaling enhances T cell infiltration in CD200 tumors. membrane glycoprotein MRC OX-2 is a member of the immunoglobulin su- Because we found that tumor-infiltrating lymphocytes do not usu- perfamily with a light-chain-like structure. Biochem. Soc. Symp. 51: 149–157. 10. Koning, N., D. F. Swaab, R. M. Hoek, and I. Huitinga. 2009. Distribution of the ally express CD200R, whereas TAMCs (MDSCs and tumor- immune inhibitory molecules CD200 and CD200R in the normal central nervous associated macrophages [TAMs]) express high levels of CD200R system and multiple sclerosis lesions suggests neuron-glia and glia-glia inter- (7), we hypothesize that tumor-expressed CD200 directly interacts actions. J. Neuropathol. Exp. Neurol. 68: 159–167. 11. Ragheb, R., S. Abrahams, R. Beecroft, J. Hu, J. Ni, V. Ramakrishna, G. Yu, and with CD200R in TAMCs and differentially regulates their expression R. M. Gorczynski. 1999. Preparation and functional properties of monoclonal of key chemokines (51), such as CXCL9, CXCL16, and, possibly, antibodies to human, mouse and rat OX-2. Immunol. Lett. 68: 311–315. CCL22 (Fig. 6E), thereby regulating T cell infiltration into tumors. 12. Dick, A. D., C. Broderick, J. V. Forrester, and G. J. Wright. 2001. Distribution of OX2 antigen and OX2 receptor within retina. Invest. Ophthalmol. Vis. Sci. 42: Intravenously injected melanoma cells do not usually grow in 170–176. liver, abdominal cavity, and kidney, as demonstrated by the tumor- 13. Rosenblum, M. D., E. B. Olasz, K. B. Yancey, J. E. Woodliff, Z. Lazarova, K. A. Gerber, and R. L. Truitt. 2004. Expression of CD200 on epithelial cells of growth pattern in B6 mice; however, in CD200R-deficient mice, the murine hair follicle: a role in tissue-specific immune tolerance? J. Invest. + we observed massive metastatic growth of CD200 melanomas in Dermatol. 123: 880–887. these locations (Fig. 4). These results suggest that CD200R sig- 14. Wright, G. J., M. Jones, M. J. Puklavec, M. H. Brown, and A. N. Barclay. 2001. The unusual distribution of the neuronal/lymphoid cell surface CD200 (OX2) naling may play a more significant role in limiting metastasis of glycoprotein is conserved in humans. Immunology 102: 173–179. + CD200 tumors to these sites. This phenomenon may be explained 15. Barclay, A. N., G. J. Wright, G. Brooke, and M. H. Brown. 2002. CD200 and by the higher expression of CD200R in myeloid cells at these membrane protein interactions in the control of myeloid cells. Trends Immunol. 23: 285–290. anatomical locations (Fig. 3B). Additionally, we hypothesize that 16. Wright, G. J., H. Cherwinski, M. Foster-Cuevas, G. Brooke, M. J. Puklavec, the local immune responses are uniquely altered by the absence of M. Bigler, Y. Song, M. Jenmalm, D. Gorman, T. McClanahan, et al. 2003. CD200R signaling. NK cells were implicated in melanoma cell Characterization of the CD200 receptor family in mice and humans and their interactions with CD200. J. Immunol. 171: 3034–3046. liver colonization (34–36). Indeed, in addition to reduced T re- 17. Jenmalm, M. C., H. Cherwinski, E. P. Bowman, J. H. Phillips, and sponses and elevated Treg responses, we observed significantly J. D. Sedgwick. 2006. Regulation of myeloid cell function through the CD200 receptor. J. Immunol. 176: 191–199. reduced NK cells in CD200R-deficient livers undergoing meta- 18. Zhang, S., H. Cherwinski, J. D. Sedgwick, and J. H. Phillips. 2004. Molecular static tumor growth (Fig. 7). Although these results could explain mechanisms of CD200 inhibition of mast cell activation. J. Immunol. 173: 6786– why CD200R-deficient liver is sensitive to metastatic melanoma 6793. 19. Mihrshahi, R., A. N. Barclay, and M. H. Brown. 2009. Essential roles for Dok2 growth, the detailed mechanisms of how CD200R deficiency af- and RasGAP in CD200 receptor-mediated regulation of human myeloid cells. fects liver local NK responses remain to be studied. J. Immunol. 183: 4879–4886. Taken together, we found a critical role for CD200R signaling 20. Minas, K., and J. Liversidge. 2006. Is the CD200/CD200 receptor interaction + more than just a myeloid cell inhibitory signal? Crit. Rev. Immunol. 26: 213–230. in limiting the growth and metastasis of CD200 tumors. Tumor- 21. Mihrshahi, R., and M. H. Brown. 2010. Downstream of tyrosine kinase 1 and 2 expressed CD200 may do so by directly interacting with TAMCs play opposing roles in CD200 receptor signaling. J. Immunol. 185: 7216–7222. 22. Hoek, R. M., S. R. Ruuls, C. A. Murphy, G. J. Wright, R. Goddard, (MDSCs and TAMs), which differentially regulates their production S. M. Zurawski, B. Blom, M. E. Homola, W. J. Streit, M. H. Brown, et al. 2000. of VEGF and the expression of key chemokines, such as CXCL9 Down-regulation of the macrophage lineage through interaction with OX2 and CXCL16; this attracts T cells, affecting tumor angiogenesis and (CD200). Science 290: 1768–1771. + 23. Snelgrove, R. J., J. Goulding, A. M. Didierlaurent, D. Lyonga, S. Vekaria, T cell infiltration into tumors. More metastatic growth of CD200 L. Edwards, E. Gwyer, J. D. Sedgwick, A. N. Barclay, and T. Hussell. 2008. A melanoma to some anatomical locations and organs, such as liver, in critical function for CD200 in lung immune homeostasis and the severity of CD200R-deficient mice suggests that CD200R signaling is more influenza infection. Nat. Immunol. 9: 1074–1083. 24. Simelyte, E., S. Alzabin, I. Boudakov, and R. Williams. 2010. CD200R1 reg- important for immune responses in these organs. Given the im- ulates the severity of arthritis but has minimal impact on the adaptive immune portant roles of CD200–CD200R interactions in regulating TAMCs response. Clin. Exp. Immunol. 162: 163–168. 25. McWhirter, J. R., A. Kretz-Rommel, A. Saven, T. Maruyama, K. N. Potter, and in inhibiting tumor formation and metastasis, enhancing C. I. Mockridge, E. P. Ravey, F. Qin, and K. S. Bowdish. 2006. Antibodies se- CD200R signaling in tumors should provide an option for the im- lected from combinatorial libraries block a tumor antigen that plays a key role in munotherapy of human cancer. Our successful treatment of B16 immunomodulation. Proc. Natl. Acad. Sci. U S A 103: 1041–1046. 26. Moreaux, J., D. Hose, T. Reme, E. Jourdan, M. Hundemer, E. Legouffe, lung metastasis using a triggering anti-CD200R mAb (8) proves P. Moine, P. Bourin, M. Moos, J. Corre, et al. 2006. CD200 is a new prognostic that this approach may be feasible. factor in multiple myeloma. Blood 108: 4194–4197. The Journal of Immunology 9

27. Tonks, A., R. Hills, P. White, B. Rosie, K. I. Mills, A. K. Burnett, and 40.Rygiel,T.P.,G.Karnam,G.Goverse,A.P.vanderMarel,M.J.Greuter, R. L. Darley. 2007. CD200 as a prognostic factor in acute myeloid leukaemia. R. A. van Schaarenburg, W. F. Visser, A. B. Brenkman, R. Molenaar, Leukemia 21: 566–568. R. M. Hoek, et al. 2012. CD200-CD200R signaling suppresses anti-tumor 28. Petermann, K. B., G. I. Rozenberg, D. Zedek, P. Groben, K. McKinnon, responses independently of CD200 expression on the tumor. Oncogene 31: C. Buehler, W. Y. Kim, J. M. Shields, S. Penland, J. E. Bear, et al. 2007. CD200 2979–2988. is induced by ERK and is a potential therapeutic target in melanoma. J. Clin. 41. Rygiel, T. P., and L. Meyaard. 2012. CD200R signaling in tumor tolerance and Invest. 117: 3922–3929. inflammation: a tricky balance. Curr. Opin. Immunol. 24: 233–238. 29. Kretz-Rommel, A., F. Qin, N. Dakappagari, E. P. Ravey, J. McWhirter, 42. Erin, N., A. Podnos, G. Tanriover, O¨ . Duymus¸, E. Cote, I. Khatri, and D. Oltean, S. Frederickson, T. Maruyama, M. A. Wild, M. J. Nolan, et al. 2007. R. M. Gorczynski. 2015. Bidirectional effect of CD200 on breast cancer de- CD200 expression on tumor cells suppresses antitumor immunity: new ap- velopment and metastasis, with ultimate outcome determined by tumor ag- proaches to cancer immunotherapy. J. Immunol. 178: 5595–5605. gressiveness and a cancer-induced inflammatory response. Oncogene 34: 30. Kretz-Rommel, A., F. Qin, N. Dakappagari, R. Cofiell, S. J. Faas, and 3860–3870. K. S. Bowdish. 2008. Blockade of CD200 in the presence or absence of antibody 43. Solinas, G., G. Germano, A. Mantovani, and P. Allavena. 2009. Tumor- effector function: implications for anti-CD200 therapy. J. Immunol. 180: 699–705. associated macrophages (TAM) as major players of the cancer-related inflam- 31. Stumpfova, M., D. Ratner, E. B. Desciak, Y. D. Eliezri, and D. M. Owens. 2010. mation. J. Leukoc. Biol. 86: 1065–1073. The immunosuppressive surface ligand CD200 augments the metastatic capacity 44. Ruffell, B., N. I. Affara, and L. M. Coussens. 2012. Differential macrophage of squamous cell carcinoma. Cancer Res. 70: 2962–2972. programming in the tumor microenvironment. Trends Immunol. 33: 119–126. 32. Tomayko, M. M., and C. P. Reynolds. 1989. Determination of subcutaneous 45. Lewis, C. E., and J. W. Pollard. 2006. Distinct role of macrophages in different tumor size in athymic (nude) mice. Cancer Chemother. Pharmacol. 24: 148–154. tumor microenvironments. Cancer Res. 66: 605–612. 33. Bai, X. F., J. Liu, O. Li, P. Zheng, and Y. Liu. 2003. Antigenic drift as a 46. Horie, S., S. J. Robbie, J. Liu, W. K. Wu, R. R. Ali, J. W. Bainbridge, mechanism for tumor evasion of destruction by cytolytic T lymphocytes. J. Clin. L. B. Nicholson, M. Mochizuki, A. D. Dick, and D. A. Copland. 2013. CD200R Invest. 111: 1487–1496. signaling inhibits pro-angiogenic gene expression by macrophages and sup- 34. Gorelik, E., R. H. Wiltrout, K. Okumura, S. Habu, and R. B. Herberman. 1982. presses choroidal neovascularization. Sci. Rep. 3: 3072. Role of NK cells in the control of metastatic spread and growth of tumor cells in 47. Rijkers, E. S., T. de Ruiter, A. Baridi, H. Veninga, R. M. Hoek, and L. Meyaard. mice. Int. J. Cancer 30: 107–112. 2008. The inhibitory CD200R is differentially expressed on human and mouse T 35. Takeda, K., M. J. Smyth, E. Cretney, Y. Hayakawa, N. Yamaguchi, H. Yagita, and B lymphocytes. Mol. Immunol. 45: 1126–1135. and K. Okumura. 2001. Involvement of tumor necrosis factor-related apoptosis- 48. Fallarino, F., C. Asselin-Paturel, C. Vacca, R. Bianchi, S. Gizzi, M. C. Fioretti, inducing ligand in NK cell-mediated and IFN-gamma-dependent suppression of G. Trinchieri, U. Grohmann, and P. Puccetti. 2004. Murine plasmacytoid den- subcutaneous tumor growth. Cell. Immunol. 214: 194–200. dritic cells initiate the immunosuppressive pathway of tryptophan catabolism in 36. Dobos, J., A. Mohos, J. To´va´ri, E. Ra´so´,T.Lorincz,} G. Za´dori, J. Tı´ma´r, and response to CD200 receptor engagement. J. Immunol. 173: 3748–3754. A. Lada´nyi. 2013. Sex-dependent liver colonization of human melanoma in SCID 49. Rygiel, T. P., E. S. Rijkers, T. de Ruiter, E. H. Stolte, M. van der Valk, mice–role of host defense mechanisms. Clin. Exp. Metastasis 30: 497–506. G. F. Rimmelzwaan, L. Boon, A. M. van Loon, F. E. Coenjaerts, R. M. Hoek, 37. Gorczynski, R. M., Z. Chen, J. Hu, Y. Kai, and J. Lei. 2001. Evidence of a role et al. 2009. Lack of CD200 enhances pathological T cell responses during in- for CD200 in regulation of immune rejection of leukaemic tumour cells in fluenza infection. J. Immunol. 183: 1990–1996. C57BL/6 mice. Clin. Exp. Immunol. 126: 220–229. 50. Gorczynski, R. M., Z. Chen, W. He, I. Khatri, Y. Sun, K. Yu, and I. Boudakov. 38. Gorczynski, R. M., D. A. Clark, N. Erin, and I. Khatri. 2011. Role of CD200 2009. Expression of a CD200 transgene is necessary for induction but not expression in regulation of metastasis of EMT6 tumor cells in mice. Breast maintenance of tolerance to cardiac and skin allografts. J. Immunol. 183: Cancer Res. Treat. 130: 49–60. 1560–1568. 39. Gorczynski, R. M., Z. Chen, J. Diao, I. Khatri, K. Wong, K. Yu, and J. Behnke. 51. Franciszkiewicz, K., A. Boissonnas, M. Boutet, C. Combadie`re, and F. Mami- 2010. Breast cancer cell CD200 expression regulates immune response to EMT6 Chouaib. 2012. Role of chemokines and chemokine receptors in shaping the tumor cells in mice. Breast Cancer Res. Treat. 123: 405–415. effector phase of the antitumor immune response. Cancer Res. 72: 6325–6332.