C57BL/6 mice were purchased from The Jackson laboratory, and CD200R^–/– mice were originally generated via a contract with Taconic Farms and were bred into the C57BL6 background for over 12 generations (Liu et al., 2016). All mice were maintained and cared for in The Ohio State University (OSU) laboratory animal facilities which are fully accredited by OSU Institutional Animal Care and Use Committee (IACUC). To establish subcutaneous (s.c.) tumors in C57BL/6 and CD200R^–/–, 5 × 10⁵ Yumm1.7 cells in 100 μl PBS were injected into each mouse s.c. Development of tumors was monitored and tumors were measured for length (a) and width (b) every 2–3 days using a digital caliper, and tumor volumes were calculated as (a^∗b²)/2.

Total RNA was extracted from fresh tumors (n = 3/group) using Trizol^® according to the manufacturer’s instructions. Approximately 50 ng of total RNA were hybridized with the mouse PanCancer immune profiling code set containing 770 unique pairs of 35–50 base pair biotin-labeled capture probes and reporter probes with internal reference controls (NanoString nCounter^® PanCancer IO360 panel, NanoString Technologies, Inc., Seattle, WA). Hybridization was performed overnight at 65°C. Unbound probes were washed away, the tripartite structure was bound to the streptavidin-coated cartridge by the biotin capture probe, aligned by an electric current, and immobilized. Degradation of fluorophore and photobleaching were prevented by adding SlowFade. Read counts from the raw data output were assessed for differential gene expression and cell type scoring after normalization using Rosalind Software.¹ Briefly, Log₂ counts were represented as z-scores in heat map to indicate alterations in gene expression and immune cell profile for each sample.

Genomic DNA was prepared from tumor tissues and submitted for the immunoSEQ assay (Adaptive Biotechnologies, Seattle, WA). The somatically rearranged mouse T cell receptor (TCR) hypervariable complementarity-determining region 3 (CDR3) was amplified from genomic DNA of tumor samples using a two-step, amplification bias-controlled multiplex PCR approach (Robins et al., 2009; Carlson et al., 2013). Specifically, the first PCR consists of forward and reverse amplification primers specific for every V and J gene segment, and amplifies the CDR3 of the TCR locus. The second PCR adds a proprietary barcode sequence and Illumina adapter sequences. CDR3 libraries were sequenced on an Illumina instrument according to the manufacturer’s instructions. Raw Illumina sequence reads were demultiplexed according to Adaptive’s proprietary barcode sequences. Demultiplexed reads were then further processed to: remove adapter and primer sequences; identify and correct for technical errors introduced through PCR and sequencing; and remove primer dimer, germline and other contaminant sequences (Robins et al., 2012). The data is filtered and clustered using both the relative frequency ratio between similar clones and a modified nearest-neighbor algorithm, to merge closely related sequences. The resulting sequences were sufficient to allow annotation of the V(N)D(N)J genes constituting each unique CDR3 and the translation of the encoded CDR3 amino acid sequence. V, D and J gene definitions were based on annotation in accordance with the IMGT database.² The set of observed TCR CDR3 sequences were normalized to correct for residual multiplex PCR amplification bias and quantified against a set of synthetic TCR CDR3 sequence analogs (Carlson et al., 2013). Data was analyzed using the immunoSEQ Analyzer toolset.

Fluorescence labeled monoclonal antibodies to mouse CD45 (30-F11), CD3 (145-2c11), CD4 (GK1.5), CD8α (53-6.7), NK1.1 (Pk136), CD11b (M1/70), Gr-1 (RB6-8C5), Ly6C (HK1.4), F4/80 (745-2342), Foxp3 (Nrrf-3c), TNF-α (XT22), IFN-γ (XMG1.2), and isotype-matched control antibodies were purchased from Biolegend or BD Biosciences. Mononuclear cells from tumors were prepared as we previously described (Liu et al., 2016). For cell surface staining, cells were incubated with antibodies in 0.1 M PBS (pH7.4) supplemented with 1% 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 analysis. For intracellular staining of TNF-α, IFN-γ or Foxp3, cells were first stimulated with cell stimulation cocktail (Invitrogen) for 4 h in the presence of Gorgi^stop (BD Biosciences). The cells were first stained for the cell surface markers (CD4/8), followed by a standard intracellular cytokine staining procedure. A Celesta flow cytometer (BD) was used to detect stained cells. Data was analyzed using the flowjo software (Tree Star, Inc., OR).

Bone marrow (BM) cells from C57BL/6 and CD200R^–/– mice were plated in 15cm culture dishes in DMEM medium containing 20% FBS and 30% culture supernatant from L929-G-CSF cells. The cells were allowed to grow in an incubator at 37°C for 3 days. On day 3, fresh culture medium was added to the plate and allowed it to grow for another 2 days. Macrophages (adherent cells) were dissociated on day 5 using 5mM EDTA and re-suspended in DMEM medium containing 10% FBS and 1% L929-G-CSF supernatant. The macrophages (1 × 10⁶ cells/ml) were then co-cultured with Yumm1.7 tumor cells at 1:1 ratio or Yumm1.7 supernatant. Supernatant from another tumor cell line (NB9464D) was used as a negative control. 24 and 48 h later, culture supernatants were collected for ELISA. The detached cells were used for flow cytometry analysis or qPCR for chemokine gene expression analysis.

MAX CCL8 (MCP-2) ELISA kit (Biolegend) was used to quantify CCL8 in culture supernatants from BMDM/Yumm1.7 co-cultures according to manufacturer’s instructions.

Quantitative real-time PCR was done using previously determined conditions (33). The following primers were used for amplifying the CCL8 gene: 5′-ACGCTAGCCTCCACTCCAAA-3′ (forward) and 5′-GAGCCTTATCTGGCCCAGTC-3′ (reverse) and HPRT gene (endogenous control): 5′-AGCCTAAGATGA GCGCAAGT-3′ (forward) and 5′-TTACTACGCAGATGGCCA CA-3′ (reverse). Each sample was assayed in triplicate, and the relative gene expression was calculated by plotting the Ct (cycle number) and the average relative expression for each group was determined using the comparative method (2^–ΔΔCt).

We first established Yumm1.7 tumors in male or female C57BL/6 mice by injecting 5 × 10⁵ Yumm1.7 cells in 100 μl PBS. When tumors are palpable (typically 7 days after tumor cell injection), mice were treated with 250 μg/mouse of the following antibodies i.p. alone or in combination: αCD200 (OX-90), αPD1 (RPM1-14) or αCTLA-4 (9H10). Mice were treated every 3 days until the end of the experiments.

Statistical analyses were performed using GraphPad Prism 9.1.1 software (GraphPad Software Inc., La Jolla, CA, United States). The differences between the treatments compared to the untreated control were analyzed by multiple t-tests without multiple comparisons correction. The nanostring data were represented as mean of log2 fold change relative to control. All other data were presented as mean ± standard error of the mean (SEM) unless otherwise indicated. The overall P-value for Kaplan-Meier analysis was calculated using the log-rank test. Analysis of differences between two normally distributed test groups was performed using an unpaired t-test assuming unequal variance and multiple t-tests. P < 0.05 was considered to be statistically significant.

The Yumm1.7 melanoma tumor cell line was derived from a spontaneous tumor originally developed in a mouse bearing Braf/Pten mutations (Meeth et al., 2016). We found that Yumm1.7 cells constitutively expressed CD200 on its cell surface (Figure 1A). To investigate the role of CD200-CD200R signaling pathway in melanoma tumor establishment and progression, C57BL/6 and CD200R^–/– C57BL/6 mice were subcutaneously inoculated with CD200⁺ Yumm1.7 melanoma cells. About 1 week after tumor cell injection, tumors started to show up in all mice regardless of CD200R-deficiency (Figure 1B). However, tumors grew faster in CD200R^–/– mice compared to their wild type counterpart as measured by higher tumor growth curve (Figure 1B) and bigger area under the tumor volume curve (AUC) (Duan et al., 2012; Figure 1C). At the end of the experiment (day 20 after tumor cell injection), we sacrificed mice and measured tumor weight. We found that tumors from CD200R^–/– mice were significantly heavier than tumors harvested from WT mice (Figure 1D). Overall, our results show that the growth of CD200⁺ Yumm1.7 melanoma is enhanced when CD200R signaling is absent.

To understand what factors were driving the enhanced Yumm1.7 tumor growth in CD200R^–/– mice, RNA samples from tumors were used for Nanostring analysis. We used the pre-formatted Pan-cancer 360 IO panel to characterize and measure the expression of about 750 vital genes in the TME. We first compared Immune cell type abundance within the two TME. As shown in Figure 2A, Nanostring analysis revealed decreased immune cell presence in the CD200R^–/– TME. In 3 out of 3 CD200R^–/– tumor samples, the abundance scores of immune cells were low, while 2 out of 3 tumor samples from WT mice exhibited higher enrichment of immune cells, including total CD45⁺ leukocytes, total T cells and CD8⁺ T cells. Strikingly, pathway analysis revealed that majorities of immune pathways including adaptive immunity/T cell functions were downregulated (Figure 2B). To validate the immune cell type abundance data from Nanostring, we analyzed immune cell subsets using flow cytometry. As seen in Figure 2C, flow cytometry confirmed significant decreases of CD45⁺ cells in tumors from CD200R^–/– mice. Additionally, we observed decreased percentages of NK cells, T cells (CD3⁺) and CD8⁺ T cells among total leukocytes (CD45⁺) in tumors from CD200R^–/– mice, while percentages of B cells (B220⁺) and CD4⁺ T cells among total leukocytes were not significantly different between tumors from CD200R^–/– and WT mice. Similarly, we did not detect differences in CD11b⁺Gr1^– tumor associated macrophage (TAM) population and CD11b⁺Gr1⁺ myeloid derived suppressor (MDSC) cells (Figure 2D). CD4⁺FoxP3⁺ regulatory T cells (Tregs) are significant players in regulating T cell responses in TME. We therefore quantified Tregs and found that Tregs were significantly increased in the CD200R^–/– tumors (Figure 2E). To determine whether the functions of immune cells such as T cells were impacted by the lack of the CD200-CD200R interaction, we quantified IFN-γ and TNF-α levels in CD4⁺ and CD8⁺ tumor infiltrating lymphocytes (TIL). We found that both IFN-γ and TNF-α production were significantly suppressed in CD4⁺ (Figure 2F) and CD8⁺ T cells (Figure 2G) within the CD200R^–/– tumors.

To assess the impact of CD200-CD200R signaling on the diversity of the immune repertoire of tumor-infiltrating T cells, we prepared genomic DNA from tumors (n = 4/group) and performed TCR-seq analysis (Adaptive Technology, Seattle, Washington). We found that in the absence of CD200R signaling, clonality of tumor infiltrating T cells were similar (Figure 2H), suggesting that lack of CD200R signaling did not significantly affect dominant T cell populations in TME.

Differential gene expression analysis of the data generated from the Nanaostring assay revealed significant alterations in TME (Figure 3A). Among the 26 differentially expressed genes, 22 were genes downregulated in the CD200R^–/– TME. Consistent with the downregulated immune pathways indicated in Figure 2A, the downregulated genes included genes important for T cell trafficking and effector functions. Notably, several chemokine and chemokine receptors (CCR6, CCR7, CCL19, CXCL13) were also downregulated. On the other end of the spectrum, there were 4 upregulated genes in TME in the absence of CD200R. Notable among them was CCL8, a chemokine ligand for murine CCR8 (Figure 3B). Our qPCR analysis verified its upregulation in tumors from CD200R-dificient mice (Figure 3C).

To determine if tumor CD200 interacts with CD200R on macrophages to regulate CCL8 production, we generated bone marrow derived macrophages (BMDM) from CD200R^–/– and WT mice, co-cultured them with Yumm1.7 tumor cells and measured CCL8 gene expression and protein secretion. As shown in Figure 3D, we found that CCL8 gene expression was significantly higher in CD200R^–/– macrophages after a 24 and 48 h co-culture with Yumm1.7 cells. At the protein level, most significant up-regulation of CCL8 was detected 48 h after co-culture (Figure 3E). Overall, these results indicate that tumor CD200 interacts with CD200R and limits the secretion of CCL8.

It has been reported (Kretz-Rommel et al., 2007; Choueiry et al., 2020) that CD200 blockade may improve the efficacy of cancer immunotherapy. To determine if CD200 blockade could inhibit the growth of CD200⁺ melanoma, we tested a monoclonal antibody (OX-90) in the Yumm1.7 tumor model. OX-90 has been shown to block CD200-CD200R interaction (Snelgrove et al., 2008; Choueiry et al., 2020). We inoculated C57BL/6 mice with 0.5 × 10⁶ Yumm1.7 melanoma cells subcutaneously. Once the tumors were established, the mice were injected with the OX-90 antibody (250 μg/mouse) i.p. every 3 day until the end of the experiment. As shown in Figure 4, treatment of mice with OX-90 failed to affect Yumm1.7 tumor growth, as measured by tumor growth kinetics (Figure 4A), mean area under curve (Figure 4B) and tumor weight at the end of the experiment (Figure 4C).

To understand if OX-90 treatment alters TIME, we analyzed the immune cell abundance using Nanostring assay. The OX-90 treated group followed a pattern similar to that seen in the CD200R^–/– mice. Specifically, we found reduction of total CD45⁺ cells, total T cells, CD8⁺ T cells and NK cells in tumors from OX-90 treated mice (Figure 5A). Consistent with this observation, our flow cytometry analysis verified that total CD45⁺ and T cells were reduced (Figure 5B), while proportions of other lymphocyte subsets were not significantly altered by OX-90 treatment (Figure 5C). Strikingly, ImmunoSEQ analysis indicated that OX-90 treatment decreased the clonality of tumor infiltrating T cells (Figure 5D), suggesting that OX-90 treatment decreases the quality of T cell responses in TME.

Since anti-CD200 monotherapy failed to show efficacy in inhibiting Yumm1.7 tumor growth, we tested if anti-CD200 in combination with other checkpoint inhibitors will result in tumor growth inhibition. We first tested if anti-CD200 treatment had synergy with anti-PD-1 therapy. As shown in the upper panel of Figure 6, neither anti-CD200 or anti-PD-1 monotherapy, nor their combination showed any efficacy, as measured by tumor growth kinetics (Figure 6A), AUC (Figure 6B) or tumor weight at the end of the experiment (Figure 6C). Since we observed increased Tregs in TME of CD200R-deficient tumors (Figure 2D), we also examined if anti-CD200 could inhibit tumor growth in combination with anti-CTLA4 antibody 9H10. 9H10 has been shown to deplete Tregs and inhibits tumor growth (Simpson et al., 2013). We found that anti-CTLA4 antibody alone significantly reduced tumor growth as measured by tumor growth kinetics (Figure 6D), AUC (Figure 6E) or tumor weight at the end of the experiment (Figure 6F). Although OX-90 + αCTLA4 also showed significant efficacy, this combination did not prove to be more effective than αCTLA4 monotherapy. Thus, anti-CD200 therapy failed to show synergy with checkpoint inhibitors.