Efferocytosis is the process by which apoptotic cells are cleared from tissue by phagocytic cells. The removal of apoptotic cells prevents them from undergoing secondary necrosis and releasing their inflammation-inducing intracellular contents. Efferocytosis also limits tissue damage by increasing immunosuppressive cytokines and leukocytes and maintains tissue homeostasis by promoting tolerance to antigens derived from apoptotic cells. Thus, tumor cell efferocytosis following cytotoxic cancer treatment could impart tolerance to tumor cells evading treatment-induced apoptosis with deleterious consequences in tumor residual disease. We report here that efferocytosis cleared apoptotic tumor cells in residual disease of lapatinib-treated HER2+ mammary tumors in MMTV-Neu mice, increased immunosuppressive cytokines, myeloid-derived suppressor cells (MDSC), and regulatory T cells (Treg). Blockade of efferocytosis induced secondary necrosis of apoptotic cells, but failed to prevent increased tumor MDSCs, Treg, and immunosuppressive cytokines. We found that efferocytosis stimulated expression of IFN-γ, which stimulated the expression of indoleamine-2,3-dioxegenase (IDO) 1, an immune regulator known for driving maternal-fetal antigen tolerance. Combined inhibition of efferocytosis and IDO1 in tumor residual disease decreased apoptotic cell- and necrotic cell-induced immunosuppressive phenotypes, blocked tumor metastasis, and caused tumor regression in 60% of MMTV-Neu mice. This suggests that apoptotic and necrotic tumor cells, via efferocytosis and IDO1, respectively, promote tumor ‘homeostasis’ and progression.

Significance:

These findings show in a model of HER2+ breast cancer that necrosis secondary to impaired efferocytosis activates IDO1 to drive immunosuppression and tumor progression.

In untransformed tissues, the immune system responds to apoptotic cells (AC) in complex ways that support wound healing, suppress inflammation, limit immune-mediated tissue damage and prevent autoimmunity (1, 2). ACs are cleared from tissues through efferocytosis, the term used to describe phagocytic AC engulfment (3, 4). The receptor tyrosine kinase MerTK is required for efferocytosis by macrophages and other phagocytes (5–7). MerTK ligands, such as Gas6, bind to MerTK on phagocytes, while simultaneously binding externalized phosphatidyl serine (PtdSer) of the AC (8). Once engaged, MerTK signaling through Rac1 initiates cytoskeletal events driving phagocytosis (9), while concurrently suppressing proinflammatory transcription factors (e.g., STAT1) to block expression of proinflammatory cytokines (e.g., IFN-γ, IL12), and activating immunosuppressive and tolerogenic cytokines (e.g., IL4, IL10, IL13, TGFβ; ref. 10). Genetic or pharmacologic MerTK blockade impairs efferocytosis, enforcing two important consequences. First, MerTK inhibition decreases tolerogenic, immune-suppressive cytokine expression. Second, ACs that are not dispatched ultimately lose membrane integrity and undergo secondary necrosis, releasing intracellular contents that trigger inflammation (11–14), tissue damage and in some cases, autoimmunity (5, 11, 15, 16). This is particularly evident in apoptosis-enriched scenarios. For example, MerTK loss during postpartum involution, during which, milk-producing mammary epithelial cells (MEC) undergo widespread apoptosis, impairs MEC efferocytosis, causing widespread necrosis, inflammation, and mammary scarring (17–19).

Molecular signatures of tolerogenic wound healing often are observed in cancers, correlating with poor outcome and disease progression (20, 21), although their origins are unclear. Nonetheless, molecular markers of wound healing and immunosuppression become exaggerated following widespread apoptosis and efferocytosis in postpartum breast cancers (ppBC) within the involuting mammary gland (22, 23), supporting a prometastatic environment enhancing the exaggerated metastasis and lethality of ppBCs. However, genetic or pharmacological MerTK blockade in ppBC models decreases efferocytosis, tolerogenic cytokines, M2-like macrophage accumulation, and efferocytosis-enhanced metastasis (22).

Widespread tumor cell apoptosis and efferocytosis may have important implications in the setting of tumor treatment, given that cytotoxic regimens induce apoptosis in tumor cells, but often without pathologic complete response. Thus, treatment-induced apoptotic tumor cells (ATC) could trigger widespread efferocytosis, increasing local immunosuppression and imparting tolerance to any tumor cells evading apoptosis. Here, we report the impact of posttreatment efferocytosis on the tumor microenvironment (TME), including efferocytosis-induced accumulation of immunosuppressive leukocytes and cytokines. However, blockade of efferocytosis failed to prevent these immunosuppressive changes, despite induction of secondary necrosis. We propose that tumor cell necrosis potently induces inflammation, which engages the inflammation resolving factor indoleamine 2,3-dioxygenase (IDO1), restoring the immunosuppressive phenotype of the TME.

Mice

Animals were housed under pathogen-free conditions. Experiments were performed in accordance with AAALAC guidelines and approved by the Vanderbilt University Institutional Animal Care and Use Committee. MMTV-Neu primary mammary tumor cells (106) in growth factor–reduced Matrigel were injected into inguinal mammary fatpads of 6-week female FVB mice. Tumor volume was calculated using (length × width2 × 0.5). Mice were treated with 100 mg/kg lapatinib (SelleckChem), 20 mg/kg BMS-777607 (SelleckChem), 20 mg/kg epacadostat (Incyte Corporation), or equal volume of vehicle (0.1% Tween-80, 0.5% methyl cellulose) by oral gavage.

Histologic analyses

Tissue processing, hematoxylin–eosin staining, Trichrome staining, and IHC was performed by VICC Breast SPORE Pathology Shared Resource. IHC antibodies included CD45 (Ab10558, 1:5,000), arginase-1 (Santa Cruz Biotechnologies N-20, 1:400), CD8 (Invitrogen 5H10, 1:100), FoxP3 (eBioscience150D/E4, 1:50), and Target Retrieval buffer pH9 (Dako) or citrate buffer pH6 (Dako), then developed with Envision (Dako). TUNEL analysis was performed with the TUNEL Kit (EMD Millipore). Photomicrographs acquired on an Olympus CK40 inverted microscope through an Optronics DEI-750C camera using CellSens capture software. Metastases were enumerated by a pathologist in 5 mm paraffin sections of lungs from tumor bearing mice.

Reverse transcription and qRT-PCR

Whole-tumor or whole cell RNA was harvested with RNeasy (Qiagen), reverse transcribed (High Capacity; Applied Biosystems) and amplified using the following primers: MRC1 (forward: 5′CCCTCAGCAAGCGATGTGC; reverse: 5′GGATACTTGCCAGGTCCCCA3′); Nos2 (forward: 5′GGAGCATCCCAAGTACGAGTGG; reverse: 5′CGGCCCACTTCCTCCAG); Il10 (forward: 5′GGCGCTGTCATCGATTTCTCC; reverse: 5′GGCCTTGTAGACACCTTGGTC); Tgfb1 (forward: 5′CGCAACAACGCCATCTATGAG; reverse: 5′CGGGACAGCAATGGGGGTTC); Il4 (forward: 5′GGTCACAGGAGAAGGGACG; reverse: 5′GCGAAGCACCTTGGAAGCC); Il12b (forward: 5′GGAGTGGGATGTGTCCTCAG; reverse: 5′CGGGAGTCCAGTCCACCTCT); and Rplp0 (forward: 5′TCCTATAAAAGGCACACGCGGGC; reverse: 5′AGACGATGTCACTCCAACGAGGACG). Target gene Ct values were normalized to mRplp0, and calculated as (Cttarget gene − CtRplp0) Sample A − (Cttarget gene − CtRplp0)Sample B.

Flow cytometry

Tumors dissociated 1 hour with collagenase (0.5 mg/mL; Roche Life Sciences) and DNAse (0.19 mg/mL; Bio-Rad) were treated with ACK lysis buffer (Thermo Fisher Scientific), filtered through 70-μm strainers. Cells (1.5 × 106) were stained 30-minute with fluorescence-conjugated rat anti-mouse BD Pharmingen antibodies diluted 1:200: CD3 (53–2.1), CD8 (53–6.7), CD4 (GK1.5), CD45 (30F11), CD11b (M1/70), or with hamster anti-mouse CD11c (HL3), and mouse IgG2a anti-mouse NK1.1 (PK136). Cells were fixed, permeabilized, and stained 1 hour with anti-FoxP3 and anti-CD206 antibodies (1:100). Stained cells were analyzed on a 3-laser BD LSRII (BD Biosciences).

Human breast cancer dataset analysis

Overall patient survival correlating with tumor IDO1 levels (>2 SD from mean expression) was analyzed within The Cancer Genome Atlas Invasive Breast Cancer Dataset (24) and the METABRIC dataset (25) using cBioPortal software (26).

Cell line authentication

Raw264.7, THP-1, and MCF7 cells were purchased in 2012 from ATCC and cultured at low passage in DMEM with 10% FCS and 1% anti-antireagent (Gibco). Cell identity was verified by ATCC using genotyping with a Multiplex STR assay. All cell lines were screened monthly for Mycoplasma. All cell lines were used for experiments within 50 passages from thawing.

Western blotting and immunoprecipitation

Cells or tumors were homogenized in ice-cold lysis buffer, resolved by SDS-PAGE, and transferred to nitrocellulose membranes as described previously. Membranes were blocked and probed with the following primary antibodies: MerTK (1:1,000; Santa Cruz Biotechnologies); α-actin, (1:10,000; Sigma-Aldrich); and the following from Cell Signaling Technologies: STAT1 (1:2,000), P-STAT1 (1:500), IDO1 (1:1,000), and Rab11 (1:1,000).

Efferocytosis assays

To induce apoptosis, cells were treated 4 hours in suspension with 1 μmol/L lapatinib (for MMTV-Neu) or 1 μmol/L BKM120 (for MCF7) plus 2 μmol/L ABT-263 (Selleck Chemicals), washed five times, and used directly for efferocytosis assays. Necrosis was induced by freezing (−80°), followed by thawing. Phagocytes seeded at 104/well in 24-well dishes were cultured 16 hours in 2% FBS prior to adding ACs or necrotic cells. Nonadherent ACs and necrotic cells were removed from cultured media after 4 hours, cultured media was collected, filtered then returned to originating phagocytes. Where indicated, phagocytes were pretreated 2 hours with BMS-777607 (1 μmol/L). Cultured media collected after an additional 48 hours was passed through 0.2-μm filters, and used neat (250 μL) for ELISA (BioLegend) according to manufacturer's protocol.

Statistical analysis

Experimental groups were compared with a control group using Student unpaired, two-tailed t test. Multiple groups were compared across a single condition using one-way ANOVA. To compare the response of two agents combined to either single agent alone, two-way ANOVA was used. P < 0.05 defined significant differences from null hypothesis.

To test the hypothesis that therapeutically induced tumor cell apoptosis might trigger an immune-suppressive response within tumor residual disease, we used an immune-competent, genetically engineered mouse model of breast cancer, MMTV-Neu (27). This model uses overexpression of the rat HER2 homologue, Neu, to generate mammary tumors. Previous studies confirmed that MMTV-Neu tumor cells undergo apoptosis in response to lapatinib, a receptor tyrosine kinase inhibitor of HER2/Neu (28). Lapatinib was chosen over a monoclonal anti-Neu antibody, given that antibody-dependent cellular cytotoxicity might cloud analysis of apoptosis-induced immune responses. In addition, chemotherapies were not ideal, as many chemotherapies reportedly affect systemic immunity, indirectly affecting immune responses to tumor cell apoptosis/efferocytosis (29). Once tumors reached 50 to 100 mm3, mice were randomized into treatment groups (Fig. 1A), receiving a single treatment with vehicle or lapatinib (100 mg/kg). Tumors collected 2 hours after treatment confirmed decreased phospho-Neu in lapatinib-treated samples (Fig. 1B; Supplementary Fig. S1A), which resumed by day 7. On day 1, lapatinib-mediated induction of tumor cell apoptosis was seen (Fig. 1C and D). but ATCs were cleared by treatment day 7, supporting use of this model to test TME changes occurring in response to treatment-induced tumor cell death and efferocytosis.

Figure 1.

A novel model to assess the immune response to treatment-induced tumor cell apoptosis and efferocytosis. A, Schematic representation of treatment groups and experimental timeline. B, IHC for phospho-Neu Y1248 in tumors collected 1 hour after lapatinib treatment (day 1, d1) and in tumors collected on day 7 (i.e., 6 days after lapatinib treatment). Representative images are shown. C and D, TUNEL analysis of tumor sections collected from mice at treatment day 2. C, Quantitation of TUNEL+ cells per field. Each data point is the average of five random fields per tumor; midlines are the average of N = 3 samples, ±SD. P values, Student unpaired two-tailed t test. D, Representative images are shown. E and F, Histologic analysis of tumor sections. Representative images are shown. N = 5. E, Hematoxylin and eosin (H&E)-stained sections. Arrows, ECM accumulation. F, Masson's Trichrome-stained sections are shown. Arrows, alcian blue+ collagen. G and H, Tumors (treatment day 7) were assessed by IHC for CD3, FoxP3, and arginase-1 (Arg-1). G, Representative images are shown. H, Quantitation of CD3+, FoxP3+, and Arg-1+ cells per field, as described in B. I, RT-qPCR analysis of tumor RNA harvested at treatment day 7 measuring relative levels of indicated gene transcripts. Values were calculated using the ddCT method. Each data point represents the average value of five technical replicates, N = 5 tumors. For each transcript, values were corrected for the average value measured in vehicle-treated samples. Student t test. n.s., nonsignificant.

Figure 1.

A novel model to assess the immune response to treatment-induced tumor cell apoptosis and efferocytosis. A, Schematic representation of treatment groups and experimental timeline. B, IHC for phospho-Neu Y1248 in tumors collected 1 hour after lapatinib treatment (day 1, d1) and in tumors collected on day 7 (i.e., 6 days after lapatinib treatment). Representative images are shown. C and D, TUNEL analysis of tumor sections collected from mice at treatment day 2. C, Quantitation of TUNEL+ cells per field. Each data point is the average of five random fields per tumor; midlines are the average of N = 3 samples, ±SD. P values, Student unpaired two-tailed t test. D, Representative images are shown. E and F, Histologic analysis of tumor sections. Representative images are shown. N = 5. E, Hematoxylin and eosin (H&E)-stained sections. Arrows, ECM accumulation. F, Masson's Trichrome-stained sections are shown. Arrows, alcian blue+ collagen. G and H, Tumors (treatment day 7) were assessed by IHC for CD3, FoxP3, and arginase-1 (Arg-1). G, Representative images are shown. H, Quantitation of CD3+, FoxP3+, and Arg-1+ cells per field, as described in B. I, RT-qPCR analysis of tumor RNA harvested at treatment day 7 measuring relative levels of indicated gene transcripts. Values were calculated using the ddCT method. Each data point represents the average value of five technical replicates, N = 5 tumors. For each transcript, values were corrected for the average value measured in vehicle-treated samples. Student t test. n.s., nonsignificant.

Close modal

Despite no apparent histologic changes at treatment day 2 (Supplementary Fig. S1B), lapatinib-treated tumors at day 7 displayed increased extracellular matrix (ECM) deposition (Fig. 1E), containing increased collagen (Fig. 1F), a hallmark of immunosuppressive wound healing. CD3+ tumor-infiltrating T lymphocytes (TIL) were similarly present in vehicle-treated and lapatinib-treated samples at treatment day 2 (Supplementary Fig. S1C), but were increased in lapatinib-treated samples on day 7 (Fig. 1G and H; Supplementary Fig. S1D). Cells expressing FoxP3, a molecular marker of immunosuppressive regulatory T cells (Treg), were increased in lapatinib-treated tumors at day 7, as were those expressing arginase-1, demarcating immune-suppressive M2-like macrophages and myeloid-derived suppressor cells (MDSC). Genes encoding immune-suppressive cytokines Il10, Il13, and Tgfb1 were upregulated in lapatinib-treated tumors at day 7 (Fig. 1I, N = 3), although Il4 was not significantly altered. In contrast, genes encoding pro-inflammatory cytokines IL1β, IL12, and IFN-γ were similar or, in the case of IFN-γ, downregulated, in lapatinib-treated and vehicle-treated tumors (Supplementary Fig. S1E).

Biomimicry of ACs in tumors induces immunosuppressive changes

To confirm that these findings were not due to lapatinib treatment per se, but rather in response to tumor cell apoptosis, we used lapatinib to treat 4T1 mouse mammary tumors, which poorly express HER2/ErbB2/Neu, and do not undergo apoptosis upon lapatinib treatment (Fig. 2A). Arginase-1 and FoxP3 IHC revealed similar staining in tumors treated with vehicle and lapatinib at day 2 and day 7 (Fig. 2B). Il10, Il13, and Tgfb1 were not upregulated in lapatinib-treated 4T1 tumors at day 7 (Fig. 2C), suggesting that lapatinib does not induce immunosuppressive changes in the TME in the absence of apoptosis.

Figure 2.

Biomimicry of efferocytosis in tumors drives immunosuppressive changes. A, Schematic representation of treatment groups and experimental timeline. B, IHC to detect Arg-1 and FoxP3 in tumors collected on days 2 and 7. Representative images are shown. C, RT-qPCR analysis of tumor RNA (treatment day 7, d7) measuring relative levels of indicated gene transcripts, calculated using the ddCT method. Each data point represents the average value of five technical replicates, N = 3 tumors. For each transcript, values were corrected for the average value measured in vehicle-treated samples. Student t test. D, Schematic representation of treatment groups and experimental timeline to measure the impact of PtdSer liposomes on 4T1 tumors. E, IHC to detect Arg-1, CD3, and FoxP3 in tumors (days 2 and 7). Representative images are shown. Quantitation of the number of positive cells per field is shown in graphs to the right of each panel. Each data point is the average of five random fields per tumor; midlines are the average of N = 3 samples, ±SD. P values, Student unpaired two-tailed t test. F, RT-qPCR analysis of tumor RNA harvested at treatment day 7 measuring relative levels of indicated gene transcripts. Values were calculated as described in C. n.s., nonsignificant.

Figure 2.

Biomimicry of efferocytosis in tumors drives immunosuppressive changes. A, Schematic representation of treatment groups and experimental timeline. B, IHC to detect Arg-1 and FoxP3 in tumors collected on days 2 and 7. Representative images are shown. C, RT-qPCR analysis of tumor RNA (treatment day 7, d7) measuring relative levels of indicated gene transcripts, calculated using the ddCT method. Each data point represents the average value of five technical replicates, N = 3 tumors. For each transcript, values were corrected for the average value measured in vehicle-treated samples. Student t test. D, Schematic representation of treatment groups and experimental timeline to measure the impact of PtdSer liposomes on 4T1 tumors. E, IHC to detect Arg-1, CD3, and FoxP3 in tumors (days 2 and 7). Representative images are shown. Quantitation of the number of positive cells per field is shown in graphs to the right of each panel. Each data point is the average of five random fields per tumor; midlines are the average of N = 3 samples, ±SD. P values, Student unpaired two-tailed t test. F, RT-qPCR analysis of tumor RNA harvested at treatment day 7 measuring relative levels of indicated gene transcripts. Values were calculated as described in C. n.s., nonsignificant.

Close modal

Next, we used an approach aimed at biomimicry of cell death in tumors, using a phospholipid bilayer vesicle, or a liposome, with surface localization of PtdSer, a major signaling component of ACs. Previous studies show that AC mimicry using liposomal PtdSer induces liposome phagocytosis, promoting tolerogenic gene expression changes in phagocytes, and dampening experimental autoimmunity in models of diabetes and encephalomyelitis (30, 31). For these experiments, we injected a single dose of liposomal PtdSer into 4T1 tumors on day 1, following the impact of this single injection through the remainder of one week (Fig. 2D). Controls included intratumoral delivery of saline, and intratumoral delivery of liposomal phosphatidyl choline (PtdChol).

The number of arginase-1+ cells and CD3+ cells were similar in PtdSer, PtdChol, and saline treatment on day 2 (Fig. 2E). By day 7, PtdSer-treated tumors, but not those treated with PtdChol or saline, harbored increased Arg-1+, CD3+, and FoxP3+ cells, and upregulated Il10, Il13, and Tgfb1 (Fig. 2F), consistent with an immunosuppressive TME following mimicry of widespread apoptosis. Similar immunosuppressive responses were seen in MMTV-Neu tumors treated with liposomal PtdSer (Supplementary Fig. S2A–S2B).

MerTK inhibition blocks clearance of therapeutically induced ATCs

We examined MerTK-mediated efferocytosis in this setting by treating MMTV-Neu tumor-bearing mice with a single lapatinib dose as described above, followed by daily treatment with the MerTK inhibitor BMS-777607 (20 mg/kg; Fig. 3A; ref. 32), which potently inhibits MerTK-mediated efferocytosis (22). BMS-777607 affected neither P-Neu, lapatinib-mediated Neu inhibition (Fig. 3B), nor lapatinib-induced apoptosis (Fig. 3C). Lapatinib-induced ATCs were cleared from tumors by day 7, but remained evident in tumors treated with lapatinib + BMS-777607 (L/B; Fig. 3C). Hyper-condensed apoptotic debris, a histologic feature arising from impaired efferocytosis (18, 19), was abundant in L/B-treated tumors.

Figure 3.

MerTK inhibition blocks efferocytosis but not immunosuppressive changes in tumor RD. A, Schematic representation of treatment groups and experimental timeline. B, IHC for phospho-Neu Y1248 in tumors, 1 hour after lapatinib treatment (day 1, d1), and day 7 (d7), 1 hour after final treatment with BMS-777607. Representative images are shown. C, TUNEL analysis of tumors (days 2 and 7). Left, quantitation of the number of TUNEL+ cells per. Each data point is the average of five random fields per tumor; midlines are the average of N = 3 samples, ±SD. P values, Student unpaired two-tailed t test. Right, representative images are shown. Original magnification, ×400. D, Tumors harvested on day 7 were assessed by IHC for FoxP3 and Arg-1. Left, quantitation of FoxP3+ and Arg-1+ cells per field, as described in C. Right, representative images are shown. E, Dissociated tumors harvested on day 7 were assessed by flow cytometry for Tregs (CD45+CD3+CD4+FoxP3+) and exhausted/anergized CD8+ T cells (CD45+CD3+CD8+LAG3+). N = 3. F, RT-qPCR analysis of tumor RNA harvested at treatment day 7 measuring relative levels of indicated gene transcripts, calculated as described above (Fig. 1I), N = 5 tumors. n.s., nonsignificant.

Figure 3.

MerTK inhibition blocks efferocytosis but not immunosuppressive changes in tumor RD. A, Schematic representation of treatment groups and experimental timeline. B, IHC for phospho-Neu Y1248 in tumors, 1 hour after lapatinib treatment (day 1, d1), and day 7 (d7), 1 hour after final treatment with BMS-777607. Representative images are shown. C, TUNEL analysis of tumors (days 2 and 7). Left, quantitation of the number of TUNEL+ cells per. Each data point is the average of five random fields per tumor; midlines are the average of N = 3 samples, ±SD. P values, Student unpaired two-tailed t test. Right, representative images are shown. Original magnification, ×400. D, Tumors harvested on day 7 were assessed by IHC for FoxP3 and Arg-1. Left, quantitation of FoxP3+ and Arg-1+ cells per field, as described in C. Right, representative images are shown. E, Dissociated tumors harvested on day 7 were assessed by flow cytometry for Tregs (CD45+CD3+CD4+FoxP3+) and exhausted/anergized CD8+ T cells (CD45+CD3+CD8+LAG3+). N = 3. F, RT-qPCR analysis of tumor RNA harvested at treatment day 7 measuring relative levels of indicated gene transcripts, calculated as described above (Fig. 1I), N = 5 tumors. n.s., nonsignificant.

Close modal

MerTK inhibition does not block immune-suppressive changes in response to tumor cell death

Tumors pulsed with lapatinib alone, but not with BMS-777607 alone, harbored increased FoxP3+ TILs and arginase-1+ cells on day 7 (Fig. 3D; Supplementary Fig. S3A). Although we predicted L/B-treated tumors would harbor fewer FoxP3+ cells than lapatinib-treated tumors, we instead found increased FoxP3+ TILs and arg-1+ cells in L/B-treated tumors. Flow cytometry confirmed that Tregs (CD45+CD3+CD4+CD8negFoxP3+) were increased in L/B-treated tumors (Fig. 3E). In addition, the percentage of CD8+ T cells (CD45+CD3+CD4negCD8+) that were LAG3+ increased L/B-treated tumors, indicating increased anergy and/or exhaustion of CD8+ T cells in efferocytosis-impaired tumors following widespread apoptosis.

Because uncleared ACs ultimately lose membrane integrity and undergo secondary necrosis (7), we assessed the necrosis-inducible cytokine, IFN-γ, in efferocytosis-competent tumors (i.e., tumors treated with lapatinib alone) versus those that were efferocytosis-impaired (i.e., L/B-treated tumors). By day 7, efferocytosis-competent tumors downregulated IFN-γ (Fig. 3F), whereas efferocytosis-impaired tumors strongly expressed IFN-γ, consistent with secondary necrosis of ATCs due to blockade of efferocytosis. Necroptosis-encoding genes were not upregulated (Supplementary Fig. S3B–S3C). Despite high IFN-γ gene expression, immune-suppressive cytokines IL4 and IL10 were increased in L/B-treated tumors. These findings suggest that, although MerTK-mediated efferocytosis increases immunosuppressive changes in tumors, its blockade does not reverse immunosuppressive changes resulting from treatment-induced tumor cell death.

Secondary necrosis of tumor cells induces expression and activity of IDO1

We tested the hypothesis that, although secondary necrosis of ATCs induces proinflammatory IFN-γ, these may concomitantly induce expression and/or activity of factors aimed at resolving inflammation (Fig. 4A). It is known that IDO1, an intracellular enzyme that converts tryptophan (Trp) to kynurenine (Kyn; refs. 33, 34), is IFN-γ inducible, dampens inflammation, increases immune tolerance, and is a known suppressor of antitumor immunity (34–36).

Figure 4.

Secondary tumor cell necrosis induces IDO1 in the posttherapeutic setting. A, Schematic model of secondary necrosis in untransformed tissues contributing to inflammation and IFN-γ expression, followed by IDO1-mediated resolution of inflammation. B, Analysis of TCGA-curated invasive breast cancer dataset using cBio software. Left, IDO1 and IFNG expression levels were assessed for correlation. Right, Kaplan–Meier analysis, comparing overall survival in patients with high IDO1 expression (>2 SD over the mean value for the entire group) versus other patients. P value, log-rank test. C, RT-qPCR analysis of tumor RNA harvested at day 7 measuring relative Ido1. Values were calculated as described above (Fig. 1I), N = 5 tumors. Values were normalized to average value in vehicle-treated samples. Student t test. D, ELISA of plasma harvested from mice on day 7, measuring Kyn and Trp. Values are Kyn:Trp ratio. Each sample was assessed in technical duplicate and N = 3 biological replicates. E, RT-qPCR analysis measuring Ido1 in Raw264.7 cells 48 hours following a 4 hours pulse with apoptotic MMTV-Neu cells. Values calculated as in Fig. 1I. Values were corrected for average value in vehicle-treated samples. Student t test. N = 8 biological replicates, each assessed in five technical replicates. F, Western blot analysis of Raw264.7 cells treated with recombinant IFN-γ (2 pmol/L) for 18 or 48 hours, ±BMS-777607 (1 μmol/L). Antibodies used as shown at the left of each panel. G and H, Differentiated THP-1 (human monocyte) cells were cocultured with live, apoptotic, or necrotic MCF7 (human breast cancer) cells. Where indicated, THP-1 cells were treated with BMS-777607 (1 μmol/L) and epacadostat (500 nmol/L). Cell lysates were assessed by Western blot analysis (G). Cultured media was assessed by ELISA to measure Kyn. N = 3 biological replicates, each assessed in three technical replicates. Student t test. n.s., nonsignificant.

Figure 4.

Secondary tumor cell necrosis induces IDO1 in the posttherapeutic setting. A, Schematic model of secondary necrosis in untransformed tissues contributing to inflammation and IFN-γ expression, followed by IDO1-mediated resolution of inflammation. B, Analysis of TCGA-curated invasive breast cancer dataset using cBio software. Left, IDO1 and IFNG expression levels were assessed for correlation. Right, Kaplan–Meier analysis, comparing overall survival in patients with high IDO1 expression (>2 SD over the mean value for the entire group) versus other patients. P value, log-rank test. C, RT-qPCR analysis of tumor RNA harvested at day 7 measuring relative Ido1. Values were calculated as described above (Fig. 1I), N = 5 tumors. Values were normalized to average value in vehicle-treated samples. Student t test. D, ELISA of plasma harvested from mice on day 7, measuring Kyn and Trp. Values are Kyn:Trp ratio. Each sample was assessed in technical duplicate and N = 3 biological replicates. E, RT-qPCR analysis measuring Ido1 in Raw264.7 cells 48 hours following a 4 hours pulse with apoptotic MMTV-Neu cells. Values calculated as in Fig. 1I. Values were corrected for average value in vehicle-treated samples. Student t test. N = 8 biological replicates, each assessed in five technical replicates. F, Western blot analysis of Raw264.7 cells treated with recombinant IFN-γ (2 pmol/L) for 18 or 48 hours, ±BMS-777607 (1 μmol/L). Antibodies used as shown at the left of each panel. G and H, Differentiated THP-1 (human monocyte) cells were cocultured with live, apoptotic, or necrotic MCF7 (human breast cancer) cells. Where indicated, THP-1 cells were treated with BMS-777607 (1 μmol/L) and epacadostat (500 nmol/L). Cell lysates were assessed by Western blot analysis (G). Cultured media was assessed by ELISA to measure Kyn. N = 3 biological replicates, each assessed in three technical replicates. Student t test. n.s., nonsignificant.

Close modal

Analysis of a publicly available clinical dataset of 837 breast tumors curated by The Cancer Genome Atlas (TCGA; ref. 24) revealed a strong direct correlation between IFNG and IDO1 expression (Fig. 4B, left). Further analysis of this same dataset showed that patients with high tumor IDO1 expression (>2.0 S.D. over the median) experience substantially diminished survival versus remaining patients (Fig. 4B, right), underscoring the key role of IDO1 in tumor progression. It should be noted that these results represent tumor samples collected from patients prior to treatment, reflecting molecular relationships between IFNG and IDO1, and between IDO1 and patient survival, but not the relationship between tumor cell necrosis, IDO1, and the TME. Notably, in our studies, Ido1 was strongly increased in L/B-treated (i.e., efferocytosis-impaired) tumors at day 7 (Fig. 4C). Further, ELISA-based plasma Kyn and Trp measurements revealed an increased Kyn:Trp ratio in L/B-treated tumor-bearing mice over those treated with lapatinib alone or BMS-777607 alone (Fig. 4D), confirming increased IDO1 activity in the context of secondary tumor cell necrosis due to impaired efferocytosis.

Because numerous events within the TME might upregulate IDO1, we used an ex vivo model to understand how secondary necrosis affects Ido1. MMTV-Neu primary mouse mammary tumor cells were treated with lapatinib (1 μmol/L) and the Bcl-2/Bcl-xL inhibitor ABT-263 (1 μmol/L; ref. 37) to induce apoptosis, then plated 5:1 with macrophage-like Raw264.7 cells. Parallel monocultures of Raw264.7 were maintained for comparison. After 4 hours, nonadherent (apoptotic) cells were removed from cocultures, and media was filtered and replaced onto Raw264.7 cells for additional 48 hours culture. Although ATC coculture had no impact on Ido1 expression in Raw264.7 cells (Fig. 4E), ATC coculture with the MerTK inhibitor BMS-777607 increased Ido1 levels approximately five fold, confirming that inhibition of efferocytosis increases expression of Ido1. In Raw264.7 cells treated with recombinant IFN-γ (2 pmol/L), tyrosine phosphorylation of the proinflammatory transcription factor STAT-1 confirmed IFN-γ-induced proinflammatory signaling 18 hours after treatment, although IDO1 expression was not upregulated at this early time-point (Fig. 4F). By 48 hours, P-STAT-1 was diminished, and IDO1 was potently upregulated, consistent with the idea that IFN-γ-induced inflammatory responses are followed by responses aimed at dampening inflammation. Interestingly, MerTK inhibition modestly prolonged IFN-γ-induced P-STAT-1 but did not affect IDO1 upregulation.

NTCs induce IDO1, which is blocked by the IDO1 inhibitor, epacadostat

We differentiated THP-1 human monocyte-like cells into macrophages using M-CSF for 3 days (Supplementary Fig. S4A) then cocultured the THP-1 cells with live, apoptotic, or necrotic MCF7 human breast cancer cells for 4 hours, removing MCF7 cells and filtering media before adding media back to the originating THP-1 cells for an additional 48 hours culture. Western blot analysis showed that ATCs, and to a lesser extent, necrotic tumor cells (NTC) generated by freeze–thaw, induced P-MerTK in THP-1 cells (Fig. 4G). Although IDO1 activity, measured as Kyn in cultured media, was unaffected by THP-1 coculture with live tumor cells, and modestly increased following coculture with ATCs, MerTK inhibition substantially increased ATC-induced Kyn production (Fig. 4H), suggesting that secondary necrosis of ATCs induces IDO1 activity. Consistent with this notion, Kyn was strongly increased in THP-1 cells cocultured with freeze–thaw induced NTCs. Whether induced by primary or secondary necrosis, IDO1 activity was blocked using the IDO1 inhibitor epacadostat (38–40). Notably, IDO1 was not upregulated in differentiated THP-1 cells cocultured with HER2-amplified human BT474 breast cancer cells rendered apoptotic by treatment with lapatinib, but was induced upon blockade of efferocytosis with BMS-777607 (Supplementary Fig. S4B).

Decreased tumor growth, metastasis, and immune-suppressive leukocytes in lapatinib-treated tumors upon combined inhibition of apoptosis-induced MerTK and necrosis-induced IDO1

Next, we investigated the role of IDO1 in establishing an immune-suppressive phenotype in the context of secondary tumor cell necrosis. We again administered a single dose of vehicle or lapatinib to tumor-bearing mice when tumors reached 100 mm3, followed by daily treatment for 7 days ± BMS-777607 and ± epacadostat (20 mg/kg; Fig. 5A). The plasma Kyn:Trp ratio was used as a measure of systemic IDO1 activity at day 7, revealing increased circulating Kyn in L/B-treated mice over untreated or lapatinib-treated mice (Fig. 5B), which was reduced to baseline in mice treated with lapatinib/BMS-777607/epacadostat (L/B/E), confirming inhibition of IDO1-mediated Kyn production.

Figure 5.

Combined inhibition of MerTK and IDO1 decreases immunosuppressive tumor leukocytes, tumor growth, and metastasis in the posttherapeutic setting. A, Schematic representation of treatment groups and experimental timeline. B, ELISA of plasma harvested from mice on treatment day 7 to measure Kyn and Trp. Values are shown as the ratio of Kyn:Trp. Each sample was assessed in technical duplicate, and N = 3 biological replicates. C–E, Quantitation of TUNEL+ tumor cells (C) and FoxP3 and Arg-1 cells (D) in tumors collected on day 7. Asterisks, areas of acellular debris. Black arrows, TILs. Yellow arrows, hyper-condensed nuclei characteristic of apoptotic bodies/debris. Quantitation of TUNEL+ (C), FoxP3+, and Arg-1+ cells (E) per field. Data points are the average of five fields per tumor; midlines are average of N = 5 tumors, ± SD. F, Tumors harvested on day 7 were assessed by flow cytometry to measure tumor Tregs (CD45+CD3+CD4+FoxP3+) and exhausted/anergized CD8+ T cells (CD45+CD3+CD8+LAG3+). N = 3. G and I, Tumor-bearing mice were treated with four consecutive cycles of the regimen shown in Fig. 4B, through day 28 (N = 15). Tumor volume measured through day 29 (G). Each data point represents average tumor volume, ± SD. Metastatic lesions were enumerated (H) in lungs harvested on day 29 in histologic sections (I). Each data point shows the number of metastatic lesions per total lung. Midlines are the average number of metastases per group, ± SD. H&E, hematoxylin and eosin. n.s., nonsignificant.

Figure 5.

Combined inhibition of MerTK and IDO1 decreases immunosuppressive tumor leukocytes, tumor growth, and metastasis in the posttherapeutic setting. A, Schematic representation of treatment groups and experimental timeline. B, ELISA of plasma harvested from mice on treatment day 7 to measure Kyn and Trp. Values are shown as the ratio of Kyn:Trp. Each sample was assessed in technical duplicate, and N = 3 biological replicates. C–E, Quantitation of TUNEL+ tumor cells (C) and FoxP3 and Arg-1 cells (D) in tumors collected on day 7. Asterisks, areas of acellular debris. Black arrows, TILs. Yellow arrows, hyper-condensed nuclei characteristic of apoptotic bodies/debris. Quantitation of TUNEL+ (C), FoxP3+, and Arg-1+ cells (E) per field. Data points are the average of five fields per tumor; midlines are average of N = 5 tumors, ± SD. F, Tumors harvested on day 7 were assessed by flow cytometry to measure tumor Tregs (CD45+CD3+CD4+FoxP3+) and exhausted/anergized CD8+ T cells (CD45+CD3+CD8+LAG3+). N = 3. G and I, Tumor-bearing mice were treated with four consecutive cycles of the regimen shown in Fig. 4B, through day 28 (N = 15). Tumor volume measured through day 29 (G). Each data point represents average tumor volume, ± SD. Metastatic lesions were enumerated (H) in lungs harvested on day 29 in histologic sections (I). Each data point shows the number of metastatic lesions per total lung. Midlines are the average number of metastases per group, ± SD. H&E, hematoxylin and eosin. n.s., nonsignificant.

Close modal

A similar number of TUNEL+ cells was seen in vehicle-treated tumors, lapatinib-treated tumors, and BMS-777607/epacadostat (B/E)-treated tumors (Fig. 5C; Supplementary Fig. S4C). However, L/B/E-treated tumors harbored substantially higher TUNEL+ content. Histologic analysis revealed solid sheets of tumor cells in samples treated with vehicle, lapatinib, and B/E, whereas L/B/E-treated tumors harbored an abundance of apoptotic bodies, acellular debris, and lymphocytic infiltrate (Fig. 5D; low power images across a larger tumor area are shown in Supplementary Fig. S5). Importantly, lapatinib-induced tumor infiltration by immune-suppressive arginase-1+ cells was abrogated in L/B/E-treated tumors (Fig. 5D and E; Supplementary Fig. S6). TILs expressing CD3 were modestly increased in tumors treated with lapatinib or B/E but were abundant in L/B/E-treated tumors (Fig. 5D and E). Despite increased CD3+ TILs in L/B/E-treated tumors, FoxP3+ T-Regs were diminished (Fig. 5F). Flow cytometry also revealed that exhausted and/or anergized LAG3+ CD8+ T cells, which increased in response to lapatinib, and increased further in response to L/B, were reduced in L/B/E-treated tumors.

Next, we compared growth of MMTV-Neu tumors treated with two consecutive cycles of the week-long treatment regimen shown in Fig. 5A, for a total treatment period of 15 days, beginning when tumors were 50 to 100 mm3, comparing the impact of L/B/E treatment to what was seen with epacadostat (E), lapatinib + epacadostat (L/E), and L/B. As expected, lapatinib treatment decreased tumor growth as a single agent (Supplementary Fig. S7). Although epacadostat as a single agent did not affect tumor growth, L/E-treated tumors grew slower than lapatinib-treated tumors. L/B-treated tumors grew at a rate similar to what was seen with lapatinib treatment, while the L/B/E combination produced the greatest inhibition of tumor growth, resulting in partial tumor regression.

This was studied in more detail using MMTV-Neu tumors treated with four consecutive cycles of the week-long treatment regimen shown in Fig. 5A, for a total treatment period of 29 days, beginning when tumors were 50 to 100 mm3. As compared with growth of vehicle-treated tumors, lapatinib used as a single agent inhibited MMTV-Neu tumor growth by approximately 60% (Fig. 5G; Supplementary Fig. S8A). Although tumors treated with B/E exhibited decreased tumor growth as compared with vehicle-treated tumors (Fig. 5G), these findings were not significant due to variable responses in this treatment group, in which 40% of the tumors grew at a rate similar to vehicle-treated tumors, while the remaining 60% exhibited marked tumor growth inhibition (Supplementary Fig. S8B). In contrast, L/B/E-treated tumors were markedly growth inhibited (Fig. 5G), due to tumor regression or stasis in 80% of tumors, and only modest tumor growth in 20% of tumors (Supplementary Fig. S8C–S8E). Lungs of tumor-bearing mice harvested on day 29 revealed histological evidence of metastases in 73.3% of vehicle-treated mice, and 100% of lapatinib-treated mice (Fig. 5H and I). The average number of lung metastases per mouse did not differ between these two groups. Remarkably, metastases were diminished in lungs collected from L/B/E-treated mice, evident in only 6.7% of. these mice.

We present evidence here that two distinct forms of tumor cell death, apoptosis, and secondary necrosis, modulate the TME through distinct pathways. MerTK-mediated efferocytosis rapidly clears therapeutically induced ATCs from tumors, induces immune-suppressive cytokines, and recruits immune suppressive leukocytes to tumors. These findings are not unexpected, given that MerTK-mediated efferocytosis is required for maintaining, and often for re-establishing, homeostasis in untransformed tissues through AC clearance, expression of immune-suppressive cytokines, and tolerance to AC-derived antigens. However, MerTK-mediated efferocytosis is understudied in the context of the TME. A previous study of ppBCs found that widespread MerTK-mediated efferocytosis, induced in response to post-partum apoptosis of MECs, increased expression of immune-suppressive cytokines and leukocytes, resulting in increased tumor metastasis (22). However, the data presented here represent the first confirmation that therapeutically induced tumor cell apoptosis is followed rapidly by efferocytosis, inducing immune-suppressing leukocytes and cytokines (Fig. 1G–I), and ultimately promoting tumor progression (Fig. 5G).

Based on these results, we anticipated that blockade of MerTK-mediated efferocytosis following cytotoxic cancer treatment would block immunosuppressive changes, in part through immunogenic effects of necrosis. Despite histologic and gene expression-based evidence of secondary necrosis, we found increased immune-suppressive Tregs, arginase-1+ cells, and immunosuppressive cytokines. NTCs induced both IFN-γ and IDO1, a potently tolerogenic and immune-suppressive factor. Notably, a similar response to necrosis is reported for sterile wounds in untransformed tissues (e.g., atherosclerosis, ischemia), including induction of IDO1, which converts Trp to Kyn, thus decreasing clonal expansion and increasing anergy and/or elimination of CD8+ T cells, Treg expansion, and M2-macrophage accumulation (36, 41–43). We tested the hypothesis that necrosis-induced IFN-γ expression may induce inflammation-resolving factors, finding that NTCs induced IDO1. Interestingly, IDO1 activity was upregulated following therapeutically induced tumor cell killing in vivo, but only when efferocytosis was blocked, suggesting that secondary necrosis increased IDO1 in vivo. Together, blockade of efferocytosis and IDO1 decreased both tumor growth and metastasis (Fig. 5).

Necrosis is often considered “immunogenic,” triggering IFN-γ, CD8+ cytotoxic T-cell expansion, MDSC suppression, and decreasing Tregs (44). Despite this, it is known that IFN-γ induces multiple factors aimed at resolving inflammation, such as PD-L1, PD-1, suppressor of cytokine signaling (SOCS)-1/3, and as discussed here, IDO1 (45–47). IFN-γ-inducible IDO1 expression and activity are seen in mouse models of melanoma, lung cancer, and breast cancer, correlating with decreased CD8+ TILs and increased MDSCs (36, 48). Thus, despite the notion that necrosis induces a proinflammatory phenotype transiently, we propose that NTCs ultimately contribute to an immunosuppressive phenotype as well.

In summary, the immune system responds to ATCs and NTCs in distinct ways, but both responses ultimately attempt to resolve inflammation and re-establish homeostasis in tumors, a scenario with deleterious consequences in the posttherapeutic setting. Efferocytosis inhibition could be combined with IDO1 inhibitors to block the immune system's response to both ATCs and NTCs. Further investigation into how the immune system responds to tumor cell death, and potential adverse effects of targeting these immune responses, are warranted.

J.M. Balko reports receiving a commercial research grant from Incyte, Genentech, and BMS; has ownership interest (including stock, patents, etc.) in patent on MHC-II expression as a predictor of response to anti-PD-1 therapy; and is an consultant/advisory board member for Novartis. Peggy Scherle is a CSO at Prelude Therapeutics and has ownership interest (including stock, patents, etc.) in Incyte. H.K. Koblish is a Director, Pharmacology at Incyte Corporation and has ownership interest (including stock, patents, etc.) in Incyte Corporation. R.S. Cook reports receiving a commercial research grant from Incyte pharmaceuticals, Inc. No potential conflicts of interest were disclosed by the other authors.

Conception and design: T.A. Werfel, J.M. Balko, R.S. Cook

Development of methodology: T.A. Werfel, D.L. Elion, V. Sanchez, M.J. Nixon, J.M. Balko, R.S. Cook

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T.A. Werfel, D.L. Elion, B. Rahman, D.J. Hicks, M.J. Nixon, J.L. James, J.M. Balko

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T.A. Werfel, D.L. Elion, B. Rahman, P.I. Gonzales-Ericsson, M.J. Nixon, J.M. Balko, P.A. Scherle, R.S. Cook

Writing, review, and/or revision of the manuscript: T.A. Werfel, J.M. Balko, P.A. Scherle, H.K. Koblish, R.S. Cook

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D.J. Hicks, H.K. Koblish

Study supervision: T.A. Werfel, R.S. Cook

We acknowledge Vanderbilt Shared Resources who contributed to studies reported herein: the VICC Breast SPORE Pathology (Dr. Melinda Sanders), Translational Pathology, Digital Histology, and VANTAGE shared resources. This work was supported by Specialized Program of Research Excellence (SPORE) grant NIH P50 CA098131 (to R.S. Cook), Cancer Center Support grant NIH P30 CA68485 (to R.S. Cook), CTSA UL1TR000445 (to R.S. Cook) from National Center for Advancing Translational Sciences, W81XWH-161-0063 (to R.S. Cook) from the Congressionally Directed Medical Research Program.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Boniakowski
AE
,
Kimball
AS
,
Jacobs
BN
,
Kunkel
SL
,
Gallagher
KA
. 
Macrophage-mediated inflammation in normal and diabetic wound healing
.
J Immunol
2017
;
199
:
17
24
.
2.
Eming
SA
,
Wynn
TA
,
Martin
P
. 
Inflammation and metabolism in tissue repair and regeneration
.
Science
2017
;
356
:
1026
30
.
3.
Gordon
S
,
Pluddemann
A
. 
Macrophage clearance of apoptotic cells: a critical assessment
.
Front Immunol
2018
;
9
:
127
.
4.
Trahtemberg
U
,
Mevorach
D
. 
Apoptotic cells induced signaling for immune homeostasis in macrophages and dendritic cells
.
Front Immunol
2017
;
8
:
1356
.
5.
Lemke
G
. 
Biology of the TAM receptors
.
Cold Spring Harb Perspect Biol
2013
;
5
:
a009076
.
6.
Elliott
MR
,
Koster
KM
,
Murphy
PS
. 
Efferocytosis signaling in the regulation of macrophage inflammatory responses
.
J Immunol
2017
;
198
:
1387
94
.
7.
Elliott
MR
,
Ravichandran
KS
. 
The dynamics of apoptotic cell clearance
.
Dev Cell
2016
;
38
:
147
60
.
8.
Nakano
T
,
Ishimoto
Y
,
Kishino
J
,
Umeda
M
,
Inoue
K
,
Nagata
K
, et al
Cell adhesion to phosphatidylserine mediated by a product of growth arrest-specific gene 6
.
J Biol Chem
1997
;
272
:
29411
4
.
9.
Mahajan
NP
,
Earp
HS
. 
An SH2 Domain-dependent, Phosphotyrosine-independent Interaction between Vav1 and the mer receptor tyrosine kinase: a mechanism for localizing guanine nucleotide-exchange factor action
.
J Biol Chem
2003
;
278
:
42596
603
.
10.
Lemke
G
,
Rothlin
CV
. 
Immunobiology of the TAM receptors
.
Nat Rev Immunol
2008
;
8
:
327
36
.
11.
Cohen
PL
,
Caricchio
R
,
Abraham
V
,
Camenisch
TD
,
Jennette
JC
,
Roubey
RA
, et al
Delayed apoptotic cell clearance and lupus-like autoimmunity in mice lacking the c-mer membrane tyrosine kinase
.
J Exp Med
2002
;
196
:
135
40
.
12.
Scott
RS
,
McMahon
EJ
,
Pop
SM
,
Reap
EA
,
Caricchio
R
,
Cohen
PL
, et al
Phagocytosis and clearance of apoptotic cells is mediated by MER
.
Nature
2001
;
411
:
207
11
.
13.
Seitz
HM
,
Camenisch
TD
,
Lemke
G
,
Earp
HS
,
Matsushima
GK
. 
Macrophages and dendritic cells use different Axl/Mertk/Tyro3 receptors in clearance of apoptotic cells
.
J Immunol
2007
;
178
:
5635
42
.
14.
Thorp
E
,
Cui
D
,
Schrijvers
DM
,
Kuriakose
G
,
Tabas
I
. 
Mertk receptor mutation reduces efferocytosis efficiency and promotes apoptotic cell accumulation and plaque necrosis in atherosclerotic lesions of apoe-/- mice
.
Arterioscler Thromb Vasc Biol
2008
;
28
:
1421
8
.
15.
Rothlin
CV
,
Ghosh
S
,
Zuniga
EI
,
Oldstone
MB
,
Lemke
G
. 
TAM receptors are pleiotropic inhibitors of the innate immune response
.
Cell
2007
;
131
:
1124
36
.
16.
Wallet
MA
,
Sen
P
,
Flores
RR
,
Wang
Y
,
Yi
Z
,
Huang
Y
, et al
MerTK is required for apoptotic cell-induced T cell tolerance
.
J Exp Med
2008
;
205
:
219
32
.
17.
Atabai
K
,
Sheppard
D
,
Werb
Z
. 
Roles of the innate immune system in mammary gland remodeling during involution
.
J Mammary Gland Biol Neoplasia
2007
;
12
:
37
45
.
18.
Monks
J
,
Henson
PM
. 
Differentiation of the mammary epithelial cell during involution: implications for breast cancer
.
J Mammary Gland Biol Neoplasia
2009
;
14
:
159
70
.
19.
Sandahl
M
,
Hunter
DM
,
Strunk
KE
,
Earp
HS
,
Cook
RS
. 
Epithelial cell-directed efferocytosis in the post-partum mammary gland is necessary for tissue homeostasis and future lactation
.
BMC Dev Biol
2010
;
10
:
122
.
doi: 10.1186/1471-213X-10-122
.
20.
Dvorak
HF
. 
Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing
.
N Engl J Med
1986
;
315
:
1650
9
.
21.
Hanahan
D
,
Coussens
LM
. 
Accessories to the crime: functions of cells recruited to the tumor microenvironment
.
Cancer Cell
2012
;
21
:
309
22
.
22.
Stanford
JC
,
Young
C
,
Hicks
D
,
Owens
P
,
Williams
A
,
Vaught
DB
, et al
Efferocytosis produces a prometastatic landscape during postpartum mammary gland involution
.
J Clin Invest
2014
;
124
:
4737
52
.
23.
Martinson
HA
,
Jindal
S
,
Durand-Rougely
C
,
Borges
VF
,
Schedin
P
. 
Wound healing-like immune program facilitates postpartum mammary gland involution and tumor progression
.
Int J Cancer
2015
;
136
:
1803
13
.
24.
Ciriello
G
,
Gatza
ML
,
Beck
AH
,
Wilkerson
MD
,
Rhie
SK
,
Pastore
A
, et al
Comprehensive molecular portraits of invasive lobular breast cancer
.
Cell
2015
;
163
:
506
19
.
25.
Curtis
C
,
Shah
SP
,
Chin
SF
,
Turashvili
G
,
Rueda
OM
,
Dunning
MJ
, et al
The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups
.
Nature
2012
;
486
:
346
52
.
26.
Cerami
E
,
Gao
J
,
Dogrusoz
U
,
Gross
BE
,
Sumer
SO
,
Aksoy
BA
, et al
The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data
.
Cancer Discov
2012
;
2
:
401
4
.
27.
Guy
CT
,
Webster
MA
,
Schaller
M
,
Parsons
TJ
,
Cardiff
RD
,
Muller
WJ
. 
Expression of the neu protooncogene in the mammary epithelium of transgenic mice induces metastatic disease
.
Proc Natl Acad Sci USA
1992
;
89
:
10578
82
.
28.
Cook
RS
,
Garrett
JT
,
Sanchez
V
,
Stanford
JC
,
Young
C
,
Chakrabarty
A
, et al
ErbB3 ablation impairs PI3K/Akt-dependent mammary tumorigenesis
.
Cancer Res
2011
;
71
:
3941
51
.
29.
Homet Moreno
B
,
Ribas
A
. 
Anti-programmed cell death protein-1/ligand-1 therapy in different cancers
.
Br J Cancer
2015
;
112
:
1421
7
.
30.
Pujol-Autonell
I
,
Serracant-Prat
A
,
Cano-Sarabia
M
,
Ampudia
RM
,
Rodriguez-Fernandez
S
,
Sanchez
A
, et al
Use of autoantigen-loaded phosphatidylserine-liposomes to arrest autoimmunity in type 1 diabetes
.
PLoS One
2015
;
10
:
e0127057
.
31.
Pujol-Autonell
I
,
Mansilla
MJ
,
Rodriguez-Fernandez
S
,
Cano-Sarabia
M
,
Navarro-Barriuso
J
,
Ampudia
RM
, et al
Liposome-based immunotherapy against autoimmune diseases: therapeutic effect on multiple sclerosis
.
Nanomedicine (Lond)
2017
;
12
:
1231
42
.
32.
Schroeder
GM
,
An
Y
,
Cai
ZW
,
Chen
XT
,
Clark
C
,
Cornelius
LA
, et al
Discovery of N-(4-(2-amino-3-chloropyridin-4-yloxy)-3-fluorophenyl)-4-ethoxy-1-(4-fluorophenyl)-2-oxo-1,2-dihydropyridine-3-carboxamide (BMS-777607), a selective and orally efficacious inhibitor of the Met kinase superfamily
.
J Med Chem
2009
;
52
:
1251
4
.
33.
Lovelace
MD
,
Varney
B
,
Sundaram
G
,
Franco
NF
,
Ng
ML
,
Pai
S
, et al
Current evidence for a role of the kynurenine pathway of tryptophan metabolism in multiple sclerosis
.
Front Immunol
2016
;
7
:
246
.
doi: 10.3389/fimmu.2016.00246
.
34.
Prendergast
GC
. 
Immune escape as a fundamental trait of cancer: focus on IDO
.
Oncogene
2008
;
27
:
3889
900
.
35.
Lob
S
,
Konigsrainer
A
,
Rammensee
HG
,
Opelz
G
,
Terness
P
. 
Inhibitors of indoleamine-2,3-dioxygenase for cancer therapy: can we see the wood for the trees?
Nat Rev Cancer
2009
;
9
:
445
52
.
36.
Smith
C
,
Chang
MY
,
Parker
KH
,
Beury
DW
,
DuHadaway
JB
,
Flick
HE
, et al
IDO is a nodal pathogenic driver of lung cancer and metastasis development
.
Cancer Discov
2012
;
2
:
722
35
.
37.
Tse
C
,
Shoemaker
AR
,
Adickes
J
,
Anderson
MG
,
Chen
J
,
Jin
S
, et al
ABT-263: a potent and orally bioavailable Bcl-2 family inhibitor
.
Cancer Res
2008
;
68
:
3421
8
.
38.
Koblish
HK
,
Hansbury
MJ
,
Bowman
KJ
,
Yang
G
,
Neilan
CL
,
Haley
PJ
, et al
Hydroxyamidine inhibitors of indoleamine-2,3-dioxygenase potently suppress systemic tryptophan catabolism and the growth of IDO-expressing tumors
.
Mol Cancer Ther
2010
;
9
:
489
98
.
39.
Liu
X
,
Shin
N
,
Koblish
HK
,
Yang
G
,
Wang
Q
,
Wang
K
, et al
Selective inhibition of IDO1 effectively regulates mediators of antitumor immunity
.
Blood
2010
;
115
:
3520
30
.
40.
Yue
EW
,
Sparks
R
,
Polam
P
,
Modi
D
,
Douty
B
,
Wayland
B
, et al
INCB24360 (Epacadostat), a highly potent and selective indoleamine-2,3-dioxygenase 1 (IDO1) inhibitor for immuno-oncology
.
ACS Med Chem Lett
2017
;
8
:
486
91
.
41.
Andersen
MH
. 
Immune regulation by self-recognition: novel possibilities for anticancer immunotherapy
.
J Natl Cancer Inst
2015
;
107
.
doi: 10.1093/jnci/djv154
.
42.
Epelman
S
,
Liu
PP
,
Mann
DL
. 
Role of innate and adaptive immune mechanisms in cardiac injury and repair
.
Nat Rev Immunol
2015
;
15
:
117
29
.
43.
Cole
JE
,
Astola
N
,
Cribbs
AP
,
Goddard
ME
,
Park
I
,
Green
P
, et al
Indoleamine 2,3-dioxygenase-1 is protective in atherosclerosis and its metabolites provide new opportunities for drug development
.
Proc Natl Acad Sci USA
2015
;
112
:
13033
8
.
44.
Kroemer
G
,
Galluzzi
L
,
Kepp
O
,
Zitvogel
L
. 
Immunogenic cell death in cancer therapy
.
Annu Rev Immunol
2013
;
31
:
51
72
.
45.
Mühlbauer
M
,
Fleck
M
,
Schütz
C
,
Weiss
T
,
Froh
M
,
Blank
C
, et al
PD-L1 is induced in hepatocytes by viral infection and by interferon-α and -γ and mediates T cell apoptosis
.
J Hepatol
;
45
:
520
8
.
46.
Liu
J
,
Hamrouni
A
,
Wolowiec
D
,
Coiteux
V
,
Kuliczkowski
K
,
Hetuin
D
, et al
Plasma cells from multiple myeloma patients express B7-H1 (PD-L1) and increase expression after stimulation with IFN-{gamma} and TLR ligands via a MyD88-, TRAF6-, and MEK-dependent pathway
.
Blood
2007
;
110
:
296
304
.
47.
Krebs Danielle
L
,
Hilton Douglas
J
. 
SOCS proteins: negative regulators of cytokine signaling
.
Stem Cells
2001
;
19
:
378
87
.
48.
Pinton
L
,
Solito
S
,
Damuzzo
V
,
Francescato
S
,
Pozzuoli
A
,
Berizzi
A
, et al
Activated T cells sustain myeloid-derived suppressor cell-mediated immune suppression
.
Oncotarget
2016
;
7
:
1168
84
.

Supplementary data