Enhancer of zeste homolog (EZH2) is a key epigenetic regulator of gene expression and is frequently overexpressed in various cancer types, suggesting a role in oncogenesis. The therapeutic potential of EZH2 inhibitors is currently being explored, but their effect on antitumor immunity is largely unknown. Here we report that suppressing EZH2 activity using EZH2 inhibitor GSK126 resulted in increased numbers of myeloid-derived suppressor cells (MDSC) and fewer CD4+ and IFNγ+CD8+ T cells, which are involved in antitumor immunity. Addition of a neutralizing antibody against the myeloid differentiation antigen GR-1 or gemcitabine/5-fluorouracil–depleted MDSCs alleviated MDSC-mediated immunosuppression and increased CD4+ and CD8+ T-cell tumor infiltration and GSK126 therapeutic efficacy. Mechanistically, we identified a novel pathway of MDSC production in cancer in which EZH2 inhibition directs myeloid differentiation from primitive hematopoietic progenitor cells. These findings suggest that modulating the tumor immune microenvironment may improve the efficacy of EZH2 inhibitors.

Significance:

This study uncovers a potential mechanism behind disappointing results of a phase I clinical trial of EZH2 inhibitor GSK126 and identifies a translatable combinational strategy to overcome it.

Epigenetic histone modifications comprise an important mechanism that controls cellular processes including tumorigenesis and immunity. Epigenetic aberrations are often associated with tumor progression and cancer development (1). Of particular interest, the methyltransferase enhancer of zeste homolog 2 (EZH2), the catalytic component of the polycomb repressive complex 2 that trimethylates lysine27 of histone H3 to promote transcriptional repression (2), is frequently overexpressed in various cancers including lung (3), colorectal (4), breast (5), pancreatic (6), and prostate (7) cancers. Aberrant EZH2 overexpression is also associated with poor prognosis (8). Importantly, EZH2 is functionally critical for the growth and metastasis of cancer cells and thus pharmacologic inhibition of EZH2 has been proposed as a targeted therapy for various cancers (9, 10). Currently, several EZH2 inhibitors including GSK126 and EPZ-6438 are in clinical trials to treat advanced solid tumors or B-cell lymphomas (11).

The tumor microenvironment (TME) plays a critical role in directing the outcome of tumor rejection versus progression. Increasing evidence suggests that to fully assess the effect of anticancer drugs and devise more effective therapies, it is imperative to understand the impact of the drug on the TME. Whereas GSK126 is a promising anticancer drug currently undergoing multiple clinical trials, the premise has been largely based on studies using immunodeficient hosts (12–16), where it is impossible to assess the effect of the drug on the TME. Because EZH2 functions in many immune cell types that might contribute to tumor immunity (17–22), it is important to address how its inhibition might affect immune cell function during tumor development, a currently unresolved question.

In this study, we administered a clinically equivalent dose of GSK126 to immunocompetent and immunodeficient hosts and examined the effect on tumor growth. Surprisingly, we found that GSK126 had no effect on tumors in immunocompetent hosts, unlike that observed in immunodeficient hosts, suggesting that GSK126 promotes immune suppression, which neutralizes its antitumor effect. Indeed, GSK126 treatment led to a dampened CD8+ T-cell response in the tumor. Interestingly, GSK126 potently promoted myeloid-derived suppressor cells (MDSC) formation during tumor growth and these MDSCs suppressed CD8+ T cell function. Importantly, MDSC depletion restored the antitumor effect of GSK126 in immunocompetent hosts, which was associated with normalization of CD8+ T cell functions. Inspired by these findings, we investigated whether clinically used chemotherapy drugs that deplete MDSCs, including 5-fluorouracil (5-FU) and gemcitabine, might have a combined effect with GSK126 on cancer. Indeed, GSK126+5-FU and GSK126+gemicitabine achieved better antitumor efficacy than any monotherapy. Therefore, this study demonstrated an unintended effect of GSK126: promotion of MDSC generation, which masked the antitumor effect of this drug. Furthermore, MDSC depletion was able to unleash the antitumor effect of GSK126 in immunocompetent hosts.

Mice and tumor models

Six to 8-week-old female C57BL/6 mice and nude mice were purchased from the Chinese Academy of Medical Sciences (Beijing, China). OT-1 transgenic mice were purchased from the Jackson Laboratory. The mice were housed and maintained in laminar flow cabinets under specific pathogen-free conditions. For all in vivo studies, GSK126 (obtained from Selleck) or vehicle was administered intravenously at a dose of 50, 150, or 300 mg/kg. Lewis lung cancer (LLC) cells or MC38 cells (1 × 106) were implanted subcutaneously in female C57BL/mice or nude mice (n = 4–6 mice/group). Tumor growth was measured using calipers every 2 days. Tumor volume was calculated as follows: V = (length × width2) × 0.5. For all drug efficacy studies, GSK126 treatment (five times/week) was initiated once tumor volume was approximately 150 mm3. For Gr-1 depletion studies, 200 μg anti-Gr-1 neutralizing antibody (obtained from Bio X Cell) was administered intraperitoneally twice weekly, beginning on the indicated days. For GSK126+gemcitabine/5-FU combined treatment studies, gemcitabine (obtained from Selleck) was administered intraperitoneally at 60 mg/kg once weekly; 5-FU (obtained from Selleck) was administered intraperitoneally at 50 mg/kg once weekly. Tumor, spleen, blood, and bone marrow samples were harvested from euthanized mice at the indicated timepoints and immune cells and immune-related molecules analyzed by flow cytometry. In addition, tumor samples were analyzed by terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) immunofluorescence (IF). The Institutional Animal Care and Use Committee of Third Military Medical University (Chongqing, China) has approved our animal studies. All animal experiments were carried out in accordance with the Animal Study Guidelines of Third Military Medical University (Chongqing, China).

Cell lines and cell culture

LLC and MC38 cells were obtained from the ATCC in 2012. MC38-OVA cells were produced in our laboratory in 2016. Cancer cells were cultured in high-glucose DMEM supplemented with 10% FBS and 100 U/mL penicillin/streptomycin. All cell lines were examined and authenticated by short tandem repeat profiling in September 2016. All cell lines were Mycoplasma negative and used within 10 passages. CD8+ T cells were cultured in RPMI1640 supplemented with 10% FBS and 100 U/mL penicillin/streptomycin. To activate CD8+ T cells, the cells were seeded in tissue culture plates with plate-bound anti-CD3 (1 μg/mL, BioLegend) and anti-CD28 (1 μg/mL, BioLegend). For T-cell suppression assays, CD8+ T cells were labeled with CFSE, stimulated with anti-CD3 and anti-CD28 as above, and cocultured at 2:1 or 4:1 ratios with spleen-derived MDSCs in 96-well flat-bottom plates. Seventy-two hours later, cells were analyzed by flow cytometry. For differentiation assays, hematopoietic progenitor cells (HPC) were cultured in StemSpan SFEM medium containing GM-CSF/IL6 (10 ng/mL, Sigma-Aldrich) and treated with 5 μmol/L GSK126 for the indicated times.

Adoptive transfer tumor experiment

MC38 cells expressing ovalbumin (OVA; 1 × 106) were implanted subcutaneously into female C57BL/6 mice. GSK126 was administered for the first 4 days. Two days later, naive CD45.1+CD8+ T cells were purified from OT-1 transgenic mice by magnetic bead sorting (Miltenyi Biotec) and 2 × 106 activated or naive CD45.1+CD8+ T cells were injected into CD45.2+C57BL/6 mice.

Cell isolation

Bone marrow cells were harvested by flushing the femurs and tibias with PBS supplemented with 2% FBS. Splenocytes were collected by mechanical disruption. Tumors were weighed and 200 mg sample was cut into small fragments, and digested in 6 mL of dissociation solution [RPMI1640 medium supplemented with 10% FBS, collagenase type I (200 U/mL) and DNase I (100 μg/mL)] for 60 minutes at 37°C. Blood was immediately placed into heparin-coated EP tubes. Erythrocytes were lysed using red blood cell lysing buffer (Bio-Rad). Cell suspensions were passed through 70-μm cell strainers and then washed and resuspended in staining buffer.

For the isolation of spleen MDSCs, splenocytes were stained with anti-CD11b mAb and anti-Gr-1 (mAb; both from BioLegend) and CD11b+Gr-1+ cells were sorted using the BD FACSAria II Special Order System. For macrophage isolation from spleens, cells were stained with anti-CD11b mAb (BioLegend) and anti-F4/80 mAb (BioLegend) and CD11b+F4/80+ cells were sorted using the BD FACSAria II Special Order System. For bone marrow HPC isolation, cells were stained with anti-Lineage mAb cocktail (BD Pharmingen), anti-Sca-1 mAb (BioLegend), and anti-c-kit mAb (BioLegend); lineageSca-1c-kit+ cells were sorted using the BD FACSAria II Special Order System. For spleen CD8+ T-cell isolation, CD8+ T cells were isolated by positive selection using the CD8+ T cell Isolation Kit (Miltenyi Biotec). For all sorted samples, a purity of greater than 95% was achieved.

Flow cytometry

All antibodies are listed in Supplementary Table S1. Cells were blocked with rat IgG (10 μg/mL; Sigma) for at least 20 minutes on ice, washed with staining media [2%(vol/vol) HI, FBS in Hanks' Balanced Salt Solution (BSS) without Ca2+ or Mg2+, denoted SM], and then stained with fluorescently conjugated antibodies in SM for 30 minutes on ice. For the analysis of intracellular cytokines, CD8+ T cells were restimulated with PMA (0.5 μg/mL, Sigma-Aldrich) and ionomycin (0.5 μg/mL, Sigma-Aldrich), or ovalbumin (1 μg/mL, Sigma-Aldrich). After 4 hours, cell suspensions were stained using the Fixable Viability Dye (eBioscience), to enable exclusion of dead cells, and were surface stained with anti-mouse CD8 antibody. For intracellular staining, cells were fixed and permeabilized in fixation/permeabilization solution (BD Pharmingen) for 30 minutes and then stained with anti-IFNγ, anti-granzyme, and anti-Ki67. For cell-cycle analysis, BrdU incorporation was analyzed using the APC-BrdU Flow Kit (BD Pharmingen). Bone marrow and tumor tissues were assayed 24 hours after a single-intraperitoneal injection of BrdU (100 μL of a 10 mg/mL solution). All labeled cells were analyzed using a Beckman flow cytometry system.

qRT-PCR

Total RNA was extracted using TRizol Reagent (Invitrogen) or RNAqueous MicroKit (Invitrogen), and cDNA was synthesized using the cDNA Reverse Transcription Kit (Takara). qPCR was performed using SYBR Green PCR Mister Mix (Takara) to quantify the relative expression of mRNA. The sequences of primers are listed in Supplementary Table S1.

Western blot analysis

LLC cells, MC38 cells, LLC tumor tissues, MC38 tumor tissues, HPCs, MDSCs, and macrophages were lysed in RIPA Lysis buffer with 1% PMSF (Beyotime Biotechnology). Then, a BCA Kit (Beyotime Biotechnology) was used to determine protein concentrations and equal amounts analyzed by Western blotting. The blots were blocked with 4% BSA for 1 hour at room temperature and then probed with anti-mouse antibodies against H3K27me3 (Cell Signaling Technology, 1:500), EZH2 (Cell Signaling Technology, 1:500), GAPDH (Beyotime Biotechnology, 1:1,000), and Histone H3 (Beyotime Biotechnology, 1:500) overnight at 4°C. After washing with 0.1% TBST, a goat anti-rabbit secondary antibody (Beyotime Biotechnology, 1:5,000) was added. Then, a Chemiluminescence Detection System (ProteinSimple) was used.

Immunofluorescence

IF analysis was performed on 6-μm frozen tumor tissues. Tumor tissues were fixed with 4% paraformaldehyde for 20 minutes, blocked with 5% BSA for 20 minutes at room temperature. Then samples were incubated with specific antibodies against mouse Gr-1 (1:200, monoclonal R&D) or mouse CD3 (1:200, polyclonal Abcam) overnight in dark at 4°C. After washing with PBS, the samples were incubated with Cy3-conjugated or 647-conjugated anti-IgG (Beyotime Biotechnology, 1:200) for 30 minutes at 37°C. After washing, nuclei were stained with DAPI. Finally, samples were mounted and imaged using a confocal laser scanning microscope.

TUNEL assay

Tumor cell apoptosis was determined by TUNEL assay, which was carried out using a DeadEnd Florometric TUNEL System (Beyotime Biotechnology) according to the manufacturer's instructions. Cell nuclei that fluorescently stained with cyanine 3 (Cy3) (red) were defined as TUNEL-positive nuclei. Slides were coverslipped with DAPI antifade-mounting medium (Beyotime Biotechnology). TUNEL-positive nuclei were monitored by laser scanning confocal microscopy. The percentage of apoptotic cells was obtained by dividing the number of apoptotic cells by the total number of cells.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 7.0 software. Two-tailed unpaired Student t tests and one-way or two-way ANOVA with Tukey multiple comparison posttest were used to compare two or more groups. Statistical significance was indicated as *, P < 0.05; **, P < 0.01; ***, P < 0.001; and ****, P < 0.0001. All experiments were independently repeated at least three times.

Inhibition of EZH2 activity with GSK126 has no effect on tumor growth in immunocompetent mice

To investigate whether the application of a clinically equivalent dose of GSK126 achieves similar or better efficacy in immunocompetent hosts, compared with that observed in immunodeficient hosts, we used immunodeficient (nude mice) and immunocompetent (C57BL/6) tumor models. First, we subcutaneously injected LLC cells into nude mice, which was followed by the injection of different doses of GSK126 (Fig. 1A). We found that whereas 50 mg/kg GSK126 did not restrain tumor growth, 150 or 300 mg/kg GSK126 limited tumor growth in nude mice, as measured by both tumor volume and weight (Fig. 1B). Surprisingly, unlike nude mice, 150 mg/kg GSK126 had no obvious effect on tumor growth in immunocompetent hosts (Fig. 1C). Similarly, GSK126 was effective in retarding tumor growth in immunodeficient but not immunocompetent hosts implanted with MC38 cells, a colon carcinoma line (Fig. 1D and E). To investigate whether GSK126 can downregulate the methyltransferase activity of EZH2 in cancer cells, we analyzed EZH2 expression and H3K27 trimethylation. We found that GSK126 treatment did not affect EZH2 protein levels but decreased H3K27 trimethylation in LLC and MC38 cells both in vitro and in vivo (Supplementary Fig. S1). These results indicate that GSK126 can inhibit the EZH2 methyltransferase activity in LLC and MC38 cells.

Figure 1.

Inhibition of EZH2 activity by GSK126 has no effect on tumor growth in immunocompetent mice. A, Schematic illustration of treatment schedule. B and C, Effect of GSK126 on the growth of LLC cells in nude mice and C57BL/6 mice. GSK126 was delivered five times per week until the end of the experiment. Mean ± SEM is shown (n = 5–6). Differences in tumor growth were assessed at the last timepoint. D and E, Effect of GSK126 on the growth of MC38 cells in nude mice and C57BL/6 mice. GSK126 was delivered five times per week until the end of the experiment. Mean ± SEM is shown (n = 5–6). Differences in tumor growth were assessed at the last timepoint. One-way ANOVA with Tukey multiple comparison posttest (B) and a two-tailed unpaired Student t test (C–E) was used to evaluate statistical significance (*, P < 0.05; ***, P < 0.001). ns, nonsignificant.

Figure 1.

Inhibition of EZH2 activity by GSK126 has no effect on tumor growth in immunocompetent mice. A, Schematic illustration of treatment schedule. B and C, Effect of GSK126 on the growth of LLC cells in nude mice and C57BL/6 mice. GSK126 was delivered five times per week until the end of the experiment. Mean ± SEM is shown (n = 5–6). Differences in tumor growth were assessed at the last timepoint. D and E, Effect of GSK126 on the growth of MC38 cells in nude mice and C57BL/6 mice. GSK126 was delivered five times per week until the end of the experiment. Mean ± SEM is shown (n = 5–6). Differences in tumor growth were assessed at the last timepoint. One-way ANOVA with Tukey multiple comparison posttest (B) and a two-tailed unpaired Student t test (C–E) was used to evaluate statistical significance (*, P < 0.05; ***, P < 0.001). ns, nonsignificant.

Close modal

GSK126 attenuates the antitumor immune response

EZH2 is essential for T-cell proliferation, differentiation, and survival; therefore, it is important for antitumor immunity (20, 21, 23–26). The findings described above promoted us to hypothesize that GSK126 treatment affects the TME in a manner favorable to the growth of inoculated tumor cells. To address this, we assessed the composition and function of tumor infiltrating immune cells in untreated and GSK126-treated immunocompetent mice. GSK126 treatment reduced the percentages and numbers of tumor-infiltrating CD4+ and CD8+ T cells (Fig. 2A and B), which was associated with decreased T-cell proliferation (Fig. 2C) and increased T-cell apoptosis (Supplementary Fig. S2A and S2B). A similar reduction in the number of T cells infiltrating the tumor was observed by CD3 IF staining upon GSK126 treatment (Supplementary Fig. S2C). Importantly, the proportion of CD8+ T cells producing IFNγ was significantly decreased upon GSK126 treatment (Fig. 2D). Furthermore, GSK126-induced T-cell suppression appeared to be systemic as T-cell percentages and function were also affected in the blood and spleen (Supplementary Fig. S2). These results agree with the previous findings that EZH2 is critical for the proliferation, survival, and function of T cells (18, 19, 21–24). Therefore, an unintended consequence of GSK126 treatment is hampering T-cell activity.

Figure 2.

GSK126 attenuates the antitumor immune response. A, The percentages of CD4+ and CD8+ T cells in CD45+ cells in tumor tissues. Mean ± SEM is shown (n = 5). B, Absolute numbers of CD4+ and CD8+ T cells in tumor tissues (n = 5). C, The percentages of proliferating CD4+ and CD8+ T cells in tumor tissues were analyzed by Ki67 staining. Mean ± SEM is shown (n = 6). D, CD8+ T cells from tumor tissues were stimulated with PMA and ionomycin for 4 hours and IFNγ and granzyme B (Gzms-B) expression assessed by intracellular staining. Both the percentage of positive cells and mean fluorescence intensity (MFI) are shown (n = 6). Significance was assessed by two-tailed unpaired Student t tests (*, P < 0.05; **, P < 0.01; ***, P < 0.001). ns, nonsignificant.

Figure 2.

GSK126 attenuates the antitumor immune response. A, The percentages of CD4+ and CD8+ T cells in CD45+ cells in tumor tissues. Mean ± SEM is shown (n = 5). B, Absolute numbers of CD4+ and CD8+ T cells in tumor tissues (n = 5). C, The percentages of proliferating CD4+ and CD8+ T cells in tumor tissues were analyzed by Ki67 staining. Mean ± SEM is shown (n = 6). D, CD8+ T cells from tumor tissues were stimulated with PMA and ionomycin for 4 hours and IFNγ and granzyme B (Gzms-B) expression assessed by intracellular staining. Both the percentage of positive cells and mean fluorescence intensity (MFI) are shown (n = 6). Significance was assessed by two-tailed unpaired Student t tests (*, P < 0.05; **, P < 0.01; ***, P < 0.001). ns, nonsignificant.

Close modal

GSK126 suppresses CD8+ T-cell activity by reprograming the TME

It has been previously reported that GSK126 treatment can directly suppress CD8+ T-cell activity by downregulating H3K27me3 levels (19). However, GSK126 might also reprogram the TME, making it more immunosuppressive and thus impairing CD8+ T-cell activity. Therefore, we investigated whether the TME is involved in GSK126-mediated suppression of T-cell function. We used an experimental model in which the direct effect of the TME on CD8+ T cells could be evaluated in vivo by transferring antigen-specific T cells into tumor-bearing mice that were previously treated with GSK126 (Fig. 3A). First, we transferred activated OT1 CD45.1+CD8+ T cells into CD45.2 tumor–bearing mice and observed a decreased percentage of OT-1 CD8+ T cells in GSK126-treated compared with untreated mice (Fig. 3B). Furthermore, the proportion of OT-1 CD8+ T cells that produced IFNγ was significantly reduced in the group previously treated with GSK126, suggesting that the function of OT-1 CD8+ T cells was impaired (Fig. 3C and D). Interestingly, when we transferred naïve, rather than activated, OT1 CD45.1+CD8+ T cells into CD45.2 tumor–bearing mice we also observed a reduction in the proportion of IFNγ producers in recipients that had been previously treated with GSK126 (Fig. 3E and F). Taken together, we concluded that the TME, reprogramed by GSK126, can indirectly suppress CD8+ T-cell activity, in addition to inhibiting T-cell activity directly.

Figure 3.

GSK126 suppresses CD8+ T-cell activity by reprograming the TME. A–F, Schematic illustration (A) of treatment schedule for mice that received activated (B–D) or naïve (E and F) OT1 CD8+ T cells. B, The percentage of OT1 CD45.1+CD8+ T cells in the spleen was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). Splenic CD8+ T cells were activated using PMA and ionomycin (C and E) or ovalbumin (OVA; D and F) for 4 hours, and IFNγ expression by CD45.1+CD8+ T cells was assessed by intracellular staining. Mean ± SEM is shown (n = 5). Significance was assessed by two-tailed unpaired Student t tests (*, P < 0.05; ***, P < 0.001).

Figure 3.

GSK126 suppresses CD8+ T-cell activity by reprograming the TME. A–F, Schematic illustration (A) of treatment schedule for mice that received activated (B–D) or naïve (E and F) OT1 CD8+ T cells. B, The percentage of OT1 CD45.1+CD8+ T cells in the spleen was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). Splenic CD8+ T cells were activated using PMA and ionomycin (C and E) or ovalbumin (OVA; D and F) for 4 hours, and IFNγ expression by CD45.1+CD8+ T cells was assessed by intracellular staining. Mean ± SEM is shown (n = 5). Significance was assessed by two-tailed unpaired Student t tests (*, P < 0.05; ***, P < 0.001).

Close modal

GSK126 results in the expansion of the MDSC population in the TME

Immunosuppressive cells in the TME, such as regulatory T cells (Treg), tumor-associated macrophages (TAM), and MDSCs, can suppress CD8+ T-cell activity. We next investigated which immunosuppressive cells mediate T-cell suppression after GSK126 treatment. Using LLC inoculation models, we examined the immunosuppressive cell populations that might be affected by GSK126. Whereas Tregs, TAMs, and dendritic cells did not appear to be affected by GSK126 (Supplementary Fig. S3), this treatment led to an increase in the percentages and numbers of CD11b+Gr-1+MDSCs in the tumor tissue (Fig. 4A). Similarly, we observed increased numbers of MDSCs by Gr-1 IF staining of tumor sections upon GSK126 treatment (Supplementary Fig. S4A). Further analysis revealed that the proportions of both G-MDSC (CD11b+Ly6G+Ly6Clow) and M-MDSC (CD11b+Ly6G Ly6Chigh) subtypes were increased upon GSK126 treatment (Fig. 4B). MDSCs from GSK126-treated mice had comparable suppressive function with those from untreated counterparts, indicating that these cells were immunosuppressive (Fig. 4C). Consistent with this, the levels of signature MDSC molecules including reactive oxygen species (ROS), arginase (Arg1), and inducible nitric oxide synthase (iNOS) were similar between MDSCs isolated from GSK126-treated and -untreated mice (Fig. 4D and E). These results suggest that GSK126 promotes the expansion of MDSCs during tumor development.

Figure 4.

GSK126 results in the expansion of the MDSC population in the TME. A, The percentage and absolute number of MDSCs in tumor tissues was analyzed by flow cytometry. Mean ± SEM are shown (n = 5). B, The percentage of G-MDSCs (CD11b+Ly6G+Ly6Clow) and M-MDSCs (CD11b+Ly6GLy6Chigh) in CD11b+ cells in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). C, CD8+ T-cell proliferation suppression assay. CFSE-labeled CD8+ T cells from wild-type (WT) mice were stimulated with anti-CD3 and anti-CD28 antibodies and cocultured for 3 days with MDSCs isolated from control and GSK126-treated mice at 1:2 or 1:4 ratios. Mean ± SEM is shown (n = 3). D, Superoxide production by CD11b+Gr-1+cells in tumors was assessed using DHR (ROS) and flow cytometry. Mean fluorescence intensity (MFI) is shown (mean ± SEM of four samples). E, qPCR analysis of ARG1 and iNOS expression was performed on MDSCs, which were sorted from tumor tissues from control and GSK126-treated mice. Mean ± SEM is shown (n = 6). Significance was assessed by two-tailed unpaired Student t tests (*, P < 0.05; **, P < 0.01). ns, nonsignificant.

Figure 4.

GSK126 results in the expansion of the MDSC population in the TME. A, The percentage and absolute number of MDSCs in tumor tissues was analyzed by flow cytometry. Mean ± SEM are shown (n = 5). B, The percentage of G-MDSCs (CD11b+Ly6G+Ly6Clow) and M-MDSCs (CD11b+Ly6GLy6Chigh) in CD11b+ cells in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). C, CD8+ T-cell proliferation suppression assay. CFSE-labeled CD8+ T cells from wild-type (WT) mice were stimulated with anti-CD3 and anti-CD28 antibodies and cocultured for 3 days with MDSCs isolated from control and GSK126-treated mice at 1:2 or 1:4 ratios. Mean ± SEM is shown (n = 3). D, Superoxide production by CD11b+Gr-1+cells in tumors was assessed using DHR (ROS) and flow cytometry. Mean fluorescence intensity (MFI) is shown (mean ± SEM of four samples). E, qPCR analysis of ARG1 and iNOS expression was performed on MDSCs, which were sorted from tumor tissues from control and GSK126-treated mice. Mean ± SEM is shown (n = 6). Significance was assessed by two-tailed unpaired Student t tests (*, P < 0.05; **, P < 0.01). ns, nonsignificant.

Close modal

MDSC depletion unmasks the antitumor effect of GSK126 in immunocompetent hosts

To investigate whether the GSK126-mediated increase in the numbers of MDSCs was responsible for the lack of efficacy of this drug in immunocompetent hosts, we depleted MDSCs by injecting anti-Gr-1 neutralizing antibodies during GSK126 treatment (Fig. 5A; refs. 27–30). We found that tumor volume and mass significantly decreased (Fig. 5B and C), which was associated with increased tumor cell death (Supplementary Fig. S5A and S5B), in GSK126-treated tumor-bearing mice that also received anti-Gr-1 compared with mice treated with GSK126 alone. We confirmed that addition of anti-Gr-1 effectively depleted MDSCs in these mice (Fig. 5D; Supplementary Fig. S5C and S5D). MDSC depletion also led to enhanced T-cell function in GSK126-treated tumor-bearing mice. The proportions of CD4+ and CD8+ T cells in the tumor were increased in mice receiving both GSK126 and anti-Gr-1 compared with mice treated with GSK126 alone (Fig. 5E; Supplementary Fig. S5E). Importantly, the proportion of CD8+ T cells that produced IFNγ was significantly increased in mice treated with both GSK126 and anti-Gr-1 compared with mice receiving GSK126 alone (Fig. 5F; Supplementary Fig. S5F). Together, these results demonstrate that GSK126 treatment in immunocompetent hosts promotes MDSC accumulation, which masks the antitumor effects of this drug, and that suppressing MDSCs is critical to uncover the antitumor effects of GSK126.

Figure 5.

MDSC depletion unmasks the antitumor effect of GSK126 in immunocompetent hosts. A, Schematic illustration of treatment schedule. Mice received anti-Gr-1 antibody or IgG and GSK126 on the indicated days. GSK126 was administered five times per week and the anti-Gr-1 antibody twice per week until the end of the experiment. B and C, Analysis of tumor growth (n = 5). D, Percentage of MDSCs in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). E, The percentages of CD4+ and CD8+ T cells in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). F, The percentage of CD8+ T cells producing IFNγ from tumor tissues was analyzed by intracellular staining. Mean ± SEM is shown (n = 5). Two-way ANOVA with Tukey multiple comparison posttest was used to evaluate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001). ns, nonsignificant.

Figure 5.

MDSC depletion unmasks the antitumor effect of GSK126 in immunocompetent hosts. A, Schematic illustration of treatment schedule. Mice received anti-Gr-1 antibody or IgG and GSK126 on the indicated days. GSK126 was administered five times per week and the anti-Gr-1 antibody twice per week until the end of the experiment. B and C, Analysis of tumor growth (n = 5). D, Percentage of MDSCs in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). E, The percentages of CD4+ and CD8+ T cells in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 5). F, The percentage of CD8+ T cells producing IFNγ from tumor tissues was analyzed by intracellular staining. Mean ± SEM is shown (n = 5). Two-way ANOVA with Tukey multiple comparison posttest was used to evaluate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001). ns, nonsignificant.

Close modal

Clinically used chemotherapy drugs rescue the antitumor efficacy of GSK126 in immunocompetent hosts

Anticancer drugs commonly used clinically, including gemcitabine and 5-FU, were previously shown to deplete MDSCs (31). We thus determined whether combining gemcitabine or 5-FU with GSK126 could improve the efficacy of GSK126 in immunocompetent hosts (Fig. 6A). We found that addition of either gemcitabine or 5-FU could potentiate the antitumor effects of GSK126 (Fig. 6B–E). In addition, both agents could deplete MDSCs in our mouse models (Fig. 6F and G; Supplementary Fig. S6A–S6D). Importantly, we found that the percentages and functions of tumor-infiltrating CD4+ and CD8+ T cells in tumor-bearing mice were enhanced when mice were treated with gemcitabine or 5-FU in addition to GSK126 (Fig. 6H–J; Supplementary Fig. S6E–S6G). These results suggested that the addition of clinically used drugs that deplete MDSCs, such as gemcitabine and 5-FU, might enhance the antitumor effects of GSK126.

Figure 6.

Clinically used chemotherapy drugs restore the antitumor efficacy of GSK126 in immune competent hosts. A, Schematic illustration of treatment schedule. Mice received gemcitabine/5-FU and GSK126 on the indicated days. GSK126 was delivered five times per week and gemcitabine (B, C, F, and H) or 5-FU (D, E, G, I, and J) was administered once per week until the end of the experiment. B and C, The effect of gemcitabine and GSK126 on tumor growth (n = 4–5). D and E, The effect of 5-FU and GSK126 on tumor growth (n = 5). F and G, The percentage of MDSCs in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 4–5). H and I, The percentages of CD4+ and CD8+ T cells in tumor tissues was analyzed by flow cytometry. Mean ± SEM are shown (n = 4–5). J, The percentage of CD8+ T cells producing IFNγ in tumor tissues was analyzed by intracellular staining. Mean ± SEM is shown (n = 5). Two-way ANOVA with Tukey multiple comparison posttest was used to evaluate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). ns, nonsignificant.

Figure 6.

Clinically used chemotherapy drugs restore the antitumor efficacy of GSK126 in immune competent hosts. A, Schematic illustration of treatment schedule. Mice received gemcitabine/5-FU and GSK126 on the indicated days. GSK126 was delivered five times per week and gemcitabine (B, C, F, and H) or 5-FU (D, E, G, I, and J) was administered once per week until the end of the experiment. B and C, The effect of gemcitabine and GSK126 on tumor growth (n = 4–5). D and E, The effect of 5-FU and GSK126 on tumor growth (n = 5). F and G, The percentage of MDSCs in tumor tissues was analyzed by flow cytometry. Mean ± SEM is shown (n = 4–5). H and I, The percentages of CD4+ and CD8+ T cells in tumor tissues was analyzed by flow cytometry. Mean ± SEM are shown (n = 4–5). J, The percentage of CD8+ T cells producing IFNγ in tumor tissues was analyzed by intracellular staining. Mean ± SEM is shown (n = 5). Two-way ANOVA with Tukey multiple comparison posttest was used to evaluate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). ns, nonsignificant.

Close modal

GSK126 promotes HPC differentiation into MDSCs

We further investigated how GSK126 promotes the accumulation of MDSCs. We found that the proliferation and apoptosis, of MDSCs were unaffected by treatment with GSK126 (Supplementary Fig. S7A–S7C). Moreover, upon GSK126 treatment we found a systemic increase in MDSCs, including in the bone marrow where MDSCs differentiate from precursor HPCs (Fig. 7A–C). Interestingly, we found that the expression of EZH2 protein was high in HPCs, the precursor of MDSCs, but was barely detectable in MDSCs, themselves, consistent with our finding that GSK126 had no effect on MDSCs (Fig. 7D). Decreased H3K27 trimethylation was also observed after GSK126 treatment of HPCs (Supplementary Fig. S7D). Furthermore, we observed higher frequencies and numbers of granulocyte/macrophage progenitors (defined as lineageSca-1+C-kitCD16/32+CD34+) within the HPC population (Fig. 7E and F) upon GSK126 treatment, suggesting that GSK126-treated HPCs had enhanced potential to differentiate into myeloid cells. We therefore hypothesized that GSK126 promotes MDSC differentiation. We then sorted HPCs and induced HPCs to differentiate into MDSCs (Supplementary Fig. S7E). Indeed, GSK126 was able to enhance differentiation of HPCs to MDSCs in vitro in the presence of GM-CSF and IL6 (Fig. 7G). These results suggest that GSK126 promotes HPC differentiation into MDSCs.

Figure 7.

GSK126 promotes differentiation of MDSCs from HPCs. The percentages of MDSCs in bone marrow (A), blood (B), and spleen (C) were analyzed by flow cytometry. Mean ± SEM is shown (n = 6). D, EZH2 expression and H3K27 trimethylation in HPCs, MDSCs, and macrophages were analyzed by qPCR and immunoblot. Gating for granulocyte/macrophage progenitors (GMP; lineageSca-1+c-kitCD16/3 2+CD34+), megakaryocyte/erythrocyte progenitors (MEP; lineageSca-1+c-kitCD16/32CD34), and common myeloid progenitors (CMP; lineage Sca-1+c-kitCD16/32CD34+; E) and percentages and total cell numbers of these HPC subpopulations (F). Mean ± SEM is shown (n = 6). G, Differentiation of MDSCs in vitro. GM-CSF/IL6 was used for inducing HPC differentiation. GSK126 was added at the time point of 0, or 24, or 48 hours later, and MDSCs were analyzed after 72 hours. Mean ± SEM is shown (n = 3). One-way ANOVA with Tukey multiple comparison posttest (D and G) and a two-tailed unpaired Student t test (A–C) were used to evaluate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). ns, nonsignificant.

Figure 7.

GSK126 promotes differentiation of MDSCs from HPCs. The percentages of MDSCs in bone marrow (A), blood (B), and spleen (C) were analyzed by flow cytometry. Mean ± SEM is shown (n = 6). D, EZH2 expression and H3K27 trimethylation in HPCs, MDSCs, and macrophages were analyzed by qPCR and immunoblot. Gating for granulocyte/macrophage progenitors (GMP; lineageSca-1+c-kitCD16/3 2+CD34+), megakaryocyte/erythrocyte progenitors (MEP; lineageSca-1+c-kitCD16/32CD34), and common myeloid progenitors (CMP; lineage Sca-1+c-kitCD16/32CD34+; E) and percentages and total cell numbers of these HPC subpopulations (F). Mean ± SEM is shown (n = 6). G, Differentiation of MDSCs in vitro. GM-CSF/IL6 was used for inducing HPC differentiation. GSK126 was added at the time point of 0, or 24, or 48 hours later, and MDSCs were analyzed after 72 hours. Mean ± SEM is shown (n = 3). One-way ANOVA with Tukey multiple comparison posttest (D and G) and a two-tailed unpaired Student t test (A–C) were used to evaluate statistical significance (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). ns, nonsignificant.

Close modal

Inhibition of EZH2 methyltransferase activity might represent a viable strategy for the treatment of cancers with high EZH2 activity. The development of EZH2-specific inhibitors has been an active area of investigation (9, 11, 32). Indeed, GSK126 is one of a number of candidate compounds that are currently being evaluated in preclinical and clinical trials. In preclinical trials, inhibition of EZH2 by GSK126 largely slowed the growth of lymphoma with EZH2-activating mutations (12) and pediatric gliomas (14). However, in a phase I clinical trial of GSK126 with 22 evaluable patients, only 1 patient with diffuse large B-cell lymphoma showed a partial response and 7 patients had stable disease (33). Compared with the superior therapeutic benefits of GSK126 observed in preclinical trials, this drug had little therapeutic effect in clinical trials. We considered that one of the possible reasons was that GSK126 might impair the antitumor immune response in humans because preclinical trials were performed in immunodeficient hosts. Therefore, we compared the therapeutic efficacy of this drug in immunocompetent and immunodeficient hosts. We found that GSK126 treatment restrained tumor growth in immune deficient, but not in immunocompetent hosts. In the immunocompetent hosts (C57BL/6 mice), GSK126 promoted MDSC generation, which suppressed antitumor T-cell immunity and masked its antitumor effect. These results suggest a possible explanation for the disappointing results from a phase I clinical trial of GSK126: that this drug might dampen antitumor immunity. However, another EZH2 inhibitor EPZ-6438 showed encouraging results; specifically, 49/203 (24%) patients responded including 14 complete responses and 35 partial responses (33). Therefore, the effects of other EZH2 inhibitors on tumor immunity remain unknown, which warrants further investigation.

It has been reported that EZH2 is essential for T-cell proliferation, differentiation, and survival, and that therefore it is important for antitumor immunity (20, 21, 23–26). In contrast, Zou and colleagues reported that a combination of GSK126 (30 mg/kg) with 5-aza-2-′deoxycytidine (0.2 mg/kg) resulted in TME reprograming, the production of Th 1–type chemokines CXCL9 and CXCL10 in the tumor, and increased tumor infiltration of CD8+ T cells (34, 35). Therefore, the effects of GSK126 on both T cells and the TME might contribute to the differences in antitumor efficacy of GSK126 between immunodeficient and -competent mice. In this study, we used adoptive transfer of naive or activated CD8+ T cells to prove the contribution of TME reprogramming to the inhibition of tumor immunity mediated by GSK126. Furthermore, we found increased MDSC accumulation in the TME upon GSK126 treatment during tumor development and that MDSC depletion by anti-Gr-1 neutralizing antibodies unmasked the antitumor effects of GSK126. Taken together, the effects of GSK-126 on both T cells and MDSCs, which are both involved in tumor immunity, should be considered when using this drug to treat patients with cancer.

We further investigated how GSK126 induces the accumulation of MDSCs in the TME. In addition to the accumulation of MDSCs in the TME, increases in MDSCs in the spleen, peripheral blood, and bone marrow was also apparent. Therefore, we focused on MDSC development in the bone marrow. We proved that GSK126 promotes HPC differentiation into MDSCs in vitro. EZH2 has been shown to maintain HPC function by stabilizing chromatin structure and repressing the expression of prodifferentiation genes (36). Therefore, we examined the expression of regulators of HPC differentiation (37). We found that the expression of Cebpe was significantly upregulated after GSK126 treatment (Supplementary Fig. S7F). Studies have shown that the CCAAT-enhancer-binding proteins (C/EBP) family, including Cebpa, Cebpb, and Cebpe, are key regulators of proliferation or differentiation of myeloid progenitors (38–40). Therefore, we speculate that GSK126 may upregulate Cebpe, which in turn activates expression of myeloid cell target genes to promote differentiation of HPCs into MDSCs. However, the underlying molecular mechanisms by which GSK126 promotes MDSC development, and whether other EZH2 inhibitors have the same effect, need to be studied further.

From a therapeutic standpoint, our study outlines a strategy that can effectively reduce the number, and abolish the suppressive function, of MDSCs during GSK126 treatment, thereby tipping the balance toward effective antitumor immunity. Here, we showed that the highly proliferative nature of MDSCs renders them susceptible to low dose gemcitabine and 5-FU. A novel finding of our study is that combined gemcitabine/5-FU +GSK126 treatment improves the efficacy of GSK126 and strongly inhibits tumor growth. These data indicate that gemcitabine/5-FU +GSK126 combination treatment could provide superior therapeutic benefits and suggests a new strategy for GSK126 clinical therapy.

In summary, we demonstrated that GSK126-mediated suppression of EZH2 activity promotes HPC differentiation to MDSCs, resulting in increased MDSCs in the TME, which suppress antitumor immunity. Importantly, we also showed that MDSC depletion with an antibody or gemcitabine/5-FU can enhance the antitumor efficacy of GSK126. Mechanistically, our results suggest that the unintended effect of GSK126 in promoting MDSC generation, which masks its antitumor effect, can be suppressed by MDSC depletion; this could be exploited clinically to unleash the antitumor effects of GSK126.

No potential conflicts of interest were disclosed.

Conception and design: S. Huang, H. Long, B. Zhu

Development of methodology: S. Huang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Zhou, J. Huang, L. Zhou

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Huang, Z. Wang, J. Luo, B. Zhu

Writing, review, and/or revision of the manuscript: S. Huang, Y.Y. Wan, H. Long, B. Zhu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): B. Zhu

Study supervision: H. Long, B. Zhu

We are extremely appreciative of Chunyan Hu for her support with FCM. We are also very grateful to Jin Peng and Qian Chen for confocal microscopy experiments. This work was supported by the National Nature Science Foundation of China (grant nos. 81472648, 81802865, and 81620108023).

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.
Jones
PA
,
Baylin
SB
. 
The epigenomics of cancer
.
Cell
2007
;
128
:
683
92
.
2.
Margueron
R
,
Reinberg
D
. 
The Polycomb complex PRC2 and its mark in life
.
Nature
2011
;
469
:
343
9
.
3.
Geng
J
,
Li
X
,
Zhou
Z
,
Wu
CL
,
Dai
M
,
Bai
X
. 
EZH2 promotes tumor progression via regulating VEGF-A/AKT signaling in non-small cell lung cancer
.
Cancer Lett
2015
;
359
:
275
87
.
4.
Kodach
LL
,
Jacobs
RJ
,
Heijmans
J
,
van Noesel
CJ
,
Langers
AM
,
Verspaget
HW
, et al
The role of EZH2 and DNA methylation in the silencing of the tumour suppressor RUNX3 in colorectal cancer
.
Carcinogenesis
2010
;
31
:
1567
75
.
5.
Kleer
CG
,
Cao
Q
,
Varambally
S
,
Shen
R
,
Ota
I
,
Tomlins
SA
, et al
EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells
.
Proc Nat Acad Sci U S A
2003
;
100
:
11606
11
.
6.
Chen
Y
,
Xie
D
,
Yin Li
W
,
Man Cheung
C
,
Yao
H
,
Chan
CY
, et al
RNAi targeting EZH2 inhibits tumor growth and liver metastasis of pancreatic cancer in vivo
.
Cancer Lett
2010
;
297
:
109
16
.
7.
Xu
K
,
Wu
ZJ
,
Groner
AC
,
He
HH
,
Cai
C
,
Lis
RT
, et al
EZH2 oncogenic activity in castration-resistant prostate cancer cells is polycomb-independent
.
Science
2012
;
338
:
1465
9
.
8.
Chase
A
,
Cross
NC
. 
Aberrations of EZH2 in cancer
.
Clin Cancer Res
2011
;
17
:
2613
8
.
9.
Kim
KH
,
Roberts
CW
. 
Targeting EZH2 in cancer
.
Nat Med
2016
;
22
:
128
34
.
10.
Crea
F
,
Fornaro
L
,
Bocci
G
,
Sun
L
,
Farrar
WL
,
Falcone
A
, et al
EZH2 inhibition: targeting the crossroad of tumor invasion and angiogenesis
.
Cancer Metastasis Rev
2012
;
31
:
753
61
.
11.
McCabe
MT
,
Creasy
CL
. 
EZH2 as a potential target in cancer therapy
.
Epigenomics
2014
;
6
:
341
51
.
12.
McCabe
MT
,
Ott
HM
,
Ganji
G
,
Korenchuk
S
,
Thompson
C
,
Van Aller
GS
, et al
EZH2 inhibition as a therapeutic strategy for lymphoma with EZH2-activating mutations
.
Nature
2012
;
492
:
108
12
.
13.
Fillmore
CM
,
Xu
C
,
Desai
PT
,
Berry
JM
,
Rowbotham
SP
,
Lin
YJ
, et al
EZH2 inhibition sensitizes BRG1 and EGFR mutant lung tumours to TopoII inhibitors
.
Nature
2015
;
520
:
239
42
.
14.
Mohammad
F
,
Weissmann
S
,
Leblanc
B
,
Pandey
DP
,
Hojfeldt
JW
,
Comet
I
. 
EZH2 is a potential therapeutic target for H3K27M-mutant pediatric gliomas
.
Nat Med
2017
;
23
:
483
92
.
15.
Bitler
BG
,
Aird
KM
,
Garipov
A
,
Li
H
,
Amatangelo
M
,
Kossenkov
AV
, et al
Synthetic lethality by targeting EZH2 methyltransferase activity in ARID1A-mutated cancers
.
Nat Med
2015
;
21
:
231
8
.
16.
Kim
E
,
Kim
M
,
Woo
DH
,
Shin
Y
,
Shin
J
,
Chang
N
, et al
Phosphorylation of EZH2 activates STAT3 signaling via STAT3 methylation and promotes tumorigenicity of glioblastoma stem-like cells
.
Cancer Cell
2013
;
23
:
839
52
.
17.
Gunawan
M
,
Venkatesan
N
,
Loh
JT
,
Wong
JF
,
Berger
H
,
Neo
WH
. 
The methyltransferase Ezh2 controls cell adhesion and migration through direct methylation of the extranuclear regulatory protein talin
Nat Immunol
2015
;
16
:
505
16
.
18.
Beguelin
W
,
Popovic
R
,
Teater
M
,
Jiang
Y
,
Bunting
KL
,
Rosen
M
, et al
EZH2 is required for germinal center formation and somatic EZH2 mutations promote lymphoid transformation
.
Cancer Cell
2013
;
23
:
677
92
.
19.
Zhao
E
,
Maj
T
,
Kryczek
I
,
Li
W
,
Wu
K
,
Zhao
L
, et al
Cancer mediates effector T cell dysfunction by targeting microRNAs and EZH2 via glycolysis restriction
.
Nat Immunol
2016
;
17
:
95
103
.
20.
DuPage
M
,
Chopra
G
,
Quiros
J
,
Rosenthal
WL
,
Morar
MM
,
Holohan
D
, et al
The chromatin-modifying enzyme Ezh2 is critical for the maintenance of regulatory T cell identity after activation
.
Immunity
2015
;
42
:
227
38
.
21.
Tumes
DJ
,
Onodera
A
,
Suzuki
A
,
Shinoda
K
,
Endo
Y
,
Iwamura
C
, et al
The polycomb protein Ezh2 regulates differentiation and plasticity of CD4(+) T helper type 1 and type 2 cells
.
Immunity
2013
;
39
:
819
32
.
22.
Yin
J
,
Leavenworth
JW
,
Li
Y
,
Luo
Q
,
Xie
H
,
Liu
X
, et al
Ezh2 regulates differentiation and function of natural killer cells through histone methyltransferase activity
.
Proc Nat Acad Sci U S A
2015
;
112
:
15988
93
.
23.
Karantanos
T
,
Chistofides
A
,
Barhdan
K
,
Li
L
,
Boussiotis
VA
. 
Regulation of T cell differentiation and function by EZH2
.
Front Immunol
2016
;
7
:
172
.
24.
Chen
G
,
Subedi
K
,
Chakraborty
S
,
Sharov
A
,
Lu
J
,
Kim
J
, et al
Ezh2 regulates activation-induced CD8(+) T cell cycle progression via repressing Cdkn2a and Cdkn1c expression
.
Front Immunol
2018
;
9
:
549
.
25.
He
S
,
Liu
Y
,
Meng
L
,
Sun
H
,
Wang
Y
,
Ji
Y
. 
Ezh2 phosphorylation state determines its capacity to maintain CD8(+) T memory precursors for antitumor immunity
. 
2017
;
8
:
2125
.
26.
Yang
XP
,
Jiang
K
,
Hirahara
K
,
Vahedi
G
,
Afzali
B
,
Sciume
G
, et al
EZH2 is crucial for both differentiation of regulatory T cells and T effector cell expansion
.
Sci Rep
2015
;
5
:
10643
.
27.
Terabe
M
,
Matsui
S
,
Park
JM
,
Mamura
M
,
Noben-Trauth
N
,
Donaldson
DD
, et al
Transforming growth factor-beta production and myeloid cells are an effector mechanism through which CD1d-restricted T cells block cytotoxic T lymphocyte-mediated tumor immunosurveillance: abrogation prevents tumor recurrence
.
J Exp Med
2003
;
198
:
1741
52
.
28.
Kawano
M
,
Mabuchi
S
,
Matsumoto
Y
,
Sasano
T
,
Takahashi
R
,
Kuroda
H
, et al
The significance of G-CSF expression and myeloid-derived suppressor cells in the chemoresistance of uterine cervical cancer
.
Sci Rep
2015
;
5
:
18217
.
29.
van Deventer
HW
,
Burgents
JE
,
Wu
QP
,
Woodford
RM
,
Brickey
WJ
,
Allen
IC
, et al
The inflammasome component NLRP3 impairs antitumor vaccine by enhancing the accumulation of tumor-associated myeloid-derived suppressor cells
.
Cancer Res
2010
;
70
:
10161
9
.
30.
Xia
S
,
Li
X
,
Cheng
L
,
Han
M
,
Zhang
M
,
Liu
X
, et al
Chronic intake of high fish oil diet induces myeloid-derived suppressor cells to promote tumor growth
.
Cancer Immunol Immunother
2014
;
63
:
663
73
.
31.
Vincent
J
,
Mignot
G
,
Chalmin
F
,
Ladoire
S
,
Bruchard
M
,
Chevriaux
A
, et al
5-Fluorouracil selectively kills tumor-associated myeloid-derived suppressor cells resulting in enhanced T cell-dependent antitumor immunity
.
Cancer Res
2010
;
70
:
3052
61
.
32.
Miranda
TB
,
Cortez
CC
,
Yoo
CB
,
Liang
G
,
Abe
M
,
Kelly
TK
, et al
DZNep is a global histone methylation inhibitor that reactivates developmental genes not silenced by DNA methylation
.
Mol Cancer Ther
2009
;
8
:
1579
88
.
33.
Gulati
N
,
Beguelin
W
,
Giulino-Roth
L
. 
Enhancer of zeste homolog 2 (EZH2) inhibitors
.
Leuk Lymphoma
2018
;
59
:
1
12
.
34.
Peng
D
,
Kryczek
I
,
Nagarsheth
N
,
Zhao
L
,
Wei
S
,
Wang
W
, et al
Epigenetic silencing of TH1-type chemokines shapes tumour immunity and immunotherapy
.
Nature
2015
;
527
:
249
53
.
35.
Nagarsheth
N
,
Peng
D
,
Kryczek
I
,
Wu
K
,
Li
W
,
Zhao
E
, et al
PRC2 epigenetically silences Th1-type chemokines to suppress effector T-cell trafficking in colon cancer
.
Cancer Res
2016
;
76
:
275
82
.
36.
Herviou
L
,
Cavalli
G
,
Cartron
G
,
Klein
B
,
Moreaux
J
. 
EZH2 in normal hematopoiesis and hematological malignancies
.
Oncotarget
2016
;
7
:
2284
96
.
37.
Paul
F
,
Arkin
Y
,
Giladi
A
,
Jaitin
DA
,
Kenigsberg
E
,
Keren-Shaul
H
, et al
Transcriptional heterogeneity and lineage commitment in myeloid progenitors
.
Cell
2015
;
163
:
1663
77
.
38.
Kim
K
,
Hwang
SM
,
Kim
SM
,
Park
SW
,
Jung
Y
,
Chung
IY
. 
Terminally differentiating eosinophils express neutrophil primary granule proteins as well as eosinophil-specific granule proteins in a temporal manner
.
Immune Network
2017
;
17
:
410
23
.
39.
Gery
S
,
Gombart
AF
,
Fung
YK
,
Koeffler
HP
. 
C/EBPepsilon interacts with retinoblastoma and E2F1 during granulopoiesis
.
Blood
2004
;
103
:
828
35
.
40.
Suh
HC
,
Benoukraf
T
,
Shyamsunder
P
,
Yin
T
,
Cao
Q
,
Said
J
, et al
LPS independent activation of the pro-inflammatory receptor Trem1 by C/EBPepsilon in granulocytes
.
Sci Rep
2017
;
7
:
46440
.