Glycolysis regulates the expansion of myeloid-derived suppressor cells in tumor-bearing hosts through prevention of ROS-mediated apoptosis

Abstract

Immunotherapy aiming to rescue or boost antitumor immunity is an emerging strategy for treatment of cancers. The efficacy of immunotherapy is strongly controlled by the immunological milieu of cancer patients. Myeloid-derived suppressor cells (MDSCs) are heterogeneous immature myeloid cell populations with immunosuppressive functions accumulating in individuals during tumor progression. The signaling mechanisms of MDSC activation have been well studied. However, there is little known about the metabolic status of MDSCs and the physiological role of their metabolic reprogramming. In this study, we discovered that myeloid cells upregulated their glycolytic genes when encountered with tumor-derived factors. MDSCs exhibited higher glycolytic rate than their normal cell compartment did, which contributed to the accumulation of the MDSCs in tumor-bearing hosts. Upregulation of glycolysis prevented excess reactive oxygen species (ROS) production by MDSCs, which protected MDSCs from apoptosis. Most importantly, we identified the glycolytic metabolite, phosphoenolpyruvate (PEP), as a vital antioxidant agent able to prevent excess ROS production and therefore contributed to the survival of MDSCs. These findings suggest that glycolytic metabolites have important roles in the modulation of fitness of MDSCs and could be potential targets for anti-MDSC strategy. Targeting MDSCs with analogs of specific glycolytic metabolites, for example, 2-phosphoglycerate or PEP may diminish the accumulation of MDSCs and reverse the immunosuppressive milieu in tumor-bearing individuals.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1

MDSC accumulation during tumor progression. (a) The IVIS images of Balb/c mice receiving 4T1-LG at indicated time points after implantation. Arrow indicated the IVIS image of the 4T1 lung metastasis in mice. (b) The tumor growth curve and total flux of luciferase activity of mice in (a) (n=3 per group). (c) The hematoxylin and eosin (H&E) staining of lung and liver sections of mice at indicated time points after inoculation of 4T1-LG tumor cells. (d) Representative flow cytometric analysis of Ly6G+ and Ly6C+ populations among gated CD11b+ cells in indicated tissues of control and 4T1-tumor-bearing mice at the third week and sixth week after tumor inoculation. (e) Absolute numbers of total CD11b+ cells, Ly6G+CD11b+ cells and Ly6C+CD11b+ cells in indicated organs (n=3 per group). (f) Immunohistochemical staining for detection of Gr-1+ and CD11b+ cells, respectively, in the liver and tumor sites of 4T1-tumor-bearing mice at indicated time points after tumor inoculation. All the data are representative of two independent experiments. ND, not done; ns, not significant. *P<0.05, **P<0.01, ***P<0.001 (unpaired Student’s t-test). (c and f) Scale bars, 100 μm. (b and e) Error bars, S.D.

Figure 2

Metabolic status in myeloid cells during tumor progression. (a) Heat map summarizing the gene expression data from cDNA microarray. mRNA levels of genes of interest in splenic neutrophils, monocytes from normal healthy mice, gMDSCs and mMDSCs from tumor site of 4T1-tumor-bearing mice were shown. (b) Quantitative real-time PCR analysis of gene expression of glycolytic enzymes in FACSorted CD11b+Ly6C+ and CD11b+Ly6G+ cells from tumor sites of 4T1-tumor-bearing mice and spleens from normal BALB/c mice (n=3 per group). (c) Quantitative real-time PCR analysis of ARG1, NOS2, PDCD1LG1 and PDCD1LG2 in FACSorted CD11b+Ly6C+ and CD11b+Ly6G+ cells from tumor sites of 4T1-tumor-bearing mice and spleens from normal BALB/c mice (n=3 per group). (d) ECAR of FACSorted splenic gMDSCs from 4T1-tumor-bearing mice and splenic neutrophils from normal BALB/c mice (left panel); calculated non-glycolytic acidification, glycolysis and glycolytic capacity of the cells (middle panel); ECAR versus OCR of the above cells under basal condition and oligomycin treatment (right panel, n=7 per group). All the data are representative of two independent experiments. *P<0.05, **P<0.01 and ***P<0.001 (unpaired Student’s t-test). (b–d) Error bars, S.D.

Figure 3

In vitro induction of MDSCs. (a) Quantitative real-time PCR analysis of gene expression of glycolytic enzymes in purified bone marrow neutrophils and monocytes after incubation with 4T1-tumor cells or with primary MECs for 40 h. (b) Representative flow cytometric zebra plots of MDSCs induced from bone marrow cells in the presence of GM-CSF for 5 days. CD11b and CD11c expression (left) of the total cells and Ly6C and Ly6G expression (right) of gated CD11b+CD11c− cells were shown. (c) Quantitative real-time PCR analysis of gene expression of glycolytic enzymes in FACSorted GM-CSF-induced (GM-MDSCs) or bone marrow (primary BM) CD11b+CD11c−Ly6C+ and CD11b+CD11c−Ly6G+ cells (n=3 per group). (d) Quantitative real-time PCR analysis of ARG1, NOS2, PDCD1LG1 and PDCD1LG2 in the cells in (c) (n=3 per group). (e) Proliferation of CD4+ (left) and CD8+ (right) T cells cocultured with different numbers of MACSorted fresh bone marrow CD11b+ (BM ctrl) or MACSorted GM-CSF-induced CD11b+ MDSCs (GM-MDSCs). The percentage of the proliferating CD4+ (left panel) and CD8+ (right panel) T cells was calculated by dividing the number of EdU+ cells of each well by the average number of EdU+ cells in wells without MDSCs (as 100%). All the data are representative of two independent experiments. *P<0.05, **P<0.01 and ***P<0.001 (unpaired Student’s t-test). (a and c–e) Error bars, S.D.

Figure 4

The in vivo accumulation of MDSCs was dependent on glycolysis. (a) Counts of CD11b+ cells in the blood of PBS-treated (ctrl) or 2-DG-treated normal (N) and 4T1-tumor-bearing (TB) mice at indicated time points after tumor inoculation. Arrow indicated the time point starting 2-DG treatment (n=3 per group). (b) Absolute numbers of CD11b+ cells in the indicated tissues of the mice in (a) at the fourth week after tumor inoculation (n=3 per group). (c) Size of tumor in the PBS-treated (ctrl) and 2-DG-treated tumor-bearing mice measured at the indicated time points (n=6–7 per group). (d) Immunohistochemical staining for detection and (e) quantification of Gr-1+ cells in the tumor and the liver of the PBS-treated (ctrl) and 2-DG-treated tumor-bearing mice at week 4 after tumor inoculation. Arrowheads indicated Gr-1+ cells (n=3 per group). Scale bars, 100 μm. All the data are representative of two independent experiments. ND, not done; ns, not significant. *P<0.05, **P<0.01 and ***P<0.001 (unpaired Student’s t-test). (a–c and e) Error bars, S.D.

Figure 5

The expansion of MDSCs was dependent on glycolysis. (a) Percentage of BrdU+ cells among mMDSCs and gMDSCs from the bone marrow (BM), spleen (Sp) and tumor site of the PBS-treated (ctrl) and 2-DG-treated 4T1-tumor-bearing mice (n=3 per group). (b) Proliferation of CD8+ T cells cocultured with different numbers of splenic PBS (ctrl)-, 2-DG-treated CD11b+ MDSCs from 4T1-tumor-bearing mice. (c) Proliferation of CD8+ T cells cocultured with different numbers of splenic PBS (ctrl)-, IA-treated CD11b+ MDSCs from 4T1-tumor-bearing mice. (d) Cell numbers of CD11b+Gr-1+ MDSCs recovered from the MDSC induction culture 5 days after induction using 20 ng/ml of GM-CSF in the presence or absence of 2-DG in 24-well plates. (e) Cell numbers of CD11b+Gr-1+ MDSCs recovered from the MDSC induction culture 3 days after induction using 20 ng/ml of GM-CSF in the presence or absence of 0.5 mM 2-DG and PEP (5 mM) in 96-well plates. (f) Percentage of proliferating CD11b+Gr-1+ MDSCs among the cells from (e). All the data are representative of two independent experiments. ns, not significant. *P<0.05, **P<0.01 and ***P<0.001 (unpaired Student’s t-test). (a–f) Error bars, S.D.

Figure 6

Glycolysis prevented apoptosis and excess ROS production of MDSCs. (a) Percentage of PI−Annexin V+ early apoptotic cells among mMDSCs and gMDSCs from the bone marrow and spleen of the PBS-treated (ctrl) and 2-DG-treated 4T1-tumor-bearing mice. 4T1-tumor-bearing mice were treated with 40 mg of 2-DG or PBS (ctrl) as described in Materials and Methods (n=3 per group). (b) Representative flow cytometric zebra plots gating on Gr-1+ population of in vitro 4T1-induced myeloid cells subjected to 2-DG or control treatment for 18 or 48 h showed PI and Annexin V staining. The bar graph of percentage of early (PI−Annexin V+) and late (PI+ Annexin V+) apoptotic cells among Gr-1+ population of in vitro 4T1-induced myeloid cells subjected to 2-DG or PBS (ctrl) treatment for 18 or 48 h was shown (n=3 per group). (c) Histogram (left) and bar graph (right) showing fluorescence intensity or mean fluorescence intensity (MFI) of FITC-dUTP-labeled or -unlabeled DNA fragments in MACSorted CD11b+Gr-1+ GM-CSF-induced MDSCs treated with 2-DG or PBS (ctrl) for 22 h and subjected to detection of DNA fragment by TUNEL assay (n=3 per group). (d) ROS production in freshly isolated neutrophils from normal BALB/c mice and in freshly isolated gMDSCs from 4T1-tumor-bearing mice. (e) Representative histograms (left two panels) of ROS production in neutrophils from normal BALB/c mice and in gMDSCs from 4T1-tumor-bearing mice treated with 2-DG or vehicle (DPBS, ctrl) for 2 h. Bar graph (right panel) showing the fold changes of MFI of DCFDA of the cells after treatment with 2-DG for 2 h. The MFI of DCFDA of the cells treated with vehicle was used as denominator (n=3 per group). (f) A representative histogram for ROS production in MACSorted GM-CSF-induced CD11b+Gr-1+ MDSCs treated with or without 10 mM of 2-DG for 6 h. All the data are representative of two independent experiments. ns, not significant. *P<0.05, **P<0.01 and ***P<0.001 (unpaired Student’s t-test). (a–c and e) Error bars, S.D.

Figure 7

PEP ameliorated apoptosis in MDSCs through prevention of excess ROS production during blockade of glycolysis. (a) Percentage of PI−Annexin V+ early apoptotic cells among CD11b+Gr-1+ GM-CSF-induced MDSCs treated with 10 mM of 2-DG or vehicle (ctrl) for 8 h in the presence or absence of PEP at the concentration of 5 mM (n=3 per group). (b) MFI of FITC-dUTP-labeled DNA fragments in MACSorted Gr-1+ GM-CSF-induced MDSCs treated with 10 mM of 2-DG or vehicle (ctrl) for 22 h in the presence or absence of PEP at the concentration of 7.5 mM (n=3 per group). (c) ROS production in CD11b+Gr-1+ GM-CSF-induced MDSCs treated with 10 mM of 2-DG or vehicle (ctrl) for 5 h in the presence or absence of PEP at the concentration of 5 mM (n=3 per group). (d) ROS production in CD11b+Gr-1+ GM-CSF-induced MDSCs pretreated with 2 μM of DPI or vehicle for 1 h before the addition of 2-DG at the concentration of 10 mM for additional incubation for 5 h (n=3 per group). (e) Percentage of PI−Annexin V+ early apoptotic cells among CD11b+Gr-1+ GM-CSF-induced MDSCs treated as in (d) (n=3 per group). All the data are representative of two independent experiments. ns, not significant. *P<0.05, **P<0.01 and ***P<0.001 (unpaired Student’s t-test). (a–e) Error bars, S.D.