SEARCH

SEARCH BY CITATION

Keywords:

  • CD8+ T-cell activation;
  • CD8+ T-cell suppression;
  • Monocytic myeloid-derived suppressor cell (MO-MDSC);
  • Nitric oxide;
  • Polymorphonuclear MDSC (PMN-MDSC)

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

Tumor growth coincides with an accumulation of myeloid-derived suppressor cells (MDSCs), which exert immune suppression and which consist of two main subpopulations, known as monocytic (MO) CD11b+CD115+Ly6GLy6Chigh MDSCs and granulocytic CD11b+CD115Ly6G+Ly6Cint polymorphonuclear (PMN)-MDSCs. However, whether these distinct MDSC subsets hamper all aspects of early CD8+ T-cell activation — including cytokine production, surface marker expression, survival, and cytotoxicity — is currently unclear. Here, employing an in vitro coculture system, we demonstrate that splenic MDSC subsets suppress antigen-driven CD8+ T-cell proliferation, but differ in their dependency on IFN-γ, STAT-1, IRF-1, and NO to do so. Moreover, MO-MDSC and PMN-MDSCs diminish IL-2 levels, but only MO-MDSCs affect IL-2Rα (CD25) expression and STAT-5 signaling. Unexpectedly, however, both MDSC populations stimulate IFN-γ production by CD8+ T cells on a per cell basis, illustrating that some T-cell activation characteristics are actually stimulated by MDSCs. Conversely, MO-MDSCs counteract the activation-induced change in CD44, CD62L, CD162, and granzyme B expression, while promoting CD69 and Fas upregulation. Together, these effects result in an altered CD8+ T-cell adhesiveness to the extracellular matrix and selectins, sensitivity to FasL-mediated apoptosis, and cytotoxicity. Hence, MDSCs intricately influence different CD8+ T-cell activation events in vitro, whereby some parameters are suppressed while others are stimulated.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

The discovery of tumor-specific antigenic peptides recognized by CD8+ T cells has laid the foundation for immunotherapeutic approaches aimed at maximizing cytotoxic T lymphocyte (CTL) mediated eradication of cancer cells. The optimization of such therapies requires a thorough knowledge of the mechanisms regulating CTL induction and activity and of the countermeasures taken by tumors to avoid destruction.

Naive CD8+ T cells constantly sample APCs in the secondary lymphoid organs. As a function of this activity, naive T cells express low levels of CD44 and high levels of the homing receptor CD62L, ensuring entry into LNs [1]. Upon activation at these sites, a series of events is initiated that dramatically alters the molecular repertoire of CD8+ T cells, enabling these cells to proliferate, migrate, and acquire effector functions [2]. Importantly, distinct features of CTL activation are obtained in different phases of the activation process and are not necessarily interdependent [3, 4]. Thus, a brief DC-T-cell encounter is sufficient to upregulate the early activation markers CD44 and CD69, but longer stable contacts are needed to initiate IL-2 and IFN-γ secretion and the abundant expression of activation markers (including CD25, generating a high-affinity trimeric IL-2R). Finally, only upon dissociation from the DC, CD8+ T cells start to proliferate vigorously [3]. Cell-fate decisions in effector and memory CD8+ T-cell differentiation are regulated by the strength and duration of IL-2 signals during the primary activation [5, 6], and the expression of the transcription factor T-bet [5, 7]. Ultimately, cytotoxicity is the main function of effector CD8+ T cells, and is predominantly regulated at the level of the cytotoxic granule content rather than the process of degranulation itself [8].

However, immunoregulatory cell types may interfere with distinct aspects of CD8+ T-cell differentiation. For example, CD4+CD25+ regulatory T cells inhibit granule exocytosis without interfering with normal effector differentiation [9]. Myeloid-derived suppressor cells (MDSCs), which consist of an immature monocytic CD11b+CD115+Ly6GLy6Chigh (monocytic (MO)-MDSC) and granulocytic CD11b+CD115Ly6G+Ly6Cint (polymorphonuclear (PMN)-MDSC) population [10-13], also hamper T-cell functions during various pathologies, including cancer [14], infections [15], and transplantation [16]. These cells expand from bone marrow (BM) progenitors under the influence of hematopoietic growth factors and inflammatory cytokines [17] and can deploy a variety of mechanisms for T-cell suppression [18], whereby the spleen appears to be an important site of action in tumor-bearers [19]. In this context, NO production via STAT-1 activation is an important suppressive mechanism employed by MO-MDSCs [11], while PMN-MDSCs are main producers of reactive oxygen/nitrogen species. These nitrate TCR and CD8 molecules at the sites of MDSC-T cell contact, thereby disrupting TCR complexes and preventing T-cell activation [20]. Other mechanisms of MDSC-mediated suppression include l-arginine depletion from the environment and the sequestration of cystine leading to a reduced availability of cysteine for T-cell activation [18].

Up to now it is unclear which aspects of CTL activation and differentiation are affected by the distinct MDSC subsets. Here, we demonstrate that splenic MDSCs not only inhibit several features of early CD8+ T-cell activation (proliferation; IL-2 secretion/responsiveness; CD44, CD162, and granzyme B expression; CD62L downregulation) — whereby MO- and PMN-MDSCs differ in their capacity to do so and differentially depend on IFN-γ, STAT-1, interferon regulatory factor 1 (IRF-1), and NO — but at the same time stimulate other activation events, such as IFN-γ production, CD69 expression, and Fas expression. Hence, MDSC-CD8+ T-cell interactions are more intricate than anticipated and include both inhibitory and stimulatory events.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

MO- and PMN-MDSCs differentially depend on IFN-γ, STAT-1, and IRF-1 to suppress T-cell proliferation

Previous studies suggested that MDSCs require IFN-γ to become T-cell suppressive [11]. To gain further insight in the dependence of MO- and PMN-MDSCs on IFN-γ and IFN-γ-activated transcription factors for their activation, EG7-OVA (where OVA is ovalbumin) tumors were grown in WT, IFN-γR−/−, STAT-1−/− (inducing “first wave” IFN-γ-dependent genes), and IRF-1−/− (inducing “second wave” IFN-γ-dependent genes [21]) mice. In all strains, splenic CD11b+CD115+Ly6GLy6Chigh MO-MDSCs and CD11b+CD115 Ly6G+Ly6Cint PMN-MDSCs were expanded in the course of tumor growth and MDSC subsets were purified from the spleen when tumors reached an approximate diameter of 15 mm (Supporting Information Fig. 1).

Upon coculture with OVA-stimulated TCR transgenic OT-1 splenocytes, WT MO-MDSCs suppressed T-cell proliferation in a dose-dependent manner, while IFN-γR−/− and STAT-1−/− MO-MDSCs almost completely lost their suppressive capacity (Fig. 1A and Supporting Information Fig. 2A for CFSE dilution). However, MO-MDSCs from IFN-γ–/– tumor-bearers were as suppressive as their WT counterparts (data not shown). These data illustrate that (i) there is an absolute requirement for IFN-γ/STAT-1 to activate the suppressive potential of splenic MO-MDSCs in vitro, and (ii) this does not rely on autocrine IFN-γ signaling. Interestingly, when treating MO-MDSCs with recombinant IFN-γ, only 72% of the population phosphorylates STAT-1, illustrating MO-MDSC heterogeneity and suggesting that only the IFN-γ-responsive part of this population mediates suppression (Supporting Information Fig. 3). Remarkably, IRF-1−/− MO-MDSCs retained a partial antiproliferative capacity (Fig. 1A), suggesting the coexistence of IRF-1-dependent and -independent suppressive mechanisms.

image

Figure 1. MO- and PMN-MDSCs differentially depend on IFN-γ, STAT-1, and IRF-1 to activate their antiproliferative capacity. (A) MO-MDSCs and (B) PMN-MDSCs were purified from WT, IFN-γR−/−, STAT-1−/−, and IRF-1−/− C57BL/6 mice bearing similarly sized EG7-OVA tumors. These MDSCs were added in various amounts to OVA-stimulated (250 μg/mL) OT-1 splenocytes and proliferation was measured after 42 h. The percentage suppression induced by various MDSC-SPC ratios is shown as individual replicates, measured via 3H-thymidine incorporation. Each symbol represents an individual replicate, bars represent means, and data are pooled from two to six independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant, Mann–Whitney test.

Download figure to PowerPoint

iNOS gene expression is IFN-γ/STAT-1/IRF-1-regulated [22]. Hence, IRF-1–/– MO-MDSCs were unable to produce NO (Fig. 2A(i)) and their T-cell suppressive capacity could not be reverted by the iNOS inhibitor l-NG-monomethyl arginine (l-NMMA) (Fig. 2A(ii)), corroborating the existence of parallel IRF-1/iNOS-dependent and -independent suppressive pathways. This conclusion is strengthened by the partial reduction in suppressive capacity by WT MO-MDSCs upon l-NMMA addition (Fig. 2A(ii)), and the fact that the NO-donor S-nitroso-N-acetyl-d,l-penicillamine (SNAP) could never decrease T-cell proliferation to the same extent as MO-MDSCs despite comparable NO concentrations in the culture (Fig. 2A(i) and (ii)).

image

Figure 2. MO-MDSC-mediated, but not PMN-MDSC-mediated, suppression of T-cell proliferation is partially IRF-1/NO-dependent. OVA-stimulated OT-1 splenocytes were cultured with medium alone (no MDSC); with 100 μM SNAP; or with WT, IFN-γR−/−, STAT-1−/−, IRF-1−/−, or iNOS−/− (A) MO- or (B) PMN-MDSCs (1:1 ratio), in the presence or absence of l-NMMA. At the end of the culture (42 h), (i) nitrite (NO2) concentrations, as read-out for NO, and (ii) corresponding percentage suppression of OVA-driven OT-1 proliferation was determined. Data in (i) and (ii) are shown as mean + SEM of at least two independent experiments, each comprising at least two individual replicates. **p < 0.005; ns, not significant, Wilcoxon-matched pairs test.

Download figure to PowerPoint

Conversely, IFN-γR−/−, STAT-1−/−, and IRF-1−/− PMN-MDSCs displayed an NO-independent suppressive capacity, which was moderately, but significantly, lower than WT cells (Fig. 1B and 2B(ii)). Again, IFN-γ−/− PMN-MDSC-mediated suppression was not hampered (data not shown). The relatively minor importance of IFN-γ is not due to a lack of IFN-γ responsiveness, since IFN-γ treatment of PMN-MDSCs uniformly phosphorylates STAT-1 (Supporting Information Fig. 3).

MO-MDSCs and especially PMN-MDSCs stimulate CD8+ T-cell IFN-γ production on a per cell basis

Though most often used as read-out for MDSC-mediated T-cell suppression, proliferation is only one aspect of early CD8+ T-cell activation. Cytokine secretion, activation marker expression, onset of proliferation, and acquisition of effector functions occur in sequential phases and are not necessarily interdependent [3, 4]. We first investigated the impact of splenic MDSC subsets on IFN-γ production by OVA-stimulated, CFSE-labeled OT-1 T cells, at 24 h (i.e. before the onset of proliferation) and 42 h following coculture initiation. By gating on viable CD8+ T cells in each proliferation cycle and intracellular IFN-γ staining (for gating strategy: Supporting Information Fig. 4A), we assessed IFN-γ production per cell, irrespective of the number of viable CD8+ T cells in the culture. At 24 h, MO-MDSCs did not influence IFN-γ production, while PMN-MDSCs significantly increased the percentage of IFN-γ+CD8+ T cells (Fig. 3A and B). Between 24 and 42 h, both MDSC subsets decreased the percentage of CD8+ T cells that have undergone cell divisions, in agreement with their antiproliferative capacity (Fig. 3A). However, the percentage of IFN-γ+CD8+ T cells in each division cycle was always significantly higher upon coculture with PMN-MDSCs and mostly also with MO-MDSCs (Fig. 3A and B). Overall, this resulted in equally high IFN-γ concentrations in the supernatant of MO-MDSC cocultures and a significantly increased IFN-γ level in PMN-MDSC cocultures at 42 h, compared with that of control cultures (Supporting Information Fig. 5). Notably, CD8+ T cells are the highest IFN-γ producers in these cocultures, while MDSCs did not produce this cytokine (data not shown). Interestingly, IFN-γR−/− and IRF-1−/− MDSCs retained their IFN-γ-inducing capacity (only MO-MDSCs from IRF-1−/− mice do not reach significance but clearly show a trend), suggesting this feature does not require prior activation by IFN-γ itself (Supporting Information Fig. 6).

image

Figure 3. MDSCs augment IFN-γ production in activated CD8+ T cells on a per cell basis. CFSE-labeled OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs at a 1:1 ratio. After 24 and 42 h of incubation, cells were subjected to intracellular IFN-γ staining. (A) The percentage of IFN-γ+ cells (rectangular gate) within the gated CD8+ population is indicated for each individual CFSE peak. The percentage of total CD8+ T cells, present in each individual CFSE peak, is indicated beneath each dot plot. Dot plots from one representative replicate out of nine are shown. (B) The percentage of IFN-γ+ CD8+ T cells in individual CFSE peaks is shown as mean + SEM of three experiments each comprising three individual replicates. *p < 0.05, **p < 0.005; ns, not significant, Wilcoxon-matched pairs test.

Download figure to PowerPoint

IL-12 and the IL-12-regulated transcription factor T-bet were shown before to enhance IFN-γ production by CD8+ T cells [7, 23-25], suggesting they could be involved in MDSC-mediated IFN-γ induction. However, IL-12 concentrations in the OVA-stimulated OT-1 cultures were low and did not increase upon addition of MO- or PMN-MDSCs (Supporting Information Fig. 7), arguing against a role for this cytokine. Moreover, PMN-MDSCs, and more variably also MO-MDSCs, repressed the activation-induced expression of T-bet in CD8+ T cells, thereby dissociating T-bet expression from IFN-γ production (Supporting Information Fig. 8).

Thus, splenic MDSCs are efficient suppressors of CD8+ T-cell proliferation, but stimulate their IFN-γ production on a per cell basis.

MO- and PMN-MDSCs differentially modulate IL-2 secretion, CD25 expression, and STAT-5 activation

Autocrine IL-2 production is essential for CD8+ T-cell activation [26], so we questioned whether this cytokine is also regulated by splenic MDSCs. IL-2 levels in the supernatant at 24 h were significantly reduced by MO-MDSCs, while, by 42 h, both IL-2 concentrations in the culture (Fig. 4A) and IL-2 production by CD8+ T cells (Supporting Information Fig. 9) were down-modulated by MO- and PMN-MDSCs. Hence, OT-1 IFN-γ and IL-2 production is oppositely regulated (upregulation of IFN-γ, downregulation of IL-2) by both MDSC subsets. However, the reduction in IL-2 availability is not sufficient to explain the antiproliferative effect of MDSCs, since recombinant IL-2 addition did not rescue T-cell proliferation (data not shown).

image

Figure 4. Both MDSC subsets reduce IL-2 production but only MO-MDSCs downregulate CD25 expression and STAT-5 phosphorylation. (A) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio). After 24 and 42 h, supernatant was collected and IL-2 levels were quantified by ELISA. Three individual experiments, each comprising two individual replicates, were performed for each condition. Bars represent means. *p < 0.05, **p < 0.005; ns, not significant, Wilcoxon matched pairs test. (B) OVA-stimulated OT-1 splenocytes were cultured with medium alone; with 100 μM SNAP; with WT, IFN-γR–/–, or iNOS–/– MDSCs (1:1 ratio); or with WT MDSCs (1:1 ratio) + 5 μM l-NMMA. After 24 and 42 h, CD8+ OT-1 T cells were gated and assayed for CD25 expression via flow cytometry. Normalized ΔMFI (median fluorescence intensity) for CD25 expression (= (MFI CD25 (MDSC) – MFI isotype (MDSC))/ (MFI CD25 (no MDSC) – MFI isotype (no MDSC)) is given for the different conditions. CD25 MFI is shown as mean ± SEM of at least three independent experiments each comprising at least two individual replicates. (C) OVA-stimulated OT-1 splenocytes were cocultured with MO- or PMN-MDSCs (1:1 ratio) and after 24 and 42 h of incubation, CD8+ OT-1 T cells were gated and assayed for pSTAT-5 expression via Phosflow. The percentage of pSTAT-5+ CD8+ T cells is given for the different conditions. MDSCs from three individual replicates were tested for each condition in a single experiment. Statistics in (B) and (C) were performed using the Mann–Whitney test. *p < 0.05, **p < 0.005, ***p < 0.0005; ns, not significant.

Download figure to PowerPoint

Besides IL-2 availability, the expression of the IL-2Rα (CD25) is needed for optimal IL-2 responsiveness [6]. MO-MDSCs, but not PMN-MDSCs, significantly downregulated CD25 expression on OVA-stimulated OT-1 CD8+ T cells at 24 and 42 h (Fig. 4B and Supporting Information Fig. 10A; for gating strategy: Supporting Information Fig. 4B). By adding l-NMMA, CD25 expression improved after 24 h and completely recovered after 42 h, illustrating a role for NO. In agreement, IFN-γR−/− and iNOS−/− MO-MDSCs did not modulate CD25 expression. Moreover, NO as single agent is sufficient to downregulate CD25 expression, since the presence of SNAP equals the effect of MO-MDSCs (Fig. 4B and Supporting Information Fig. 10A). Finally, in line with the effects on CD25 expression, MO-MDSCs, but not PMN-MDSCs, strongly diminish STAT-5 phosphorylation in CD8+ T cells after 24 and 42 h of stimulation (Fig. 4C and Supporting Information Fig. 10B).

MO- and PMN-MDSCs differentially affect the expres-sion of lymphocyte homing and adhesion molecules

We next evaluated whether activation/differentiation markers are differentially regulated by splenic MDSC subsets in activated CD8+ T cells, and whether, in analogy with cytokine secretion, the expression of some molecules is counteracted by MDSCs while others might be stimulated. CD69 and CD62L are both involved in the homing of T lymphocytes to lymphoid organs [1, 27]. CD69, the prototypic early activation marker, is not affected by MO- or PMN-MDSCs at 24 h, but is significantly enhanced by MO-MDSCs at 42 h as compared with that of control cultures (Fig. 5A and Supporting Information Fig. 10A). Moreover, both MO- and PMN-MDSCs at least partially prevent the CD62L downregulation normally seen upon CD8+ T-cell activation (Fig. 5B(i) and Supporting Information Fig. 11B(i)). Remarkably, addition of l-NMMA to WT MO-MDSCs or the use of IFN-γR−/− or iNOS−/− MO-MDSCs even further augmented CD62L expression, while SNAP strongly lowered CD62L levels (Fig. 5B(i) and Supporting Information Fig. 11B(i)). These data demonstrate that MO-MDSCs are intrinsically strong inhibitors of activation-induced CD62L downregulation, a feature that is somewhat tempered by their high secretion of the CD62L-lowering molecule NO. PMN-MDSCs, which do not produce NO, prevent CD62L downregulation to the same extent as MO-MDSCs.

image

Figure 5. MDSCs alter the expression levels of cell adhesion molecules and the adhesive properties of CD8+ T cells. (A) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio) and after 24 and 42 h, CD69 expression levels were measured on viable CD8+ T cells. Normalized ΔMFI for CD69 is shown for the different conditions and presented as the mean ± SEM of four independent experiments, each comprising three individual replicates. (B) OVA-stimulated OT-1 splenocytes were cultured with medium alone; with 100 μM SNAP; with WT, IFN-γR−/−, or iNOS−/− MDSCs (1:1 ratio); or with WT MDSCs (1:1 ratio) + 5 μM l-NMMA. After 42 h, viable CD8+ T cells were gated and the expression of (i) CD62L, (ii) CD44, and (iii) CD162 was measured. Normalized ΔMFI for the indicated markers is shown for the different conditions as the mean ± SEM of at least three independent experiments comprising at least one replicate. (C) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio) for 66 h, after which P-selectin binding activity was assessed via incubation with P-selectin-IgG followed by PE-anti-human IgG. (i) Representative flow cytometry histograms showing soluble P-selectin-IgG binding by viable CD8+ T cells in the presence of the indicated MDSC types. (ii) Normalized ΔMFI for P-selectin-IgG binding is shown as mean ± SEM of two independent experiments, each comprising three individual replicates. (D) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio) for 66 h, and CD8+ T cells were purified by FACS sorting. These cells were fluorescently labeled with DiD and added to P-selectin-IgG or HA-coated wells, after which unbound cells were removed by gentle washing and the remaining fluorescence (bound cells) was measured. Data were plotted as the percent change in fluorescence intensity comparing MDSC-conditioned T cells versus control T cells. Mean ± SEM of three independent experiments is shown. *p < 0.05, **p < 0.005, ***p < 0.0005; ns, not significant, Mann–Whitney test.

Download figure to PowerPoint

Other important adhesion molecules on activated CD8+ T cells are the hyaluronic acid (HA) receptor CD44, which mediates extravasation of activated T cells from blood to inflamed tissues [28], and CD162 (also known as PSGL-1), which functions as ligand for P- and E-selectin and contributes to T-cell rolling and entry into inflammatory sites [29]. While PMN-MDSCs do not affect CD44 expression, MO-MDSCs strongly inhibit its surface expression level (Fig. 5B(ii) and Supporting Information Fig. 11B(ii)). This is functionally relevant, since MO-MDSC-treated, but not PMN-MDSC-treated, CD8+ T cells show significantly reduced adhesion to HA (Fig. 5D). NO is partly responsible for this, as illustrated by a partial CD44 recovery upon addition of l-NMMA or the use of IFN-γR−/− or iNOS−/− MO-MDSCs. SNAP does not lower CD44 to the same extent as MO-MDSCs, corroborating the existence of other regulatory mechanisms (Fig. 5B(ii)). For CD162, MO-MDSCs suppress its surface expression in an entirely NO-dependent fashion, while PMN-MDSCs actually increase the expression of this molecule (Fig. 5B(iii)). These data are confirmed by labeling of the CD8+ T cells with a P-selectin-IgG construct (Fig. 5C). Moreover, MO-MDSC-treated T cells adhere less efficiently, while PMN-MDSC-treated cells increase their retention on coated P-selectin (Fig. 5D).

Hence, also at the level of activation/adhesion marker expression, splenic MDSC effects are complex and can be either inhibitory or stimulatory.

MO-MDSCs, but not PMN-MDSCs, render CD8+ T cells more susceptible to Fas-mediated apoptosis

Persistent TCR stimulation, together with IL-2 signals, can promote apoptosis of T cells, mainly through Fas-FasL (CD95-CD95L) interactions [6]. We therefore investigated whether splenic MDSC subsets are able to regulate Fas-mediated cell death in CD8+ T cells.

PMN-MDSCs did not modify Fas expression, while MO-MDSCs firmly increased its expression after 42 h (Fig. 6A and Supporting Information Fig. 12). In the absence of NO (l-NMMA, IFN-γR−/−, iNOS−/– MO-MDSCs), Fas is not induced. Conversely, SNAP increased Fas expression, although never to the levels seen with MO-MDSCs, suggesting that NO is crucial but not sufficient to maximize Fas expression on CD8+ T cells. To assess whether MO-MDSCs sensitize T cells to Fas-mediated apoptosis, the Fas agonistic antibody Jo2 or control antibody were added to the cocultures. Fas ligation massively induces CD8+ T-cell death in the presence of MO-MDSCs at 42 h, but not in any other condition, in agreement with the Fas expression data (Fig. 6B). These findings clearly illustrate that splenic MO-MDSCs further augment the activation-induced upregulation of Fas and sensitize CD8+ T cells to Fas-mediated apoptosis.

image

Figure 6. MO-MDSCs render activated CD8+ T cells more susceptible to Fas-mediated apoptosis by augmenting Fas expression. (A) OVA-stimulated OT-1 splenocytes were cultured with medium alone; with 100 μM SNAP; with WT, IFN-γR−/−, or iNOS−/− MDSCs (1:1 ratio); or with WT MDSCs (1:1 ratio) + 5 μM l-NMMA. After 42 h, viable CD8+ T cells were gated and CD95 expression was measured. Normalized ΔMFI for CD95 is shown for the different conditions as mean ± SEM of three individual replicates from at least one experiment. *p < 0.05, **p < 0.005, ***p < 0.0005; ns, not significant, Mann–Whitney test (B) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio). After 24 h, 1 μg/mL Fas agonistic Jo2 antibody or control antibody was added and after an additional 18 h, CFSE+ CD8+ T cells were gated and the level of apoptosis was measured. (i) Representative flow cytometry dot plots showing Annexin V and 7-amino-actinomycin staining on gated CFSE+CD8+ T cells upon mAb treatment. (ii) The percentage of viable, early apoptotic, and late apoptotic CD8+ cells upon treatment with either control antibody (white bars) or Jo2 antibody (black bars) is shown as mean ± SEM of six individual replicates, tested in two independent experiments. *p < 0.05, **p < 0.005, ***p < 0.0005; ns, not significant, Wilcoxon-matched pairs test.

Download figure to PowerPoint

MO-MDSCs, but not PMN-MDSCs, interfere with the differentiation of mature CTLs

Finally, we analyzed to which extent splenic MDSC subsets affect the cytotoxic activity of CD8+ T cells. One of the major pathways utilized by CTLs to eliminate target cells is via granzyme B exocytosis [8]. Following 3 days of OVA stimulation, PMN-MDSCs had no effect on the presence of granzyme B in the remaining viable OT-1 T cells, while MO-MDSCs significantly reduced its expression in those cells (Fig. 7A), suggesting that MO-MDSC-treated CD8+ T cells have a diminished killing capacity. Therefore, viable CD8+ T cells were purified from OVA-stimulated cocultures and their cytotoxic activity was assessed against EG7-OVA and control EL-4 cells. In agreement with the granzyme B data, only MO-MDSCs were able to strongly reduce antigen-specific cytotoxicity (Fig. 7B). When MO-MDSCs were only added during the 4 h effector phase, neither the effect on CTL cytotoxicity could be recorded (Supporting Information Fig. 13A), nor were the MO-MDSCs from EG7-OVA tumor bearers killed by the OVA-specific CTLs (Supporting Information Fig. 13B). These data show that, although both splenic MDSC subsets diminish the number of CTLs due to their antiproliferative effect, only MO-MDSCs also actively impede the formation of mature CTLs, but cannot obstruct the cytotoxic activity of existing mature CTLs.

image

Figure 7. MO-MDSCs, but not PMN-MDSCs, prevent the differentiation ofCTLs. (A) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio) and after 66 h, granzyme B content was measured in viable CD8+ T cells via intracellular flow cytometry. (i) Representative flow cytometry histograms showing granzyme B content in nonstimulated T cells (no OVA, no MDSC), OVA-stimulated T cells (no MDSC), and OVA-stimulated T cells in the presence of MDSCs (MDSC). (ii) Average ΔMFI for granzyme B as observed in four independent experiments, each comprising at least two individual replicates. Bar represents the mean. *p < 0.05; ns, not significant, Wilcoxon-matched paired test. (B) OT-1 splenocytes were stimulated with OVA in the presence of MO- or PMN-MDSCs (1:1 ratio) and after 66 h, CD8+ T cells were purified via FACS sorting. Purified cells were then cocultured for 4 h with 51Cr-labeled EL-4 or EG7-OVA target cells at a 1:50 target:effector ratio. The percent cytotoxicity is shown as mean + SEM from three independent experiments. *p < 0.05; ns, not significant, Mann–Whitney test.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

CD8+ T-cell activation and differentiation is a tightly regulated process, involving massive alterations in surface marker expression, cytokine secretion, and proliferative, migratory, and cytotoxic potential. Evidence exists that these features can be regulated independently from each other [3, 4], for example, upon interaction with immunoregulatory cells such as Treg cells [9]. MO- and granulocytic (PMN-) MDSCs both interfere with CD8+ T-cell proliferation [11, 12], but their effects on other features of early CD8+ T-cell activation are largely unknown. Here, we show that splenic MDSC subsets differentially modulate multiple aspects of CD8+ T-cell activation, encompassing both inhibitory and stimulatory effects, resulting in a distinct functional outcome (for overview: Supporting Information Table 1).

We first assessed whether MO- and PMN-MDSCs require IFN-γ-dependent triggering to become immunosuppressive (evaluating in first instance T-cell proliferation as most commonly used read-out) as this was claimed by some studies [11, 30, 31] but contested by others [32], a difference that might be explained by the tissue of origin [33] (inflammatory site or blood versus spleen). The use of splenic MDSCs from tumor-bearers throughout this study is in line with the central importance of this organ for inducing tolerance to tumor antigens [19]. IFN-γR−/− and STAT-1−/−, but not IFN-γ−/−, splenic MO-MDSCs induced by EG7-OVA largely lost their antiproliferative capacity, illustrating that suppression is entirely dependent on IFN-γ-mediated triggering by activated T cells, but not on IFN-γ production by the MDSC — as claimed before [31] — or IFN-γ priming in vivo. Interestingly, IRF-1-deficiency uncovers the existence of parallel IRF-1-dependent and -independent suppressive mechanisms in MO-MDSCs, both of which are needed to maximize suppression. IRF-1-dependent NO production is responsible for at least 50% of the suppression, but the IRF-1-independent mechanism remains unknown. Remarkably, also the PMN-MDSC-mediated suppressive mechanism is heterogeneous, with a minor IFN-γ/STAT-1/IRF-1-dependent component and a major IFN-γ-independent mechanism. Since different pathological conditions — including different tumor types [12] — preferentially expand one or the other MDSC subset, these data suggest that different intervention strategies might be needed to ablate suppression in different settings. In the case of EG7-OVA, the total splenic MDSC population contains approximately 40% MO-MDSCs (both before and after purification), but these cells appear to dominate since the suppressive mechanism of unseparated MDSCs largely depends on NO (Supporting Information Fig. 14). Finally, it should be noted that these findings are not confined to the EG7-OVA model. Indeed, RMA-OVA-induced splenic MO-MDSCs from WT mice suppress T-cell proliferation in a dose-dependent and largely NO-dependent fashion, while IFN-γR−/− MO-MDSCs lack this activity (Supporting Information Fig. 15). PMN-MDSCs display a lower T-cell antiproliferative capacity in this model, which is partly dependent on IFN-γ signaling and independent from NO.

Proliferation is a relatively late event in the course of CD8+ T-cell activation, preceded by the secretion of cytokines such as IL-2 and IFN-γ, and the expression of early activation markers such as CD69 and CD25 [3]. Our data now demonstrate that MDSCs manipulate early activation events in an intricate way — suppressing some aspects, while stimulating others — to optimize T-cell suppression. Most literature, with some exceptions [31], suggests that MDSCs suppress IFN-γ production, but those data are often confounded by the antiproliferative effect of MDSCs resulting in lower T-cell numbers. Via intracellular IFN-γ staining, we demonstrated that IFN-γ production by CD8+ T cells is enhanced on a per cell basis in the presence of splenic PMN-MDSCs already before the initiation of proliferation, and the percentage of IFN-γ+CD8+ T cells remains enhanced throughout each division cycle. This makes sense from the MDSC point of view, since IFN-γ initiates their antiproliferative program. PMN-MDSCs are superior IFN-γ inducers, while the MO-MDSC-mediated suppression of proliferation depends more on IFN-γ, hinting to a collaborative effort between the subsets. Notably, the IFN-γ-inducing effect of splenic MDSCs is also clearly visible upon polyclonal (anti-CD3 + anti-CD28) T-cell activation, again with a predominant role for PMN-MDSCs, illustrating that antigen-specific contacts between MDSCs and T cells are not required (Supporting Information Fig. 16). Interestingly, however, the IFN-γ induction by MDSCs might be more prominent in the spleen as compared with that at the tumor site. Indeed, employing the Lewis Lung Carcinoma (LLC) model, tumor-infiltrating MO-MDSCs were shown to be strongly antiproliferative (to a large extent in an NO-independent fashion, data not shown) and did not allow for IFN-γ production (Supporting Information Fig. 17). By contrast, their splenic counterparts stimulated IFN-γ production on a per cell basis, even though being antiproliferative through NO, thus phenocopying EG7-OVA-induced splenic MO-MDSCs. Along the same line, splenic MDSCs (both MO- and PMN-MDSCs) induced by RMA-OVA tumor growth tended to induce IFN-γ production by OT-1 CD8+ T cells (Supporting Information Fig. 15). Finally, unseparated MDSCs from EG7-OVA tumor-bearers also enhanced IFN-γ production at an early time point (Supporting Information Fig. 14). The exact mechanism of splenic MDSC-mediated IFN-γ induction remains speculative at present, but seems not to be mediated by IL-12 or T-bet. Other IFN-γ-inducing cytokines include IL-18, IL-23, IL-15, and IL-21 and could be tested for their involvement in future experiments. Alternatively, monocytes and neutrophils might provide costimulatory signals for CD8+ T cells [34], as such contributing to the induction of IFN-γ.

Interestingly, IL-2 secretion is lowered by both MDSC types from the spleen. Since IL-2 is critical for primary T-cell expansion, this strategy also fits in the antiproliferative program of MDSCs. In addition, downstream events of IL-2, such as CD25 expression and STAT-5 phosphorylation, are significantly inhibited by MO-, but not PMN-MDSCs, in an NO-dependent fashion, possibly explaining MO-MDSC's superior antiproliferative capacity. Previously, immortalized myeloid suppressor lines were reported to affect IL-2R signaling [35], and our data extend these findings to primary MDSCs.

Moreover, we report an influence of splenic MO-MDSCs on the expression of several functionally important CD8+ T-cell activation markers, with a varying implication of NO. Of note, some activation markers are not affected by the presence of MDSCs, indicating that these cells do not cause an overall shut-down of T-cell activation, but rather target certain aspects of the T cell. For example, upregulation of the early activation marker CD69 is not prevented, and in the case of MO-MDSCs even stimulated at later time points. CD69 prevents T-cell egress from lymphoid organs by inhibiting the sphingosine-1-phosphate receptor-1 [27], suggesting that MO-MDSCs might hamper activated T cells in their migration to inflammatory sites (such as a tumor). CD62L also favors homing of T cells to lymphoid organs, and its downregulation accompanies T-cell activation and entry into nonlymphoid tissues [36]. Earlier findings reported that MDSCs could downregulate CD62L expression to some extent on naive T cells [37], but their effect on activated T cells was not reported. Both MDSC subsets partially counteract CD62L shedding on Ag-stimulated CD8+ T cells, again suggesting that these cells might lower the emigration of (tumor-reactive) CD8+ T lymphocytes from the spleen or LNs. Notably, NO strongly favors CD62L downregulation, suggesting that MO-MDSCs utilize a mechanism that counteracts their own NO production. In addition, MO-, but not PMN-MDSCs, cause a downregulation of CD44 and CD162 expression and a reduced adhesion to HA and P-selectin, which are both required for entry of effector cells into the inflammatory site [28, 29]. CD44 expression is only partly recovered when MO-MDSCs are unable to produce NO (l-NMMA, iNOS−/−) or are unresponsive to IFN-γ (IFN-γR−/−), while CD162 downregulation is entirely NO-dependent. Possible working mechanisms of NO include tyrosine nitrosylation or guanylate cyclase activation in T cells [38]. Another level of NO activity is its inactivation of the transcription repressor Yin-Yang 1, thereby releasing Fas expression, for example, in cancer cells [39]. Similarly, MO-MDSCs upregulate Fas expression on activated CD8+ T cells, sensitizing them to Fas-mediated apoptosis. This proapoptotic mechanism might be complementary to the reported NO-dependent cytochrome c release, which also induces apoptosis [40]. Together, these data could explain the increased level of T-cell apoptosis seen in the presence of MO-MDSCs or their progeny [41, 42]. Of note, several of these effects (CD25 downregulation in an NO-dependent fashion, CD44 downregulation in an NO-independent fashion, CD95 upregulation in an NO-dependent fashion) were recapitulated using (i) unseparated EG7-OVA-induced splenic MDSCs (Supporting Information Fig. 14), and (ii) LLC-induced splenic MO-MDSCs and their tumor-infiltrating counterparts, although the latter depended less on NO, despite their equally high NO production level (Supporting Information Fig. 17). Moreover, also RMA-OVA-induced splenic MO- and PMN-MDSCs regulated CD25, CD44, and CD95 in a similar way as EG7-OVA-induced MDSCs, providing evidence that this mechanism can be extrapolated to several models (Supporting Information Fig. 15). Importantly, upon polyclonal T-cell stimulation, MO-MDSCs produce less NO and do not affect CD25 and CD95 expression, suggesting that either threshold levels of NO or antigen-driven T-cell activation are required for these effects to take place (Supporting Information Fig. 16). Finally, we demonstrated that viable CD8+ T cells that had been conditioned by EG7-OVA-induced MO-, but not PMN-MDSCs, completely lost their cytotoxic capacity. At least part of this defect is due to a significantly reduced level of granzyme B in secretory vesicles, although we cannot exclude additional defects at the level of degranulation.

Overall, we demonstrate that splenic MO-MDSCs affect multiple aspects of early CD8+ T-cell activation: reduced T-cell proliferation, enhanced IFN-γ production, reduced IL-2 responsiveness, enhanced expression of lymphoid organ retention signals, reduced expression of extravasation signals, enhanced sensitivity for apoptosis, and reduced expression of cytotoxic molecules. PMN-MDSCs have more subtle effects, the most prominent of which being the stimulation of IFN-γ production by CD8+ T cells. These results demonstrate that MDSCs are fully equipped to efficiently reduce CTL-mediated antitumor immunity.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

Mice

Female C57BL/6 mice were from Janvier. IFN-γR−/− and IRF-1−/− mice were a gift of Dr. Peter Brouckaert (UGent, Belgium). STAT-1−/− and OT-1 TCR transgenic mice were provided by Dr. Chantal Mathieu (KULeuven, Belgium) and Dr. Kristiaan Thielemans (VUB, Belgium). Procedures followed the guidelines of the Belgian Council for Laboratory Animal Science.

Cell line and tumor growth

EG7-OVA is an OVA-transfected EL-4 thymoma and RMA-OVA is an OVA-transfected RMA thymoma. Cells were cultured in RPMI with 10% FCS, 0.03% l-glutamine, 100 mg/mL streptomycin, 100 mg/mL penicillin (Invitrogen).

Mice were injected subcutaneously with 3 × 106 EG7-OVA, RMA-OVA, or LLC and sacrificed when average tumor diameters reached 15 mm.

Flow cytometry

Antibodies are presented in Supporting Information Table 2. Dead cells were excluded via 7-amino-actinomycin (BD Bioscience). P-selectin-IgG stainings were performed by resuspending the cells in IMDM + 2% FCS. Intracellular pSTAT-1 and pSTAT-5 stainings were performed using Phosflow Perm buffer III, according to the manufacturer's instructions (BD Bioscience). Intracellular IFN-γ, IL-2, T-bet, and granzyme B stainings were performed using Cytoxic/Cytoperm (BD Biosciences) following the manufacturer's instructions (BD Bioscience). For IFN-γ and IL-2, the cells were pretreated with Brefeldin A (4 h). Data were acquired on a FACSCanto II (BD Biosceince) and analyzed by FlowJo (Tree Star).

Cell isolation

MDSC subsets or unseparated MDSCs were purified from the spleen of tumor-bearers as described [11]. To purify tumor-infiltrating MO-MDSCs, LLC tumors were dissociated with 10 U/mL collagenase I, 400 U/mL collagenase IV, and 30 U/mL DNaseI (Worthington). Density gradients (Axis-Shield) were used to remove debris and dead cells. Next, CD11b+ cells were MACS-enriched (anti-CD11b microbeads, Miltenyi Biotec) followed by FACSorting of MO-MDSCs using a BD FACSAria II (BD Biosciences). OT-1 T cells were purified from MDSC/OT-1 cocultures using FACS sorting.

OT-1 T-cell activation and suppression

OT-1 splenocytes were stained with 0.2 μM CFSE (Molecular Probes) following the manufacturer's instructions. Purified MDSCs were added to 2 × 105 OT-1 splenocytes and stimulated with 250 μg/mL OVA (Sigma-Aldrich) ± l-NMMA (NG-monomethyl-l-arginine, Alexis Biochemicals, 5 μM) or SNAP, Molecular Probes, 100 μM added after 0 and 24 h). For 3H-thymidine incorporation, 1 μCi 3H-thymidine (Amersham) was added after 24 h and cells proliferated for another 18 h. For transwell assays (0.4 μm pores, Sigma-Aldrich), either 3 × 105 CFSE-labeled splenocytes were in plate wells and 3 × 105 MDSCs in transwell inserts; or splenocytes were in plate wells and MDSCs + splenocytes (1:1) in transwell inserts. After 42 h, proliferation of T-cells in the plate wells was measured.

Antibody-induced apoptosis

Fas-agonistic Jo2 or control mAb (1 μg/mL) (BD Biosciences) were added to cultures after 18 h. After another 24 h, apoptosis was determined using 7-amino-actinomycin and AnnexinV staining (BD Bioscience).

Cytokines, NO2 quantification

IFN-γ and IL-2 were quantified using sandwich ELISAs (PharMingen), IL-12p70 by the Bio-plex ProTM kit (Bio-Rad) on the Bioplex 200 system (Bio-Rad). NO2 was measured using Greiss reagent as described [43].

OT-1 adhesion

Ninety-six-well microtiter plates (Nunc) were coated overnight (4°C) with HA (50 μg/mL) (Sigma-Aldrich), P-selectin-IgG (BD Pharmingen), or control IgG (10 μg/mL). Wells were blocked with 1% dry milk (2 h, 37°C). DiD-labeled CD8+ OT-1 T cells were resuspended in appropriate binding buffer (P-selectin-binding: IMDM + 2% FCS; HA-binding: RPMI1640, 40 mM Hepes, 0.1% BSA, 2 mM MgCl2), added to the plates, subjected to a short spin, and incubated (30 min, 37°C). Nonadherent cells were removed by gentle washing. Bound cells were quantified by a FLUOstar OPTIMA fluorescence plate reader (BMG Labtech).

T-cell cytotoxicity

Cytotoxicity of CD8+ T cells was tested using a 4-h 51Cr-release assay. Spontaneous lysis was measured by incubating target cells only with medium, maximal lysis by incubating with 10% saponin.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

This work was supported by a doctoral grant from FWO-Vlaanderen to E.S. and K.M., by a doctoral grant from IWT-Vlaanderen to D.L. and Y.M., and by research grants from “Stichting tegen Kanker” and “Vlaamse Liga tegen Kanker” to P.D.B. and J.A.V.G. The authors also thank Ella Omasta, Marie-Thérèse Detobel, Maria Slazak, Nadia Abou, and Eddy Vercauteren for technical and administrative assistance.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information
  • 1
    Arens, R. and Schoenberger, S. P., Plasticity in programming of effector and memory CD8+ T-cell formation. Immunol. Rev. 2010. 235: 190205.
  • 2
    Lefrançois, L. and Obar, J. J., Once a killer, always a killer: from cytotoxic T cell to memory cell. Immunol. Rev. 2010. 235: 206218.
  • 3
    Mempel, T. R., Henrickson, S. E. and Von Andrian, U. H., T-cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature 2004. 427: 154159.
  • 4
    Carr, J. M., Carrasco, M. J., Thaventhiran, J. E. D., Bambrough, P. J., Kraman, M., Edwards, A. D., Al-Shamkhani, A. et al., CD27 mediates interleukin-2-independent clonal expansion of the CD8+ T cell without effector differentiation. Proc. Natl. Acad. Sci. USA 2006. 103: 1945419459.
  • 5
    Kaech, S. M. and Wherry, E. J., Heterogeneity and cell-fate decisions in effector and memory CD8+ T-cell differentiation during viral infection. Immunity 2007. 27: 393405.
  • 6
    Boyman, O. and Sprent, J., The role of interleukin-2 during homeostasis and activation of the immune system.Nat. Rev. Immunol. 2012. 12: 180190.
  • 7
    Joshi, N. S., Cui, W., Chandele, A., Lee, H. K., Urso, D. R., Hagman, J., Gapin, L. et al., Inflammation directs memory precursor and short-lived effector CD8+ T cell fates via the graded expression of T-bet transcription factor. Immunity 2007. 27: 281295.
  • 8
    Wolint, P., Betts, M. R., Koup, R. A. and Oxenius, A., Immediate cytotoxicity but not degranulation distinguishes effector and memory subsets of CD8+ T cells. J. Exp. Med. 2004. 199: 925936.
  • 9
    Mempel, T. R., Pittet, M. J., Khazaie, K., Weninger, W., Weissleder, R., Boehmer von, H. and Andrian von, U. H., Regulatory T cells reversibly suppress cytotoxic T cell function independent of effector differentiation. Immunity 2006. 25: 129141.
  • 10
    Van Ginderachter, J. A., Meerschaut, S., Liu, Y., Brys, L., De Groeve, K., Hassanzadeh Ghassabeh, G., Raes, G. et al., Peroxisome proliferator-activated receptor gamma (PPARgamma) ligands reverse CTL suppression by alternatively activated (M2) macrophages in cancer. Blood 2006. 108: 525535.
  • 11
    Movahedi, K., Guilliams, M., Van den Bossche, J., Van den Bergh, R., Gysemans, C., Beshin, A., De Baetselier, P. et al., Identification of discrete tumor-induced myeloid-derived suppressor cell subpopulations with distinct T cell-suppressive activity. Blood 2008. 111: 42334244.
  • 12
    Youn, J. I., Nagaraj, S., Collazo, M. and Gabrilovich, D. I., Subsets of myeloid-derived suppressor cells in tumor-bearing mice. J. Immunol. 2008. 181: 57915802.
  • 13
    Youn, J. I., Collazo, M., Shalova, I. N., Biswas, S. K. and Gabrilovich, D. I., Characterization of the nature of granulocytic myeloid-derived suppressor cells in tumor-bearing mice. J. Leukoc. Biol. 2012. 91: 167181.
  • 14
    Gabrilovich, D. I., Ostrand-Rosenberg, S. and Bronte V., Coordinated regulation of myeloid cells by tumours. Nat. Rev. Immunol. 2012. 12: 253268.
  • 15
    Van Ginderachter, J. A., Beschin, A., De Baetselier, P. and Raes, G., Myeloid-derived suppressor cells in parasitic infections. Eur. J. Immunol. 2010. 40: 29762985.
  • 16
    Luyckx, A., Schouppe, E., Rutgeers, O., Lenaerts, C., Koks, C., Fevery, S., Devos, T. et al., Subset characterization of myeloid-derived suppressor cells arising during induction of BM chimerism in mice. Bone Marrow Transplant. 2012. 47: 985992.
  • 17
    Marigo, I., Bosio, E., Solito, S., Mesa, C., Fernandez, A., Dolcetti, L., Ugel, S. et al., Tumor-induced tolerance and immune suppression depend on the C/EBPbeta transcription factor. Immunity 2010. 32: 790802.
  • 18
    Ostrand-Rosenberg, S., Myeloid-derived suppressor cells: more mechanisms for inhibiting antitumor immunity. Cancer Immunol. Immunother. 2010. 59: 15931600.
  • 19
    Ugel, S., Peranzoni, E., Desantis, G., Chioda, M., Walter, S., Weinschenk, T., Ochando, J. C. et al., Immune tolerance to tumor antigens occurs in a specialized environment of the spleen. Cell Rep. 2012. 2: 628639.
  • 20
    Nagaraj, S., Schrum, A. G., Cho, H.-I., Celis, E. and Gabrilovich, D. I., Mechanism of T cell tolerance induced by myeloid-derived suppressor cells. J. Immunol. 2010. 184: 31063116.
  • 21
    Kröger, A., Köster, M., Schroeder, K., Hauser, H. and Mueller, P. P., Activities of IRF-1. J. Interferon Cytokine Res. 2002. 22: 514.
  • 22
    Kamijo, R., Harada, H., Matsuyama, T., Bosland, M., Gerecitano, J., Shapiro, D., Le, J. et al., Requirement for transcription factor IRF-1 in NO synthase induction in macrophages. Science 1994. 263: 16121615.
  • 23
    Sullivan, B. M., Juedes, A., Szabo, S. J., von Herrath, M. and Glimcher, L. H., Antigen-driven effector CD8 T cell function regulated by T-bet. Proc. Natl. Acad. Sci. USA 2003. 100: 1581815823.
  • 24
    Wilson, D. C., Matthews, S. and Yap, G. S., IL-12 signaling drives CD8+ T cell IFN-γ production and differentiation of KLRG1+ effector subpopulations during Toxoplasma gondii infection. J. Immunol. 2008. 180: 59355945.
  • 25
    Yeo, C. J. J. and Fearon, D. T., T-bet-mediated differentiation of the activated CD8+ T cell. Eur. J. Immunol. 2011. 41: 6066.
  • 26
    Feau, S., Arens, R., Togher, S. and Schoenberger, S. P., Autocrine IL-2 is required for secondary population expansion of CD8+ memory T cells. Nat. Immunol. 2011. 12: 908913.
  • 27
    Shiow, L. R., Rosen, D. B., Brdickova, N., Xu, Y., An, J., Lanier, L. L., Cyster, J. G. et al., CD69 acts downstream of interferon-alpha/beta to inhibit S1P1 and lymphocyte egress from lymphoid organs. Nature 2006. 440: 540544.
  • 28
    DeGrendele, H. C., Estess, P. and Siegelman, M. H., Requirement for CD44 in activated T cell extravasation into an inflammatory site. Science 1997. 278: 672675.
  • 29
    Alon, R., Rossiter, H., Wang, X., Springer, T. A. and Kupper, T. S., Distinct cell surface ligands mediate T lymphocyte attachment and rolling on P and E selectin under physiological flow. J. Cell Biol. 1994. 127: 14851495.
  • 30
    Kusmartsev, S., Li, Y. and Chen, S.-H., Gr-1+ myeloid cells derived from tumor-bearing mice inhibit primary T cell activation induced through CD3/CD28 costimulation. J. Immunol. 2000. 165: 779785.
  • 31
    Gallina, G., Dolcetti, L., Serafini, P., De Santo, C., Marigo, I., Colombo, M. P., Basso, G. et al., Tumors induce a subset of inflammatory monocytes with immunosuppressive activity on CD8+ T cells. J. Clin. Invest. 2006. 116: 27772790.
  • 32
    Sinha, P., Parker, K. H., Horn, L. and Ostrand-Rosenberg, S., Tumor-induced myeloid-derived suppressor cell function is independent of IFN-γ and IL-4Rα. Eur. J. Immunol. 2012. 42: 20522059.
  • 33
    Haverkamp, J. M., Crist, S. A., Elzey, B. D., Cimen, C. and Ratliff, T. L., In vivo suppressive function of myeloid-derived suppressor cells is limited to the inflammatory site. Eur. J. Immunol. 2011. 41: 749759.
  • 34
    Sandilands, G. P., McCrae, J., Hill, K., Perry, M. and Baxter, D., Major histocompatibility complex class II (DR) antigen and costimulatory molecules on in vitro and in vivo activated human polymorphonuclear neutrophils. Immunology 2006. 119: 562571.
  • 35
    Mazzoni, A., Bronte, V., Visintin, A., Spitzer, J. H., Apolloni, E., Serafini, P., Zanovello, P. et al., Myeloid suppressor lines inhibit T cell responses by an NO-dependent mechanism. J. Immunol. 2002. 168: 689695.
  • 36
    Galkina, E., Tanousis, K., Preece, G., Tolaini, M., Kioussis, D., Florey, O., Haskard, D. O. et al., L-selectin shedding does no regulate constitutive T cell trafficking but controls the migration pathways of antigen-activated T lymphocytes. J. Exp. Med. 2003. 198: 13231335.
  • 37
    Hanson, E. M., Clements, V. K., Sinha, P., Ilkovitch, D. and Ostrand-Rosenberg, S., Myeloid-derived suppressor cells downregulate L-selectin expression on CD4+ and CD8+ T cells. J. Immunol. 2009. 183: 937944.
  • 38
    Bingisser, R. M., Tilbrook, P. A., Holt, P. G. and Kees, U. R., Macrophage-derived nitric oxide regulates T cell activation via reversible disruption of the Jak3/STAT5 signaling pathway. J. Immunol. 1998. 160: 57295734.
  • 39
    Garban, H. J. and Bonavida, B., Nitric oxide inhibits the transcription repressor Yin-Yang 1binding activity at the silencer region of the Fas promoter: a pivotal role for nitric oxide in the upregulation of Fas gene expression in human tumor cells. J. Immunol. 2001. 167: 7581.
  • 40
    Ushmorov, A., Ratter, F., Lehmann, V., Dröge, W., Schirrmacher, V. and Umansky, V., Nitric-oxide-induced apoptosis in human leukemic lines requires mitochondrial lipid degradation and cytochrome c release. Blood 1999. 93: 23422352.
  • 41
    Kusmartsev, S. and Gabrilovich, D. I., STAT1 signaling regulates tumor-associated macrophage-mediated T cell deletion. J. Immunol. 2005. 174: 48804891.
  • 42
    Apolloni, E., Bronte, V., Mazzoni, A., Serafini, P., Cabrelle, A., Segal, D. M., Young, H. A. et al., Immortalized myeloid suppressor cells trigger apoptosis in antigen-activated T lymphocytes. J. Immunol. 2000. 165: 67236730.
  • 43
    Liu, Y., Van Ginderachter, J. A., Brys, L., De Baetselier, P., Raes, G. and Geldhof, A. B., Nitric-oxide independent CTL suppression during tumor progression: association with arginase-producing (M2) myeloid cells. J. Immunol. 2003. 170: 50645074.
Abbreviations
HA

hyaluronic acid

IRF-1

interferon regulatory factor 1

LLC

Lewis lung carcinoma

l-NMMA

l-NG-monomethyl arginine

MDSC

myeloid-derived suppressor cell

MO-MDSC

monocytic MDSC

PMN-MDSC

polymorphonuclear MDSC

SNAP

S-nitroso-N-acetyl-d,l-penicillamine

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
eji2738-sup-0001-S1.pdf2687K

Table S1. Overview of the effects of MO- and PMN-MDSCs on various aspects of CD8+ T-cell activation.

Table S2. List of commercial antibodies used for flow cytometry

Figure S1. MO- and PMN-MDSCs were purified from the spleens of EG7-OVA tumor-bearing WT, IFN-γR-/-, STAT-1-/- and IRF-1-/- mice.

Figure S2. MO- and PMN-MDSCs differentially depend on IFN-γ, STAT-1 and IRF-1 to activate their anti-proliferative capacity.

Figure S3. MO-MDSCs respond more heterogeneously to an IFN-γ challenge than PMN-MDSCs, as measured by STAT-1 phosphorylation.

Figure S4. Gating strategy on CD8+ OT-1 T cells after 24 h and 42 h culture.

Figure S5. PMN-MDSCs increase IFN-γ secretion levels upon co-culture with OVA-stimulated OT-1 splenocytes.

Figure S6. IFN-γR-/- and IRF-1-/- MDSCs enhance IFN-γ production by activated CD8+ T cells on a per cell basis.

Figure S7. MO- and PMN-MDSCs do not augment IL-12 levels upon co-culture with OVA-stimulated OT-1 splenocytes.

Figure S8. MO- and especially PMN-MDSCs suppress T-bet expression in activated CD8+ T cells.

Figure S9. MDSCs down-modulate IL-2 production by activated CD8+ T cells.

Figure S10. MO-MDSCs down-regulate CD25 expression and STAT5 phosphorylation.

Figure S11. MDSCs alter the expression levels of cell adhesion molecules on CD8+ T cells.

Figure S12. MO-MDSCs augment Fas expression on activated CD8+ T cells.

Figure S13. Neither MO- nor PMN-MDSCs are targets for OVA-specific CTLs, nor do they affect the cytotoxic activity of mature CTLs.

Figure S14. Unseparated splenic MDSCs affect CD8+ T-cell activation events.

Figure S15. RMA-OVA-induced splenic MDSCs affect CD8+ T-cell activation events.

Figure S16. MDSCs differentially affect CD8+ T-cell activation events upon polyclonal stimulation.

Figure S17. Tumor-infiltrating MO-MDSCs are strongly anti-proliferative and recapitulate only some aspects of their splenic counterparts.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.