TLR9 activation of Stat3 constrains its agonist-based immunotherapy (original) (raw)

Cancer Res. Author manuscript; available in PMC 2010 Mar 15.

Published in final edited form as:

PMCID: PMC2657819

NIHMSID: NIHMS89579

Marcin Kortylewski,1,3,4 Maciej Kujawski,1,3 Andreas Herrmann,1 Chunmei Yang,1 Lin Wang,1 Yong Liu,1 Rosalba Salcedo,2 and Hua Yu1,4

Marcin Kortylewski

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

Maciej Kujawski

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

Andreas Herrmann

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

Chunmei Yang

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

Lin Wang

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

Yong Liu

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

Rosalba Salcedo

2 Cancer Inflammation Program, SAIC-NCI, Frederick, MD21701, USA

Hua Yu

1 Division of Cancer Immunotherapeutics & Tumor Immunology, Beckman Research Institute at City of Hope, Duarte, CA91010, USA

2 Cancer Inflammation Program, SAIC-NCI, Frederick, MD21701, USA

4Correspondence should be addressed to Hua Yu or Marcin Kortylewski. Hua Yu, Ph.D., 1500 East Duarte Rd., Duarte, CA91010, Phone: (626)256-4673 ext.63365, Fax: (626)256-8708, E-mail: gro.hoc@uyh, Marcin Kortylewski, Ph.D., 1500 East Duarte Rd., Duarte, CA91010, Phone: (626)256-4673 ext.64120, Fax: (626)256-8708, E-mail: gro.hoc@ikswelytrokm

3These authors contributed equally

Supplementary Materials

figure 1.

GUID: 82FC1EE4-3CED-418B-ACF8-1A61EE5F8B3C

figure 2.

GUID: 0EEB5401-8A80-4848-A728-2405DE3F34CC

figure 3.

GUID: 8B86B8AC-2DEF-4975-A5BF-7C88AE5DECD5

figure 4.

GUID: FCFC25AD-7F93-42DD-ADE0-42A5218AB3D8

figure 5.

GUID: E1635363-FCD5-48B5-8716-8D81DBA456D8

figure 6.

GUID: 39F2FEFA-D702-439F-8D48-9B0FAE4684FE

figure 7.

GUID: C31AF0A6-EDD2-4F94-BB5A-93107D057ED6

Abstract

Although toll-like receptor (TLR) agonists, such as CpG, are used as immunotherapeutic agents in clinical trials for cancer and infectious diseases, their effects are limited and the underlying mechanism(s) that restrains CpG efficacy remains obscure. Here we demonstrate that signal transducer and activator of transcription 3 (Stat3) plays a key role in downmodulating CpG’s immunostimulatory effects. In the absence of IL-6 and IL-10 induction, CpG directly activates Stat3 within minutes through TLR9. Ablating Stat3 in hematopoietic cells results in rapid activation of innate immunity by CpG, with enhanced production of interferon-γ, tumor necrosis factor-α, interleukin-12, and activation of macrophages, neutrophils and natural killer cells marked with Stat1 activation. Innate immune responses induced by CpG in mice with a _Stat3_-ablated hematopoietic system cause potent antitumor effects, leading to eradication of large (> 1 cm) B16 melanoma tumors within 72h. Moreover, ablating Stat3 in myeloid cells increases CpG-induced dendritic cell maturation, T cell activation, generation of tumor antigen-specific T cells and long-lasting antitumor immunity. A critical role of Stat3 in mediating immunosuppression by certain cytokines and growth factors in the tumor microenvironment has been recently documented. By demonstrating direct and rapid activation of Stat3 by TLR agonists, we identify a second level of Stat3-mediated immunosuppression. Our results further suggest that targeting Stat3 can drastically improve CpG-based immunotherapeutic approaches.

Keywords: Stat3, TLR9, CpG, tumor, immunotherapy

INTRODUCTION

Toll-like receptor (TLR) activation, originally studied in the context of microbial infections, is an attractive strategy for boosting prophylactic vaccines and for anticancer therapies (1). The simplicity of the natural ligands for TLR9 - fragments of unmethylated double-stranded DNA, allows for large scale synthesis of TLR9 agonists optimized for stability, cell type- and species-specific activation (2). Based on numerous studies in mice demonstrating their efficacy in breaking immune tolerance (3, 4) and promoting Th1-dependent immunity (2), several TLR9 agonists are currently used to treat solid and hematologic malignancies (1). While TLR9 agonist-based therapies have no major toxicity and show some promise, the efficacy of the TLR9-mediated anti-tumor immune response is limited, which is largely attributed to immunosuppression observed in cancer patients (1, 5).

It has been reported that antibody-mediated blockade of IL-10 restores TLR9-induced activation of tumor-infiltrating dendritic cells (DCs), IL-12 production and generation of antitumor immune memory (6). IL-10 is frequently overly produced in tumors and is a key activator of signal transducer and activator of transcription 3 (Stat3), which is activated in numerous cancers, promoting expression of diverse molecules important for tumor cell proliferation, survival, invasion and metastasis (7). Recently, Stat3 was identified as an important mediator of tumor-induced immunosuppression at many levels (8–10). Several immunosuppressive factors commonly found in the tumor microenvironment, such as IL-10, IL-6 and VEGF (9, 11), are known to activate Stat3, leading to propagation of Stat3 activity from tumor cells to diverse immune cells infiltrating the tumor, and affecting innate and adaptive antitumor immunity (9, 12, 13). Elevated Stat3 activity in tumor stromal immune cells further promotes expression of cytokines and growth factors, including but not limited to IL-10, IL-6 and VEGF, which in turn further activate Stat3 in tumor cells (9, 12, 13). This multi-directional Stat3 propagation facilitates the crosstalk between tumor cells and diverse immune cells in the tumor milieu, forming an immunosuppressive network (13). Dendritic cells are among the immune cells affected by Stat3 activity, which not only inhibits expression of Th1 cytokines such as IL-12 and IFNγ, but also reduces the levels of MHC Class II, CD86 and CD80, contributing to the accumulation of toleragenic DCs and Tregs in the tumor (9, 12, 13). While these observations document an important role for Stat3 in mediating immunosuppression in the tumor microenvironment, it remains to be further explored whether or not Stat3 restrains TLR9 activation.

What has become evident is that TLR agonists, such as CpG oligodeoxynucleotides (ODN), can downmodulate their own activity, which likely represents a physiological mechanism for limiting collateral damage during infection (14). However, the mechanisms by which TLR agonists restrain TLR activity remain poorly defined. The current study tests the hypothesis that Stat3 constrains TLR activity and that targeting Stat3 can improve the efficacy of TLR9-based immunotherapy. We demonstrate that both TLR9 and TLR4 agonists activate Stat3 rapidly and continuously, which can occur in the absence of IL-6 and IL-10 induction. Furthermore, activation of Stat3 by CpG is abrogated in _TLR9_−/− immune cells. Removing Stat3 in hematopoietic cells in vivo allows much greater induction by CpG of several key Th1 cytokines, such as IFNγ and IL-12, activating components of innate immunity. Stat3 ablation combined with local CpG treatment activates innate and adaptive antitumor immune responses, leading to rapid and long-term regression of large B16 melanoma tumors. To demonstrate the feasibility of future clinical applications, we show that systemic Stat3 targeting with a small-molecule drug synergizes peritumoral CpG treatment. These results indicate that Stat3 restrains CpG-induced immune responses not only via tumor-induced immunosuppression, but also by TLR9 activation itself. As such, targeting Stat3 is a potential viable strategy to improve various TLR9-based immunotherapeutic approaches.

MATERIALS AND METHODS

Cells

Mouse B16 melanoma cells were purchased from ATCC. Mouse C4 melanoma cells were generous gifts from Dr. J. Fidler (M. D. Anderson Cancer Center, Houston, TX).

In vivo experiments

Mouse care and experimental procedures were performed under pathogen-free conditions in accordance with institutional guidance from Research Animal Care Committees of City of Hope. _Mx1_-Cre mice were purchased from the Jackson Laboratory. Stat3flox/flox mice were kindly provided by Dr. S. Akira and Dr. K. Takeda (University of Osaka, Osaka Japan). Generation of mice with _Stat3_−/− hematopoietic cells by inducible _Mx1_-Cre recombinase system has been reported (12, 15). For s.c. tumor challenge, B16 tumor cells (1×105) were injected into 7–8 week old _Mx1_-Cre/Stat3flox/flox mice and Stat3flox/flox littermates 4d post poly(I:C)-treatment to induce Stat3 ablation. 7–10d later mice were injected peritumorally with 5 μg of phosphothioated CpG (CpG1668: TCCATGACGTTCCTGATGCT) or control GpC (GpC1668: TCCATGAGCTTCCTGATGCT) ODN and tumor growth was monitored trice weekly. For studies on CpG effects, mice were sacrificed at 1, 2 or 3d post-CpG treatment and spleens, lymph nodes and tumor specimens were harvested. For immune cell depletion, mice were pretreated with anti-CD8 plus anti-CD4 antibodies (clone 2.43 and GK1.5, respectively) before tumor inoculation then injected weekly. For treatment using Stat3 inhibitor CPA7 (12, 16), B16 and C4 tumors were implanted into male C57BL/6 or C3H mice, respectively, and allowed to grow until 5–7 mm in diameter. Mice were given i.v. injections of vehicle (10% DMSO/PBS) or CPA7 (5 μg/kg) twice weekly and treated with 5 μg CpG injected peritumorally on following days.

Nitric oxide measurements

CD11b-positive myeloid cells were magnetically separated from splenocyte suspensions derived from tumor-bearing mice using EasySep positive selection kits (StemCell Technologies) and incubated (48h) with or without CpG (5 μg/ml). Supernatants collected from cultured cells were treated with 1% sulfanilamide and 0.1% N-1-naphtylethylenediamine dihydrochloride. The absorbance was measured, within 30 min, on a microplate reader (520–550 nm band pass filter) normalizing for the basal nitrate content in the RPMI 1640 medium. Nitrate concentration in each sample was calculated in relation to a standard curve.

Flow cytometry

Single cell suspensions of spleen, lymph node or tumor tissues were prepared by mechanic dispersion followed by collagenase D/DNase I treatment as described (12). For extracellular staining, freshly prepared cells suspended in PBS/2% FCS/0.1% w/v sodium azide were stained with fluorochrome-coupled antibodies to CD11c, I-Ab (MHCII), CD40, CD80, CD86, CD11b, CD49b, CD3, CD8, CD4 or CD69 (BD Biosciences). Prior to intracellular staining with antibodies to phosphotyrosine-Stat3 and -Stat1 (BD Biosciences), cells were fixed in paraformaldehyde and permeated in methanol. For intracellular cytokine staining using antibodies specific for IL-6 and IL-10 (BD Pharmingen), cells were incubated (4h) with Leukocyte Activation Cocktail (BD Pharmingen) and stained according to manufacturer’s protocol. Fluorescence data were collected on a FACScalibur (BD Biosciences) and analyzed using FlowJo software (Tree Star).

ELISPOT assays

Lymph nodes were harvested from Stat3+/+ and _Stat3_−/− mice two weeks after B16 tumor challenge and 3d after CpG treatment. Lymph node cells were seeded into 96-well filtration plate (5×105 cells/well) in the presence or absence of p15E peptide (10 μg/ml) and incubated (37°C, 24h). Peptide-specific IFN-γ-positive spots were detected following manufacturer’s protocol (Cell Sciences), scanned and quantified using Immunospot Analyzer (Cellular Technology Ltd).

Immunohistochemistry, immunofluorescence and intravital two-photon microscopy

Immunohistochemical staining with hematoxylin/eosin (DAKO) was performed on formalin-fixed and paraffin embedded tissue sections (5 μm). Flash-frozen tumor specimens were fixed in acetone, permeabilized with methanol and stained with antibodies specific to neutrophils (7/4, Cedarlane), macrophages (Mac3), iNOS (BD Biosciences) and detected with fluorochrome-coupled secondary antibodies (Invitrogen). After staining the nuclei (Hoechst 33342, Invitrogen), slides were mounted and analyzed by fluorescence microscopy. For intravital two-photon imaging, CpG-treated Stat3+/+ and _Stat3_−/− B16 tumor-bearing mice received a single retroorbital injection of dextran-rhodamine (Invitrogen) 1h before imaging. Mice were anesthetized and intravital two-photon microscopy was carried out using equipment and software from Ultima Multiphoton Microscopy Systems.

Electromobility shift assay (EMSA), Western blotting and antibody arrays

EMSA to detect Stat3 DNA-binding and Western blot analyses were performed as described (12). Cytokine antibody arrays for detecting cultured DC protein secretion were developed according to manufacturer’s protocol (Panomics).

Statistical analysis

Unpaired t-test was used to test differences between two treatment groups. Statistically significant p values were labeled as follows: ***; p<0.001; **, p<0.01 and *, p<0.05. Data were analyzed using Prism software (GraphPad).

RESULTS

Direct and rapid induction of Stat3 activity in DCs by TLR agonists

While it has recently become evident that TLR agonists can induce their own inhibition (17–19), the role played by TLRs in mediating this feedback mechanism remains poorly understood. Because Stat3 is known to inhibit expression of multiple Th1 immunostimulatory cytokines (9, 13), we tested the possibility that certain TLR agonists might directly activate Stat3, thereby leading to downmodulation of TLR activity. We first tested Stat3 activation by TLR agonists in freshly prepared splenic CD11c+ DCs. At 3h post treatment, both CpG and TLR4 agonist, LPS, induced tyrosine phosphorylation of Stat3, whereas poly(I:C), a TLR3 agonist shown to have potent antitumor immunostimulatory effects (20), did not (Fig. 1A). We further assessed the effects of CpG on activation of Stat3 as well as Stat1 and Stat5 (Fig. 1B). Like Stat3, both Stat1 and Stat5 were phosphorylated on tyrosine residues critical for their dimerization within the first 2h after CpG treatment. However, their activation was transient in comparison to phospho-Stat3 levels, which still remained elevated after 24h. Both the brief Stat1 and prolonged Stat3 activation induced by CpG within the first 24h was confirmed by testing DNA binding activity in primary splenocytes (Supplementary Fig. 1).

TLR9 receptor activation induces immediate and direct activation of Stat3 in dendritic cells. A, TLR ligand-specific Stat3 activation in primary splenic DCs. CD11c+ DCs freshly isolated from pooled splenocytes (_n=_3) were treated (3h) with TLR ligands (1 μg/ml) as indicated. Stat3 activation was assessed by Western blotting to detect phospho-Stat3 or total Stat3 and β-actin for loading control. B, Stimulation with CpG induced rapid and prolonged activation of Stat3 but not Stat1 or Stat5 in splenic DCs (enriched from pooled splenocytes; _n_=3). Kinetics of Stat phosphorylation was assessed by Western blotting as in Fig. 1A. C, Delayed kinetics of CpG-induced IL-6 expression in splenic DCs. Intracellular cytokine staining and FACS analysis of IL-6 and IL-10 production in splenic CD11c+ DCs after 4 or 24h of in vitro incubation in the presence or absence of CpG (1 μg/ml). D, Stat3 activation is mediated by TLR9 independent of IL-10 expression. Western blots detecting tyrosine-phosphorylated Stat3 in DCs enriched from pooled _TLR9_−/− and _IL-10_−/− splenocytes or from the matching WT controls (WT1, C57BL/6 strain; WT2, Balb/C) (n=4 mice/group). All results were confirmed by at least two independent experiments.

It has been suggested that CpG can activate IL-10, a known Stat3 activator, through a negative feedback mechanism (18). Because the kinetics of Stat3 activation by CpG occur rapidly, it is highly unlikely that Stat3 activation involves induced expression of IL-6 or IL-10. To rule out the participation of these Stat3 activating cytokines in mediating CpG-induced rapid Stat3 activation, we measured IL-6 and IL-10 levels in splenic CD11c+ DCs at 4 and 24h after CpG stimulation. Our data indicated that while CpG treatment induced higher levels of IL-6 at 24h, there was no detectable induction within 4h (Fig. 1C), in contrast to CpG-induced rapid Stat3 activation (Fig. 1B). The induction of IL-10 within the tested time points was negligible. Finally, we also observed induction of Stat3 activity in response to CpG treatment in splenic DCs isolated from _IL-10_-deficient mice, but not in _TLR9_-deficient DCs (Fig. 1D). These results demonstrate direct TLR9-mediated Stat3 activation triggered by CpG ODN.

Stat3 signaling in DCs restrains TLR9 agonist-induced expression of Th1 type cytokines and chemokines

To directly test Stat3’s role in restraining TLR9 agonist-induced Th1 type immune responses, we examined the effects of Stat3 ablation on CpG-induced Th1 cytokines and chemokine expression in Stat3+/+ and _Stat3_−/− DCs. Wild-type (WT) splenic DCs stimulated in vitro with CpG ODN secreted several proinflammatory mediators, including IL-12, MIG (CXCL9), MIP1α (CCL3), RANTES (CCL5), IL-6 as well as low levels of IFNγ and TNFα (Fig. 2A). Both IFNγ and TNFα expression levels stimulated by CpG were drastically increased upon Stat3 ablation, by 8- and 26-fold, respectively (Fig. 2B). In addition, following TLR9 activation, _Stat3_−/− DCs secreted immunostimulatory IP-10 (CXCL10). Results from real-time PCR confirmed very strong induction of innate immunity activators, such as IFNβ, IFNγ, TNFα, and IL-6, in splenic _Stat3_−/− DCs after 4h of CpG stimulation ex vivo, compared to WT DCs (Supplementary Fig. 2A). Similarly, the mRNA expression of IL-12, RANTES and IL-6 was upregulated in vivo within 18h post peritumoral injection of CpG, as determined in DCs freshly isolated from tumors (Supplementary Fig. 2B) or from mouse tumor-draining lymph nodes (Supplementary Fig. 2C).

Stat3 ablation sensitizes DCs to CpG stimulation resulting in increased production of Th1 proinflammatory mediators. A, Splenic DCs were incubated (48h) with or without CpG. Conditioned culture media were analyzed using antibody arrays to detect cytokine and chemokine secretion. B, Quantification of antibody array analyses showing increased CpG-induced cytokine and chemokine secretion in _Stat3_−/− mice compared to Stat3+/+ counterparts. Dot intensities were measured by densitometry and normalized to positive controls (shown in the box), mean±SEM, (3 independent experiments, n=3–4 mice/group).

Targeted Stat3 ablation allows potent antitumor innate immune responses by TLR9 triggering

In order to assess whether Stat3 restrained CpG-induced innate immunity against tumors, we induced Stat3 allele truncation in hematopoietic cells of adult mice using the Mx1-Cre-loxP system (12). In the Cre-loxP system, in addition to hematopoietic cells, some organs also undergo partial Stat3 truncation upon poly(I:C) treatment (21). To avoid interference from poly(I:C) treatment, subcutaneous B16(F10) tumor challenge was performed 4d after the last poly(I:C) administration. Established B16 tumors (>1 cm diameter) were treated 10d later with a single peritumoral injection of CpG1668 oligonucleotide (Fig. 3A). Our previous study showed that Stat3 ablation in myeloid cells leads to activation of tumor-infiltrating immune cells and antitumor immune response (12). In order to assess the combined effects of Stat3 targeting and CpG treatment on established tumors, the experimental conditions were setup to allow tumor-induced immunosuppression to mask the effects of Stat3 ablation. Relative to our previous study, we doubled the number of tumor cells for the initial challenge and allowed B16 tumors to reach larger size before treatment (diameter exceeding 1 cm vs. 3–5 mm). Under these conditions no significant difference in tumor size in Stat3+/+ and _Stat3_−/− mice was noted (Fig. 3B). Although CpG treatment did not show significant antitumor activity in control littermates (Stat3+/+) with heavy tumor load, the same treatment resulted in eradication of large B16 tumors (some reaching 1.5 cm in diameter), within three days post injection in mice lacking intact Stat3 alleles (Fig. 3A, B). In contrast, in WT mice with smaller initial B16 tumors (4–6 mm diameter at day 7), single treatment with CpG only reduced tumor growth similar to the effect of Stat3 truncation alone (Fig. 3C). However, combination of Stat3 ablation with local TLR9 stimulation led to complete regression of rapidly growing B16 tumors (Fig. 3C). To gain a glimpse of the tumor in vivo after CpG treatment/Stat3 ablation, we examined tumor vasculature by intravital two-photon imaging. While tumors treated only with CpG displayed relatively intact vasculature, peritumoral injection of CpG into mice with Stat3 truncation in hematopoietic cells led to a complete disruption of tumor vasculature within 18h (Fig. 3D). A closely located tumor-draining lymph node served as a positive control and showed a normal vasculature pattern. These in vivo data further implicated rapid antitumor effects by CpG in the absence of Stat3.

CpG ODN triggers rapid eradication of large B16 tumors in mice with Stat3 ablation. Mice with Stat3+/+ or _Stat3_−/− hematopoietic cells were challenged with B16 melanoma cells (1×105, s.c.). Mice with established tumors were treated with a single peritumoral injection of CpG ODN (5 μg). A, Tumor size before and after CpG treatment. B-C, Changes in tumor volume within 3d post-CpG treatment (2 independent experiments, n=4 mice/group, with either larger, ~10 mm (B) or smaller, ~5 mm (C), average tumor diameter at the time of CpG ODN injection, mean±SEM. D, Two-photon intravital imaging of vasculature stained using dextran-rhodamine within tumor or tumor-draining lymph node in Stat3+/+ or _Stat3_−/− mice 18h post CpG ODN treatment. Images composed of multiple scans from 52 μm z-dimensional stack (2 independent experiments; bar=100 μm).

Cellular mechanisms for CpG/Stat3 ablation-induced anti-tumor innate immune responses

Because CpG treatment in mice with conditional Stat3 knockout led to rapid tumor regression, we assessed the underlying cellular mechanism(s) responsible for the antitumor innate immunity. We observed that within the first 18h after CpG injection, the inflammatory cell populations migrating into tumor tissues of _Stat3_-deficient mice were predominantly neutrophils (Fig. 4A and Supplementary Fig. 3) and to a lesser extent macrophages (Fig. 4B). Immunohistochemical staining of B16 tumor specimen using apoptosis marker (active caspase 3) revealed induction of cell death in tumor areas strongly infiltrated by neutrophils within 24h after CpG injection, and necroses by 48h (Supplementary Fig. 4). Because increased NO synthesis by activated macrophages is known to induce antitumor innate immune response, we also assessed the level of inducible nitric oxide synthase (iNOS) expression in tumor-infiltrating Mac3+ cells within 24h post CpG treatment (Fig. 4B). iNOS expression was detectable only in _Stat3_−/− mice treated with CpG and mostly overlapped with the staining for macrophage-specific (Fig. 4B) but not neutrophil-specific markers (unpublished data). In vitro experiments confirmed that Stat3 ablation led to increased nitrate levels secreted by CD11b+ myeloid cells in response to TLR9-stimulation (Fig. 4C). These results demonstrated that innate immunity caused by Stat3 ablation contributed to the observed rapid tumor regression after CpG ODN injection.

Local CpG treatment and Stat3 ablation induce potent antitumor innate immunity. A–B, B16 tumors show strongly increased infiltration by neutrophils (A) and iNOS-expressing macrophages (B) in _Stat3_−/− relative to Stat3+/+ mice. Fluorescence microscopy of cryosections of tumor specimens detected with antibodies specific for neutrophils (7/4, green) or macrophages (Mac3, red) together with iNOS (green), 2 independent experiments, n=3–4 mice/group. C, CpG induces high levels of NO secretion from _Stat3_-negative myeloid cells. Splenic Stat3+/+ or _Stat3_−/− CD11b+ myeloid cells isolated from tumor-bearing mice were treated with CpG in vitro. Nitrate concentrations in culture media were assessed after 48h; mean±SEM, combined from 2 independent experiments performed in triplicate. D, Stat3 deletion promotes CpG-induced tumor infiltration by activated NK lymphocytes. Left two panels - FACS analysis of CD49b+ NK cell numbers among tumor-infiltrating lymphocytes after excluding CD4+ and CD8+ T cell populations (2 independent experiments, n=3–4 mice/group). Right two panels, Stat1 activation in NK cells isolated from tumor-draining lymph nodes collected from 3–4 mice, 24h after peritumoral ODN injection, as assessed by FACS, using intracellular staining with antibodies specific for tyrosine-phosphorylated Stat1 (2 independent experiments).

In addition to macrophages and neutrophils, natural killer (NK) cells can also play an important role in antitumor innate immunity. To determine if ablating Stat3 promotes CpG-induced NK cells activity, we used flow cytometry to assess the presence of NK cells in tumor tissues 24h after CpG injection. In vivo treatment with CpG enhanced the number of tumor-infiltrating NK cells in mice with a _Stat3_-negative hematopoietic system but failed to stimulate NK activity in WT tumor-bearing controls (Fig. 4D, left). Stat1 activation has been shown to be critical for NK antitumor activity (22). Flow cytometric analysis of phospho-Stat1 in NK cells prepared from tumor-draining lymph nodes indicated that Stat3 ablation allowed strong Stat1 activation by CpG ODN (Fig. 4D, right).

Ablating Stat3 sensitizes DCs to TLR9-mediated priming and augments tumor antigen-specific immunity

We next assessed if Stat3 inhibition affects CpG-induced DC activation in tumor-bearing mice, as DCs are critical for the regulation of T-cell antitumor immunity (20, 23). Flow cytometric analysis of CD11c+ DCs isolated from tumor-draining lymph nodes of mice lacking Stat3 in hematopoietic cells showed enhanced DC activation as measured by increased expression of major histocompatibility complex (MHC) class II together with co-stimulatory molecules CD86 (Fig. 5A, two top rows) or CD80 and CD40 (Fig. 5A, two bottom rows) 2d after peritumoral injection of CpG, but not control GpC ODN. Control in vitro experiments on CpG-treated splenic DCs (with low endogenous Stat3 activity), further confirmed that TLR9-induced Stat3 limits DC activation (Supplementary Fig. 5). Furthermore, while tumor-isolated WT CD11c+ DCs displayed little Stat1 activation after CpG, their _Stat3_−/− counterparts showed high levels of phospho-Stat1 (Fig. 5B).

Ablating Stat3 promotes TLR9-induced dendritic cells maturation and antigen-specific T cell responses. Stat3 deletion enhances CpG-induced tumor DC maturation in vivo. A, The phenotypic analysis of CD11c+ DCs accumulated in tumor-draining lymph nodes of Stat3+/+ and _Stat3_−/− mice 48h post-CpG injection. The maturation of CD11c+ DCs was increased by Stat3 ablation as shown by a greater percentage of double-positive MHC class IIhi and CD86hi DCs (top two rows), as well as higher expression of CD80 and CD40 on DCs (bottom two rows) as seen by FACS analyses (3 independent experiments, n=3–4 mice/group). B, Stat1 is activated in _Stat3_−/− DCs from tumor-draining lymph nodes following CpG but not control GpC ODN treatment. Intracellular flow cytometric analysis of DCs isolated from tumor-draining lymph nodes using an antibody to detect tyrosine-phosphorylated Stat1 (phospho-Stat1) (3 mice/group). C, Stat3 deletion promotes tumor antigen-specific T cell responses.Expression of the early lymphocyte activation marker CD69 was analyzed by flow cytometry on CD8+ T cells 24 h after injecting CpG or control GpC ODNs; 3 independent experiments using lymph node cell suspensions, n=3–4 mice/group. _D, Stat3_−/− mice mount a stronger T-cell response against an endogenous B16 tumor-antigen than their Stat3+/+ counterparts, following treatment with CpG ODN. IFNγ production in T cells derived from tumor-draining lymph node was assessed by ELISPOT assay; mean±SEM of p15E-specific IFNγ-producing cells (2 independent experiments, cells pooled from n=4 mice/group).

To evaluate the effect of Stat3 ablation on CpG-induced effector lymphocyte activity, we analyzed CD8+ T cells within tumor-draining lymph nodes of _Stat3_-positive and _Stat3_-negative mice after CpG-ODN treatment. CD8+ lymphocytes in tumor-draining lymph nodes of _Stat3-_ablated mice showed high levels of CD69 activation marker 24h after peritumoral injection of CpG (Fig. 5C). Importantly, 10d after CpG-treatment, _Stat3_-deficient mice displayed enhanced ability to mount tumor antigen-specific T cell immune response. ELISPOT assays, following ex vivo exposure to the B16 tumor-specific p15E peptide antigen, indicated significantly higher numbers of IFNγ-secreting T cells in tumor-draining lymph nodes derived from _Stat3_−/− mice treated with CpG, relative to their WT or untreated counterparts (Fig. 5D).

Ablating Stat3 in myeloid cells strongly promotes CpG-mediated antitumor immune responses

We next assessed whether Stat3 restrained the long-term effects of TLR9 triggering on tumor growth. CpG treatment prevented tumor outgrowth for over three weeks in mice with _Stat3-_deficient myeloid cells (Fig. 6A). Importantly, the prolonged antitumor immunity was abrogated after _Stat3_−/− mice were depleted of CD4+ and CD8+ T cells. In the absence of CD4+ and CD8+ T cells, tumors reoccurred within the primary tumor site about 10d later, suggesting the role of T cells in the development of antitumor memory responses. Lack of both lymphocyte populations did not prevent the initial robust tumor regression, confirming the critical role of innate immunity in eliminating established tumors. This is consistent with the fact that, regardless of Stat3 status, only a few CD8+ T cells were present in tumor specimen within 18h after CpG stimulation (Supplementary Fig. 6). In agreement with our previous study (12), Stat3 ablation alone reduced tumor growth rate, but did not lead to complete tumor regression (Fig. 6A). Treatment with control GpC oligonucleotide did not significantly inhibit tumor progression (unpublished data). These results suggested that Stat3 restrained CpG-induced adaptive antitumor immune responses.

Stat3 targeting in vivo improves TLR9-mediated antitumor immunity. A, Single peritumoral injection of CpG results in long-term complete rejection of tumors in _Stat3_−/− but not in Stat3+/+ mice through CD4- and CD8-mediated immunity. Mice with established B16 tumors were treated with CpG ODNs. Depleting antibodies against CD4+ and CD8+ T cells or control rat IgG were given to the indicated group of mice (3 independent experiments, n=6 mice/group). Statistically significant differences between the CpG-treated _Stat3_−/− group of mice with or without CD4/CD8 T cell depletion are indicated with asterisks; **, p<0.01 and *, p<0.05. B_–_D, Growth of B16F10 (B) and C4 (C) tumors is significantly inhibited when peritumoral CpG ODN treatment is combined with systemic inhibition of Stat3 activity by a Stat3 inhibitor, CPA7. Mice with established tumors (average diameter 4–8 mm) were treated with CPA7, followed by peritumoral CpG injection a day later. The treatment was repeated twice weekly. Shown are the results of three (B) and two (C) independent experiments. D, Combination of Stat3 targeting with CpG-treatment generates concomitant antitumor immunity. Surviving mice treated as in Fig. 6B, were re-challenged with the same number of B16F10 cells injected in the opposite flank two weeks after the first tumor challenge. Statistically significant differences between the group treated with combination of CPA7 with CpG comparing to controls treated only with CpG only are indicated with asterisks; **, p<0.01 and *, p<0.05.

Superior antitumor efficacy by combining small-molecule Stat3 inhibitor with TLR9 agonist

Our genetic studies demonstrated that eliminating Stat3 signaling in myeloid cells allowed potent local and systemic antitumor responses induced by peritumoral injection of CpG ODN. We then used a small-molecule inhibitor, CPA7 (12, 16), to confirm the therapeutic potential of such a strategy. The therapeutic effect of systemic Stat3 inhibition followed by local treatment with CpG was tested using two metastatic tumor models, B16F10 (Fig. 6B) and C4 melanoma (Fig. 6C). In both tumor models, CpG/Stat3 inhibitor combination yielded the strongest tumor growth retardation compared to the effects of each reagent alone. In addition, only co-treatment with the Stat3 inhibitor and CpG produced sufficient concomitant antitumor immunity to prevent B16F10 tumor outgrowth at the site of the secondary challenge (Fig. 6D). Similar to our genetic studies, the combinatorial therapy might have co-activated mechanisms of innate and adaptive immunity since CD8-depletion only partially affected the initial antitumor effect of CPA7/CpG treatment but clearly prevented the long-term mice survival (Supplementary Fig. 7).

DISCUSSION

We and others have identified a critical role for Stat3 in inducing immunosuppression, in the tumor microenvironment, by inhibiting the expression of Th1 immunostimulatory molecules and promoting the expression of several immunosuppressive factors (9, 24). In the tumor setting, activation of Stat3 is induced by cytokines, such as IL-10, IL-6, and growth factors, including VEGF and bFGF, among many other tumor-associated factors. Our current study demonstrated that Stat3 was a key molecule limiting the efficacy of TLR-agonists. We showed that both TLR9 and TLR4 agonists activated Stat3 rapidly, and that CpG ODN activated Stat3 directly through TLR9. Moreover, targeted deletion or pharmacologic targeting of Stat3 allowed potent antitumor innate and adaptive immune responses after CpG treatment, leading to regression of established tumors. These findings revealed a second mechanism by which Stat3 mediates immunosuppression: activation by TLRs without the induction of cytokines and growth factors. Our results also demonstrated that targeting Stat3 can drastically improve CpG-based immunotherapies.

Although signaling through TLR9 also involves two other Stat factors, Stat1 and Stat5, their early immediate activation is transient. The sustained Stat activity apparently requires the production of secondary mediators like IL-6 and/or IL-10 for Stat3 (Supplementary Fig. 2) (18) or IFNβ and IFNγ for Stat1 (Supplementary Fig. 2) (25). However, since Stat3 strongly inhibits IFNs expression, it is likely that immunostimulation by the secondary effect of TLR9 activation is affected by Stat3 activity.

Currently, CpG-based immunotherapies involve both prophylactic vaccines and therapeutic antitumor approaches. Our results suggest that injection of CpG ODN in individuals without chronic pathogen infection or cancer may also lead to Stat3 activation, which limits CpG Th1 immunostimulatory effects. For cancer therapies, previous studies in mice demonstrated improved TLR9-mediated therapeutic activity when CpG treatment was combined with blocking the immunosuppressive effects of IL-10 (6) or regulatory T cells (26) using neutralizing antibodies. The role of Stat3 as a critical mediator of tumor-induced immunosuppression is supported by numerous studies demonstrating Stat3-mediated inhibition of antigen-presentation (13). Additionally, IL-10 is a key factor responsible for activating Stat3 in immune cells in the tumor milieu. Because IL-10 activates Stat3 and Stat3 is required for IL-10 expression, the anti-IL-10 antibody-mediated enhanced antitumor efficacy induced by CpG is likely due to reduction in Stat3 signaling. A role of TLR agonists in inducing Treg cells in tumors has also been shown, and recent reports indicate a possible function for Stat3 in promoting Treg function (27). Since both TLR4 and TLR9 agonists activate Stat3, it is possible that Tregs induced by TLR are in part mediated by Stat3 activation. In addition, as recently demonstrated by Cheng et al. (28) and by our own studies (12, 29), Stat3 is critical for the accumulation and function of CD11b+Gr1+ myeloid-derived suppressor cells (MDSCs) in tumor-bearing mice. Our results indicate that of Stat3 ablation and TLR9 activation may convert immunosuppressive/proangiogenic MDSCs into cytotoxic neutrophils and stimulate their massive recruitment into tumor site.

A role for TLRs in recognizing a variety of pathogen-associated patterns, and thereby initiating innate and adaptive immunity, has been well appreciated. However, recent evidence indicates that TLRs are also expressed by tumor cells that promote tumor cell proliferation and survival, as well as invasion and metastasis. A critical role of Stat3 in upregulating a large number of genes critical for proliferation, survival, angiogenesis and metastasis has been widely documented (7). Recent research in our laboratory further indicated that activated Stat3 in tumor-associated diverse immune cells, especially in myeloid cells, stimulated the expression, in these cells, of many factors involved in immune suppression, tumor invasion and metastasis (29). Because CpG and LPS directly activate Stat3, it is likely that the observed tumor-promoting effects of TLRs (and their agonists) (14) are partially mediated by Stat3 activation. This is consistent with an emerging role for Stat3 in wound healing (30, 31), a physiological process involving immune suppression, cell proliferation, invasion and angiogenesis and characteristic of cancer development. However, while wound healing is self-limiting, cancer progression is not. Taken together, blocking Stat3 not only lifts the “brake” on TLR-mediated immune responses, but also reduces the oncogenic potential mediated by TLR activation. It is also interesting to note that, unlike LPS and CpG, poly(I:C) did not rapidly induce Stat3 activity. A recent publication supports its use for cancer immunotherapies (20).

Despite several challenges, TLR9 agonists remain attractive anticancer agents, since they have shown some efficacy and are well tolerated in patients, with relatively minor adverse effects (1). Our study strongly supports the use of combination therapies in which TLR9 activation is preceded by or simultaneous with Stat3 targeting. Although the potential side effects resulting from Stat3 targeting remain to be fully determined, first insights into possible effects of Stat3 deletion in the human system were revealed in patients with hyper-IgE syndrome (32, 33). These patient studies suggested that long-term Stat3 inhibition could cause immune disorders and complications due to microbial infections, that is, manageable chronic diseases. Therefore, temporary Stat3 inhibition by small-molecule drugs or siRNA may not generate severe undesirable effects. Our studies support the use of Stat3 targeting to improve the outcome of immunotherapies based on TLR activation.

Supplementary Material

figure 1

figure 2

figure 3

figure 4

figure 5

figure 6

figure 7

Acknowledgments

We would like to thank Dr. H. Kay of Paradigm Biotech, LLC (Tampa, FL) for providing CPA7, Dr. S. Akira and Dr. K. Takeda of University of Osaka, Japan for Stat3flox mice and the Pathology Core at City of Hope for tissue processing. We also acknowledge the dedication of staff members at the animal facilities and flow cytometry cores at City of Hope. We are grateful to Dr. S. Da Costa for editing the manuscript. This work is supported in part by grants from Harry Lloyd Charitable Trust. The authors have no conflicting financial interests.

References

1. Krieg AM. Toll-like receptor 9 (TLR9) agonists in the treatment of cancer. Oncogene. 2008;27:161–7. [PubMed] [Google Scholar]

2. Klinman DM. Immunotherapeutic uses of CpG oligodeoxynucleotides. Nat Rev Immunol. 2004;4:249–58. [PubMed] [Google Scholar]

3. Pasare C, Medzhitov R. Toll pathway-dependent blockade of CD4+CD25+ T cell-mediated suppression by dendritic cells. Science. 2003;299:1033–6. [PubMed] [Google Scholar]

4. Yang Y, Huang CT, Huang X, Pardoll DM. Persistent Toll-like receptor signals are required for reversal of regulatory T cell-mediated CD8 tolerance. Nat Immunol. 2004;5:508–15. [PubMed] [Google Scholar]

5. Appay V, Jandus C, Voelter V, et al. New generation vaccine induces effective melanoma-specific CD8+ T cells in the circulation but not in the tumor site. J Immunol. 2006;177:1670–8. [PubMed] [Google Scholar]

6. Vicari AP, Chiodoni C, Vaure C, et al. Reversal of tumor-induced dendritic cell paralysis by CpG immunostimulatory oligonucleotide and anti-interleukin 10 receptor antibody. J Exp Med. 2002;196:541–9. [PMC free article] [PubMed] [Google Scholar]

7. Yu H, Jove R. The STATs of cancer--new molecular targets come of age. Nat Rev Cancer. 2004;4:97–105. [PubMed] [Google Scholar]

8. Takeda K, Clausen BE, Kaisho T, et al. Enhanced Th1 activity and development of chronic enterocolitis in mice devoid of Stat3 in macrophages and neutrophils. Immunity. 1999;10:39–49. [PubMed] [Google Scholar]

9. Wang T, Niu G, Kortylewski M, et al. Regulation of the innate and adaptive immune responses by Stat-3 signaling in tumor cells. Nat Med. 2004;10:48–54. [PubMed] [Google Scholar]

10. Welte T, Zhang SS, Wang T, et al. STAT3 deletion during hematopoiesis causes Crohn’s disease-like pathogenesis and lethality: a critical role of STAT3 in innate immunity. Proc Natl Acad Sci U S A. 2003;100:1879–84. [PMC free article] [PubMed] [Google Scholar]

11. Gabrilovich DI, Chen HL, Girgis KR, et al. Production of vascular endothelial growth factor by human tumors inhibits the functional maturation of dendritic cells. Nat Med. 1996;2:1096–103. [PubMed] [Google Scholar]

12. Kortylewski M, Kujawski M, Wang T, et al. Inhibiting Stat3 signaling in the hematopoietic system elicits multicomponent antitumor immunity. Nat Med. 2005;11:1314–21. [PubMed] [Google Scholar]

13. Yu H, Kortylewski M, Pardoll D. Crosstalk between cancer and immune cells: role of STAT3 in the tumour microenvironment. Nat Rev Immunol. 2007;7:41–51. [PubMed] [Google Scholar]

14. Conroy H, Marshall NA, Mills KH. TLR ligand suppression or enhancement of Treg cells? A double-edged sword in immunity to tumours. Oncogene. 2008;27:168–80. [PubMed] [Google Scholar]

15. Lee CK, Raz R, Gimeno R, et al. STAT3 is a negative regulator of granulopoiesis but is not required for G-CSF-dependent differentiation. Immunity. 2002;17:63–72. [PubMed] [Google Scholar]

16. Turkson J, Zhang S, Mora LB, Burns A, Sebti S, Jove R. A novel platinum compound inhibits constitutive Stat3 signaling and induces cell cycle arrest and apoptosis of malignant cells. J Biol Chem. 2005;280:32979–88. [PubMed] [Google Scholar]

17. Akira S, Takeda K. Toll-like receptor signalling. Nat Rev Immunol. 2004;4:499–511. [PubMed] [Google Scholar]

18. Samarasinghe R, Tailor P, Tamura T, Kaisho T, Akira S, Ozato K. Induction of an anti-inflammatory cytokine, IL-10, in dendritic cells after toll-like receptor signaling. J Interferon Cytokine Res. 2006;26:893–900. [PubMed] [Google Scholar]

19. Hu X, Chen J, Wang L, Ivashkiv LB. Crosstalk among Jak-STAT, Toll-like receptor, and ITAM-dependent pathways in macrophage activation. J Leukoc Biol. 2007;82:237–43. [PubMed] [Google Scholar]

20. Trumpfheller C, Caskey M, Nchinda G, et al. The microbial mimic poly IC induces durable and protective CD4+ T cell immunity together with a dendritic cell targeted vaccine. Proc Natl Acad Sci U S A. 2008;105:2574–9. [PMC free article] [PubMed] [Google Scholar]

21. Kuhn R, Schwenk F, Aguet M, Rajewsky K. Inducible gene targeting in mice. Science. 1995;269:1427–9. [PubMed] [Google Scholar]

22. Lee CK, Rao DT, Gertner R, Gimeno R, Frey AB, Levy DE. Distinct requirements for IFNs and STAT1 in NK cell function. J Immunol. 2000;165:3571–7. [PubMed] [Google Scholar]

23. Schuler G, Schuler-Thurner B, Steinman RM. The use of dendritic cells in cancer immunotherapy. Curr Opin Immunol. 2003;15:138–47. [PubMed] [Google Scholar]

24. Sumimoto H, Imabayashi F, Iwata T, Kawakami Y. The BRAF-MAPK signaling pathway is essential for cancer-immune evasion in human melanoma cells. J Exp Med. 2006;203:1651–6. [PMC free article] [PubMed] [Google Scholar]

25. Flores RR, Diggs KA, Tait LM, Morel PA. IFN-gamma negatively regulates CpG-induced IL-10 in bone marrow-derived dendritic cells. J Immunol. 2007;178:211–8. [PubMed] [Google Scholar]

26. Davila E, Kennedy R, Celis E. Generation of antitumor immunity by cytotoxic T lymphocyte epitope peptide vaccination, CpG-oligodeoxynucleotide adjuvant, and CTLA-4 blockade. Cancer Res. 2003;63:3281–8. [PubMed] [Google Scholar]

27. Pallandre JR, Brillard E, Crehange G, et al. Role of STAT3 in CD4+CD25+FOXP3+ regulatory lymphocyte generation: implications in graft-versus-host disease and antitumor immunity. J Immunol. 2007;179:7593–604. [PubMed] [Google Scholar]

28. Cheng P, Corzo CA, Luetteke N, et al. Inhibition of dendritic cell differentiation and accumulation of myeloid-derived suppressor cells in cancer is regulated by S100A9 protein. J Exp Med. 2008;205:2235–49. [PMC free article] [PubMed] [Google Scholar]

29. Kujawski M, Kortylewski M, Lee H, Herrmann A, Kay H, Yu H. Stat3 mediates myeloid-cell-dependent tumor angiogenesis. J Clin Invest. 2008;118:3367–77. [PMC free article] [PubMed] [Google Scholar]

30. Sano S, Itami S, Takeda K, et al. Keratinocyte-specific ablation of Stat3 exhibits impaired skin remodeling, but does not affect skin morphogenesis. Embo J. 1999;18:4657–68. [PMC free article] [PubMed] [Google Scholar]

31. Dauer DJ, Ferraro B, Song L, et al. Stat3 regulates genes common to both wound healing and cancer. Oncogene. 2005;24:3397–408. [PubMed] [Google Scholar]

32. Minegishi Y, Saito M, Tsuchiya S, et al. Dominant-negative mutations in the DNA-binding domain of STAT3 cause hyper-IgE syndrome. Nature. 2007;448:1058–62. [PubMed] [Google Scholar]

33. Holland SM, DeLeo FR, Elloumi HZ, et al. STAT3 mutations in the hyper-IgE syndrome. N Engl J Med. 2007;357:1608–19. [PubMed] [Google Scholar]