IL-10: Master Switch from Tumor-Promoting Inflammation to Antitumor Immunity

Many approaches have been pursued to improve the arsenal of T-cell–stimulating agents in the fight against cancer, with proinflammatory cytokines such as IFN-α and interleukin (IL)-2 being among the most clinically successful agents for the treatment of melanoma (1). With the discovery of inhibitory molecules, cytotoxic T-lymphocyte antigen 4 (CTLA-4) and programmed death-1 (PD-1) on T cells as “immune checkpoints” and the development of inhibitors for these immune checkpoints, immunotherapy as therapeutic targets in cancer treatment has finally emerged into mainstream oncology. Anti–CTLA-4 antibodies were the first to be approved by the U.S. Food and Drug Administration for the treatment of melanoma. Soon thereafter, antibodies inhibiting the PD-1 receptor or its ligands showed tremendous success in the treatment of several solid malignancies.

Over the past decade, many studies demonstrated that infiltration of T cells is the single most important prognostic factor for the majority of human tumor types (2). Consequently, inhibition of PD-1, which effectively stimulates CD8⁺ T cells to enter the tumor, achieves durable clinical responses in patients with melanoma (3). T cells specific to tumor-associated antigens (TAA) are rare in the blood and tumors of patients with cancer. Anti–CTLA-4 but not anti–PD-1 treatment increases TAA-specific CD8⁺ T-cell responses in responding patients with melanoma (4, 5). The combination of T-cell checkpoint inhibitors, anti–PD-1 and anti–CTLA-4, revealed a stunning improvement over each therapy alone, with 53% of patients with melanoma receiving the combined regimen experiencing complete responses (6). However, additional therapeutic approaches are needed for patients not responding to these therapies and for the treatment of other tumor types.

IL-10 is the founding member of the IL-10 family of cytokines. It is a noncovalent homodimeric alpha helical cytokine with structural similarities to IFN-γ. The IL-10 receptor (IL10R) consists of two molecules of an IL-10–specific chain IL10R1 and two molecules of IL10R2 that is shared with other cytokines. IL10R is expressed on the surface of most hematopoietic cells, including T cells, B cells, and macrophages. Genetically engineered mouse models lacking IL-10 or IL10R have revealed that the immunologic and physiologic functions of IL-10 are nonredundant. In support of the anti-inflammatory and tumor-inhibitory function, IL-10–deficient mice and humans develop inflammatory bowel disease (IBD) and cancer (7–9).

Originally, IL-10 was identified as a factor produced by Th2-polarized CD4⁺ T cells, which suppresses the proliferation of CD4⁺ T cells and the secretion of cytokines in Th1 helper cells, in particular IFN-γ (10). Shortly thereafter, IL-10 was shown to induce thymocyte proliferation by elevating the expression of CD3 and CD8. Moreover, IL-10 was shown to induce the cytotoxicity of CD8⁺ T cells (11). In preclinical tumor models, IL-10 induces the rejection of tumors invoking both its proimmunity and its anti-inflammatory functions. The anti-inflammatory role of IL-10 has been explored in many studies, whereas the well-documented IL-10–dependent stimulation of CD8⁺ T-cell cytotoxicity in the immune response to cancer has received considerably less attention. In this Crossroads article, these two aspects of IL-10 will be reconciled with a particular focus on understanding the use of the recombinant pegylated cytokine as an anticancer agent and its ability to stimulate TAA-specific CD8⁺ T cells.

IL-10 is a potent anti-inflammatory factor in bacterial endotoxemia. In a mouse model of endotoxemia, repeated injections of bacterial lipopolysaccharides (LPS) induce an acute cytokine release syndrome, leading to a vascular shock and death. LPS binds to Toll-like receptor 4 (TLR4), a pattern recognition receptor (PRR) expressed on many myeloid cells. PRRs alert the innate immune system of foreign molecules such as bacterial cell wall constituents or DNA, and of damaged host cells. The stimulation of PRRs is not antigen specific and leads to an inflammatory reaction, largely driven by the innate immune system. However, LPS also induces inflammatory cytokines IL-12/IL-23, which stimulate inflammatory T cells. IL-10 is induced by LPS as a negative feedback; it inhibits IL-12/IL-23 expression and the expression and signaling of other proinflammatory cytokines as well as the signaling of the PRRs. As a consequence, IL-10^−/− mice are exquisitely sensitive to the LPS-induced shock (12). Treatment with IL-10 protects mice from experimental endotoxemia (13). Mechanistically, IL-10 induces the phosphorylation of STAT3 in macrophages and T cells, and the deletion of STAT3 in these cells leads to IBD. IL-10–activated STAT3 induces the transcriptional repressor nuclear factor, interleukin 3 regulated (NFIL3), which in turn inhibits the transcription of the shared p40 subunit of IL-12 and IL-23, thereby inhibiting both key inflammatory cytokines (14, 15). IL-10–activated STAT3 also induces the suppressor of cytokine signaling 3 (SOCS3), which binds and inhibits the signaling of the proinflammatory IL-6 and IL-12/IL-23 receptors. IL-10 thereby intercepts the inflammatory responses to LPS or other PRR stimulants at various levels.

Many preclinical models of inflammatory diseases stimulate antigen- or organ-specific immunity by vaccinations using LPS or other PRR stimulants. Such vaccinations are exacerbated in IL-10 knockout mice and are suppressed in the presence of IL-10. The IL-10–mediated suppression of IL-12/IL-23–dependent inflammation frequently is interpreted as a general immunosuppressive function. When LPS or other PRR ligands are injected directly into a tumor, the inflammatory response locally can result in sufficient tissue damage to reject a tumor, in particular if the inflammation is unrestricted, as in IL-10 knockout mice (16). This tumor rejection is primarily driven by inflammatory T cells and macrophages. In line with this reasoning, a TLR7 agonist (imiquimod) is in clinical use for the treatment of localized precancerous lesions in the skin. Imiquimod creates a local inflammatory reaction with the intent to eliminate malignant cells. The exacerbation of local inflammation in the absence of IL-10 also supported the call for the development of IL-10–neutralizing antibodies for cancer therapy (17). However, systemic applications of TLR agonists lead to a life-threatening systemic and nondiscriminating inflammatory reaction. Consequently, and due to limiting toxicities, this approach has little or no systemic antitumor application. In contrast, IL-10 induces systemic but tumor-specific immunity by enhancing antigen-specific CD8 T cells largely without non–tumor-directed toxicity (see below).

The inflammation which, when induced locally, may lead to the elimination of a local tumor appears to be detrimental in patients with cancer. Inflammatory cytokines and chronic inflammation correlate with increased tumor incidence and a worsened prognosis for patients with cancer (18). The development and progression of tumors in chronically inflamed tissue is thought to be the consequence of increased compensatory cell proliferation, a tumor-promoting microenvironment with increased angiogenesis and proteolytic activity, an increased mutation rate due to the release of oxygen radicals, and the conspicuous absence of tumor immunosurveillance by cytotoxic T cells.

The balance between proinflammatory and anti-inflammatory signals provided by different T-cell populations appears to be crucial for the maintenance of normal physiology as well as the suppression of cancer development. IL-10 is an essential cytokine for the homeostasis of anti-inflammatory regulatory T cells (Tregs) and the suppression of proinflammatory IL-17–expressing T cells (Th17). The anti-inflammatory role of IL-10 is demonstrated in genetically engineered mouse models deficient for IL-10 and in patients carrying a somatic mutation in the IL-10 or the IL10R gene, both of whom spontaneously develop IBD. Furthermore, IL-10 knockout mice develop colon cancer (9) and humans deficient in IL-10 signaling develop lymphomas at a young age (8); these consequences of IL-10 deficiency provide support for the concept of tumor-promoting inflammation and for an essential role of IL-10 in its control.

In IL-10 knockout mice, proinflammatory Th17 cells expand and Tregs are suppressed. A Treg-specific deletion of IL-10 or the IL10R enhances inflammation and eventually leads to colitis, indicating that Tregs function both as essential source and recipient of the cytokine (19). IL-10 induces STAT3 phosphorylation in Tregs, and Tregs lacking STAT3 fail to expand in the inflamed gut even if they are still capable of suppressing CD4 T-cell proliferation in vitro (20). This suggests that IL-10–mediated homeostasis of Tregs is essential for its anti-inflammatory function. Treg homeostasis is mediated by STAT3, a transcription factor well known for its proliferative and antiapoptotic functions.

Mice with a mutation in the adenomatous polyposis coli gene (APC^Δ468) develop intestinal tumors. T-cell–specific ablation of IL-10 in these mice dramatically changes the inflammatory milieu and increases the tumor burden (21). Conversely, adoptive transfer of Tregs into mice with colonic tumors leads to an IL-10–dependent reduction of tumor burden (22). Collectively, these data support a role of IL-10 and Tregs in the control of inflammation-driven tumors.

Conversely, proinflammatory Th17 cells are essential for the development of many autoinflammatory diseases (23). Polarization and activity of Th17 cells is stimulated by the myeloid-derived cytokine IL-23. Crohn disease and other autoinflammatory diseases in humans are genetically linked to the IL23R gene (24). However, IL-23 knockout mice are resistant to experimentally induced autoinflammatory diseases, and they are also highly resistant to tumor development. The inhibition of tumors in these mice is enabled by a striking deficiency of inflammatory mediators such as IL-17 and tumor-promoting inflammation. Simultaneously, tumor-infiltrating CD8⁺ T cells and their cytotoxic mediators and IFN-γ are highly prevalent (25). The mutual exclusivity of “inflammatory” and “cytotoxic immunity” mediating cells is explained by the signature effector cytokines, IL-17 or IFN-γ. IL-17 attracts and activates granulocytes and myeloid cells promoting angiogenesis and wound repair. IFN-γ induces antigen presentation and the development of adaptive T cells.

The transcription factor RORγt defines the proinflammatory Th17 cells but is also expressed in a subset of proinflammatory Tregs (26). T-cell–specific deletion of RORγt inhibits both inflammatory T-cell populations, suppresses tumor development in APC^Δ468 mice, and increases the expression of IL-10 (26). In the absence of RORγt, cytotoxic granzymes and perforin-positive cells are increased in the gut, indicating the activity of cytotoxic immunosurveillance. Interestingly, the expression of RORγt is also repressed by NFIL3 (27), a transcriptional repressor mediating the IL-10–induced inhibition of IL-12/23p40 expression (14). These data demonstrate the intricate antagonism of proinflammatory and proimmunity signaling and the importance of IL-10 in the control of Th17 cells that drive tumor-promoting inflammation.

Shortly after its cloning, IL-10 was described as a B-cell–derived T-cell growth factor (B-TCGF) for its ability to stimulate CD8⁺ T cells (11, 28). IL-10 was found to induce the expression of CD3 and CD8 molecules on thymocytes and to promote the cytotoxic activity of CD8⁺ T cells (29). Moreover, IL-10 enhanced the proliferation of CD8⁺ T cells upon direct stimulation of the T-cell receptor (TCR) signaling using anti-CD3 monoclonal antibodies (mAb). IL-10 functionally replaces IL-2 in the proliferation of CD8⁺ T cells stimulated with anti-CD3 mAbs. In contrast, due to the reduction of MHC molecules on monocytes in vitro, IL-10 suppressed the proliferation of CD8⁺ T cells in allogeneic monocyte cocultures (30). In retrospect, the latter result needs to be interpreted with caution given the induction of IFN-γ by IL-10 in humans and mice. IFN-γ strongly induces the expression of MHC and costimulatory molecules, with both molecules stimulating TCR signaling and thereby enabling T-cell proliferation. Such a paracrine loop would not be accurately established in in vitro culture conditions.

Tumor cells engineered to express IL-10 were rejected in immunocompetent hosts (31). Moreover, transplantable tumors injected into mice genetically engineered to express human IL-10 in myeloid cells (IL-10TG) were rejected, which was dependent on the presence of CD8⁺ T cells in the host (32). Furthermore, treatment of syngeneic mouse tumors with recombinant human IL-10 induced CD8⁺ T-cell–dependent tumor rejection (33). IL-10 application was efficacious in the effector phase of tumor rejection, whereas application during tumor vaccination failed to significantly enhance tumor rejection (34).

IL-10^−/− mice were found to be particularly susceptible to chemically induced skin cancers. Tumors developing in IL-10^−/− mice progressed and metastasized prematurely with all IL-10^−/− mice succumbing to widespread metastasis (35). Intratumoral CD8⁺ T cells, MHC molecules, and granzymes were suppressed in tumors developing in IL-10^−/− mice, but were increased in IL-10 transgenic mice. A single injection of IL-10 or of a long-lived, pegylated version of IL-10 (PEG-IL-10) systemically into tumor-bearing mice induced the expression of IFN-γ and granzymes in the tumor (see Fig. 1). Continuous treatment with PEG-IL-10 led to the CD8⁺ T-cell–dependent rejection of large endogenous breast cancers in Her2 transgenic mice and eliminated lung metastasis in aggressive transplanted tumor models. PEG-IL-10 treatment also led to a durable immune memory against tumors, with mice remaining resistant to tumor challenge up to 8 months after the initial tumor rejection. This durable tumor immunity likely is mediated by an IL-10 signaling pathway specific to intratumoral CD8⁺ T cells (36). IL-10 induced the phosphorylation of STAT1 and STAT3 and the expression of IFN-γ in intratumoral CD8⁺ T cells. In CD4⁺ or CD8⁺ T cells from lymphoid organs, IL-10 induced only STAT3 phosphorylation and failed to induce IFN-γ (36). STAT1 is particularly important in the induction of IFN-γ, because it induces T-bet, a transcription factor defining the Th1 and T_c1 lineages of IFN-γ–producing T cells (37). T-bet coordinates with the TCR-activated transcription factor NFAT for the induction of IFN-γ and granzymes in cytotoxic T cells (38). Accordingly, IFN-γ, granzymes, and perforin were increased in cytotoxic T cells in PEG-IL-10 treatment only upon TCR stimulation. The intratumoral IFN-γ produced by CD8⁺ T cells induced both MHC class I and II molecules, potentially allowing antigen presentation within the tumor (See Fig. 1). In human tumors, expression of MHC molecules is correlated with improved prognosis for the patients (39). Consequently, PEG-IL-10 also led to a dramatic surge of tumor cell–specific, intratumoral CD8⁺ T cells. The proliferation of tumor-directed T cells generally is thought to occur in the tumor-draining lymph node, with the postproliferative T cells returning to the tumor through the blood stream. In contrast, IL-10 induced the accumulation of tumor-resident CD8⁺ T cells and tumor rejection, even when T-cell trafficking to and from the tumor site was inhibited. IL10R-expressing tumor-resident CD8⁺ T cells expanded within the tumor tissue, at the expense of T cells not carrying the IL-10 receptor (36). Similar to the proliferation-inducing and antiapoptotic function of STAT3 in Tregs and other cell types, the strong IL-10–mediated STAT3 phosphorylation observed specifically in tumor-infiltrating CD8⁺ T cells may be essential for their expansion within tumors. In support of this concept, the in vivo maintenance of antiviral memory CD8⁺ T cells also requires STAT3 and IL-10 (40). In summary, PEG-IL-10 treatment induces the stimulation of tumor-resident, tumor-specific CD8⁺ T cells, driven by the IFN-γ–mediated upregulation of antigen presentation within the tumor (Fig. 1).

Clinical trials with recombinant IL-10 were initiated with a focus on the anti-inflammatory function of IL-10. Indeed, genome-wide association studies confirmed the association of IL-10 with various inflammatory diseases such as Crohn disease, ulcerative colitis, and Behçet disease (41). Several thousand patients with inflammatory disease were treated with injections of IL-10 (42). These clinical trials were performed before a detailed understanding of the molecular mechanism of IL-10′s anti-inflammatory function was established and it was therefore difficult to optimize the treatment. Disease-associated proinflammatory cytokines such as TNF-α, IL-1β, IL-12, and IL-17 were reduced by half in some studies, indicating a pharmacodynamic activity (43). However, the suppression of IL-1 and TNF-α lasted only as long as the serum concentration of IL-10 was sufficiently elevated, returning to high values 24 hours after the IL-10 injection (44). Nevertheless, in early clinical studies encouraging efficacy of IL-10 was observed in psoriasis, hepatitis, and Crohn disease. However, in pivotal trials, frequently using three-times-weekly dosing, statistically significant disease modification has not been observed. Taken together, the variable concentration of IL-10 due to the short half-life of IL-10 in patients may have severely limited the therapeutic benefit.

However, signs of the IL-10–mediated stimulation of adaptive immunity were observed frequently. IL-10 increased the presence of HLA-DR⁺ CD14⁺ monocytes in the blood, indicating a propensity for antigen-presenting cells (45). Peripheral blood mononuclear cells of patients with Crohn disease, treated daily with higher doses of IL-10, secreted elevated levels of IFN-γ compared with the controls (46). A single bolus of IL-10, given to healthy volunteers before or after a challenge with LPS, induced the cytotoxicity-mediating granzyme B and IFN-γ in the serum, but decreased IL-12 p40 (47). IFNγ-producing CD4 T cells were increased in patients with psoriatic arthritis (48). These results suggest that therapeutic doses of IL-10 induced key elements of a CD8⁺ T-cell response in humans, including IFN-γ and granzymes. They also confirm the independence of IFN-γ induction by IL-10 from the regulation of the proinflammatory IL-12.

The immune system plays a dual role in cancer initiation and progression. Inflammatory responses are abundant in tumors and promote tumor growth while tumor antigen-specific, cytotoxic T cells eliminate tumor cells, but are rare in human tumors. Both divergent responses appear to antagonize each other. IL-10 is at the crossroads of this regulation. IL-10 inhibits the inflammatory Th17 T cells and macrophages and a cytokine milieu rich in TNF-α and IL-23 that promotes tumor incidence and progression. Conversely, IL-10 directly induces the expansion of tumor-specific CD8⁺ T cells in the tumor and their cytotoxic activity. In contrast with other therapeutic modalities, PEG-IL-10 appears to induce the key immunostimulatory cytokine IFN-γ only in CD8⁺ T cells, and this is strictly dependent upon antigen-stimulated TCR signaling within the tumor.

In contrast, the therapeutic stimulation with inflammatory cytokines such as IL-2 and IL-12 induces the systemic release of high levels of IFN-γ, leading to systemic immune toxicity. Inhibition of immune checkpoints bypasses the restrictions of the TCR signal on T cells, thereby allowing T cells to be activated by weaker antigen and MHC stimulation. Enlisting an increased repertoire of T cells in the antitumor response leads to enhanced tumor immunity but may potentially also explain autoimmune-related side effects of checkpoint inhibitors (6). These therapeutically successful approaches may therefore be mechanistically linked to the observed autoimmune toxicities.

PEG-IL-10 may occupy a unique position in the therapeutic arsenal against cancer, correcting both tumor-associated inflammation and a lack of tumor immunity at once. This may represent a promising novel therapeutic avenue for the treatment of patients with cancer.

No potential conflicts of interest were disclosed.