Dual biological effects of the cytokines interleukin-10 and interferon-γ

Abstract

It is generally thought that each cytokine exerts either immune stimulatory (inflammatory) or immune inhibitory (antiinflammatory or regulatory) biological activities. However, multiple cytokines can enact both inhibitory and stimulatory effects on the immune system. Two of these cytokines are interleukin (IL)-10 and interferon-gamma (IFNγ). IL-10 has demonstrated antitumor immunity even though it has been known for years as an immunoregulatory protein. Generally perceived as an immune stimulatory cytokine, IFNγ can also induce inhibitory molecule expression including B7-H1 (PD-L1), indoleamine 2,3-dioxygenase (IDO), and arginase on multiple cell populations (dendritic cells, tumor cells, and vascular endothelial cells). In this review, we will summarize current knowledge of the dual roles of both of these cytokines and stress the previously underappreciated stimulatory role of IL-10 and inhibitory role of IFNγ in the context of malignancy. Our progressive understanding of the dual effects of these cytokines is important for dissecting cytokine-associated pathology and provides new avenues for developing effective immune therapy against human diseases, including cancer.

IL-10

IL-10 was first identified in 1988 as a product of T helper type 2 (Th2) cells that inhibited cytokine production from T helper type 1 (Th1) cells [1]. In addition to Th2 cells, IL-10 is produced by other cell populations, including macrophages and dendritic cells (DCs), B cells [2], keratinocytes, mast cells [3, 4], and several subsets of T cells, including regulatory T cells (Tregs) [5, 6], and T helper 17 cells (Th17) [7]. Tregs release IL-10 as one of their many immunosuppressive activities [6, 8, 9]. IL-10 targets T cells and antigen-presenting cells (APC) and mediates immune regulation [10–12]. Here, we provide a summary of the defined immunoregulatory roles of IL-10 and emphasize the heretofore unappreciated immunostimulatory roles of IL-10 (Table 1).

Table 1.

Interleukin-10

Effect	Mechanism	Context	References
Immune inhibitory effects
Prevents DC trafficking to lymph node	Autocrine signaling in DCs	Mycobacterium infection	[18]
Inhibition of TAA responses	CD8+ Treg production of IL-10	Cancer	[5]
Increased susceptibility to re-infection, chronic infection, transplanted tumors	IL-10 in lungs, systemic IL-10, IL-10 from Tregs, exogenous/ectopically expressed IL-10	Health, infection, cancer	[25–30]
Suppression of NK and NKT cell activity	αβ and γδ T-cell production of IL-10	Cancer	[37]
Decreasing CTL, NK killing	IL-10 from tumor cells and other populations, downregulation of MHC on tumors, increase in non-classical MHC on tumors	Cancer	[38–41]
Impedes T-cell proliferation	Tumor-infiltrating IL-10+ monocytes	Cancer	[35]
Increases in CTLA-4 expression	IL-10 in tumor environment	Cancer	[42]
Maintenance of FoxP3 expression and Treg suppressive function	Myeloid-derived IL-10	Colitis	[20]
Maintenance of intestinal homeostasis	Treg and mucosal IL-10	Health	[6, 21–24]
Constrains Th17 populations	IL-10 modulation of IL-1 expression in DC	Health, colitis	Wilke, under review
Decreases TAA-specific killing, but increases activation of iNKT cells	IL-10+ neutrophils in tumor-bearing hosts	Cancer	[36]
Non-immune effects
Protects neurons from toxin-induced damage	IL-10 signaling in spinal cord neurons		[60]
Inhibition of autophagy in macrophages	Exogenous IL-10 treatment		[61]
Increases angiogenesis	Possibly through regulation of VEGF, NO expression	Hypoxia	[62]
Immune stimulatory effects
Increases MHC II expression on B cells, B cell survival, proliferation, Ig production	Exogenous IL-10	Health	[46–49]
Supports growth and proliferation of T cells, induces CTL precursors and increases CTL function	In combination with IL-2 and IL-4, IL-10 promotes CTL	Health	[51, 52]
Tumor cells rejected, tumor metastases decreased	IL-10-transfected tumor cells; decrease in tumor MHC, increased NK lysis of tumor	Cancer	[54–56]
Increased activated memory T cells, Ag-specific T cell proliferation and IFNγ production	IL-10 administered to tumor-bearing mice	Cancer	[58]
Limits immunosuppressive cell populations in the tumor environment	Limits IL-1 expression in MDSC, Treg induction	Cancer	[59]

IL-10 and immunoregulation

IL-10 enacts most of its immunosuppressive activity indirectly, via effects on APCs. IL-10 downregulates major histocompatibility molecule (MHC) [13] and B7 co-stimulatory molecule expression [14] and promotes inhibitory B7-H1 expression on APCs [15]. It limits proinflammatory cytokine and chemokine expression in APCs but can also directly affect CD4⁺ T cells by limiting their activation, proliferation, and cytokine production [10, 11, 16, 17]. Interestingly, a study with mycobacterium demonstrated that autocrine IL-10 signaling in DCs can prevent their trafficking to lymph nodes [18]. This impedes the recruitment of naïve T cells to the draining lymph nodes of infected tissue, as well as polarization of these same T cells to a Th1 phenotype.

It has recently been shown that IL-10 is associated with the suppressive effects and development of Tregs. Human tumor-associated CD8⁺ Tregs produce IL-10 and inhibit tumor-associated antigen (TAA)-specific T-cell responses. We have shown that these CD8⁺ Tregs may be induced by plasmacytoid DCs [5]. CD4⁺ Tregs also produce IL-10 and mediate immune suppression through IL-10 [6, 8, 9, 19]. In addition to the importance of IL-10 in Treg-mediated suppression, myeloid-derived IL-10 maintains Forkhead box P3 (FoxP3) expression and suppressive function of Tregs in mice with colitis [20]. It indicates that IL-10 is involved in Treg development in the gut. We have also observed that IL-10 is essential in constraining Th17 cell population development in mice and in patients with Crohn’s disease through its control of IL-1 expression (Wilke et al., unpublished data). Indeed, several studies have revealed the roles of IL-10 in mediating intestinal homeostasis through its immune regulatory function [6, 21–24].

Although much of our current knowledge regarding the in vivo immune regulatory functions of IL-10 arises from studies of infectious disease models [25–30], the immune suppressive effects of IL-10 have also been documented in tumor-bearing hosts. In several preclinical models, blockade of IL-10 signaling in tumor or T cells has been shown to enhance antitumor immunity [31–33]. Tumor-infiltrating myeloid cells are capable of producing large amounts of IL-10 [34, 35]. In addition to myeloid cells [34, 35], IL-10 may be derived from tumor-associated neutrophils. A very recent manuscript demonstrated the ability of the acute phase protein serum amyloid A-1 (SAA-1) to induce IL-10-secreting neutrophils that inhibited antigen-specific CD8⁺ T-cell expansion in melanoma patients [36]. IL-10 can suppress NK and NKT cell activity and downregulate MHC molecule expression on tumor cells and DCs, thereby hindering CTL-mediated killing [37–40].Tumor cell-derived IL-10 has been associated with an increased expression of the non-classical HLA class Ib molecule HLA-G. This also obstructs host CTL cytolytic activity [41]. Furthermore, it has been reported that the suppressive effects of IL-10 may be linked to the action of CTLA-4 [42] and cyclooxygenase (COX)-2 [43–45], two other molecules of interest in the study of tumor immunity. Altogether, it is evident that IL-10 can target multiple immune cell subsets and mediate immune suppression through several modes of action.

IL-10 and immunostimulation

The first immune stimulatory function of IL-10 was observed in B cells. Mouse splenic B cells upregulate their expression of MHC II upon treatment with either human or mouse rIL-10 [46]. IL-10 also serves as a survival factor for B cells and increases their antibody production. Early experiments on human B cells demonstrated that IL-10 served as a co-stimulatory factor for B cell proliferation and synergized with IL-4 to expand B cell cultures [47]. Additionally, IL-10 treatment stimulated B cell production of the immunoglobulins (Ig)M, IgG, and IgA, including autoimmune antibodies [48]. A subsequent study showed that IL-10 supported viability of human germinal center B cells and induced synthesis of B-cell lymphoma 2 (Bcl-2) [49], a protein that was already known to play a key role in the rescue of germinal center B cells from apoptosis [50].

IL-10 can also mediate immune stimulatory effects on T cells. MacNeil et al. first showed that IL-10, along with IL-2 and/or IL-4, supported the growth and proliferation of both mature and immature T cells [51]. Shortly thereafter, more specific effects of IL-10 on CD8⁺ T cells were explored. IL-10 was found to be capable of inducing cytotoxic lymphocyte precursors and augmenting cytotoxic function [52]. IL-10 is therefore a growth and differentiation factor for CD8⁺ T cells.

A few laboratories have investigated the stimulatory role of IL-10 in the context of tumor immunity [53]. Suzuki et al. showed that tumor cells transfected with murine IL-10 grew more slowly in vivo and were frequently rejected by the host animals [54]. The following year, IL-10 was discovered to inhibit tumor metastasis in both experimental and spontaneous tumor models via effects on natural killer (NK) cells [55]. Although the precise mechanisms involved were not explored in this paper, a subsequent study from the Fulton group demonstrated that IL-10 downregulated MHC I expression on tumor cells and in doing so, supported NK cell-mediated tumor cell lysis [56]; without expression of self MHC on target cells, the inhibitory signal to NK cells is abolished and lysis ensues [57]. A more recent paper found that recombinant human (rh)IL-10 treatment in immunized mice after tumor challenge significantly enhanced antitumor immunity and vaccine efficacy. Three weeks after IL-10 administration, splenic CD8⁺CD44^hiCD122⁺ (activated memory) T cells had increased and antigen-specific in vitro proliferation was enhanced. Additionally, antigen-specific IFNγ production was increased in animals challenged with tumor and then treated with IL-10 when compared to animals only given tumor challenge. Interestingly, the effect of IL-10 on CTL function could be enhanced by CD4⁺ T-cell depletion. This suggests that IL-10 may have opposing effects on CD8⁺ and CD4⁺ T cells in a tumor model [58] or that the stimulatory effect of IL-10 may be more pronounced in the absence of CD4⁺ Tregs.

IL-10 may also target myeloid-derived suppressor cells (MDSCs) and support tumor immunity. We have recently observed that chemically induced tumors, transplanted tumor growth, and metastasis are increased in IL-10-deficient mice. Ablation of endogenous IL-10 increased the expression of MHC on APCs and promoted the development of MDSCs and Tregs in tumor-bearing hosts; this may explain the surprising phenotype of IL-10-deficient mice [59]. It is of note, then, to recognize that the functions of IL-10 are complex and context-dependent, and in a given system, we may only see the net outcome of multiple effects.

IL-10 and “non-immune effects”

IL-10 has recently been shown to protect neurons from toxin-induced damage. Activation of the IL-10R in spinal cord neurons allowed Il-10 to signal through JAK/STAT and PI3-K pathways, upregulating the antiapoptotic proteins Bcl-2 and Bcl-xL, blocking cytochrome C release and cleavage of caspases. These effects were observed using the in vitro model of glutamine-induced apoptosis [60]. IL-10 is also involved in the process of autophagy, a mechanism whereby cells maintain homeostasis. Starvation and exposure to certain stimuli can induce autophagy or “self-eating” of excess macromolecules or organelles in the cytosol. Park and colleagues recently investigated the effects of IL-10 treatment on starved murine macrophages and demonstrated that IL-10 and IL-10 receptor signaling through the PI3 K pathway inhibited the in vitro induction of autophagy [61]. Further studies are required to determine whether this is also the case in vivo. Studies in our laboratory have demonstrated that IL-10 can suppress tumor angiogenesis through its control of the pro-angiogenic factor IL-1 [59]. However, there is evidence that IL-10 may be pro-angiogenic [62]. In a model of ischemia-induced pathological angiogenesis in mice, IL-10 deficiency resulted in reduced retinal blood vessel formation. Macrophages from IL-10^−/− mice expressed decreased levels of VEGF and NO in comparison with those from wild-type mice [62]. Therefore, IL-10 can directly and indirectly mediate non-immune effects.

IFNγ

IFNγ, originally termed “macrophage activating factor,” was first described (along with IFN-α and IFN-β) as a mediator that interfered with viral replication [63, 64]. IFNγ is produced primarily by NK cells, CD4⁺ and CD8⁺ T cells, and NKT cells [65, 66]. In many of these populations, IL-12 and IL-18 can induce or further increase the production of IFNγ [65, 67–70]. IFNγ plays key roles in host defense (including antiviral and antibacterial defense) and antitumor immunity. Notably, there are different interpretations of how IFNγ contributes to antitumor effects. While some scientists believe that antigen-specific T cells control tumors via secretion of IFNγ [71], others argue that it is rather an innate immune response involved in the control of carcinogen-induced tumors [72]. In contrast to its immune stimulatory role, IFNγ is also capable of downregulating immune responses; in this way, it can minimize immune-mediated tissue and organ damage in the context of autoimmunity and infectious immunity. However, the immune regulatory roles of IFNγ may result in immune evasion and hamper antitumor immunity in the context of malignancy (Table 2).

Table 2.

Interferon-γ

Effect	Mechanism	Context	References
Immune inhibitory effects
Decreases effector T-cell generation and function	Stimulates B7-H1 expression on APC, endothelial and epithelial cells	Cancer	[15, 86, 121–124]
Limits T-cell activation and function	Induces non-cognate MHC class I on tumor cells	Cancer	[150]
Induces T-cell dysfunction	Induces IDO expression in APC and other cell types	Cancer	[125–132]
Supports immunosuppressive functions of tumor-associated myeloid cells	May induce arginase expression in these populations	Cancer	[134–138, 141–145]
Increases tumor evasion of CTL lysis	Increases expression of non-classical MHC and decreases antigen expression on tumor cells	Cancer	[150–152]
Prolongs survival and immune evasion of melanoma cells	IFNγ+ macrophages invade irradiated skin, genes induced in melanocytes by IFNγ include CTLA-4 and non-classical MHC	Cancer	[153–155]
Undermines secondary antitumor CD8+ T response	Mediates contraction of CD4+ T-cell population via apoptosis	Cancer	[154]
Non-immune effects
Tumor suppressor functions, minimizes subsequent sensitivity to IFNg, increasing otential for metastasis	Signaling induces expression of SOCS1	Cancer	[147, 148]
Inhibits vascularization	IFNγ stimulates IP-10 and MIG, angiogenesis inhibitors	Cancer, organ transplantation	[165–168]
Regulates HSC repopulation	IFNγ secretion by immune and other cells	Homeostasis, infection	[96]
Immune stimulatory effects
Decreases susceptibility to infection	Maintenance of TNFα signaling, Ag-specific IgG2a responses, Ag-specific proliferation	Bacterial and viral infection	[90–95]
Improves resistance to chemically induced and spontaneous tumors	IFNγ signaling in tumor cells increases immunogenicity	Cancer	[98–101]
Activates antitumor potential of macrophages	Stimulates IL-12 secretion and tumor cell killing by macrophage	Cancer	[104–106]
Supports proliferation, activation, lytic capability, and production of effector cytokines in NK and NKT cells	IFNγ production by activated NKT cells	Cancer	[107, 108]
Recruits CTLs to tumor environment	Polarizes Th0 to Th1, which recruit CTL to tumor	Cancer	[113]
Blocks Treg expansion	Causes cell cycle arrest in Tregs	Cancer	[116]
Mediates T cell and NK cell tumor infiltration	Stimulates productions of chemokines MIG and IP-10	Cancer	[113, 117, 118]
Decreases tumor growth, promotes antitumor immunity	Mediates angiostasis, maintains tumor cell immunogenicity (dying tumor cells release TAAs)	Cancer	[98, 162, 165–169]
Promotes Th17 cell expansion and function	Induces IL-1 and IL-23	Autoimmune	[86]

IFNγ and immunostimulation

It is well known that IFNγ stimulates MHC and associated chaperone molecule expression on target cells [73–76]. IFNγ induces the replacement of proteasome subunits by “immunoproteasome” subunits, thereby increasing the quantity and diversity of peptides presented on MHC I complexes to CD8⁺ T cells [77]. IFNγ activates the lysosome and increases antibacterial activity in macrophages [78]. It serves as a chemoattractant and repellant for lymphocytes [79], controls isotype switching [80] and antibody production in B cells, and directs the growth and differentiation of several cell types [81–83]. IFNγ is a well-known product of Th1 cells and plays a key role in skewing of naïve T cells to a Th1 phenotype, both through inhibition of IL-4 production [84] and induction of IL-12 secretion [85]. Interestingly, although IFNγ may suppress Th17 cell development from naïve T cells in mice, IFNγ can strongly stimulate the expression of IL-1 and IL-23 by myeloid APCs and promote memory Th17 cell expansion in humans [86]. This may explain why Th1 and Th17 cells coexist in the microenvironments of chronic inflammation and cancer [87]. This phenomenon is functionally linked to the development of human autoimmune pathologies, including psoriasis and chronic bowel diseases [88, 89].

IFNγ is, perhaps, best known for its antiviral and antibacterial effects in infected hosts. In 1993, it was observed that IFNγ receptor-deficient mice had an increased susceptibility to Listeria monocytogenes and vaccinia virus infection [90]. It is clear that IFNγ plays a vital role in viral control of infected hosts [91] and in the establishment of a long-term antiviral state [92]. Perhaps, not surprisingly, patients with IFNγ signaling deficiency are more susceptible to infection [93–95]. This increased sensitivity is associated with disrupted TNFα production, which would normally result from intact IFNγ signaling. Furthermore, IFNγ signaling is required to mobilize hematopoietic stem cells (HSCs) to proliferate and repopulate immune populations in mice with long-term systemic Mycobacterium avium infection. IFNγ-deficient HSCs proliferate more slowly. This indicates that homeostatic levels of IFNγ regulate HSC activity and in turn affect immune cell repopulation and immune responses [96].

There is an increasing body of evidence demonstrating varied roles of IFNγ in antitumor immunity, reviewed in [97] and elsewhere. Early studies showed that progressively growing Meth A fibrosarcoma cells were rejected in lipopolysaccharide-treated mice and that this rejection was abrogated following inhibition of IFNγ signaling [98]. Subsequently, it was found that IFNγ-deficient mice and IFNγ receptor-deficient mice were more susceptible than their wild-type counterparts to methylcholanthrene (MCA)-induced tumor formation [99, 100]. It has been suggested that the cooperation between IFNγ and lymphocytes is crucial for host protection from carcinogen-induced and spontaneous tumorigenesis [101]. However, there is controversy regarding the contribution of lymphocytes and T-cell-derived IFNγ to tumor initiation and immunosurveillance [102, 103]. Nonetheless, IFNγ is a powerful, multi-effector cytokine, derived from innate immune cells and T cells, that plays different roles in the progressive stages of tumor development.

Of course, many contributions of IFNγ to antitumor immunity are due to its direct effects on components of innate and adaptive immunity. As it was originally described, IFNγ has potent macrophage-activating capabilities; it can induce macrophage killing of multiple tumor cell lines [104] and production of IL-12 by macrophages exposed to bacterial components [105, 106]. Interestingly, NK and NKT cells serve as both sources and targets of IFNγ during the antitumor response. Both IFNγ and IL-12 stimulate the proliferation, lytic capability, and production of effector cytokines in NK and NKT cells. IFNγ secreted from activated NKT cells led to NK cell activation in vivo [107, 108]. In further support of this, blockade of IFNγ signaling in RM-1-bearing mice led to enhanced tumor metastasis, while administration of the same antibody to mice deficient in IL-12p40 or NKT cells had no significant effect on the same [100]. Several reports have documented marked effects of IFNγ on components of the adaptive immune system, including CD8⁺ T cells [101, 109, 110]. IFNγ also has a key role in promoting the Th1 polarization of CD4⁺ T helper cells [111, 112], which are crucial in recruiting CTLs to the tumor microenvironment [113]. Early experiments had shown that Stat1^−/− mice could not (in contrast to wild-type mice) reject the growth of a P815 mastocytoma subclone or control the development of another subclone engineered to express IL-12 [114]. Stat1^−/− T cells exposed to IL-12 produced only half as much IFNγ as wild-type T cells and could not lyse tumor cells. Importantly, Stat6^−/− mice could spontaneously reject the same poorly immunogenic subclone of P815, while wild-type mice could not. This rejection was associated with an increased T-cell cytotoxicity and IFNγ production. Because Stat6^−/− mice preferentially polarize T cells toward a Th1 phenotype, these data strongly suggest that IFNγ signaling is crucial in determining the Th1 versus Th2 helper T-cell balance and subsequently the efficacy of the in vivo antitumor immune response [115]. The Ley laboratory more recently documented another antitumor function for IFNγ. Their 2009 report showed that IL-12 treatment blocked in vitro Treg expansion in an IL-12 receptor-dependent fashion. IL-12 stimulated IFNγ-mediated inhibition of Treg cell proliferation and decreased IL-2 production. In these experiments, IFNγ signaling caused cell cycle arrest in Treg cells. The data suggest that this IFNγ-dependent mechanism could counteract the ability of Tregs to protect tumors in cancer patients [116]. In addition to its direct effects on immune activation, IFNγ stimulates the production of chemokines, including IP-10 and MIG. IP-10 and MIG are powerful T and NK cell chemoattractants [117, 118], mediate T cell and NK cell tumor infiltration, and indirectly enhance antitumor immunity [113].

IFNγ and immunoregulation

At the same time that IFNγ was being touted as a promising antitumor agent, evidence began to accumulate in support of regulatory roles of IFNγ. IFNγ appears capable of driving several cellular and molecular mechanisms that may underlie tumor initiation and immunosuppression (Table 2). As the clearest example of the duality of IFNγ occurs in malignancy, we review the regulatory mechanisms and roles of IFNγ in the context of tumor.

Effects of IFNγ on B7-H1 (PD-L1)

We have studied ovarian cancer and ovarian cancer-associated DCs and found that both of these cell types express the inhibitory molecule B7-H1 [15, 119]. IFNγ was the most important stimulus of B7-H1 expression on human APC, epithelial cells, and vascular endothelial cells [15, 86]. It is generally assumed that IFNγ is predominantly produced by the Th1 cell subset. However, we observed that Th17 cells also express high levels of IFNγ [113]. It is possible that IFNγ derived from Th1 cells, Th17 cells, and other populations stimulate B7-H1 expression in the tumor microenvironment, thereby hampering antitumor immunity. We then examined the hepatocellular carcinoma (HCC) environment to determine whether this could be verified. We found an increased expression of the inhibitory molecule B7-H1 on Kupffer cells (KC) within the tumor in comparison with the surrounding tissue and an increased PD-1 expression on CD8⁺ T cells within the tumor microenvironment [120]. Interestingly, B7-H1⁺ KC and CD8⁺ T cells were co-localized in HCC stroma. PD-1⁺CD8⁺ T cells had decreased proliferative ability and effector function when compared with PD-1⁻ T cells, but B7-H1 neutralization recovered effector T-cell function. These data indicate that B7-H1/PD-1 interactions contribute to immune suppression in human HCC, and it is likely that IFNγ signaling contributes to B7-H1 expression in this setting [121, 122]. A 2007 report showed that B7-H1 expression in CD138⁺ multiple myeloma cells was induced or augmented by IFNγ and that stimulation by IFNγ inhibited the generation of CTLs. Interestingly, this expression could be reduced via interruption of MyD88, TRAF6, or STAT1 signaling [123]. An increased B7-H1 expression has also been found on cells of high-risk patients with myelodysplastic syndromes (MDS), which are hematologic stem cell disorders characterized by cytopenia, excessive hematopoetic cell apoptosis, and a high risk of progression to acute myeloid leukemia. IFNγ synergized with TNFα to induce B7-H1 expression on MDS blasts, and these blasts proliferated better than B7-H1⁻ blasts. It should be noted that T cells from MDS patients also expressed an increased PD-1. This suggests that the B7-H1 inhibitory signal can be accurately propagated in such patients, where the interaction of these two molecules may contribute to disease progression [124].

Effects of IFNγ on indoleamine-2,3-dioxygenase (IDO)

IDO function has been well studied and defined in the context of tumors [125–129]. IDO is expressed in several cell types in the tumor microenvironment, where it minimizes effector T-cell access to tryptophan, an essential amino acid. IDO-expressing DCs can induce Treg-mediated expression of B7-H1 on target DCs and effectively downregulate the T-cell arm of the antitumor response [130]. Interestingly, IFNγ strongly induces IDO production, and both DC- and Treg-derived IFNγ can induce IDO [131] and mediate T-cell dysfunction [128, 132].

Effects of IFNγ on arginase

Arginase, an enzyme, metabolizes the amino acid arginine. Environmental arginine plays a central role in the proliferation and expression of the CD3ζ chain in T cells and is crucial in cytokine production [133, 134]. Some studies have revealed a role of IFNγ in the induction of arginase expression [135–138], while others have shown that IFNγ inhibits arginase [139, 140]. While arginase induction by IFNγ may, in fact, be context-dependent, it is important to point out that this enzyme is centrally involved in tumor immunology, where it serves to support the suppressive functions of tumor-associated macrophages [134], DCs [141], and MDSCs [142–145]. Perhaps, it is not surprising that arginase expression in MDSCs can also play a beneficial role in immune modulation; MDSC are capable of mediating the suppression of graft-versus-host disease [146].

IFNγ and SOCS1

One of the most important regulatory roles of IFNγ is through its relationship with suppressor of cytokine signaling-1 (SOCS1). SOCS1 expression is induced by a number of cytokines, including IFNγ. Exposure to IFNγ strongly upregulates SOCS1, which renders cells refractory to further cytokine signaling through its ability to inhibit all the janus kinases (JAKs). SOCS1 acts as a tumor suppressor via various modes of oncogene inhibition, including mediation of the ubiquitination and subsequent degradation of potential cellular oncogenes like Vav1, a guanine nucleotide exchange factor [147]. More recent work has demonstrated that SOCS1 can bind to tyrosine 441 (Y441) of the IFNγ receptor subunit 1(IFNγR1), where it mediates some of its negative effects on IFNγ signal transduction. Mice with mutated Y441 are rendered hypersensitive to IFNγ signaling, displaying greater resistance to lung metastases and a more significant antitumor effect when treated with alpha-galactosylceramide (αGalCer) [148]. These mice also had increased MIG serum levels, which may also have contributed to their antitumor phenotype. Via specific modulation of IFNγ signaling, this study both demonstrates one of the traditional, immunostimulatory roles of the cytokine in tumor biology and offers new insight into the control of IFNγ signaling by SOCS1.

Effects of IFNγ on non-classical MHC, antigen presentation, and APCs

An early study showed that IFNγ-treated tumor cells are more aggressive in vivo in mouse models [149]. Recent studies have demonstrated that this is due to the loss of tumor antigen expression and the induction of non-classical MHC expression, which enable tumor cells to evade CTL-mediated lysis [150–152]. Furthermore, a recent study has shown that UVB irradiation induces an IFNγ signature in murine melanocytes—a signature that includes multiple genes (like non-classical MHC class Ib antigens) known to participate in immune evasion. The source of IFNγ in this study was a population of pro-tumorigenic macrophages infiltrating irradiated skin; IFNγ⁺ macrophages have also been found in a majority of human melanoma cases [153]. These tumor-associated macrophages and myeloid cells function as APCs, mediating immune suppression and prolonging the survival of melanoma cells.

Effects of IFNγ on T cells

The studies described above demonstrate the immunoregulatory roles of IFNγ on innate immune components including APCs. While IFNγ is broadly known as a cytokine that supports T-cell immunity in the context of malignancy, there is evidence that this cytokine can also undermine the antitumor response by negatively affecting CD4⁺ T cells. In 2007, the laboratory of William Murphy reported that after various immunotherapies, IFNγ mediated contraction of the CD4⁺ T-cell population via the induction of apoptosis [154]. Because CD4⁺ T cells are essential for the most effective secondary antitumor CD8⁺ T-cell response, this intriguing observation should be considered in the context of both patient therapy and clinical trial design.

These immune regulatory roles are consistent with several studies in patients with cancer. Intratumoral expression of IFNγ was shown to be associated with a more aggressive phenotype in human melanomas [155]. Most clinical trials of IFNγ treatment did not demonstrate clinical efficacy in patients with melanoma [156–159]. Furthermore, multiple clinical trials have demonstrated that IFNγ-treated patients fare worse than untreated patients [160, 161]. Therefore, it is reasonable to point out that the “dark side” of IFNγ action is operative in patients with cancer (Table 2).

IFNγ and “non-immune effects”

Some functions of IFNγ occur outside of the immune system—these include the arrest of cell growth, induction of apoptosis, and inhibition of angiogenesis. Many of these functions target tumor cell lines: cells made unresponsive to IFNγ via genetic manipulation grew more rapidly in vitro and in vivo [98, 162]. This was partially due to a reduction in tumor cell immunogenicity, but also to a decrease in IFNγ-induced angiostasis within the tumor microenvironment. IFNγ has been shown to inhibit proliferation of both human fibrosarcomas and mouse fibroblasts [163, 164], but had no effect on human or murine cells deficient in the Stat1 component of IFNγ signaling pathway. IP-10, the first angiostatic chemokine identified as a product of IFNγ signaling, is a powerful T-cell chemoattractant [117, 118] that works in combination with lymphocytes and other IFNγ-induced chemokines to inhibit tumor vasculogenesis and effect transplanted tumor rejection [165–168]. It is interesting to postulate that a subset of progressive tumors develops insensitivity to IFNγ: Certain human tumor cell lines are less sensitive or completely non-reactive toward this cytokine [99]. Quite recently, a report by Briesemeister and colleagues established that tumor initiation can be prevented by local production of IFNγ, but that vascularized tumors are harder to reject the longer they are established. This finding was explained by the reduction in endothelial cells mediated by IFNγ in the local environment and the resulting necrosis of the tumor in question [169]. Therefore, IFNγ- and IFNγ-induced type-1 chemokine IP-10 can directly inhibit tumor angiogenesis and result in tumor cell death. This effect would subsequently promote antitumor immunity due to the increased TAA release from dying tumor cells.

Concluding remarks

In summary, it is important to realize the pluripotency of certain cytokines and refrain from categorizing them according to their main or “classical” function. Because prototypical cytokines involved in inflammation may, in other environments, serve to downregulate immune components, and signature immunosuppressive proteins may also activate certain facets of immunity, it is important to consider specific environmental and chronological contexts. For example, IL-10 is predominantly immunosuppressive in many disease models, particularly infectious disease, where APCs serve as its principal targets [170]. Toll-like receptors are activated, and antigen-specific T-cell priming is induced. In a tumor-bearing host, however, when there is no obvious acute phase of immune response (tumor development is chronic) [127], TAA-specific T cells may slowly be primed, activated, and expanded, along with the accumulation of immune suppressive cell populations, and IL-10 may be gradually induced, suppressing inflammatory cytokines and MDSC accumulation. In this context, IL-10 may promote the antitumor immune response. Although IFNγ is a potent immune effector cytokine, it also induces B7-H1, IDO, arginase, and non-classical MHC expression, which may participate in a feedback mechanism to efficiently downregulate the immune response and avoid tissue and organ damage. In this way, it is easier to imagine such contrasting effects of a particular protein from one disease to another and from the initial stages to the endgame of a single malady. An appreciation for the varied capabilities of individual cytokines will no doubt aid in a more holistic understanding of several human illnesses and perhaps contribute to chronological changes in some cytokine-specific clinical therapies.