Interferon-Regulatory Factors Determine Macrophage Phenotype Polarization

The mononuclear phagocyte system regulates tissue homeostasis as well as all phases of tissue injury and repair. To do so changing tissue environments alter the phenotype of tissue macrophages to assure their support for sustaining and amplifying their respective surrounding environment. Interferon-regulatory factors are intracellular signaling elements that determine the maturation and gene transcription of leukocytes. Here we discuss how several among the 9 interferon-regulatory factors contribute to macrophage polarization.


Introduction
During development mononuclear phagocyte progenitors populate most tissues where they differentiate into transcriptionally and functionally diverse phenotypes [1][2][3]; for example, bone marrow, liver, and lung harbor macrophages with an enormous capacity to clear airborne particles, gutderived pathogens, or cell nuclei expelled from erythroblasts, respectively [4]. In contrast, skin, kidney, and brain host a dense network of dendritic cells [4,5]. Upon tissue injury M-CSF drives resident mononuclear phagocyte to proliferate [6] or circulating monocytes recruit to the site of injury. It is the local microenvironment that then determines mononuclear phagocyte polarization to distinct phenotypes, which can vary between disorders or between the different stages of a disease process [7]. Several factors mediate mononuclear phagocyte polarization, as being mostly described by in vitro experiments [7,8]. However, attempts to translate this simplistic model to disease states in vivo often failed to cover all aspects of heterogeneous and changing tissue environments. For example, ischemia-reperfusion injury induces transient sterile inflammation because dying tissue cells release damage-associated molecular patterns (DAMPs) that polarize macrophages toward a classically activated M1-like phenotype [9,10]. This process is associated with NF-B and STAT1 pathway activation [2]. Macrophages apoptosis or their phenotype switches towards alternatively activated, M2-like macrophages that produce IL-10 and TGF-, induce resolution of inflammation, and enforce tissue regeneration [11][12][13][14][15]. Failure of this phenotype switch leads to persistent tissue inflammation, atrophy, and fibrosis [16]. The uptake of neutrophils, epithelium-derived alarmins, and Th2 cytokines IL-4 and IL-13 supports this phenotype switch [11]. As disease processes do not always occur in a serial manner, concomitant proinflammatory and anti-inflammatory macrophages infiltrates often populate organs affected by persistent injury, for example, in slowly progressive lesions of organ transplants [17,18].
Current data suggest that the family of the interferonregulatory factors (IRFs) plays an important role in regulating macrophage polarization. IRFs are intracellular proteins that regulate immune cell maturation [19]. Here we provide a summary on IRF biology that is focused on the IRF's role in macrophage phenotype control and the associated contributions to tissue inflammation and remodeling.

The Family of Interferon-Regulatory Factors
The IRFs were discovered in search of transcription factors that bind to the conserved virus response elements within the promoters of type I IFN genes [19]. It was found that both NF-B and IRF-3 activate IFN-gene transcription while IFNgene expression is entirely based on IRFs [19]. The generation of Irf -deficient mice led to the discovery of additional regulatory roles of the IRFs for cell growth, for immune cell maturation and activation, and for apoptosis. In mammals the IRF gene family consists of nine members: IRF-1, IRF-2, IRF-3, IRF-4, IRF-5, IRF-6, IRF-7, IRF-8/ICSBP, and IRF-9. Their respective IRF proteins share significant homologies at the N-terminal 115 amino acids where they share a conserved tryptophan pentad repeat DNA-binding domain [20]. These include a DNA-binding domain of five tryptophan repeats of which three recognize the GAAA and AANNNGAA sequence motifs, that is, the IFN-stimulated response elements [20]. However, the variable domains at the C-terminus determine the functional specificity of the nine IRFs, their potential to interact with each other via IRF-association domains, and their cell type-specific actions [21] (Figure 1). Accordingly, the IRFs have been subdivided into the "interferonic" IRFs (IRF-2, -3, -7, and -9), the "stress-responsive" IRFs (IRF-1 and -5), the "hematopoietic" IRFs (IRF-4 and -8), and the "morphogenic" IRF-6 [22]. The genetic and biological characteristics of the IRF family members are listed in Table 1.
3.3. IRF-3. IRF-3 was discovered by searching genes with homology sequences with IRF-1 and IRF-2 [50]. This 427amino acid protein shares a number of characteristics with IRF-7 [51]. Unlike IRF-7, that confers MyD88 signaling, IRF-3 is involved in TRIF-dependent signaling pathways. After binding pathogens, pattern-recognition receptors like TLR-3, TLR-4, or RIG-I recruit TRIF to trigger an IRF-3-mediated induction of type I IFNs [52][53][54][55]. Additional cytoplasmic DNA recognition receptors use the STING pathway to activate IRF-3 [56]. The transcriptional activation of the IFNgene requires an enhanceosome of 7 additional proteins that create a continuous surface that recognizes the DNA-binding element [57]. Phosphorylation of TLR3's specific tyrosine residues can initiate two distinct signaling pathways. One activates TBK-1 and the other activates PI3 kinase and Akt for full phosphorylation and activation of IRF-3 [58,59]. Cytoplasmic IRF-3 is inactive unless phosphoactivation of IRF-3 triggers unfolding of the autoinhibitory elements and exposes the hydrophobic surface to interaction with CREBBP to translocate to the nucleus [60]. By contrast, ubiquitination inactivates IRF-3 [61]. GM-CSF-primed M1-like macrophages display a diminished IRF-3 axis and enhanced activation of MyD88. In contrast, M-CSF stimulated macrophages that develop an M2-like phenotype show defective NF-B activation and enhanced TRIF-mediated IRF-3 induction upon LPS stimulation [62,63]. Hence, the IRF-3 axis is rather enabled in M2-like macrophages than in M1-like macrophages. But does IRF-3 also contribute to the development of an alternatively activated macrophage phenotype? One study transduced IRF-3 into primary human microglia. Stimulation with IFN-/IL-1 suppressed proinflammatory mediators like IL-6, TNF-, or IL-1 , whereas anti-inflammatory mediators including IL-10 were enhanced [64]. Altogether the data suggest that IRF-3 is associated with anti-inflammatory microenvironments and contributes to the polarization toward a M2 macrophage phenotype. However, IRF-3 also induces a number of inflammatory cytokines such as CCL5 and IFN- [65].
3.5. IRF-5. IRF-5 is a 504-amino acid-long stress-responsive IRF [22]. IRF-5 is required for TLR-mediated induction of IL-6, TNF, IL-12, and other proinflammatory cytokines [74]. IRF-5 competes with IRF-4 for binding to the signaling adapter MyD88 and the downstream subsequent activation of proinflammatory transcription factors [27]. Its capacity to induce inflammatory cytokines and B cell transcription factors implies its role in host defense and autoimmune disorders [75,76]. This competitive interaction involves IRF-5 in the polarization into M1 macrophages [68]. In fact, the balance between IRF-4 and IRF-5 seems to be a major determinant of M1 versus M2 macrophage polarization. For example, M-CSF induces IRF-4 in human monocyte-derived macrophages while GM-CSF induces IRF-5, which results in two phenotypically different macrophage phenotypes [77]. M1 macrophages express high levels of IRF-5 where it not only mediates the expression of proinflammatory cytokines but also suppresses the immunoregulatory cytokine IL-10 [68].
IRF-5 itself is regulated by the transcriptional corepressor KAP1/TRIM28 to avoid overshooting secretion of TNF and other mediators that induce immunopathology [78]. KAP1/TRIM28 regulates IRF-5 by recruiting histone deacetylases and methyltransferases that can silence IRF-5-related gene expression [78]. IRF-5-mediated polarization of monocytic phagocytes involves the secretion of various IL-12 family members including IL-12p35 and IL-23p19, which support Th17 T cell immunity, an element of adaptive immunity that contributes to autoimmune disorders [79]. In fact, gain of function mutations in the IRF-5 gene exists that increases TLR-or NOD-mediated secretion of proinflammatory cytokines [80]. Such variants also predispose to autoimmune diseases like systemic lupus erythematous [81][82][83], which may be related to these phenomena.
3.6. IRF-6. IRF-6 is a so-called "morphogenic" IRF of 467 amino acid length. IRF-6 has a large structural homology with IRF-5 but does not seem to share its functional properties or contribute to macrophage biology, which is related to the tissue-specific expression of IRF-6. IRF-6 mutations rather predispose to cleft lip or palate and other abnormalities of limb, skin, and craniofacial morphogenesis [84,85]. [19,86,87]. IRF-7 drives the differentiation of monocytes to macrophages but a direct role in macrophage polarization has not been reported.
3.9. IRF-9. IRF-9 is a 424-amino acid-long regulator of type I IFN signaling. It forms a DNA-binding complex with  the STAT1 homodimer, for example, for the induction of CXCL10 [95]. A specific role in macrophage polarization has not been reported.

Summary and Perspective
Macrophages contribute to tissue homeostasis and all phases of tissue injury and repair. Tissue environments prime macrophages to distinct phenotypes to assure that their functional properties enforce the surrounding environment, whether this may be inflammation, the resolution of inflammation, tissue repair (fibrosis), or the resolution of extracellular matrix. Members of the IRF family are an integral component of the macrophage polarization process and, hence, regulate the phenotypic plasticity and heterogeneity of tissue macrophages. Research in this area is still in progress, but our present working model refers to IRF-1, IRF-5, and IRF8 as factors driving the proinflammatory, classically activated (M1) macrophage phenotype, while IRF-3 and IRF-4 promote anti-inflammatory, alternatively activated (M2) macrophages ( Figure 2). Future work in this area will certainly refine this concept and define additional functions of the IRF's in this context and elucidate additional mechanisms of how changing tissue environments shape immune effector cells to meet the tissue needs in homeostasis and disease.

Conflict of Interests
All authors do not have competing interests and do not have a financial relation to profiting companies or commercial products.