Frequent Loss of IRF2 in Cancers Leads to Immune Evasion through Decreased MHC Class I Antigen Presentation and Increased PD-L1 Expression.

To arise and progress, cancers need to evade immune elimination. Consequently, progressing tumors are often MHC class I (MHC-I) low and express immune inhibitory molecules, such as PD-L1, which allows them to avoid the main antitumor host defense, CD8+ T cells. The molecular mechanisms that led to these alterations were incompletely understood. In this study, we identify loss of the transcription factor IRF2 as a frequent underlying mechanism that leads to a tumor immune evasion phenotype in both humans and mice. We identified IRF2 in a CRISPR-based forward genetic screen for genes that controlled MHC-I Ag presentation in HeLa cells. We then found that many primary human cancers, including lung, colon, breast, prostate, and others, frequently downregulated IRF2. Although IRF2 is generally known as a transcriptional repressor, we found that it was a transcriptional activator of many key components of the MHC-I pathway, including immunoproteasomes, TAP, and ERAP1, whose transcriptional control was previously poorly understood. Upon loss of IRF2, cytosol-to-endoplasmic reticulum peptide transport and N-terminal peptide trimming become rate limiting for Ag presentation. In addition, we found that IRF2 is a repressor of PD-L1. Thus, by downregulating a single nonessential gene, tumors become harder to see (reduced Ag presentation), more inhibitory (increased checkpoint inhibitor), and less susceptible to being killed by CD8+ T cells. Importantly, we found that the loss of Ag presentation caused by IRF2 downregulation could be reversed by IFN-stimulated induction of the transcription factor IRF1. The implication of these findings for tumor progression and immunotherapy are discussed.

Differential requirements for IRF-2 in generation of CD1d-independent T cells bearing NK cell receptors.

NK cell receptors (NKRs) such as NK1.1, NKG2D, and Ly49s are expressed on subsets of CD1d-independent memory phenotype CD8(+) and CD4(-)CD8(-) T cells. However, the mechanism for the generation and functions of these NKR(+) T cells remained elusive. In this study, we found that CD1d-independent Ly49(+) T cells were reduced severely in the spleen, bone marrow, and liver, but not thymus, in mice doubly deficient for IFN regulatory factor-2 (IRF-2) and CD1d, in which the overall memory phenotype T cell population was contrastingly enlarged. Because a large fraction of Ly49(+) T cells coexpressed NK1.1 or NKG2D, the reduction of Ly49(+) T cells resulted indirectly in underrepresentation of NK1.1(+) or NKG2D(+) cells. Ly49(+) T cell deficiency was observed in IRF-2(-/-) mice additionally lacking IFN-α/ßR α-chain (IFNAR1) as severely as in IRF-2(-/-) mice, arguing against the involvement of the accelerated IFN-α/ß signals due to IRF-2 deficiency. Rather, mice lacking IFN-α/ßR alone also exhibited relatively milder Ly49(+) T cell reduction, and IL-2 could expand Ly49(+) T cells from IFNAR1(-/-), but not from IRF-2(-/-), spleen cells in vitro. These results together indicated that IRF-2 acted in Ly49(+) T cell development in a manner distinct from that of IFN-α/ß signals. The influence of IRF-2 deficiency on Ly49(+) memory phenotype T cells observed in this study suggested a unique transcriptional program for this T cell population among other NKR(+) T and memory phenotype T cells.

Interferon regulatory factor-2 protects quiescent hematopoietic stem cells from type I interferon-dependent exhaustion.

Type I interferons (IFNs), a family of cytokines, orchestrate numerous biological and cellular processes1, 2, 3. Although it is well known that type I IFNs are essential for establishing the host antiviral state4, their role in hematopoietic homeostasis has not been studied. Here we show that type I IFNs induce proliferation and exhaustion in hematopoietic stem cells (HSCs) and that interferon regulatory factor-2 (IRF2), a transcriptional suppressor of type I IFN signaling5, 6, preserves the self-renewal and multilineage differentiation capacity of HSCs. HSCs were substantially less abundant in the bone marrow of Irf2-/- as compared to Irf2+/- mice. Irf2-/- HSCs showed enhanced cell cycling status and failed to produce hematopoietic cells in competitive repopulation assays, and the reconstituting capacity of Irf2-/- HSCs was restored by disabling type I IFN signaling in these cells. In wild-type mice, injection of poly(I:C), an inducer of type I IFN signaling, or IFN- induced HSC proliferation, and chronic type I IFN signaling further reduced the number of quiescent HSCs. Notably, combined poly(I:C) and 5-fluorouracil (5-FU) treatment allowed exogenous HSC engraftment and hematopoietic reconstitution in WT mice. Our findings provide insight into the molecular basis for the maintenance of HSC quiescence and may lead to improvements in bone marrow transplantation and type I IFN-based therapies for viral infection and cancer.

Homeostatic erythropoiesis by the transcription factor IRF2 through attenuation of type I interferon signaling.

OBJECTIVE: Erythrocyte production is tightly regulated by cytokines, particularly erythropoietin (EPO), which affects expansion and viability of erythroid lineage cells via induction of several factors, including Bcl2-like 1 (Bcl-XL). Because type I interferon (IFN) is known to inhibit erythropoiesis, we studied mice deficient in the gene for interferon regulatory factor 2 (IRF2), which functions as a negative regulator of type I IFN signaling, in the context of erythropoiesis regulation. MATERIALS AND METHODS: We performed hematologic analyses and detected normocytic anemia in Irf2-deficient mice. RESULTS: Assessment of the maturation of erythroid progenitors in Irf2-deficient bone marrow by flow cytometry revealed a decreased number of late erythroblasts accompanied by an increased number of early erythroid progenitors. Irf2-deficient mice manifested elevated serum EPO levels, decreased Bcl-XL expression levels and enhanced apoptosis of erythroblasts, which may account for the decreased number of late erythroblasts. We further assessed the role of IRF2 in the regulation of type I IFN signaling during erythropoiesis, and found that additional homozygous mutation of IFNAR1, a subunit of type I IFN receptor complex, led to rescue of the defect of erythropoiesis in Irf2-deficient mice. CONCLUSIONS: Impaired erythropoiesis in Irf2-deficient mice results from excessive type I IFN signaling, which inhibits Bcl-XL expression in erythroid lineage cells. Our present study provides a mechanistic understanding of the potential cross-talk between type I IFN and EPO signaling pathways during erythropoiesis and may offer therapeutic insights into anemia.

IFN regulatory factor-2 regulates macrophage apoptosis through a STAT1/3- and caspase-1-dependent mechanism.

IFN regulatory factor (IRF)-2(-/-) mice are significantly more resistant to LPS challenge than wild-type littermates, and this was correlated with increased numbers of apoptotic Kupffer cells. To assess the generality of this observation, and to understand the role of IRF-2 in apoptosis, responses of peritoneal macrophages from IRF-2(+/+) and IRF-2(-/-) mice to apoptotic stimuli, including the fungal metabolite, gliotoxin, were compared. IRF-2(-/-) macrophages exhibited a consistently higher incidence of apoptosis that failed to correlate with caspase-3/7 activity. Using microarray gene expression profiling of liver RNA samples derived from IRF-2(+/+) and IRF-2(-/-) mice treated with saline or LPS, we identified >40 genes that were significantly down-regulated in IRF-2(-/-) mice, including Stat3, which has been reported to regulate apoptosis. Compared with IRF-2(+/+) macrophages, STAT3alpha mRNA was up-regulated constitutively or after gliotoxin treatment of IRF-2(-/-) macrophages, whereas STAT3beta mRNA was down-regulated. Phospho-Y705-STAT3, phospho-S727-STAT1, and phospho-p38 protein levels were also significantly higher in IRF-2(-/-) than control macrophages. Activation of the STAT signaling pathway has been shown to elicit expression of CASP1 and apoptosis. IRF-2(-/-) macrophages exhibited increased basal and gliotoxin-induced caspase-1 mRNA expression and enhanced caspase-1 activity. Pharmacologic inhibition of STAT3 and caspase-1 abolished gliotoxin-induced apoptosis in IRF-2(-/-) macrophages. A novel IFN-stimulated response element, identified within the murine promoter of Casp1, was determined to be functional by EMSA and supershift analysis. Collectively, these data support the hypothesis that IRF-2 acts as a transcriptional repressor of Casp1, and that the absence of IRF-2 renders macrophages more sensitive to apoptotic stimuli in a caspase-1-dependent process.