Epigenetic and Transcriptional Regulation of IRAK-M Expression in Macrophages Free

Effective recognition and elimination of pathogenic microorganisms is vital in mammals, and tight control of these mechanisms is essential for homeostasis. Hence, innate immune responses have evolved to protect organisms from bacteria, fungi, and viruses using TLRs that recognize pathogen-associated molecular patterns on microorganisms and initiate proinflammatory responses (1). Bacterial LPS is sensed by TLR4 (2) on the surface of phagocytes to initiate the signaling pathway that leads to the activation of MyD88/PI3K/AKT and NF-κB (3).

TLR4 signal transduction is mediated by a family of serine/threonine kinases called IL-1R–associated kinases (IRAKs) (4). IRAK-3, also known as IRAK-M, is the only member of the IRAKs that lacks kinase activity; hence, it acts as a suppressor of the TLR4 signaling pathway (5). This renders IRAK-M important for dampening the inflammatory response after pathogen elimination and is indispensable for the establishment for endotoxin tolerance (5). The role of IRAK-M in human disease also was established. The importance of IRAK-M in disease development and regulation was highlighted in studies using IRAK-M^−/− mice. In addition to its contribution to endotoxin tolerance, IRAK-M is a key player in different host defense mechanisms, including viral and bacterial pneumonia (6–8). In humans, IRAK-M was shown to be a key player in autoimmune diseases, such as systemic lupus erythematosus (9). IRAK-M^−/− NOD mice are more prone to the development of type I diabetes and the related glucose intolerance as a result of the altered function of dendritic cells (10). IRAK-M is also involved in obesity-induced atherogenesis (11), obesity-induced metabolic inflammation (12), and adiponectin-induced regulation of macrophage sensitivity to LPS (13), suggesting that it can be activated under different conditions and diverse downstream signaling cascades.

Despite its importance and the relatively good understanding of IRAK-M function, little is known about the transcriptional regulation of the IRAK-M gene. The IRAK-M promoter is not fully characterized, and the only firm association between specific transcription factors and IRAK-M expression was shown in promoter-cloning assays (14). According to that report, the AP1 complex and NF-κB, both central transcriptional complexes induced upon TLR4 activation, were shown to bind to and activate the IRAK-M promoter.

Transcriptional regulation of genes involves transcription factors (activators and repressors) (15), as well as chromatin-modifying enzymes (histone methyltransferases, histone acetyl transferases, histone deacetylases, and histone demethylases) that introduce epigenetic marks on chromatin to activate or suppress transcription (16). One major silencing epigenetic mark is the trimethylation of lysine 27 on histone 3 (H3K27me3), which is catalyzed by the enhancer of Zeste 2 (Ezh2) methyl transferase, part of the polycomb recessive complex 2 (PRC2) (17). In contrast, this methylation mark can be removed by the histone demethylase ubiquitously transcribed tetratricopeptide repeat X chromosome (UTX) (18–21), thus activating transcriptionally silenced regions. Macrophage activation by LPS promotes genome-wide chromatin modifications to allow transcriptional profile changes that promote endotoxin tolerance (22, 23). No information is available on potential epigenetic regulation of the IRAK-M gene.

In this study, we used an small interfering RNA (siRNA)-knockdown approach to analyze the transcriptional and epigenetic regulation of the IRAK-M gene. siRNA knockdown and chromatin immunoprecipitation (ChIP) experiments showed that CCAAT/enhancer binding protein (C/EBP)β, a master regulator of macrophage function, is a transcriptional activator of IRAK-M. In addition, we demonstrated that, in naive unstimulated macrophages, the IRAK-M promoter is methylated on histone 3 lysine 27 (H3K27), and C/EBPβ did not bind on the promoter. Upon LPS stimulation, the H3K27me3 silencing mark was removed, and C/EBPβ was recruited on the IRAK-M promoter to induce its transcription.

RAW 264.7 murine macrophages were cultured in DMEM (Life Technologies, Carlsbad, CA) supplemented with 10% (v/v) heat-inactivated FBS, 10 mM l-glutamine, 100 IU/ml penicillin, and 100 μg/ml streptomycin.

Escherichia coli–derived LPS (100 ng/ml) (O111:B4, catalog no. L2630; Sigma-Aldrich) was used as described in the 10Results. Cells were stained by immunofluorescence or harvested, and cell pellets were washed in PBS and stored at −80°C for total RNA extraction or Western blot analysis.

A total of 5000 RAW 264.7 mouse macrophages was plated per well in 384-well plates (Greiner Bio-One) containing 30 μl of DMEM. Cells were starved for 10–12 h, transfected with 30 nM of the specified siRNA (Supplemental Table I) using Lipofectamine RNAi Max (Invitrogen-Life Technologies, Carlsbad, CA), according to manufacturer’s instructions for 24 h, and incubated in the presence or absence of 100 ng/ml LPS (L2630; Sigma-Aldrich) for the indicated times. Cells were fixed with 4% paraformaldehyde, blocked with blocking buffer (5% FBS, 0.1% Triton X-100, 0.01% saponin, 1× PBS) for 30 min, and incubated with primary Ab overnight at 4°C and with secondary Ab (including counterstain with DAPI-Hoechst) for 2 h at room temperature (RT). Cells were washed five times with 1× PBS between each step using a microplate washer. Images were acquired with Autoscope (MetaMorph Software), and signal intensities (sum of all positive pixels divided by the number of cells) were measured and analyzed using Definiens Developer XD (24). Finally, scatter plots were generated using TIBCO Spotfire software.

One million RAW 264.7 mouse macrophages per sample were plated in 24-well plates containing 0.5 ml of DMEM. Cells were starved for 10–12 h, transfected with 30 nM of a specified siRNA using Lipofectamine RNAi Max, according to the manufacturer’s instructions (Invitrogen-Life Technologies, Carlsbad, CA) for 24 h, and incubated in the presence or absence of 100 ng/ml of LPS for the indicated times. Wells were washed with ice-cold PBS, and RNA was isolated with the InviTrap RNA Cell HTS 96 Kit, according to the manufacturer’s instructions (STRATEC). Reverse transcription was performed with an Applied Biosystems High-Capacity cDNA Reverse Transcription Kit (using 1 μg of isolated RNA as template), according to the manufacturer’s instructions. Real-time PCR was performed with the SensiMix SYBR Hi-ROX Kit (Bioline), according to the manufacturer’s instructions. Amplification was performed in an ABI PRISM 7900HT Fast Real-Time PCR System with 384-Well Block Module for a maximum of 45 cycles, as follows: start steps of 30 min at 50°C and 15 min at 95°C and repeat steps of 15 s at 95°C, 30 s at 60°C, and 30 s at 72°C. The primers used were EZH2 forward: 5′-GTGCAGTTATTCCTTCCATGC-3′ and reverse: 5′-ACGCTCAGCAGTAAGAGCAG-3′; UTX forward: 5′-GCACCACCTCCAGTAGAACAA-3′ and reverse: 5′-GTCTCATTTGGTGTTGCTGCAT-3′; C/EBPb forward: 5′-GACAAGCTGAGCGACGAGTA-3′ and reverse: 5′-GCTTGAACAAGTTCCGCAGG-3′; actin-b forward: 5′-GTCATCACTATTGGCAACGAGC-3′ and reverse: 5′-GCACTGTGTTGGCATAGAGGTC-3′; and IRAK-M forward: 5′-TTCCTGGCACGTTCGAATCA-3′ and reverse: 5′-CGCTGCAGCAAAATCCGTTA-3′.

Chromatin from 15 × 10⁶ RAW 264.7 cells or primary peritoneal thioglycollate-elicited macrophages was prepared by fixation of the cell culture with 1/10th volume of formaldehyde-containing buffer (11% formaldehyde, 100 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 50 mM HEPES [pH 8]) and incubation for 10 min at RT. For quenching the cross-linking, glycine was added for 5 min at RT, to a final concentration of 125 mM, directly to the culture media. Scraping and transfer of cells to a 15-ml conical tube and two washes with ice-cold 1× PBS (supplemented with 1 mM PMSF) followed, and the cell pellet was incubated with 10 ml of cell lysis buffer (5 mM PIPES [pH 8], 85 mM KCl, 0.5% Nonidet P-40, 1 mM PMSF, complete protease inhibitors) for 10 min on ice. Cells were lysed with 800 μl of SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl [pH 8.1]) for 10 min on ice. Chromatin was sonicated to 500–1000 bp long. Immunoprecipitation (IP) was performed with the equivalent of 3–4 × 10⁶ cells per sample, diluted 10 times with ChIP dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl [pH 8], 167 mM NaCl supplemented with protease inhibitors), and 5 μg of each Ab. Samples were rotated at 4°C overnight (1% of chromatin input was kept). The following day, each sample was mixed with 20 μl of magnetic beads and rotated for 2 h at 4–C. Immunoprecipitated material was incubated for 5 min with each of the following buffers: low-salt Wash Buffer A (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8], 150 mM NaCl supplemented with protease inhibitors), high-salt Wash Buffer B (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.8], 500 mM NaCl supplemented with protease inhibitors), Buffer C (20 mM Tris HCl [pH 8], 250 mM LiCl, 1 mM EDTA, 0.5% Nonidet P-40, 0.5% Na-Deoxycholate, 0.5 M PMSF supplemented with protease inhibitors), and Tris/EDTA (pH 8). After the last wash, samples were incubated with proteinase K (200 mg/ml), 0.5% SDS in Tris/EDTA for 2 h at 55°C and incubated overnight at 65°C for the reversion of formaldehyde cross-links. DNA was purified with phenol/chloroform extraction and was precipitated with ethanol (1% glycogen and 10% CH₃COONa). Immunoprecipitated DNA was resuspended in 40 μl of 100 mM Tris-HCl (pH 7.5). A total of 5% of the immunoprecipitated DNA was used in quantitative PCR analysis, and data were normalized using the following formula; 100*2^(adjusted input − cycle threshold IP). The following Abs were used in the ChIP experiments: C/EBPβ Ab (C-19) (sc-150X; Santa Cruz Biotechnology), anti–histone H3 Ab (D2B12) (#4620 Cell Signaling Technology; chip formulated), anti–trimethyl-Histone H3 (Lys27) Ab (cat. no. 07-449; Upstate), anti-SMAD4 Ab (sc-7966X; Santa Cruz Biotechnology), anti-RelB Ab (sc-48366X; Santa Cruz Biotechnology), and anti-p65 Ab (sc-8008X; Santa Cruz Biotechnology). The primers used for ChIP PCR were IRAK-M promoter (distal region) forward: 5′-GCCCGATTGAGAGTAGGGTAAG-3′ and reverse: 5′-TCTAACAGAAGGGTATGCGAGC-3′; IRAK-M promoter (proximal region) forward: 5′-AACCTCTTAGATCCATCGTGGC-3′ and reverse: 5′-TGTTATGAACCGTTCTGTCCGT-3′; and IRAK-M promoter (middle region) forward: 5′-GTGGGTACTGAATGCCCAGG-3′ and reverse: 5′-GTGGGAAAATGAAGTGGGAGA-3′.

RAW 264.7 cells were lysed in RIPA buffer (10 mM Tris [pH 8], 10 mM EDTA [pH 8], 140 mM NaCl, 1% Triton, 1% Na deoxycholate, 0.1% SDS) containing protease inhibitors (Roche). Lysates were electrophoresed through a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane. Membranes were processed according to standard Western blotting procedures: 1 h blocking in 5% BSA, overnight incubation of primary Abs, 1 h incubation with HRP-conjugated secondary Abs, visualization with ECL system (Pierce) using a ChemiDoc XRS+ System (Bio-Rad), and quantitation of protein by band intensity using Image Lab Software (Bio-Rad). Abs used were anti-EZH2 (5246S, clone D2C9; Cell Signaling Technology) and anti-tubulin (clone 1A2; Sigma-Aldrich).

HEK 293T cells were cultured in DMEM supplemented with 10% v/v heat-inactivated FBS (10270-106) and 1% penicillin/streptomycin (15070-063; all from Life Technologies). Cells were grown at 37°C in a 5% CO₂ incubator. HEK 293T cells growing on a 10-cm plate were transfected using 15 μl TurboFect Transfection Reagent (R0531; Thermo Scientific) and a mix of packaging, envelope, and transfer DNA vector (2, 1, and 3 μg respectively). pVSVG envelope plasmid and pΔ 8.1 packaging plasmids were kindly provided by Dr. C. Stournaras. lentiCRISPRv2 plasmid (cat. no. 52961; Addgene) was used as control transfer DNA vector. Oligo1 (5′-CACCGTGAGACTGAGACAGCTCAAG-3′) and oligo2 (5′-AAACCTTGAGCTGTCTCAGTCTCAC-3′) were hybridized and cloned in lentiCRISPRv2 vector in the BsmbI cloning position, according to the manufacturer, and used as small-guide RNA Ezh2 containing transfer DNA vector. After 24 h of transfection, the medium was replaced with normal growth medium. At 48 and 72 h posttransfection, the medium containing the viral particles was collected and filtered using a 0.45-μm filter and used right away or stored at −80°C. The collected virus-containing medium was used 1:1 with normal growth medium supplemented with 10 μg/ml Polybrene to infect RAW 264.7 macrophages. After 24 h of infection, medium was replaced with normal growth medium. Forty-eight hours after the infection, 2 μg/ml puromycin was added to the medium. Infected macrophages were selected following exposure to puromycin for 5–8 d and then used for treatments.

Data are expressed as mean ± SD from independent experiments. Statistical analysis was performed using GraphPad Prism version 5.00 for Windows (GraphPad, San Diego, CA). Groups were compared using the t test. The p values < 0.05 were considered statistically significant.

IRAK-M is induced upon endotoxin tolerance as a negative regulator of TLR4 signaling (5), but the molecular mechanism of its transcriptional activation is not fully understood. RAW 264.7 macrophages were used to identify potential regulators of IRAK-M in the naive state and upon LPS stimulation. To determine the optimal dose for IRAK-M induction, we treated macrophages with different doses of LPS for 48 h. The results showed that IRAK-M was potently induced at concentrations > 100 ng/ml (Fig. 1E), indicating that it contributes primarily to repression of macrophage responses rather than training (25, 26). Time-course stimulation showed that IRAK-M mRNA was induced after 6 h, peaking at 12 and 24 h following treatment with 100 ng/ml of LPS (Fig. 1A). IRAK-M protein levels were also induced following treatment with 100 ng/ml LPS (Fig. 1B) or 1 μg/ml LPS (Fig. 1C) compared with unstimulated cells (6-fold), as determined by a protein-quantitation method using Definiens XD Software following immunofluorescence (Fig. 1B–D). To determine the contribution of epigenetic regulators and transcription factors to IRAK-M expression, we performed a targeted siRNA screening in naive and LPS-activated RAW 264.7 macrophages. For this purpose, 46 previously validated siRNA sequences (Cenix BioScience) targeting 34 molecules, including key transcription factors, epigenetic regulators, and signaling molecules (Table I), all participating in macrophage activation and inactivation, were transfected into RAW 264.7 macrophages. Twenty-four hours following transfection, cells were treated with LPS for 48 h, and IRAK-M expression was quantified by fluorescent microscopy using Definiens Developer XD software for image analysis, as used previously (24). The results identified several siRNAs that affect IRAK-M expression, either positively or negatively (Fig. 2). Among the factors affecting IRAK-M expression was the transcription factor C/EBPβ; epigenetic repressors, such as components of the PRC2; and epigenetic activators, such as UTX.

C/EBPβ is a master regulator of macrophage differentiation and function (27) that is required for macrophage polarization toward the M2 state (28), for the induction of proinflammatory cytokines upon TLR stimulation (29), and for the induction of endotoxin tolerance (30). We hypothesized that IRAK-M transcriptional activation might be regulated by C/EBPβ. To test this hypothesis, we used siRNA-mediated knockdown of C/EBPβ and monitored IRAK-M induction upon LPS stimulation. Two siRNA sequences targeting the C/EBPβ mRNA (i.e., p1 and p2) were used independently to knockdown C/EBPβ. Both of them were effective at reducing C/EBPβ mRNA levels down to 60% (p1) and to 50% (p2) (Fig. 3A). Knockdown of C/EBPβ in unstimulated mouse macrophages reduced basal levels of IRAK-M mRNA, revealing a novel role for C/EBPβ in maintaining expression of IRAK-M (Fig. 3B). Suppression of C/EBPβ also resulted in reduced LPS-induced IRAK-M at the mRNA (Fig. 3B, Supplemental Fig. 1) and protein (Fig. 3C) levels. These data support that C/EBPβ is essential for basal and LPS-induced IRAK-M expression.

C/EBPβ is a member of the C/EBP family of transcription factors that contains a basic-leucine zipper domain to bind directly to DNA sequences (31). Because the siRNA screening showed that C/EBPβ was required for IRAK-M transcriptional activation, we investigated whether it bound on the promoter of the IRAK-M gene following LPS stimulation. Computational analysis for identifying putative transcription factor binding sites on the promoter of IRAK-M confirmed previously identified transcription factor binding sites (14) (Fig. 3D) and revealed two putative C/EBPβ binding sequences ∼60 and 100 bp upstream of the transcriptional initiation site (Fig. 3D). Induction of IRAK-M by LPS occurred 6 h following stimulation, peaking at 12 and 24 h (Fig. 1A). ChIP experiments showed that, shortly after LPS stimulation of macrophages, C/EBPβ was recruited to and bound on the IRAK-M promoter. Binding occurred 2 h following LPS stimulation and appeared to be transient because it decreased to less than basal levels at 12 h and peaked again at 24 h (Fig. 3E). These data demonstrated that, upon LPS stimulation of mouse macrophages, C/EBPβ is transiently recruited to a regulatory region proximal to the transcription initiation site of IRAK-M earlier than the accumulation of mRNA, suggesting that C/EBPβ binding is an early event, and additional transcription factors contribute to sustaining IRAK-M expression. A second wave of C/EBPβ binding occurred at 24 h, possibly due to the persistence of LPS stimulation that occurs in culture.

To determine which transcription factor(s) contribute to IRAK-M induction following binding of C/EBPβ, we examined binding of NF-κB components on the IRAK-M promoter. Earlier studies showed that NF-κB bound on the IRAK-M promoter (14). In addition, it was shown that the NF-κB component RelB contributes to endotoxin tolerance (32, 33), a condition promoted by IRAK-M. Our in silico and ChIP analyses confirmed NF-κB binding on the IRAK-M promoter and showed that RelB transiently bound 2–6 h following LPS stimulation (Fig. 3G), whereas NF-κB p65 bound on the IRAK-M promoter at 6 and 12 h (Fig. 3F), possibly allowing IRAK-M mRNA expression to reach its peak. These data suggest that, upon LPS stimulation, C/EBPβ and RelB were recruited to the IRAK-M promoter and were later replaced by NF-κB p65 and that a second wave of CEBP/β binding occurred 24 h following stimulation to sustain its levels when LPS signals persist.

Earlier reports demonstrated the contribution of SMAD4 to IRAK-M expression in human THP1 macrophages in response to TGF-β or in LPS-tolerized cells (34). Nevertheless, direct binding of SMAD4 on its promoter and a binding site were not identified (34). Therefore, we performed in silico analysis to determine potential SMAD-binding elements and found one located at −1052 to −1061 bp from the transcriptional initiation site (Fig. 3D). ChIP analysis showed that SMAD4 was present on the IRAK-M promoter under unstimulated conditions but was not further recruited after LPS stimulation (Fig. 3H), suggesting that it may be important for maintaining its basal levels but not for the initial induction by LPS because it depends on autocrine action of TGF-β, which is secreted upon restimulation of endotoxin-tolerant macrophages (34, 35).

Ezh2 is a methyl transferase and a component of PRC2 that is responsible for adding the silencing H3K27me3 mark on genomic regions (36). Specifically, Ezh2 catalyzes H3K27me3 (37). LPS stimulation of RAW 264.7 macrophages resulted in reduction of EZH2 mRNA levels (Fig. 4A). Suppression of Ezh2 using siRNA (Fig. 4A) resulted in augmented induction of IRAK-M following LPS stimulation. This was evident at the mRNA (Fig. 4B) and protein (Fig. 4C) levels, as seen by real-time RT-PCR and quantitative immunofluorescence analyses, respectively.

To further confirm the involvement of EZH2 in LPS-induced IRAK-M induction, we performed CRISPR-Cas9–mediated knockout of the Ezh2 gene. Successful Ezh2 targeting by Ezh2 guide RNA was evaluated by Western blot (Fig. 4E). LPS stimulation of cells lacking the Ezh2 gene resulted in enhanced induction of IRAK-M, in accordance with the siRNA-mediated knockdown of EZH2 mRNA (Fig. 4F). These findings suggest that EZH2 exerts a suppressive role on the IRAK-M gene, indicating that PRC2 negatively regulates IRAK-M expression.

Epigenetic silencing always acts in concert with epigenetic activators on regulatory regions poised to react to a stimulus. The UTX gene encodes for the UTX protein, a histone demethylase that primarily removes the H3K27me3 mark upon transcriptional derepression. Because EZH2 was shown to suppress IRAK-M, we asked whether UTX, which has an antagonizing function to EZH2, would have opposite effects on IRAK-M transcriptional activation. siRNA screening analysis indicated that suppression of UTX reduced IRAK-M induction by LPS (Fig. 2). To confirm the role of UTX in IRAK-M regulation, we transfected RAW 264.7 macrophages with two siRNA sequences to suppress UTX expression (Fig. 5A). As expected, UTX knockdown in RAW 264.7 macrophages inhibited IRAK-M induction, which was evident at the mRNA (Fig. 5B) and protein (Fig. 5C, 5D) levels.

To confirm that IRAK-M is regulated at the level of histone 3 methylation on its promoter, we investigated whether the epigenetic mark of PRC2 (H3K27me3) was found on the IRAK-M promoter region. ChIP experiments were performed with an anti-H3K27me3 Ab and using two sets of primers: a proximal primer detecting a region 200 bp upstream of the transcription initiation site and a distal primer 2 kb upstream of the same site. The results showed that both regions carried H3 with methylation at K27 at the naive state, which, upon LPS stimulation, was reduced (Fig. 6), coinciding with maximal induction of IRAK-M. At later time points, H3K27me3 marks were restored, potentially to avoid prolonged expression of IRAK-M and persistent immunosuppression (Fig. 7).

Molecules that mediate TLR4 signaling are tightly regulated to effectively induce or suppress TLR4 signals. Sustained and potent TLR4 signaling can result in excessive inflammation, which may lead to tissue destruction and septic shock. IRAK-M, a homolog of IRAKs lacking its catalytic domain, is induced upon activation of macrophages and binds TRAF-6, thus inhibiting TLR4 signaling and suppressing macrophage activation. IRAK-M is induced by the same TLR signals that promote activation of macrophages and by signals that control or fine-tune macrophage activation. In the current study, we used an siRNA screening approach to investigate the contribution of a series of transcription factors and epigenetic modifiers to the regulation of IRAK-M expression following LPS stimulation in macrophages. The results identified C/EBPβ as a key transcription factor for induction of IRAK-M and revealed that H3 modifications on the IRAK-M promoter participate in its regulation.

Promoter analyses showed that expression of IRAK-M depends on NF-κB, AP1, and CREB (14), factors that are induced by different macrophage-activation signals, including LPS (13, 14, 38). Computational analysis and ChIP experiments revealed that C/EBPβ also bound on the proximal region of the IRAK-M promoter and that suppression of C/EBPβ abolished basal and LPS-induced ΙRΑΚ-M expression. C/EBPβ is a transcription factor that is known to contribute to macrophage inactivation and M2-type polarization by inducing the expression of several genes, including Arginase 1 and miR-146a, whereas deletion of C/EBPβ inhibited M2 polarization (28, 39–42). We showed recently that C/EBPβ also regulates miR-155 and miR-146a at the stage of endotoxin tolerance (30), a condition that also depends on IRAK-M (43). Our results indicated that IRAK-M is regulated by C/EBPβ, providing additional evidence for the importance of C/EBPβ as a central regulator of macrophage inactivation. Binding of C/EBPβ occurred early after LPS stimulation following a reduction below the basal binding levels and a second wave of binding at 24 h. At 6 and 12 h post-LPS stimulation, NF-κB p65 binds on IRAK-M, potentially securing potent and transient induction of IRAK-M transcription. At 24 h and if TLR stimulation persists, C/EBPβ binds again to secure persistence of IRAK-M expression. In vivo, LPS is neutralized in the liver, and stimulation of TLR4 is sustained only during active bacteremia. Thus, temporal C/EBPβ and NF-κB binding secures an initial induction of IRAK-M to block potential cytokine storm and prevent sepsis, whereas a second wave of C/EBPβ binding occurs only when LPS (and bacteremia) persists. This temporal induction of C/EBPβ may be important for avoiding persistent immunosuppression when the pathogen (and subsequent LPS stimulus) is eliminated.

Macrophage inactivation was associated with histone modifications, which change the expression pattern of genes in macrophages to promote endotoxin tolerance and prevent endotoxin shock (22, 23). These changes promote silencing of a set of genes that mediate proinflammatory events and allow the expression of another group of genes that promote suppression of proinflammatory signals and resolution of inflammation. Histone modifications are regulated by different methyl transferases and demethylases controlling H3 lysine methylation levels. PRC2 was implicated in macrophage differentiation (44) and myeloproliferation (45), but no information is available on its role in macrophage activation or inactivation. Our data showed that suppression of EZH2, a methyl transferase that is a key component of PRC2, augmented LPS-induced IRAK-M expression. In contrast, suppression of UTX, a demethylase of the KDM6 subfamily that is involved in proinflammatory responses of macrophages (46), suppressed LPS-induced IRAK-M expression.

PRC2 via EZH2 mediates H3 methylation at K27, a modification that renders chromatin inactive and suppresses transcriptional activation (47), whereas UTX demethylates K27, reversing the effect (18). In this study, we demonstrated that IRAK-M was also regulated at the chromatin level and that histones bound on its promoter acquired modifications upon LPS stimulation. Accordingly, H3 bound on the IRAK-M promoter was trimethylated at K27 in the naive state when IRAK-M was not expressed. Upon LPS stimulation, H3K27 on the IRAK-M promoter was demethylated, thus allowing its transcription, a condition required for endotoxin tolerance.

Regulation of inflammatory genes is epigenetically controlled to avoid a cytokine storm and secure the state of endotoxin tolerance (48), a condition to which IRAK-M contributes by suppressing IRAK-mediated signals that activate the TRAF6/NF-κB cascade. Thus, induction of IRAK-M may be part of the feed-forward mechanism involved in establishing endotoxin tolerance. In this process, proinflammatory genes, such as TNF-α or IL-1β, are also regulated at the level of H3 modification to induce their expression at the activation state. Indeed, the TNF-α gene is epigenetically controlled (49) and primed by IFN-γ by the aposition of H3K27me3 and aposition of acetylation of lysine 27 on histone 3 on its promoter (50), allowing early induction of its expression. Acetylation of the promoter is regulated by SIRT1 and SIRT6 and highlights activation of the promoter, allowing recruitment of NF-κB and transcriptional activation (51, 52), and it provide a potential link between inflammatory response and metabolism in the context of sepsis (53). Upon macrophage activation, dimethylation of lysine 9 on histone 3 was removed from the TNF-α promoter, and H3K4me3 was added (32, 33). IL-1b, another cytokine central for the acute inflammatory response, is also silenced by epigenetic mechanisms, including trimethylation of lysine 9 on histone 3 (54, 55). Similar mechanisms occur in the regulation of inflammatory microRNAs, for which we showed previously that miR-155 and miR-146a were regulated at the H3 methylation level by obtaining trimethylation of lysine 4 on histone 3 marks at the activation state and H3K9me3 at the endotoxin tolerance state (30). In this study, we identified that PRC2 epigenetically regulated IRAK-M expression, providing an additional feed-forward mechanism for endotoxin tolerance. IRAK-M possessed H3K27me3 marks that are characteristic of silenced chromatin (56). Following LPS activation, these marks were removed to allow IRAK-M transcription. It remains to be elucidated whether additional methyl transferase, demethylase, or acetylase complexes affect IRAK-M expression, such as ones adding or removing K4 or K9 methyl marks or K24 acetyl marks.

Overall, our findings demonstrate novel transcriptional and epigenetic regulation of IRAK-M expression and highlight the importance of C/EBPβ and PRC2 for its regulation. According to this model (Fig. 7), in naive macrophages a transcriptional complex that keeps IRAK-M repressed and contains EZH2 is present on its promoter; upon activation by LPS, it is replaced by a complex containing UTX that removes H3K27me3 and allows recruitment of transcription factors, including C/EBPβ, to initiate its transcription. C/EBPβ binding is transient and temporal, coming in a second wave at 24 h. The removal of H3K27me3 marks is also transient; they return at 24 h. Thus, in the context of acute inflammation, IRAK-M is induced within 24 h; however, its transcriptional activation is quickly silenced to avoid prolonged expression, which results in the inability of the cells to return to homeostasis and responsive status. At the acute stage, NF-κB p65 contributes to IRAK-M expression and secures its potent induction. When LPS stimulation is sustained, which is the case in our experimental setting because LPS remains in the medium, a second wave of C/EBPβ binding may secure prolonged expression of IRAK-M to avoid a cytokine storm and sepsis. Responsiveness of macrophages to LPS is controlled at the epigenetic level, which includes control of inflammatory cytokines and microRNAs (30, 33, 54), as well as cell metabolism (52), another mechanism that regulates macrophage responsiveness. These two mechanisms are potentially interconnected at the level of histone-modification enzymes (53). A limitation of the study is that these observations were made in a cell culture model, which necessitates the exploration of the proposed mechanism in vivo in animal models. Moreover, it remains to be determined whether the same transcriptional and epigenetic regulators control the IRAK-M expression, endotoxin tolerance, and sepsis-induced immunosuppression observed in immunoparalyzed sepsis patients.

The authors have no financial conflicts of interest.