Pellino-1 Positively Regulates Toll-like Receptor (TLR) 2 and TLR4 Signaling and Is Suppressed upon Induction of Endotoxin Tolerance

Background: Dysregulated Toll-like receptor (TLR) signaling is a hallmark of endotoxin-tolerized macrophages in immunosuppression stages of sepsis but Pellino-1 involvement is unknown.

Results: Endotoxin tolerization suppresses LPS-induced Pellino-1, Pellino-1 potentiates TLR2/4-driven NF-κB, cytokines and K63-linked IRAK1, TBK1, TAK1, and TRAF6 polyubiquitination.

Conclusion: Pellino-1 potentiates TLR2/4 signaling and is decreased by endotoxin tolerization.

Significance: Uncovering mechanisms of endotoxin tolerance is critical to understand sepsis-associated immunosuppression.

Abstract

Endotoxin tolerance reprograms Toll-like receptor (TLR) 4-mediated macrophage responses by attenuating induction of proinflammatory cytokines while retaining expression of anti-inflammatory and antimicrobial mediators. We previously demonstrated deficient TLR4-induced activation of IL-1 receptor-associated kinase (IRAK) 4, IRAK1, and TANK-binding kinase (TBK) 1 as critical hallmarks of endotoxin tolerance, but mechanisms remain unclear. In this study, we examined the role of the E3 ubiquitin ligase Pellino-1 in endotoxin tolerance and TLR signaling. LPS stimulation increased Pellino-1 mRNA and protein expression in macrophages from mice injected with saline and in medium-pretreated human monocytes, THP-1, and MonoMac-6 cells, whereas endotoxin tolerization abrogated LPS inducibility of Pellino-1. Overexpression of Pellino-1 in 293/TLR2 and 293/TLR4/MD2 cells enhanced TLR2- and TLR4-induced nuclear factor κB (NF-κB) and expression of IL-8 mRNA, whereas Pellino-1 knockdown reduced these responses. Pellino-1 ablation in THP-1 cells impaired induction of myeloid differentiation primary response protein (MyD88), and Toll-IL-1R domain-containing adapter inducing IFN-β (TRIF)-dependent cytokine genes in response to TLR4 and TLR2 agonists and heat-killed Escherichia coli and Staphylococcus aureus, whereas only weakly affecting phagocytosis of heat-killed bacteria. Co-expressed Pellino-1 potentiated NF-κB activation driven by transfected MyD88, TRIF, IRAK1, TBK1, TGF-β-activated kinase (TAK) 1, and TNFR-associated factor 6, whereas not affecting p65-induced responses. Mechanistically, Pellino-1 increased LPS-driven K63-linked polyubiquitination of IRAK1, TBK1, TAK1, and phosphorylation of TBK1 and IFN regulatory factor 3. These results reveal a novel mechanism by which endotoxin tolerance re-programs TLR4 signaling via suppression of Pellino-1, a positive regulator of MyD88- and TRIF-dependent signaling that promotes K63-linked polyubiquitination of IRAK1, TBK1, and TAK1.

Introduction

Host initiates immune defense against invading pathogens and detects danger signals during sterile inflammation by sensing pathogen-associated molecular patterns or endogenous “alarmins” via pattern recognition receptors, including Toll-like receptors (TLRs)⁴ (1). TLRs are expressed on the cell surface (TLR1, TLR2, TLR4, TLR5, and TLR6), or in intracellular endosomes (TLR3, TLR7–9, and TLR11) and sense pathogen-associated molecular patterns (e.g. lipopolysaccharide (LPS) by TLR4) or endogenous alarmins (e.g. high mobility group box-1 by TLR2 and TLR4) (2). Ligand sensing by TLR ectodomains induces receptor dimerization that brings together intracellular Toll-IL-1 receptor (TIR) domains, creating docking platforms to enable recruitment of adapters (3, 4). All TLRs except TLR3 utilize the myeloid differentiation primary response protein (MyD) 88 that associates with TLRs via TIR domains and recruits IL-1 R-associated kinases (IRAKs) via death domain interactions (5). Clusterization of IRAK4 activates its kinase activity via trans-autophosphorylation (6), inducing IRAK4-mediated phosphorylation, K63-linked ubiquitination, and activation of IRAK1 (7, 8). TNFR-associated factor (TRAF) 6 interacts with phosphorylated IRAK1 and undergoes K63-linked ubiquitination and activation, resulting in recruitment and activation of TGF-β-activated kinase (TAK) 1 (1, 8, 9). Inhibitor of nuclear factor κB kinase (IKK)-γ binds to K63-ubiquitinated IRAK1 via its ubiquitin (Ub) recognition domains and undergoes conformational changes leading to activation of the associated kinase IKK-β (7, 10). In addition, TAK1 activates IKK-β and mitogen-activated protein kinases (MAPKs), culminating in phosphorylation and nuclear translocation of transcription factors, such as nuclear factor κB (NF-κB) and activator protein-1, which drive transcription of inflammatory cytokine genes (1, 2, 4). TLR3 and TLR4 signal from endosomes, utilizing TIR domain-containing adapter inducing IFN-β (TRIF) and TRAF3 to activate TANK-binding kinase (TBK) 1 and IKK-ϵ that mediate phosphorylation, activation, and nuclear translocation of IFN regulatory factor (IRF) 3, inducing transcription of type I IFN genes (1, 2). Engagement of MyD88- and TRIF-dependent pathways by macrophages and dendritic cells induces inflammatory cytokines and IFNs, up-regulates MHC and co-stimulatory molecules, and primes adaptive immune responses (1, 11, 12).

Pellino was first identified in Drosophila as the protein interacting with Pelle, an ortholog of IRAK4, to activate production of antimicrobial peptides and to protect against infection with Gram-positive bacteria (13, 14). Mammalian Pellino-1, Pellino-2, and Pellino-3 share a common structural organization, expressing N-terminal forkhead-associated domains that promote Pellino interactions with phospho-Thr residues of their substrates, and C-terminal RING-like domains responsible for E3 Ub ligase activity (15). Pellino proteins interact with intermediates shared by all Pellinos (e.g. IRAK kinases) but can also target substrates specific individual Pellinos (e.g. SMAD6 for Pellino-1) and regulate IL-1R, TLR, and nucleotide binding and oligomerization domain-like receptor signaling (15). Despite common domain organization and a set of common interacting partners, Pellinos exert non-redundant receptor-, cell-, and species-specific effects whose mechanisms are poorly understood, may be mediated in E3 Ub ligase-dependent- or independent manner (15), and require further studies.

Septic patients surviving the initial cytokine storm become immunocompromised and unable to mount an effective response against secondary infections that often results in fatal outcomes (16, 17). Monocytes from immunocompromised septic patients show altered TLR4 responses to lipopolysaccharide (LPS) (18), which is reminiscent of TLR4 responses observed in endotoxin-tolerant cells (19). Endotoxin tolerance is described as re-programming of TLR4 responses to LPS challenge after prior exposure to endotoxin, and is manifested by suppressed expression of proinflammatory cytokines, without inhibition of anti-inflammatory cytokines, antimicrobial effectors, or phagocytosis (16, 20, 21). It acts as a double-edged sword: on one hand, endotoxin tolerance limits excessive cytokine production during systemic inflammatory response syndrome, sparing organs and tissues from damage, whereas on the other, it is responsible for the failure of the immunocompromised host to counteract secondary infections (21). We previously identified impaired activation of MyD88- and TRIF-dependent signaling pathways, with deregulated IRAK4, IRAK1, and TBK1, as key hallmarks of endotoxin tolerance (22–24). In view of the importance of Pellino-1 in regulating IL1R/TLR signaling (15, 25), we sought to determine the significance of Pellino-1 in endotoxin tolerance and to elucidate its role in TLR2 and TLR4 signaling. To the best of our knowledge, we reported for the first time that induction of endotoxin tolerance in mouse and human macrophages impairs LPS indelibility of Pellino-1 mRNA and protein expression. Using overexpression and gene silencing, Pellino-1 was found to act as a positive regulator of TLR2- and TLR4-mediated signaling within MyD88-IRAK1-TAK1-TRAF6 and TRIF-TBK1 axes, and to regulate expression of MyD88- and TRIF-dependent cytokines. Mechanistically, we report that Pellino-1 plays a critical role to promote K63-linked polyubiquitination of IRAK1, TBK1, and TAK-1.

Experimental Procedures

Reagents and Cell Culture

Highly purified, protein-free Escherichia coli K235 LPS was kindly provided by Dr. Stefanie Vogel (University of Maryland School of Medicine, Baltimore, MD), and rabbit antibody (Ab) against human Pellino-1 (26) was a gift from Dr. Peter Cheung (Nanyang Technological University, Singapore). S-[2,3-Bis(palmitoyloxy)-(2-R,S)-propyl]-N-palmitoyl-(R)-Cys-Ser-Lys⁴-OH (Pam3Cys) and heat-killed Staphylococcus aureus were obtained from Invivogen, heat-killed E. coli were prepared as published (27), anti-phospho (p)-p38, anti-p-p65, anti-p38, anti-p65, and anti-tubulin Abs were from Cell Signaling Technologies. THP-1, human embryonic kidney (HEK)293T, and HEK293 cells were obtained from ATCC, and MonoMac6 cells (66) were provided by Dr. Jorge Cervantes (University of Connecticut Health Center, Farmington, CT). HEK 293 cells stably expressing yellow fluorescent protein (YFP)-TLR2 (293/TLR2) or YFP-TLR4 and MD2 (293/TLR4/MD2) were kindly provided by Dr. Douglas Golenbock (University of Massachusetts Medical School, Worcester, MA). HEK293 cells were maintained in DMEM supplemented with 10% FBS (HyClone), 2 mm l-glutamine, 100 units/ml of penicillin, and 100 μg/ml of streptomycin (Life Technologies) (complete (c) DMEM), 293/TLR2, and 293/TLR4/MD2 cells were cultured in cDMEM containing 5 μg/ml of puromycin or 1 mg/ml of G418, respectively (28, 29). The myelomonocytic cell lines THP1 and MonoMac6 were maintained in RPMI 1640 medium supplemented with 10% FBS (HyClone), 5 × 10⁻⁵ m β-mercaptoethanol, 2 mm l-glutamine, 100 units/ml of penicillin, and 100 μg/ml of streptomycin. Human monocytes were kindly provided by Dr. Larry Wahl (NIDCR/NIH) as de-identified samples prepared by counterflow elutriation of blood from healthy human volunteers, or obtained from Lonza. Studies with human monocytes were approved by the Institutional Review Boards of the University of Connecticut Health Center and University of Maryland School of Medicine. For functional assays, THP-1 cells were incubated for 72 h with 20 ng/ml of PMA to attain macrophage characteristics (plastic adherence, CD14 expression and phagocytosis), as reported (30).

Mice and Macrophage Isolation

C57BL/6J mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Induction of in vivo endotoxin tolerance was carried out as described (23). In brief, peritoneal exudate cells were isolated from mice 4 days after intraperitoneal injection with sterile 3% thioglycollate (REMEL Inc., Lenexa, KS), and 24 h before isolation of cells, mice were administered saline (control) or LPS (20 μg/mouse, in vivo induction of endotoxin tolerance). Peritoneal macrophages were obtained by peritoneal lavage and subsequent adherence to plastic as described (23). Cells were plated into 6-well plates (4 × 10⁶ cells per well) and cultured in cRPMI 1640. All animal procedures were carried out with institutional IACUC approval.

Recombinant Plasmids and Transfection

pcDNA3-CD14, pELAM-luciferase (Luc), pTK-Renilla-Luc, pEFBOS-FLAG-MD2, pcDNA3-AU1-MyD88, pcDNA3-TRIF, pRK5-IRAK1, pcDNA3-FLAG-TRAF6, pCMV1-FLAG-TAK1, pcDNA3-HA-TAB1, pcDNA3-FLAG-p65, and pRK5-hemagglutinin (HA)-Ub-K63 only (plasmid number 17606) were described previously (22, 24, 29, 31–33), and pSuper-TBK1 (plasmid number 26210) was from Addgene. FLAG-Pellino-1 and pSuper vectors encoding scrambled and Pellino-1 shRNA were kindly provided by Dr. Xiaoxia Li (Lerner Research Institute, Cleveland Clinical Foundation, Cleveland, OH). For transient transfection, HEK293T, 293/TLR2, and 293/TLR4/MD2 cells were plated in 100-mm dishes (for immunoprecipitation), 6- or 24-well plates (for gene expression studies and reporter assays, respectively), cultured overnight, and transfected for 3 h with plasmids, using Lipofectamine 2000 transfection reagent (Life Technologies). To obtain stable transfectants, 293/TLR4/MD2 cells transfected with the plasmids encoding pSuper-scrambled shRNA or pSuper-Pellino-1 shRNA, along with pBabePuro, were selected in cDMEM containing 5 μg/ml of puromycin (Sigma), as reported (33).

Lentiviral Transduction of THP1 Cells

GIPZ Lentiviral scrambled or Pellino-1 shRNA were obtained from GE Health Care/Dharmacon, and packaging vectors pSPAX2 (plasmid number 12260) and pMD2.G (plasmid number 12259) were from Addgene. Lentiviral transduction of THP1 cells was performed as published previously (34). In brief, to generate lentiviral particles, HEK293T cells were plated in 6-well plates (2 × 10⁶ cells/well), cultured overnight, and transfected with GIPZ shRNA (1 μg/well), pSPAX2 (0.75 μg/well), and pMD2G (0.25 μg/well) using Lipofectamine 2000, according to the manufacturer's instruction. After 48 h, medium containing viral particles was collected and fresh cDMEM was added to the wells, cells were cultured for additional 48 h, and viral particle-containing medium was collected. After filtration through 0.45-μm filters, medium was centrifuged (16,000 × g, 2 h), the pellet was resuspended in 0.25 ml of cRPMI with Polybrene (2 μg/ml), and the mixture was added to wells of 24-well plates containing THP1 cells. Plates were centrifuged for 45 min at 1200 × g, medium was aspirated and fresh cRPMI was added (0.3 ml/well) to cells. After culturing for 24 h, plates were centrifuged for 10 min at 1200 × g, cells were resuspended in cRPMI containing 5 μg/ml of puromycin, and stable bulk transfectants were selected for at least 2 weeks.

Isolation of RNA and Real-time Quantitative PCR Analysis

Total RNA was isolated using TRIzol (Life Technologies), residual genomic DNA was digested with DNase, and RNA was re-purified, as recommended by the manufacturer. cDNA was prepared from 1 μg of total RNA using Reverse Transcription System (Promega), and examined by qPCR with primers for the following genes: human hypoxanthine phosphoribosyltransferase (HPRT), 5′-ACCAGTCAACAGGGGACATAAAAG-3′ (forward), 5′-GTCTGCATTGTTTTGCCAGTGTC-3′ (reverse); human IL-8, 5′-CACCGGAAGGAACCATCTCACT-3′ (forward), 5′-TGCACCTTCACACAGAGCTGC-3′ (reverse); human CCL5, 5′-TTTGTCACCCGAAAGAACCG-3′ (forward), 5′-CAAGGACTCTCCATCCTAGCTCAT-3′ (reverse); human IFN-β, 5′-ACTGCCTCAAGGACAGGATG-3′ (forward), 5′-AGCCAGGAGGTTCTCAACAA-3′ (reverse); human PELLINO-1, 5′- CCAAGCCTGGAATATGGAGA-3′ (forward), 5′-TGCTTCACGGTAGGAGTGTG-3′ (reverse); mouse Hprt, 5′-ACCAGTCAACAGGGGACATAAAAG-3′ (forward), 5′-GTCTGCATTGTTTTGCCAGTGTC-3′ (reverse); mouse IL-6, 5′-TCAGGAAATTTGCCTATTGAAAATTT-3′ (forward), 5′-GCTTTGTCTTTCTTGTTATCTTTTAAGTTGT-3 (reverse); mouse Pellino-1, 5′-CTTTATCTCGAGCCCAGACG-3′ (forward), 5′-CTGACTGCGTGTCGGAATTA-3′ (reverse), on a MyIQ real-time qPCR machine (BioRad). The data were analyzed as reported (35).

Nucleofection

pSuper expression vectors encoding scrambled or Pellino-1 shRNA variants were introduced into THP1 cells by nucleofection, using the Nucleofector I device (Lonza) and the cell line nucleofection kit V (Lonza), as recommended by the manufacturer. After recovery for 24 h, cells were treated for 3 h with medium or LPS (100 ng/ml), RNA was isolated, reverse transcribed, and analyzed by real-time PCR using gene-specific primers.

Co-immunoprecipitation and Immunoblotting

Cell lysates were prepared as described previously (22) and precleared with protein G-agarose beads (Roche Applied Science) for 4 h at 4 °C upon rotation. Precleared cell extracts were incubated overnight at 4 °C with the respective Abs in lysis buffer containing 20 mm HEPES (pH 7.4), 0.5% Triton X-100, 150 mm NaCl, 12.5 mm β-glycerophosphate, 50 mm NaF, 1 mm DTT, 1 mm sodium orthovanadate, 2 mm EDTA, 1 mm PMSF, and protease inhibitor mixture (Roche Applied Science). Protein G-agarose beads were added (45 μl/sample) and incubation was continued for 4 h. Beads were washed 5 times with lysis buffer, and resuspended in Laemmli sample loading buffer (50 mm Tris-Cl, pH 6.8, 10% glycerol, 2% SDS, 0.1% bromphenol blue, 5% 2-mercaptoethanol). For ubiquitination assays, SDS (1% final concentration) was added to precleared samples before immunoprecipitation to disrupt non-covalent protein-protein interactions, samples were boiled for 10 min, cooled down, and diluted 10 times with ice-cold buffer (final concentration of 0.1% SDS) (7, 22). Thereafter, the respective Abs were added and immunoprecipitation and sample preparation were carried out as described above. Proteins were separated by SDS-polyacrylamide gel electrophoresis on 4–20% mini-gels (Life Technologies), electrotransferred to Immobilon-P membranes, blocked, and probed with the respective Abs, as described (22, 24, 28, 29, 32). Quantification was performed using the NIH ImageJ program.

NF-κB Reporter Assays

Reporter assays were performed as reported previously (22, 24, 28, 29, 32). In brief, cells were transfected with plasmids, using Lipofectamine 2000, recovered for 48 h, and treated for 5 h with medium or stimuli. Cells were lysed in a passive lysis buffer (Promega), and firefly luciferase versus Renilla luciferase activities were measured using dual luciferase reporter assay system (Promega) on a Berthold LB 9507 luminometer (Berthold Technologies).

Phagocytosis Assay and Flow Cytometry

THP-1 cells were plated in 24-well plates (0.5 × 10⁶ cells per well), incubated with 20 ng/ml of PMA for 72 h, washed, and resuspended with cRPMI with or without pHrodo Red-conjugated heat-killed E. coli or S. aureus bioparticles (Life Technologies) at a bacteria per cell ratio of 50:1. After incubation for the various times on ice or at 37 °C, cells were detached, washed in ice-cold PBS containing 3% FBS, stained with LIVE/DEAD® Fixable Aqua Dead Cell Stain Kit (Life Technologies), fixed for 15 min in 3.7% formaldehyde, and analyzed by FACS on a LSRII Flow Cytometer (BD Biosciences) to measure pHRodo fluorescence. The data were analyzed using the FlowJo software (Tree Star).

Statistical Analysis

Statistical analysis was performed using the GraphPad Prism 5 program for Windows (GraphPad Software Inc.). Statistical differences were evaluated using Student's t test with the confidence interval set at 95% level. The results were expressed as mean ± S.D. values.

Results

Induction of Endotoxin Tolerance Impairs LPS Inducibility of Pellino-1

In the first series of experiments, we examined the impact of in vitro induction of endotoxin tolerance on basal- and LPS-induced expression of Pellino-1 in human primary monocytes and in macrophage-like cell lines, THP-1 and MonoMac-6 cells. THP-1 cells were differentiated for 72 h with 20 ng/ml of PMA to attain macrophage-like characteristics, e.g. plastic adherence, CD14 expression, and phagocytosis (30). In line with previous studies by us and others (22–24, 32, 36–38), prior exposure to endotoxin significantly decreased induction of TNF-α gene expression to subsequent LPS challenge, documenting the induction of endotoxin tolerance in all three cell types (Fig. 1, A–C, left column). LPS stimulation of medium-pretreated monocytes, THP1, and MonoMac-6 cells resulted in robust induction of Pellino-1 mRNA, which was the strongest at 3 h, declined by 8 h, and further decreased by 16 h post-LPS challenge (Fig. 1, A–C, right columns). Notably, at all times, endotoxin-tolerized cells exhibited 65–95% inhibition of LPS-mediated induction of Pellino-1 mRNA compared with control medium-pretreated cells (Fig. 1, A–C, right column).

FIGURE 1.

The impact of endotoxin tolerance on basal and LPS-inducible Pellino-1 mRNA expression. Primary human monocytes (A), THP-1 (B), and MonoMac6 (C) cells were exposed for 24 h with medium or 100 ng/ml of LPS. Thereafter, cells were washed three times and treated with medium or 100 ng/ml of LPS for the indicated times. D, C57BL/6J mice were intraperitoneally administered thyoglycollate, and 72 h later, were intraperitoneally injected with saline or LPS (20 μg/mouse). Peritoneal exudate cells were isolated, and plastic-adherent macrophages were treated for 1 and 3 h with medium or 100 ng/ml of LPS. RNA was isolated, reverse-transcribed, and analyzed by real-time qPCR with primers specific for HPRT (housekeeping gene used for normalization), TNF-α (left columns), and Pellino-1 (right columns) (A–C). D, real-time qPCR was performed with primers specific for mouse IL-6, Pellino-1, and Hprt. A and B, the summary of seven (A) and three (B and C) experiments are depicted (mean ± S.D.). D, the results (mean ± S.D.) of a representative experiment (n = 5) are shown. *, p < 0.05.

Next, we employed an in vivo model of endotoxin tolerance (23, 36), whereby C57BL/6J mice were injected intraperitoneally with LPS (20 μg/mouse, endotoxin tolerization in vivo) or PBS (controls), and thioglycollate-elicited peritoneal macrophages were exposed to medium or LPS in vitro, followed by qPCR analyses of cytokine and Pellino-1 gene expression. Macrophages obtained from PBS-injected mice showed 4–8-fold increases in IL-6 mRNA levels following LPS stimulation for 1 and 3 h, respectively, whereas macrophages from LPS-treated animals had only 2-fold maximal induction of IL-6 mRNA (Fig. 1D, left), demonstrating endotoxin-tolerant phenotype. LPS caused only 1.4–2.4-fold increases in the induction of Pellino-1 mRNA in macrophages from mice previously administered with LPS, in contrast to robust LPS induction observed in macrophages from control, PBS-injected mice (Fig. 1D, right).

It was important to confirm inhibited LPS inducibility of Pellino-1 in endotoxin-tolerized macrophages at the protein level. As shown in Fig. 2, LPS stimulation of control, medium-pretreated human monocytes (A), THP-1 (B), and MonoMac-6 (C) cells led to strong induction of phosphorylation of p38, p65, and degradation of IκB-α within 30–60 min, whereas LPS responses in endotoxin-pretreated cells were inhibited by 74–88% (quantification panels), documenting LPS tolerance. Although LPS stimulation substantially increased Pellino-1 protein expression in medium-pretreated monocytes, THP1, and MonoMac-6 cells at 3 and 8 h (∼2–6.4-fold increases, Fig. 2, A–C), with subsequent decline by 16 h, endotoxin-tolerized cells failed to respond to LPS by up-regulation of Pellino-1 protein. Taken collectively, these results indicate that the induction of endotoxin tolerance in vitro and in vivo attenuates LPS-mediated induction of Pellino-1 expression.

FIGURE 2.

Endotoxin-tolerized human monocytes, THP-1, and MomoMac6 cells fail to up-regulate Pellino-1 protein expression in response to LPS challenge. Human monocytes (A), THP-1 (B), and MonoMac-6 (C) cells were pretreated for 20 h with medium or 100 ng/ml of LPS, washed, and restimulated with or without 100 ng/ml of LPS for the indicated times. Cell lysates were examined by Western blot analyses with Abs against Pellino-1, p-p65, IκB-α, phospho- and total p38, and tubulin, as shown. Panels on the right depict quantification of Western blot results using the NIH ImageJ software package. Shown are the data of a representative (n = 4) experiments.

Pellino-1 Is a Positive Regulator of TLR2- and TLR4-driven NF-κB Activation and Induction of IL-8

To delineate the functional significance of Pellino-1 in TLR2 and TLR4 signaling, we determined the consequences of Pellino-1 overexpression and knock-down in 293/TLR2 and 293/TLR4/MD2 cells on Pam3Cys- or LPS-driven NF-κB activation and IL-8 gene expression. As shown in Fig. 3, transfection of Pellino-1 led to constitutive degradation of IκB-α (A), p65 phosphorylation (B), and induction of an NF-κB-dependent pELAM-Luc reporter activity (C and D). Overexpressed Pellino-1 significantly increased Pam3Cys-mediated degradation of IκB-α (Fig. 3A), LPS-mediated phosphorylation of p65 (Fig. 3B), and markedly up-regulated TLR2- and TLR4-driven NF-κB reporter activation, compared with responses seen in pcDNA3-transfected cells. To confirm our results obtained by overexpression, we next performed shRNA-based knockdown of Pellino-1 in 293/TLR4/MD2 cells and analyzed LPS-induced NF-κB activation. As shown in Fig. 4, 293/TLR4/MD2 cells stably expressing Pellino-1 shRNA had significantly reduced basal and LPS-driven levels of Pellino-1 protein compared with cells expressing non-targeting scrambled shRNA (A). In contrast, we observed equal expression of Pellino-3 and tubulin proteins (Fig. 4A), showing the specificity of Pellino-1 knock-down. 293/TLR4/MD2 cells stably expressing Pellino-1 shRNA exhibited significantly impaired LPS-mediated phosphorylation of p65 (Fig. 4B) and induction of NF-κB reporter (Fig. 4C) compared with cells transfected with scrambled shRNA. These results demonstrate that Pellino-1 positively regulates TLR2- and TLR4-mediated NF-κB activation.

FIGURE 3.

Overexpression of Pellino-1 activates NF-κB and up-regulates TLR2- and TLR4-mediated NF-κB activation and IL-8 gene expression. 293/TLR2 (A, C, and E) and 293/TLR4 (B, D, and F) cells were co-transfected with pcDNA3-CD14, pEFBOS-MD2, and the FLAG-Pellino-1-encoding plasmid. For reporter assays (C and D), cells were also co-transfected with pELAM-Luc and pTK-Renilla Luc. After recovery for 48 h, cells were treated with 1 μg/ml of Pam3Cys (A, C, and E) or 100 ng/ml of LPS (B, D, and F) for the indicated times (A and B), for 5 (C and D) or 3 h (E and F). Cell lysates were analyzed by Western blotting with Abs for the indicated proteins (A and B), or assayed for firefly versus Renilla luciferase activities (C and D). RNA was isolated, reverse-transcribed, and analyzed by real-time qPCR with primers specific for IL-8 and HPRT. The results of a representative experiment (n = 3) are presented. *, p < 0.05.

FIGURE 4.

Pellino-1 knockdown in 293/TLR4/MD2 cells impairs LPS-mediated activation of NF-κB and IL-8 gene expression. 293/TLR4 cells stably expressing scrambled or Pellino-1 shRNA were co-transfected with pcDNA3-CD14, pEFBOS-MD2 alone (A, B, and D), or together with pELAM-Luc and pTK-Renilla Luc (C). After recovery for 24 h, cells were treated with medium or 100 ng/ml of LPS for 10 min (A), 60 min (B), 5 h (C), or for the indicated times (D). Cells lysates were analyzed by Western blotting with Abs against the indicated proteins (A and B) or assayed for firefly versus Renilla Luc activities (C). RNA was isolated, reverse transcribed, and examined by real-time qPCR with the respective gene-specific primers (D). The data of a representative experiment (n = 3) are shown.

Because NF-κB is critically involved in activating cytokine gene expression (39), we next studied whether Pellino-1 regulates TLR2- and TLR4-inducible expression of IL-8 mRNA. Fig. 3 shows that stimulation of pcDNA3-transfected cells with Pam3Cys (293/TLR2 cells, E) or LPS (293/TLR4/MD2 cells, F) led to 82- and 17-fold induction in the expression of IL-8 mRNA, whereas overexpression of Pellino-1 increased these responses to 117- and 58-fold, respectively. LPS caused a 12-fold increase in the levels of IL-8 mRNA in 293/TLR4/MD2 cells expressing scrambled shRNA, whereas Pellino-1 shRNA-expressing cells showed only a 1.2-fold increase (Fig. 4D). Taken together, these overexpression and gene ablation data indicate that Pellino-1 acts as a positive regulator of TLR2- and TLR4-driven NF-κB activation and IL-8 gene expression.

Pellino-1 Significantly Up-regulates TLR-mediated Induction of MyD88- and TRIF-dependent Cytokine Genes in Response to LPS, Pam3Cys, and Heat-killed E. coli or S. aureus but Only Weakly Affects Phagocytosis of Heat-killed Bacteria

Next, we sought to extend our findings on a positive regulatory role of Pellino-1 in TLR2 and TLR4 signaling obtained in HEK293 transfectants to cells with macrophage-like phenotype. We were also interested to define how Pellino-1 affects MyD88- and TRIF-dependent signaling pathways. To this end, we employed nucleofection- or lentiviral-based introduction of Pellino-1 or control scrambled shRNA species into THP-1 cells, cells were differentiated with PMA to attain macrophage phenotype, and their responses to TLR2 and TLR4 agonists, as well as to heat-killed E. coli and S. aureus were analyzed. Nucleofection of Pellino-1 shRNA into THP-1 cells decreased Pellino-1 mRNA levels by 52% compared with scrambled shRNA-transfected cells, whereas not affecting Pellino-3 gene expression (Fig. 5A, left column), demonstrating specific ablation of Pellino-1. THP1 cells nucleofected with scrambled shRNA responded to LPS by robust induction of TNF-α and IL-6 mRNA, increasing their expression by 36- and 92-fold, respectively, whereas induction of TRIF-dependent IFN-β and CCL5 was manifested to a lesser extent (4.2- and 2.5-fold increase, Fig. 5A). Pellino-1 knock-down achieved by nucleofection of Pellino-1 shRNA significantly suppressed LPS inducibility of genes encoding TNF-α (54% inhibition), IL-6 (89% inhibition), IFN-β (57% inhibition), and CCL5 (33% inhibition) (Fig. 5).

FIGURE 5.

The impact of Pellino-1 ablation on MyD88- and TRIF-dependent cytokine genes in THP-1 cells. THP-1 cells were nucleofected (A) or subjected to lentivirus-based transduction of scrambled and Pellino-1 shRNA to obtain stably transfected cell lines (B). After differentiation with PMA for 72 h, cells were washed and treated for 1 (IFN-β) and 3 h (all other genes) with medium, 100 ng/ml of LPS, or 1 μg/ml of Pam3Cys. RNA was reverse-transcribed and subjected to real-time qPCR to examine levels of the indicated cytokine genes and HPRT. Pellino expression is shown as relative values normalized to HPRT controls, and cytokine gene expression is depicted as fold of TLR induction versus medium-treated cells. Presented are the results (mean ± S.D.) of one of three experiments.

To confirm and extend these results, we obtained THP-1 cell lines stably expressing scrambled or Pellino-1 shRNA after lentivirus transduction and selection in puromycin, and analyzed their responses to TLR2 (Pam3Cys) and TLR4 (LPS) agonists as well as to heat-killed bacteria utilizing TLR2 (S. aureus) and TLR4 (E. coli). Incorporation of Pellino-1 shRNA by lentivirus transduction led to an even higher decrease (66% inhibition) in Pellino-1 mRNA levels compared with scrambled shRNA-expressing cells, whereas not affecting Pellino-3 mRNA expression (Fig. 5B, left column). Pellino-1 ablation decreased LPS-inducible TNF-α and IFN-β gene expression by 60 and 50%, respectively (Fig. 5B, top, middle, and right panels), and reduced Pam3Cys-mediated expression of TNF-α and IL-6 mRNA by 57 and 36% (Fig. 5B, bottom middle and right panels). Heat-killed E. coli caused 29- and 10-fold increases in TNF-α and pro-IL-1β gene expression in THP-1 cells transduced with scrambled shRNA, whereas expression of these cytokine genes in Pellino-1 shRNA-expressing cells was reduced by 38 and 50%, respectively (Fig. 6A, top panel). Similarly, whereas scrambled shRNA-THP1 cells responded to stimulation with heat-killed S. aureus by 14.9- and 4-fold induction of TNF-α and pro-IL-1β mRNA, Pellino-1 knockdown led to 50 and 60% reduction in their inducibility (Fig. 6A, bottom panel).

FIGURE 6.

Pellino-1-knockdown in THP-1 cells inhibits induction of cytokine genes in response to heat-killed E. coli and S. aureus but only weakly affects phagocytosis of pHRodo-labeled heat-killed bacteria. THP1 stably expressing scrambled or Pellino-1 shRNA were differentiated for 72 h with 20 ng/ml of PMA, washed, and resuspended in medium. A, cells were treated for 3 h with medium or heat-killed E. coli or S. aureus (bacteria per cell ratio = 50:1), RNA was isolated, reverse transcribed, and analyzed by real-time qPCR with gene-specific primers. B, cells were exposed for the indicated times to medium (light gray shaded histograms) or pHrodo Red-conjugated E. coli or S. aureus bioparticles (bacteria per cell ratio = 50:1) at 37 °C (phagocytosis, dark gray shaded histograms) or on ice (control, dotted line histograms), fixed, and analyzed by FACS. C, quantification of phagocytosis measured in bacteria-cell triplicate well cultures. The results are presented as MFI values calculated by subtracting MFI values measured in bacteria-cell cultures incubated on ice from MFIs detected in bacteria-THP1 cell cultures incubated at 37 °C. The results of a representative experiment (n = 3) are shown.

Because Pellino proteins regulate TLR signaling (33, 40–44) and TLR signaling modulates phagocytosis (45–47), we studied the impact of Pellino-1 deficiency on phagocytosis of heat-killed E. coli and S. aureus by PMA-differentiated THP-1 cells. We used pHrodo Red-conjugated heat-killed bacteria and phagocytosis was judged by fluorescence of pH-sensitive pHrodo Red upon its localization in the phagolysosome compartment undergoing acidification, whereas under basal conditions, this fluorophore does not emit fluorescence (48). As controls, we measured autofluorescence of cells incubated in the absence of bacteria and in bacteria-cell cultures incubated on ice (to inhibit phagocytosis). Incubation of THP1 cells expressing scrambled shRNA with pHrodo Red-conjugated E. coli and S. aureus for 2 h at 37 °C markedly increased fluorescence compared with values observed in identical cultures incubated on ice (Fig. 6, B, top, and C). THP1 cells expressing Pellino-1 shRNA and incubated with heat-killed pHrodo Red-conjugated E. coli and S. aureus at 37 °C exhibited 19–25% statistically significant increases in mean fluorescence channel (MFI) values compared with scrambled shRNA-transduced THP-1 cells (Fig. 6, B, bottom, and C). Collectively, these results indicate that Pellino-1 positively regulates expression of MyD88- and TRIF-dependent cytokine genes but only weakly affects phagocytosis of heat-killed E. coli and S. aureus.

Pellino-1 Modulates NF-κB Reporter Activation Driven by Overexpression of MyD88, IRAK1, TRAF6, TAK1, or by Transfection of TRIF and TBK1 but Does Not Affect p65-driven Responses

To position Pellino-1 within the MyD88- and TRIF-dependent pathways, we initiated activation of a transfected NF-κB luciferase reporter in HEK293T cells by overexpression of a series of adapters or kinases, and studied the impact of co-expressed Pellino-1 on these processes. Overexpressed signaling intermediates initiate signaling downstream in a ligand-independent manner, because of their ability to dimerize, bypassing the requirement for agonist-inducible dimerization, and this is a commonly used approach to discern the location of a molecule in question within signaling axes (22, 24, 32, 33, 44, 49). Fig. 7A demonstrates that ectopic expression of Pellino-1 led to a synergistic up-regulation of MyD88- and TRIF-driven NF-κB reporter activation by 2.8–4.5- and 2.2- 3.4-fold (top two panels on the left), and increased NF-κB transactivation driven by IRAK1, TAK1, and TBK1 by 3.7-, 3.1-, and 1.5-fold, respectively (bottom panels). Pellino-1 transfection also increased the ability of overexpressed TRAF6 to mediate activation of the pELAM-Luc NF-κB luciferase reporter, but did not affect p65-driven NF-κB transactivation (Fig. 7A, top two panels on the right). As evidenced by co-immunoprecipitation, expressed Pellino-1 associated with transfected and, to a lesser extent, endogenous IRAK1 and TBK1 (Fig. 7B, top panel). Transfected IRAK1 and TBK1 exhibited patterns of post-translational modifications and phosphorylation (Fig. 7B, bottom panel), reflecting constitutive activation of overexpressed kinases, in line with reported data (32, 33, 50, 51). Overexpression of Pellino-1 led to increased abundance of modified species of IRAK1 and the phosphorylated form of TBK1 (Fig. 7B, bottom panel). Next, we studied the impact of Pellino-1 overexpression on basal and LPS-mediated phosphorylation of IRAK1 (an MyD88-dependent outcome), TBK1 and IRF3 (TRIF-dependent outcomes). Transfection of Pellino-1 constitutively induced phosphorylation of IRAK1, TBK1, and IRF3, and potentiated LPS-driven phosphorylation of these intermediates (Fig. 7C). These data indicate that Pellino-1 potentiates NF-κB activation driven by overexpression of MyD88, TRIF, IRAK1, TBK1, TAK1, and TRAF6 and increases basal and LPS-inducible phosphorylation of IRAK1 and TBK1, although not affecting p65-induced NF-κB.

FIGURE 7.

Pellino-1 acts within MyD88-IRAK1-TRAF6-TAK1 and TRIF-TBK1 signaling axes but does not affect p65-induced NF-κB reporter activation. A, HEK293T cells were plated in 24-well plates and co-transfected with or without FLAG-Pellino-1 (0.8 μg/well) in the presence of pELAM-Luc (0.4 μg/well) and pTK-Renilla Luc (20 ng/well) along with the indicated amounts of pcDNA3-AU1-MyD88 or pcDNA3-TRIF, or with expression vectors encoding IRAK1, TBK1, TRAF6, TAK1, or p65 (10 ng/well each). After recovery for 48 h, cell lysates were assayed for firefly versus Renilla luciferase activities, and data are presented as fold-induction compared with values in cells transfected with pcDNA3, pELAM-Luc, and pTK-Renilla Luc taken as 1. B, HEK293T cells were plated in 6-well plates and transfected with pcDNA3, pRK5-IRAK1, or pSuper-TBK1 (1 μg/well) along with pcDNA3 or FLAG-Pellino-1 (1 μg/well). After recovery for 48 h, cell lysates were prepared and immunoprecipitated (IP) with anti-Pellino-1 Ab followed by immunoblotting (IB) with anti-IRAK1 (left) or anti-TBK1 (right) Abs (top panels); or examined by Western blot analyzes with Abs for the indicated proteins (bottom panels). C, 293/TLR4 cells were plated in 6-well plates and transfected with pcDNA3-CD14, pEFBOS-MD2 (0.5 μg/well each), along with pcDNA3 or FLAG-Pellino-1 (1 μg/well each). After recovery for 48 h, cells were treated with medium or 100 ng/ml of LPS for the indicated times. Cell lysates were immunoprecipitated with anti-IRAK1 Ab followed by Western blot analyses with anti-p-IRAK1 or anti-IRAK1 Abs (top two panels) or analyzed by immunoblotting with Abs against the indicated proteins. The data of a representative (n = 3) experiment are shown.

Pellino-1 facilitates LPS-driven K63-linked Polyubiquitination of IRAK1, TRAF6, TAK1, and TBK1

Pellino-1 is an E3 Ub ligase capable of attaching K63-linked and K48-linked Ub residues to its protein targets and interacting with many signaling intermediates, including IRAK1, TBK1, TAK1, and TRAF6. Since our previous results demonstrated the ability of Pellino-1 to interact with IRAK1 and TBK1 and potentiate kinase post-translational modifications and signaling, we next examined whether Pellino-1 induces K63-linked polyubiquitination of IRAK1, TBK1, and downstream intermediates TAK1 and TRAF6. 293/TLR4/MD2 cells were transfected with K63-only HA-tagged Ub along with an empty vector or FLAG-Pellino-1, and basal as well as LPS-inducible K63-linked ubiquitination of endogenous IRAK1, TBK1, TAK1, and TRAF6 were analyzed by co-immunoprecipitation. LPS stimulation of cells transfected with pcDNA3 led to attachment of K63-only HA-Ub moieties to immunoprecipitated endogenous IRAK1, TBK1, TAK1, and TRAF6, whereas no or very little ubiquitination was observed in unstimulated cells (Fig. 8, A–D). Transfected Pellino-1 induced attachment of K63-only HA-Ub to all studied intermediates, in line with constitutive activation of overexpressed Pellino-1 (Figs. 2–7) (33), and markedly enhanced LPS-inducible K63-linked polyubiquitination of IRAK1, TBK1, TAK1, and TRAF6 (Fig. 8, A–D). Similarly designed experiments employing transfection of K48-only HA-Ub demonstrated only weak K48-linked polyubiquitination of IRAK1 and TBK1, which was not up-regulated by LPS, and showed the failure of transfected Pellino-1 to affect this signaling outcome (data not shown). Collectively, these results revealed the ability of transfected Pellino-1 to mediate basal and potentiate LPS-mediated K63-linked ubiquitination of IRAK1, TBK1, TAK1, and TRAF6.

FIGURE 8.

Pellino-1 promotes basal and LPS-inducible K63-linked polyubiquitination of IRAK1, TBK1, TAK1, and TRAF6. 293/TLR4 cells were plated in 100-mm² tissue culture dishes and transfected with pcDNA3-CD14, pEFBOS-MD2, and pEFBOS-HA-Ub K63 only (1 μg/dish each) along with either pcDNA3 or FLAG-Pellino-1 (2 μg/dish each). After recovery for 48 h, cells were treated with medium or 100 ng/ml of LPS for the indicated times, cell lysates were prepared and subjected to immunoprecipitation (IP) with Abs against IRAK1, TBK1, TAK1, or HA, as indicated. Immune complexes and whole cell lysates were analyzed by immunoblotting (IB) with anti-HA, anti-FLAG, anti-IRAK1, anti-TBK1, anti-TAK1, anti-TRAF6, and anti-tubulin Abs, as depicted. Shown are the data of a representative (n = 3) experiments.

Discussion

Previous studies by us and others demonstrated inhibited activation of the MyD88- and TRIF-dependent pathways in endotoxin-tolerized human monocytes and mouse macrophages, including suppressed induction of MyD88-dependent (pro-IL-1β, TNF-α, IL-6, and IL-12 p40) and TRIF-dependent (IFN-β) cytokine genes (22–24, 32, 36, 52–55). We previously reported that endotoxin tolerance impairs LPS-inducible activation of IRAK4, IRAK1, TBK1, and TAK1 (23, 24, 54, 55), and, in the present study, sought to analyze molecular mechanisms of this phenomenon. Because Pellino-1 has been reported to interact with IRAK4 and IRAK1 kinases (33, 56), and is capable of forming a signalosome with TBK1, TAK1, and TRAF6 (26, 33, 57), we wanted to analyze the role of Pellino-1 in endotoxin tolerance and TLR2/4 signaling. To the best of our knowledge, this paper demonstrates for the first time that in vivo induction of endotoxin tolerance significantly blunted LPS-inducible Pellino-1 gene expression in mouse peritoneal macrophages. Similarly, in vitro endotoxin tolerization of human primary monocytes and macrophage-like MonoMac-6 and THP-1 cells abrogated the capacity of LPS to up-regulate Pellino-1 mRNA and protein expression. The induction of Pellino-1 in endotoxin-tolerized cells was markedly inhibited at all times (by 75–96%). Notably, reduction in the Pellino-1 expression upon Pellino-1 knockdown in THP-1 cells is manifested to a lower extent compared with residual Pellino-1 present in LPS-tolerant monocytes and macrophages, yet it was sufficient to markedly attenuate TLR2/4 signaling. Thus, decreased Pellino-1 levels and, possibly, its deficient activation, dependent on phosphorylation of several Ser and Thr residues, ubiquitination and sumoylation (15), could underlie compromised TLR-inducible activation of NF-κB and inflammatory cytokine genes in endotoxin-tolerized cells. Interestingly, Pellino-1 has been reported to interact with positive (IRAK1, IRAK4, TBK1, TRAF6, TAK1, and deformed epidermal autoregulatory factor-1) and negative (SMAD6 and SMAD7) TLR regulators (reviewed in Ref. 15). Whether Pellino-1 can also interact with IRAK-M, SHIP1, or A20, known negative regulators of TLR signaling (reviewed in Ref. 4), is yet to be determined. It is tempting to speculate that altered expression and, possibly, activity of Pellino-1, imposes changes in other regulators of TLR signaling, initiating complex regulatory networks resulting in altered post-translational modifications of TLR intermediates, transcriptional changes, and re-programming of TLR signaling in endotoxin-tolerized cells.

Although the molecular mechanisms behind reduced LPS inducibility of Pellino-1 in endotoxin tolerance is unknown, it is possible that decreased activation of TBK1 and IRF3 in LPS-tolerant cells is responsible (24), because the TBK-IRF3 axes induces Pellino-1 expression and TBK1 up-regulates its activity (26, 57). Interestingly, Pellino-1 interacts with inhibitory Smad6 and Smad7 involved in the TGF-β pathway, leading to reduced interactions of IRAK1 with IL1R/TLR, impaired NF-κB activation, and cytokine expression (58, 59). Such a mechanism could also play a role during endotoxin tolerance, given up-regulation of TGF-β expression upon endotoxin tolerization (60). Studies are in progress to dissect the molecular mechanisms responsible for decreased LPS inducibility of Pellino-1 in endotoxin-tolerized cells.

Controversial results have been published regarding the role of Pellino-1 in MyD88- and TRIF-driven cascades in macrophages and other cell types (reviewed in Ref. 15). To establish the role of Pellino-1 in TLR2 and TLR4 signaling, we employed Pellino-1 overexpression and shRNA-mediated ablation strategies. Pellino-1 overexpression constitutively induced NF-κB and synergistically up-regulated Pam3Cys- or LPS-inducible NF-κB activation and expression of IL-8 mRNA in HEK293 cells expressing TLR2 or TLR4/MD2, whereas Pellino-1 knockdown decreased these responses. We confirmed these results in macrophage-like THP-1 cells, where Pellino-1 ablation resulted in decreased expression of MyD88- (TNF-α, IL-6) and TRIF-dependent (IFN-β, CCL5) cytokine genes in response to TLR2 (Pam3Cys) and TLR4 (LPS) agonists and heat-killed S. aureus and E. coli. The involvement of Pellino-1 in MyD88- and TRIF-dependent signaling was also supported by our results showing that ectopic overexpression of Pellino-1 synergistically increased NF-κB reporter activation driven by transfected MyD88 or IRAK1, on one hand, and by TRIF or TBK1, on the other. Furthermore, co-immunoprecipitation experiments revealed that Pellino-1 interacts with transfected IRAK1 and TBK1 upon their co-expression, and potentiates post-translational modifications of IRAK1 and phosphorylation of TBK1. Subsequent analyses of downstream components of the MyD88 and TRIF pathways showed that Pellino-1 up-regulated NF-κB reporter activation driven by overexpression of TAK1 and TRAF6, whereas Pellino-1 did not affect p65-induced NF-κB.

Consistent with the ability of Pellino-1 to interact with multiple TLR intermediates, i.e. IRAK1, TBK1, RIP-1, and TRAF6 (26, 33, 56, 57), Pellino-1 could promote the assembly of a multiprotein complex of these intermediates and activate signaling capacities of each in an E3 Ub ligase-dependent and/or independent manner (reviewed in Ref. 15). Conversely, Pellino-1 could primarily target and activate upstream kinases IRAK1 and TBK1. In this scenario, even if signaling is initiated at the level of downstream molecules (e.g. TAK1 and TRAF6) rendered constitutively active by transfection, Pellino-1-activated upstream kinases could potentiate such responses. For instance, this could occur by promoting TRAF6 and TAK1 recruitment to K63-polyubiquitinated IRAK1 (7). Further studies are required to uncover which mechanism is utilized by Pellino-1.

Although in vitro Pellino-1 mediates both K48-linked, “degradative” and K63-linked, “signaling-promoting” polyubiquitination of its targets, such as IRAK1 (61), in vivo studies demonstrated primarily Pellino-1-mediated K63-linked polyubiquitination (33, 50) that promotes protein-protein interactions and enhances downstream signaling (15). This paper shows that co-expression of Pellino-1 exclusively up-regulates K63-linked polyubiquitination of IRAK1, TAK1, and TBK1 in 293/TLR4/MD2 cells stimulated with LPS, whereas it failed to affect weak K48-linked polyubiquitination of these molecules (data not shown), confirming previously published findings (33, 56, 62). Of importance, we demonstrate for the first time that Pellino-1 also enhances LPS-driven K63-linked polyubiquitination of TBK1, which might play a role in promoting the ability of TBK1 to form downstream signalosome complexes and activate its kinase activity. It would be interesting to delineate whether Pellino-1 alone acts as a positive regulator of IRAK1- and TBK1-driven pathways or it can co-opt other members of the Pellino family implicated in positive regulation of IL-1R/TLR pathways, e.g. Pellino-2 (63), and other E3 Ub ligases fine-tuning TBK1 signaling, e.g. Nrdp1 (64).

Modulation of phagocytosis by TLR signaling has been a subject of controversy (33, 40–44, 65). Because Pellino-1 positively controls TLR2/4-mediated signaling outcomes, we examined whether it regulates phagocytosis. Our data show only a weak impact of Pellino-1 ablation on phagocytosis of heat-killed, pHrodo Red-conjugated S. aureus and E. coli by THP-1 cells, as judged by phagolysosome acidification-induced fluorescence of bacteria-conjugated, pH-sensitive fluorophore. Thus, despite that Pellino-1 acted as a positive regulator of TLR-inducible K63-linked polyubiquitination of IRAK1, TBK1, and activator of MyD88/IRAK1- and TRIF/TBK1-driven cytokine genes, it only weakly affected phagocytosis of heat-killed bacteria. These results may suggest that Pellino-1 is unlikely to affect host innate immune responses during infections by regulating microbial uptake or phagocytosis, but rather acts via potentiation of production of MyD88- and TRIF-dependent inflammatory cytokines. Notably, Pellino-1-deficient and sufficient human pulmonary bronchial epithelial cells had comparable replication of rhinovirus upon in vitro infection, whereas significant reduction of CXCL8 expression was noted in Pellino-1^−/− cells (40). Further studies are required to determine a currently unknown impact of Pellino-1 deficiency on microbial loads during infection in vivo and phagocytosis of live bacteria.

In summary, we report herein for the first time that induction of endotoxin tolerance in vivo and in vitro down-regulated LPS inducibility of Pellino-1. Pellino-1 regulated both MyD88- and TRIF-dependent signaling pathways via association with IRAK1 and TBK1 and by positively regulating their LPS-mediated K63-linked polyubiquitination, leading to increased expression of inflammatory cytokines. We also report that Pellino-1 increased NF-κB activation driven by the MyD88-IRAK1-TRAF6-TAK1 and TRIF-TBK1 signaling axes and up-regulated K63-linked polyubiquitination of TAK1 and TRAF6, but did not affect p65-induced NF-κB. In contrast to its prominent regulation of TLR2/4-driven cytokines, Pellino-1 only weakly affected phagocytosis of heat-killed S. aureus and E. coli. Given that Pellino-1 deficiency ameliorates poly(I:C)- and LPS-induced septic shock (25), it would be important to determine consequences of Pellino-1 knock-out on susceptibility to microbial infections in vivo, infection morbidity and mortality, and to assess responses of macrophages and dendritic cells. It is tempting to speculate that small molecule- or peptide-based pharmacological inhibition of Pellino-1 signaling at early stages of sepsis or in patients with autoimmune and inflammatory diseases could provide beneficial therapeutic outcomes. In this respect, it is noteworthy that Pellino-1-interacting SMAD6 blocking peptides were shown to inhibit manifestations of lethal inflammatory responses in mice in sepsis models of cecal ligation and puncture and endotoxemia (61).

Author Contributions

A. E. M. conceived and coordinated the study and wrote the paper. M. M. designed, performed, and analyzed the experiments shown in Figs. 1, 2, and 4–8, Y. X. designed, performed, and analyzed the experiments shown in Figs. 1, 2, 3, and 7, G. P. designed, performed, and analyzed the experiments shown in Figs. 1, 2, 6, and 8, and F. Q. designed, performed, and analyzed the experiments shown in Figs. 1 and 2. All authors reviewed the results and approved the final version of the manuscript.