Deubiquitinase CYLD acts as a negative regulator for bacterium NTHi-induced inflammation by suppressing K63-linked ubiquitination of MyD88

Significance

Despite the critical role for myeloid differentiation factor 88 (MyD88) in mediating pathogen-induced innate and adaptive immune responses, how functional activity of MyD88 is tightly controlled remains unknown. Among various types of posttranslational modification, ubiquitination and deubiquitination of host signaling molecules play an important role in tightly regulating immune response. However, inducible ubiquitination, in particular proteolysis-independent polyubiquitination as well as deubiquitination of MyD88, remains largely unclear. Here, we demonstrate that deubiquitinase CYLD negatively regulates MyD88-mediated inflammation by directly interacting with MyD88 and deubiquitinating nontypeable Haemophilus influenzae-induced lysine 63-linked polyubiquitination of MyD88. Thus, our studies may not only bring insights into the negative regulation of Toll-like receptor–MyD88-dependent signaling but may also lead to the development of a previously unidentified therapeutic strategy for uncontrolled inflammation.

Abstract

Myeloid differentiation factor 88 (MyD88) acts as a crucial adaptor molecule for Toll-like receptors (TLRs) and interleukin (IL)-1 receptor signaling. In contrast to the well-studied positive regulation of MyD88 signaling, how MyD88 signaling is negatively regulated still remains largely unknown. Here, we demonstrate for the first time to our knowledge that MyD88 protein undergoes lysine 63 (K63)-linked polyubiquitination, which is functionally critical for mediating TLR–MyD88-dependent signaling. Deubiquitinase CYLD negatively regulates MyD88-mediated signaling by directly interacting with MyD88 and deubiquitinating nontypeable Haemophilus influenzae (NTHi)-induced K63-linked polyubiquitination of MyD88 at lysine 231. Importantly, we further confirmed this finding in the lungs of mice in vivo by using MyD88^−/−CYLD^−/− mice. Understanding how CYLD deubiquitinates K63-linked polyubiquitination of MyD88 may not only bring insights into the negative regulation of TLR–MyD88-dependent signaling, but may also lead to the development of a previously unidentified therapeutic strategy for uncontrolled inflammation.

Toll-like receptors (TLRs) are the major receptors to recognize and respond to microbial components. TLRs have a conserved cytoplasmic Toll/IL-1R (TIR) domain (1–5). Bacterial products are extremely complex and induce TLR-triggered innate immune responses through only four TIR motif adaptor molecules such as MyD88 (myeloid differentiation factor 88), TIRAP/MAL (MyD88-adaptor-like/TIR-associated protein), TICAM1/TRIF (Toll-receptor-associated activator of IFN), and TICAM2/TRAM (Toll-receptor-associated molecule). The activation of the downstream protein kinases and the transcription factors NF-κB, AP-1, IRF-3, and IRF7 leads to the up-regulation of proinflammatory cytokines, chemokines, and type 1 IFN (2, 6, 7). MyD88 is an essential adaptor molecule for TLRs except TLR3 and interleukin (IL)-1 receptor (1–4). Despite the prominent role for MyD88 in mediating pathogen-induced innate and adaptive immune responses, how functional activity of MyD88 is tightly controlled remains largely unknown.

Negative feedback regulation of host signaling is critical for controlling overactive immune/inflammatory response via posttranslational modification of the target proteins (8–11). Among various types of posttranslational modification, ubiquitination and deubiquitination of host signaling molecules play an important role in tightly regulating immune/inflammatory response (12–19). Deubiquitinase CYLD has been shown to play a critical role in tightly controlling inflammation by negatively regulating TLR/NF-κB pathway (9, 10, 20–25). Despite the recent identification of Smurf as lysine 48 (K48)-linked E3 ubiquitin ligase of MyD88, the molecular mechanism underlying proteolysis-independent lysine 63 (K63)-linked polyubiquitination of MyD88 remains largely unknown.

Nontypeable Haemophilus influenzae (NTHi), a Gram-negative bacterium, is the leading cause of exacerbation of chronic obstructive pulmonary disease (COPD) and otitis media (OM). NTHi induces inflammatory response via activation of TLR-dependent TRAF6–IKK–NF-κB and TRAF6–MKK3/6–p38 signaling pathways (26–30). In addition, phosphodiesterase 4B (PDE4B) also plays a key role in regulating NTHi-induced inflammation (26, 31). Moreover, deubiquitinase CYLD acts as a critical negative regulator for NTHi-induced inflammation via negative cross-talk with TRAF6 and TRAF7 (10, 32). However, it still remains unclear whether CYLD negatively regulates NTHi-induced inflammation via direct inhibition of MyD88 signaling.

In the present study, we show that NTHi induces K63-linked polyubiquitination of MyD88 in vitro and in vivo. CYLD acts as a negative regulator for NTHi-induced inflammation by suppressing K63-linked polyubiquitination of MyD88.

Results and Discussion

Bacterial Pathogen NTHi Induces K63-Linked Polyubiquitination of MyD88 in Vitro and in Vivo.

We initially sought to determine whether NTHi, a major respiratory bacterial pathogen, induces polyubiquitination of MyD88 in human epithelial cells. As shown in Fig. 1A, NTHi induced polyubiquitination of MyD88 in a time-dependent manner in HeLa CCL2 cells cotransfected with Myc-tagged full-length MyD88 (Myc-MyD88 WT) and HA-tagged ubiquitin with all lysine residues (HA-WT Ub). We next examined the polyubiquitination of the endogenous MyD88 protein in NTHi-stimulated HeLa CCL2 cells and in the lungs of mouse in vivo. NTHi also induced polyubiquitination of the endogenous MyD88 proteins in HeLa CCL2 cells and in the lungs of mouse in vivo (Fig. 1 B and C). Because K63-linked polyubiquitination plays an important role in activation of target proteins, we determined whether NTHi induces K63-linked polyubiquitination of endogenous MyD88. Interestingly, NTHi induced the K63-linked polyubiquitination of endogenous MyD88 (Fig. 1D). Consistent with the in vitro results, NTHi also induced the K63-linked polyubiquitination of the endogenous MyD88 protein in the lungs of mouse in vivo (Fig. 1E). Together, our results demonstrate for the first time to our knowledge that bacterium NTHi induces the K63-linked polyubiquitination of endogenous MyD88 protein in vitro and in vivo.

Fig. 1.

NTHi induces lysine 63 (K63)-linked polyubiquitination of MyD88 in vitro and in vivo. (A) Myc-tagged full-length of MyD88 (Myc-MyD88 WT) was cotransfected into HeLa CCL2 cells with a HA-tagged ubiquitin with all lysine residues (HA-WT Ub). Cells were treated with NTHi for the indicated time periods. Cell lysates were then immunoprecipitated (IP) with anti-HA antibody and immunoblotting (IB) was then performed with the anti-Myc antibody. (B) HeLa CCL2 cells were treated with NTHi for 1, 2, 4, or 6 h, and cell lysates were immunoprecipitated with anti-Ub antibody and analyzed by immunoblotting with anti-MyD88 antibody. (C) Anesthetized C57BL/6 mice were intratracheally inoculated with NTHi for 1, 2, 4, or 6 h. Lung lysates from mice were immunoprecipitated with anti-Ub antibody and analyzed by immunoblotting with anti-MyD88 antibody. Lysates of HeLa CCL2 (D) and mouse lung (E) were immunoprecipitated with anti-lysine 63 Ubiquitin (K63 Ub) antibody and analyzed with anti-MyD88 antibody. All data are representative of at least three independent experiments.

CYLD Deubiquitinates K63-Linked Polyubiquitinated MyD88.

Having demonstrated NTHi-induced K63-linked polyubiquitination of MyD88, we next sought to determine the deubiquitinase for K63-ubiquitinated MyD88. Among a number of deubiquitinases, CYLD has been identified as a critical deubiquitinase for a number of signal transducers including TRAF2, TRAF6, NEMO, and Bcl-3 (9, 10, 20, 24, 33). We recently also showed that CYLD negatively regulates transforming growth factor (TGF)-β signaling via deubiquitinating Akt (8). In addition, there is also a study showing that CYLD deubiquitinates Lck in T cells (34). Moreover, MyD88 polyubiquitination has been suggested to play a critical role in TLR signaling pathway (35, 36). To examine whether CYLD deubiquitinates MyD88, HeLa CCL2 cells were transfected with Myc-MyD88 WT, HA-WT Ub, or Flag-CYLD WT. As shown in Fig. 2A, NTHi induced the polyubiquitination of MyD88 in a time-dependent manner. Interestingly, NTHi-induced polyubiquitination of MyD88 was inhibited in cells cotransfected with Flag-CYLD WT. We next investigated whether CYLD specifically deubiquitinates K63-linked polyubiquitinated MyD88 by cotransfecting HeLa CCL2 cells with plasmids encoding Myc-MyD88 WT, HA-WT Ub (ubiquitin with all lysine residues), HA-K63 Ub (ubiquitin with K63 only), HA-K48 Ub (ubiquitin with K48 only), or HA-KO Ub (ubiquitin with no lysine residues). Coexpressing Flag-CYLD WT markedly reduced K63-linked polyubiquitination of MyD88 (Fig. 2B). We further determined whether CYLD inhibits K63-ubiquitinated MyD88 in a deubiquitinating enzyme (DUB) activity-dependent manner by evaluating the effects of CYLD WT, DUB-deficient CYLD mutant (CYLD H/N), and CYLD N-terminal deletion mutant (CYLD D/N) as well as pSuper-shCYLD (shCYLD) on K63-linked polyubiquitination of MyD88. As shown in Fig. 2C, expressing CYLD WT markedly decreased K63-linked polyubiquitination of MyD88. In contrast, DUB-deficient CYLD mutant (CYLD H/N) and shCYLD greatly enhanced K63-linked polyubiquitination of MyD88. As expected, CYLD D/N was unable to suppress polyubiquitination compared with CYLD WT. To further confirm the negative role of CYLD in suppressing MyD88 polyubiquitination, we examined the polyubiquitination of MyD88 in the lungs of both WT and CYLD^−/− mice inoculated with NTHi. NTHi-induced MyD88 polyubiquitination was enhanced in CYLD^−/− mice compared with that in WT mice (Fig. 2D). Collectively, these data demonstrate that CYLD is a deubiquitinase for NTHi-induced K63-linked polyubiquitination of MyD88.

Fig. 2.

CYLD deubiquitinates K63-linked polyubiquitinated MyD88. (A) HeLa CCL2 cells were cotransfected with Myc-MyD88 WT, HA-WT Ub, or Flag-CYLD WT, and then treated with NTHi for 1, 2, 4, and 6 h. Cell lysates were immunoprecipitated with anti-HA antibody and analyzed with anti-Myc antibody. (B) Cells were cotransfected with Myc-MyD88 WT, HA-WT Ub, HA-K63 Ub, HA-K48 Ub, HA-KO Ub, or Flag-CYLD WT plasmids and then treated with NTHi for 4 h. Cell lysates were immunoprecipitated with anti-HA antibody and analyzed by immunoblotting with anti-Myc antibody. (C) Cells were cotransfected with Myc-MyD88 WT, HA-K63 Ub, Flag-CYLD WT, Flag-CYLD N-terminal deletion mutant (CYLD D/N), Flag-CYLD deubiquitinating (DUB) enzyme deficient (CYLD H/N), or pSuper-shCYLD plasmids and then treated with NTHi for 4 h. Cell lysates were immunoprecipitated with anti-HA antibody and analyzed by immunoblotting with anti-Myc antibody. (D) C57BL/6 and CYLD^−/− mice were intratracheally inoculated with NTHi for the indicated times, and the lung lysates were then collected. Lung lysates were immunoprecipitated with anti-Ub antibody and analyzed by immunoblotting with anti-MyD88 antibody. All data are representative of at least three independent experiments.

CYLD Directly Interacts with MyD88 and Deubiquitinates NTHi-Induced K63-Linked Polyubiquitination of MyD88 at K231.

We further investigated which domains of MyD88 participate in interaction with CYLD. To determine the region of MyD88 binding to CYLD, Flag-CYLD WT was cotransfected into HeLa CCL2 cells with plasmids encoding the full length of MyD88 (Myc-MyD88 WT) or Myc-MyD88 (1-160) (amino acids 1–160), which contains death domain (DD) and intermediate domain (ID), or Myc-MyD88 (161-296) (amino acids 161–296) that contains TIR domain. As shown in Fig. 3A, CYLD directly interacts with Myc-MyD88 WT and Myc-MyD88 (161–296), but not with Myc-MyD88 (1–160) as assessed by performing coimmunoprecipitation (Co-IP) analysis. To confirm the direct interaction between CYLD and MyD88 under endogenous conditions, cells were stimulated with NTHi for 4 h. Interaction between CYLD and MyD88 was then assessed by performing Co-IP analysis. As shown in Fig. 3B, CYLD also interacts with MyD88 under endogenous conditions. These results demonstrate that TIR domain of MyD88 is critical for its interaction with CYLD. We next determined which region of MyD88 undergoes polyubiquitination in response to NTHi. Interestingly, NTHi-induced K63-linked polyubiquitination of MyD88 was observed in HeLa CCL2 cells cotransfected with HA-K63 Ub and MyD88 WT and MyD88 (161–296), but not with MyD88 (1–160) (Fig. 3C). We next sought to determine which lysine residue in this region is involved in K63-linked polyubiquitination of MyD88. Among several lysine residues in TIR domain of MyD88, lysine 231 (K231) has recently been shown to play a key role in MyD88-dependent TLR4 signaling. We thus determined whether K231 of TIR domain is involved in NTHi-induced K63-linked polyubiquitination of MyD88 by using a MyD88 mutant construct with point mutation at K231 to arginine (R) (MyD88 K231R). Indeed, K231R, but not K214R, markedly reduced K63-linked polyubiquitination of MyD88 compared with MyD88 WT (Fig. 3D). We next determined whether K231 residue in MyD88 is functionally involved in mediating CYLD-induced inhibition of MyD88 signaling. As shown in Fig. 3E, coexpressing MyD88 K231R, but not K214R, induced NF-κB promoter activity to a lesser extent compared with expressing MyD88 WT. We further demonstrated that NTHi induces K63-linked polyubiquitination of MyD88 but not MyD88 K231R in a time-dependent manner (Fig. 3F). Next, we determined whether K231 lysine of MyD88 is involved in CYLD-mediated inhibition of K63-linked polyubiquitination of MyD88 by cotransfecting the cells with Flag-CYLD WT or shCYLD and HA-K63 Ub, Myc-MyD88 WT, or Myc-MyD88 K231R. As shown in Fig. 3G, CYLD WT inhibited, whereas shCYLD enhanced, NTHi-induced K63-linked polyubiquitination of MyD88. In contrast, NTHi-induced K63-linked polyubiquitination of MyD88 was absent in cells cotransfected with Myc-MyD88 K231R compared with cells cotransfected with Myc-MyD88 WT. Flag-CYLD WT or shCYLD was unable to either suppress or enhance NTHi-induced K63-linked polyubiquitination of MyD88 in cells cotransfected with Myc-MyD88 K231R. Taken together, our data indicate that MyD88 lysine 231 is critical for CYLD-mediated inhibition of NTHi-induced K63-linked polyubiquitination of MyD88.

Fig. 3.

CYLD directly interacts with MyD88 and deubiquitinates NTHi-induced K63-linked polyubiquitination of MyD88 at lysine 231 (K231). (A) Plasmids encoding full-length MyD88 (Myc-MyD88 WT), Myc-MyD88 (1–160), or Myc-MyD88 (161–296) were cotransfected into HeLa CCL2 cells with Flag-CYLD WT, and the cells were then treated with NTHi for 4 h. Cell lysates were immunoprecipitated with anti-Myc antibody and analyzed by immunoblotting with anti-CYLD antibody. (B) Cells were stimulated with NTHi for 4 h. Cell lysates were then immunoprecipitated with anti-MyD88 antibody and analyzed by immunoblotting with anti-CYLD antibody. (C) Cells were cotransfected with Myc-MyD88 WT, Myc-MyD88 (1–160), Myc-MyD88 (161–296), or HA-K63 Ub plasmids, and cells were then treated with NTHi for 1, 2, 4, and 6 h. Immunoprecipitation (IP) and immunoblotting (IB) were performed with the indicated antibodies. (D) The point mutants of MyD88 in which the indicated lysine (K) residue is substituted with arginine (R) were cotransfected into HeLa CCL2 cells with HA-K63 Ub plasmid. Cells were then treated with NTHi for 4 h. Cell lysates were immunoprecipitated with anti-HA antibody and analyzed by immunoblotting with anti-Myc antibody. (E) NF-κB promoter activity was determined in Myc-MyD88 WT, Myc-MyD88 K214R, Myc-MyD88 K231R, or pSuper-shCYLD-transfected cells stimulated with NTHi for 5 h. NF-κB promoter activity for shCYLD-transfected group was normalized by only pSuper-shCYLD–transfected cells. *P < 0.01. Values are the mean ± SD (n = 3). (F) Cells were cotransfected with Myc-MyD88 WT, Myc-MyD88 K231R, or HA-K63 Ub plasmids and then treated with NTHi for 1, 2, 4, and 6 h. Cell lysates were immunoprecipitated with anti-HA antibody and analyzed by immunoblotting with anti-Myc antibody. (G) Cells were cotransfected with Myc-MyD88 WT, Myc-MyD88 K231R, Flag-CYLD WT, pSuper-shCYLD, or HA-K63 Ub plasmids and then treated with NTHi for 4 h. Cell lysates were immunoprecipitated with anti-HA antibody and analyzed by immunoblotting with anti-Myc antibody. All data are representative of at least three independent experiments.

CYLD Acts as a Negative Regulator for NTHi-Induced MyD88-Dependent Inflammation in Vitro and in Vivo.

Next, we sought to determine if CYLD indeed negatively regulates bacterium NTHi-induced inflammatory response via MyD88 in vitro and in vivo. We first investigated if depletion of CYLD is still able to enhance NTHi-induced up-regulation of proinflammatory mediators in MyD88-depleted HeLa CCL2 cells. As shown in Fig. 4 A–D, depletion of CYLD using shCYLD no longer enhanced NTHi-induced up-regulation of hIL-1β, hIL-6, hTNF-α, and hMIP-2 in MyD88-depleted cells using shMyD88 as assessed by performing quantitative real-time RT-PCR analyses. We further confirmed this finding in the lungs of mice in vivo using MyD88^−/−CYLD^−/− mice. As expected, CYLD deficiency markedly enhanced NTHi-induced up-regulation of mIL-1β, mIL-6, mTNF-α, and mMIP-2 in the lungs of CYLD^−/− mice compared with those in WT mice (Fig. 4 E–H). In contrast, CYLD deficiency no longer enhanced NTHi-induced up-regulation of proinflammatory mediators in the lungs of MyD88^−/−CYLD^−/− mice. To further determine the sources of inflammatory cytokines in the lungs of NTHi-stimulated mice, we performed immunofluorescence staining using anti-CD45 antibody (Red) as a leukocyte marker, anti-IL-1β antibody (Green) and DAPI (Blue) for the nuclei. As shown in Fig. 4I, CD45-negative alveolar epithelial cells and also CD45-positive polymorphonuclear neutrophils (PMN) were strongly stained with anti-IL-1β antibody. Consistent with the finding shown in Fig. 4 A and E, IL-1β-positive cells were markedly increased in the lungs of CYLD^−/− mice along with increased CD45/IL-1β double-positive cells. In accordance with these results, CYLD deficiency also no longer enhanced NTHi-induced PMN infiltration in bronchoalveolar lavage (BAL) fluid of MyD88^−/−CYLD^−/− mice (Fig. 4J). Similar results were also confirmed by performing histological analysis to evaluate the NTHi-induced inflammation in the lungs of MyD88^−/−CYLD^−/− mice (Fig. 4K). Taken together, these data demonstrate that CYLD indeed acts as a negative regulator for MyD88-mediated inflammatory response.

Fig. 4.

CYLD acts as a negative regulator for NTHi-induced MyD88-dependent inflammation in vitro and in vivo. (A–D) HeLa CCL2 cells were transfected with pSuper-shControl (shCON), pSuper-shCYLD (shCYLD), or pRS-shMyD88 (shMyD88) as indicated in the figure and then treated with NTHi for 6 h. The expression of proinflammatory mediators were analyzed by quantitative real-time RT-PCR (Q-PCR). (E–J) WT, MyD88 knockout (MyD88^−/−), CYLD knockout (CYLD^−/−), or MyD88/CYLD double knockout (MyD88^−/−CYLD^−/−) mice were intratracheally inoculated with NTHi for 9 h. (E–H) mRNA expression of proinflammatory mediators in the lungs of mice was analyzed by Q-PCR. *P < 0.01, **P < 0.05. Values are the mean ± SD (n = 3) in A–H. (I) The lung sections were stained by using anti-CD45 antibody (red) as a leukocyte marker, anti-IL-1β antibody (green) and DAPI (blue) for the nuclei. (Scale bar: 50 μm. Magnification: 400x.) (J) Mice were intratracheally inoculated with NTHi for 9 h. BAL fluid was harvested 9 h after NTHi inoculation, and the cells from BAL fluid were cytocentrifuged and stained with Diff-Quik staining kit. (Scale bar: 100 μm. Magnification: 100×.) The number of PMN cells from BAL fluid was counted by using a hemocytometer under the microscope (n = 3). (K) The lung tissues were dissected 9 and 24 h after NTHi inoculation. Lung tissues were fixed with formalin, embedded with paraffin, and H&E stained. (Scale bar: 100 μm. Magnification: 100x.) Blinded histopathologic scoring of lung inflammation was performed on H&E-stained lung sections in a grade 0–3 (n = 5). (L) A schematic model illustrating CYLD as a novel deubiquitinase for K63-linked polyubiquitination of MyD88. All data are representative of at least three independent experiments.

Excessive or uncontrolled inflammatory response leads to a variety of diseases such as septic shock, asthma and cancer (37, 38). Thus, inflammatory response must be tightly regulated. In the present study, we demonstrate for the first time that CYLD acts as a novel deubiquitinase for MyD88-mediated inflammation by directly interacting with and deubiquitinating bacteria-induced K63-linked polyubiquitination of MyD88 at lysine 231 (Fig. 4L). Despite the critical role MyD88 plays in mediating pathogen-induced inflammatory response, how MyD88 signaling is negatively regulated still remains largely unknown. Negative feedback regulators, e.g., deubiquitinase CYLD, are crucial for tightly controlling inflammation. Understanding how CYLD deubiquitinates K63-linked polyubiquitination of MyD88 may not only bring insights into the negative regulation of TLR-MyD88-dependent signaling but may also lead to the development of a previously unidentified therapeutic strategy. Future studies should focus on investigating which E3 ligase is involved in K63-linked polyubiquitination of MyD88.

Materials and Methods

Bacterial Culture.

Clinically isolated NTHi strain 12 was used in this study. As described (26), bacteria were grown on chocolate agar plate in 37 °C, 5% (vol/vol) CO₂ incubator and then inoculated in brain heart infusion broth supplemented with 3.5 μg/mL NAD and 10 μg/mL hemoglobin (BD Biosciences). After overnight, bacteria were grown until OD₆₀₀ = 0.3∼0.4 and then washed with PBS and resuspended in PBS for in vitro cell experiments and in saline for in vivo animal experiments.

Cell Culture.

Human cervical epithelial HeLa CCL2 cells were maintained in MEM medium (Cellgro) with 10% (vol/vol) FBS (Sigma-Aldrich) and Pen/Strep (100 U/mL penicillin and 0.1 mg/mL streptomycin; Life Technologies). Cells were cultured in 5% (vol/vol) CO₂ incubator at 37 °C.

Real-Time Quantitative RT-PCR Analysis.

Total RNA was isolated with TRIzol reagent (Life Technologies) by following the manufacturer’s instruction. For the reverse transcription reaction, TaqMan reverse transcription reagents (Life Technologies) were used as described (8, 26). Quantitative RT-PCR (Q-PCR) analyses were performed by using SYBR Green Universal Master Mix (Life Technologies). In brief, reactions were performed in triplicate. The reactions contain 2× Universal Master Mix, 1 μL of template cDNA, 500 nM primers in a final volume of 12.5 μL, and were then analyzed in a 96-well optical reaction plate (USA Scientific). Reactions were amplified and quantified by using a StepOnePlus Real-Time PCR System and the manufacturer’s corresponding software (StepOnePlus Software v2.3; Life Technologies). The relative quantities of mRNAs were determined by using the comparative Ct method and were normalized by using mouse glyceraldehydes-3-phosphate dehydrogenase (GAPDH) and human cyclophilin as an endogenous controls for human and mouse, respectively. The primers were used as described (26, 39).

Plasmids, Transient Transfection, and Reporter Assay.

The expression plasmids Flag-CYLD wild type (WT), Flag-CYLD H/N, Flag-CYLD N-terminal deletion mutant (CYLD D/N), NF-κB Luc, pRK5-HA-WT Ub (ubiquitin with all lysine residues), pRK5-HA-K63 Ub (ubiquitin with K63 only), pRK5-HA-K48 Ub (ubiquitin with K48 only), and pRK5-HA-KO Ub (ubiquitin with no lysine residues) were described (8). Plasmids encoding MyD88 WT, MyD88 (1–160), and MyD88 (161–296) were amplified from pCMV4-HA-MyD88 WT plasmid by PCR and subcloned into the BamH1 and Sal1 sites of the pcDNA3-Myc vector. The lysine (K) to alanine (R) mutants of MyD88 were generated with Myc-MyD88 WT by using QuikChange XL Site-Directed Mutagenesis kit (Stratagene). Empty vector was used as a control and was also added where necessary to ensure an equivalent amount of input DNA. All transient transfections were carried out in triplicate by using TransIT-LT1 reagent (Mirus) following the manufacturer’s instruction. HeLa CCL2 cells were transiently transfected with plasmids and then treated with NTHi for 1, 2, 4, or 6 h. Cell lysates and RNA were isolated for immunoprecipitation, immunoblotting, and Q-PCR. NF-κB luciferase activity was measured by luciferase assay and normalized with respect to β-galactosidase activity as described previously (31).

RNA-Mediated Interference.

RNA-mediated interference for down-regulating CYLD expression was carried out by using pSuper-shCYLD (8). Human shRNAs for MyD88 were purchased from Origene. Human shMyD88 consists of four shRNAs and sequences for the shRNAs are as follows: Human MyD88 (5′-AGGAGATGATCCGGCAACTGGAACAGACA-3′, 5′-CCTGAGGTTCATCACTGTCTGCGACTACA -3′, 5′-CTGAGCGTTTCGATGCCTTCATCTGCTAT-3′, 5′-AGGAGGCTGAGAAGCCTTTACAGGTGGCC -3′).

Western Blot and Ubiquitination Experiments.

Western blot analysis, immunoprecipitation, and ubiquitination experiments were performed as follows (8). Western blots were performed by using whole cell lysates, separated on 10% (wt/vol) SDS/PAGE gels, and transferred to polyvinylidene difluoride membranes (PVDF). The membrane was blocked with 5% (wt/vol) BSA in PBS buffer. The membrane was then incubated in a 1:1,000 dilution of a primary antibody in 5% (wt/vol) BSA–PBS-T. After three washes in PBS-T, the membrane was incubated with 1:5,000 dilution of the corresponding secondary antibody in 5% (wt/vol) nonfact skim milk–PBS-T. Respective proteins were visualized by using secondary HRP-conjugated mouse or rabbit IgG antibody (Cell Signaling Technology) and the ECL detection system (Amersham ECL Prime Western blotting Detection Regent; GE Healthcare). For ubiquitination assays, cells were harvested and washed with PBS and then collected in 1 mL of PBS containing 5 mM N-ethylmaleimide (NEM) and centrifuged for 5 min at 800 × g. To dissociate the noncovalent protein interactions, lysates were suspended with 100 μL of 1% SDS and boiled for 10 min. Samples were diluted with 900 μL of lysis buffer [PBS containing 0.5% Triton X-100, 150 mM NaCl, 50 mM Tris⋅HCl (pH 7.4), 10 mM NaF, 1 mM Na₃VO₄, 1 mM EDTA, 1 mM PMSF, 10 mM NEM, and protease inhibitor mixture]. The samples were centrifuged at 16,000 × g for 15 min. To conduct immunoprecipitation analysis, cell lysates were incubated with 1 μg of primary antibody and protein G Plus agarose beads (Santa Cruz; sc-2002) for overnight at 4 °C. The beads were washed three times with lysis buffer, and samples were then suspended with sample buffer, boiled, and separated on 10% (wt/vol) SDS/PAGE gels. Immunoblot analysis was subsequently performed by using the indicated antibodies. Antibodies for HA (sc-805), Myc (sc-40), CYLD (sc-74435), MyD88 (sc-74532), Ubiquitin (sc-8017), Tubulin (sc-69969), and isotype-matched control IgG (sc-2025) were purchased from Santa Cruz Biotechnology. Antibodies for Lys-63 Ub (5621), MyD88 (4283), Rabbit-HRP (7074), and Mouse-HRP (7076) were purchased from Cell Signaling.

Mice and Animal Experiments.

For NTHi-induced inflammation, C57BL/6 WT (7∼8 wk old), MyD88^−/−, CYLD^−/−, or MyD88^−/−CYLD^−/− mice were intratracheally inoculated with NTHi at a concentration of 5 × 10⁷ colony forming units (cfu) per mouse and saline was inoculated as control. The inoculated mice were then killed 1, 3, 6, 9, and 24 h after NTHi inoculation. For PMN analysis, BAL fluid was harvested by cannulating the trachea and performing lavage two times with 1 mL saline in mice followed by staining with Diff-Quik staining kit (modified Giemsa staining). All animal experiments were approved by the Institutional Animal Care and Use Committee at Georgia State University.

Histology and Immunostaining.

For histological analysis, lung tissues were dissected from the mice, and formalin-fixed paraffin-embedded lung tissue was sectioned at 4 μm thick. Lung tissue sections were stained with hematoxylin and eosin (H&E) to visualize lung inflammation. Stained sections were then imaged, and images were then recorded under light microscopy systems (AxioVert 40 CFL, AxioCam MRC, and AxioVision LE Image system; Carl Zeiss). Inflammation score in H&E staining (Grade; 0–3) were validated in a blinded fashion (39). To determine the source of IL-1β in the lungs of mice, lung tissues were stained with rabbit anti-IL-1β antibody (Abcam; ab9722, 10 μg/mL) and rat anti-CD45 antibody (Abcam; ab23910, 5 μg/mL) and signals were probed with donkey anti-rabbit-FITC (Santa Cruz; sc-2090, 5 μg/mL) and goat-anti-rat IgG H&L (Alexa Fluor 647) (Abcam; ab150159, 5 μg/mL). Nuclei was stained with DAPI.

Statistical Analysis.

All experiments were repeated at least three times. Data in bar graphs were shown as mean ± SD. Statistical analysis was assessed by unpaired two-tailed Student’s t test, and P < 0.05 were considered statistically significant.