Curcumin suppresses NTHi-induced CXCL5 expression via inhibition of positive IKKβ pathway and up-regulation of negative MKP-1 pathway

Otitis media (OM) is the most common childhood bacterial infection, and leading cause of conductive hearing loss. Nontypeable Haemophilus influenzae (NTHi) is a major bacterial pathogen for OM. OM characterized by the presence of overactive inflammatory responses is due to the aberrant production of inflammatory mediators including C-X-C motif chemokine ligand 5 (CXCL5). The molecular mechanism underlying induction of CXCL5 by NTHi is unknown. Here we show that NTHi up-regulates CXCL5 expression by activating IKKβ-IκBα and p38 MAPK pathways via NF-κB nuclear translocation-dependent and -independent mechanism in middle ear epithelial cells. Current therapies for OM are ineffective due to the emergence of antibiotic-resistant NTHi strains and risk of side effects with prolonged use of immunosuppressant drugs. In this study, we show that curcumin, derived from Curcuma longa plant, long known for its medicinal properties, inhibited NTHi-induced CXCL5 expression in vitro and in vivo. Curcumin suppressed CXCL5 expression by direct inhibition of IKKβ phosphorylation, and inhibition of p38 MAPK via induction of negative regulator MKP-1. Thus, identification of curcumin as a potential therapeutic for treating OM is of particular translational significance due to the attractiveness of targeting overactive inflammation without significant adverse effects.


NTHi induces CXCL5 expression in middle ear epithelial cells in vitro and in vivo. Epithelial cells
act as the first line of defense against injurious stimuli by mediating inflammatory responses. We sought to determine if NTHi induces CXCL5 expression in human middle ear epithelial cells (HMEECs). NTHi induced CXCL5 mRNA expression in a dose- (Fig. 1a) and time-dependent (Fig. 1b) manner in HMEECs. NTHi induced up-regulation of CXCL5 protein expression in HMEECs as quantified by ELISA (Fig. 1c). Effect of NTHi on CXCL5 expression was also confirmed in human airway epithelial BEAS-2B cells, human lung epithelial A549 cells and human cervical epithelial HeLa cells, suggesting the generalizability of this phenomenon to multiple epithelial cells (Fig. 1d). We further explored the generalizability of NTHi-induced CXCL5 expression by employing two additional commonly used clinical NTHi strains 2627 and 9274 known to cause OM 28,29 . NTHi strains 2627 and 9274 also induced CXCL5 mRNA expression at levels comparable to that of NTHi strain 12 (used throughout the study) in HMEECs (Fig. 1e). Consistent with the in vitro findings, NTHi also induced up-regulation of CXCL5 mRNA in mouse middle ear (Fig. 1f).

TLR2-MyD88-TRAF6-TAK1 signaling axis is required for NTHi-induced CXCL5 expression.
Toll-like receptors (TLRs) are cell surface receptors that play a critical role in mounting early innate immune responses against invading pathogens. TLRs recognize conserved motifs known as pathogen-associated microbial patterns (PAMPs), expressed on microbial pathogens and initiate signaling cascades leading to the production of pro-inflammatory mediators. To date, at least 11 human TLRs have been identified. Among them, TLR2 is known to recognize lipopolysaccharide (LPS), characteristic of Gram-negative bacteria. Based on our previous finding that TLR2 mediated NTHi-induced pro-inflammatory signaling cascades, we sought to determine its role in CXCL5 chemokine production. HMEECs were transfected with TLR dominant-negative mutants TLR2-DN and TLR4-DN. Overexpression of TLR2-DN significantly decreased NTHi-induced CXCL5 mRNA expression, whereas TLR4-DN had no significant effect on CXCL5 expression (Fig. 2a). Next, we sought to determine the signaling molecules downstream of TLR2, involved in mediating CXCL5 expression. Following recognition of NTHi by TLR2, myeloid differentiation factor 88 (MyD88) adaptor protein is recruited to the receptor. MyD88 then recruits and activates IL-1 receptor-associated kinases (IRAKs), which further lead to the recruitment and activation of tumor-necrosis factor-receptor-associated factor 6 (TRAF6). IRAK-TRAF6 complex dissociates from the receptor bound complex and further interacts with transforming growth factor-β -activated kinase 1 (TAK1). Activation of TAK1 leads to initiation of further downstream signaling pathways, resulting in nuclear translocation of transcription factors, and in turn regulates the expression of pro-inflammatory mediators. To determine the involvement of MyD88 and TRAF6 in CXCL5 expression, HMEECs were transfected with dominant-negative mutants MyD88-DN and TRAF6-DN. Overexpression of MyD88-DN and TRAF6-DN significantly suppressed NTHi-induced CXCL5 expression (Fig. 2b). Depletion of endogenous TAK1 with TAK1 siRNA also decreased CXCL5 mRNA expression (Fig. 2c). TAK1 siRNA knockdown efficiency was confirmed by Q-PCR. Therefore, these results suggest that TLR2-MyD88-TRAF6-TAK1 signaling axis is required for NTHi-induced CXCL5 expression.

Activation of IKKβ-IκBα signaling pathway is required for NTHi-induced CXCL5 expression.
Previous studies have shown that IKKβ signaling axis is crucial for NTHi-induced inflammatory responses 30 . Therefore, we examined the role of IKKβ in up-regulation of CXCL5. We first confirmed that NTHi activates IKKβ in HMEECs. IKKα /β phosphorylation was observed at 15 minutes, followed by a peak at 30 minutes, and declined after that (Fig. 3a). To determine if IKKβ is required for NTHi-induced CXCL5 expression, we used multiple approaches. IKKβ inhibitor significantly suppressed NTHi-induced CXCL5 mRNA expression in a dose-dependent manner (Fig. 3b). To identify the major IKK isoform involved in NTHi-induced CXCL5 regulation, HMEECs were transfected with IKK dominant-negative (DN) mutants IKKα -DN and IKKβ -DN. Overexpression of IKKβ -DN significantly decreased NTHi-induced CXCL5 mRNA expression,   whereas IKKα -DN had no significant effect (Fig. 3c). Consistent with this result, depletion of endogenous IKKβ with IKKβ siRNA also decreased CXCL5 mRNA expression (Fig. 3d). IKKβ siRNA knockdown efficiency was confirmed by Western blot. To further confirm that activated IKKβ induces CXCL5 expression, HMEECs were transfected with constitutively active (CA) form IKKβ -CA. Overexpression of IKKβ -CA markedly induced CXCL5 expression in a dose-dependent manner (Fig. 3e). Phosphorylation and proteasomal degradation of Iκ Bα , by IKKβ , are required for signal 31 . Overexpression of a trans-dominant mutant form of Iκ Bα (Iκ Bα -DN) suppressed NTHi-induced CXCL5 expression (Fig. 3f). To further confirm that Iκ Bα degradation is essential for NTHi-induced CXCL5 up-regulation, we used MG-132 proteasome inhibitor. MG-132 significantly suppressed NTHi-induced CXCL5 mRNA expression in a dose-dependent manner (Fig. 3g). Thus, these results suggest that IKKβ -Iκ Bα signaling pathway is required for CXCL5 induction by NTHi.

Activation of p38 signaling is required for NTHi-induced CXCL5 expression. NTHi has been
shown to mediate inflammatory responses via activation of p38 MAPK signaling axis in addition to activation of IKKβ -Iκ Bα pathway 31 . Therefore, we examined the role of p38 in up-regulation of CXCL5. We first confirmed that NTHi activates p38 MAPK in HMEECs. p38 phosphorylation was observed at 15 minutes, followed by a peak at 30 minutes, and declined after that (Fig. 3h). To determine if p38 MAPK activation is essential for NTHi-induced CXCL5 expression, multiple approaches were used. SB203580, a specific inhibitor of p38 activation, significantly suppressed NTHi-induced CXCL5 mRNA expression in a dose-dependent manner (Fig. 3i). To identify the major p38 isoform involved in NTHi-induced CXCL5 regulation, HMEECs were transfected with p38 DN mutants p38α -DN and p38β -DN. Over-expression of either or both p38α , p38β DN forms significantly decreased NTHi-induced CXCL5 expression (Fig. 3j), consistent with SB203580 data. Together, these data suggest that p38 pathway is required for induction of CXCL5.
To determine whether p38 MAPK induces CXCL5 expression via p65, HMEECs transfected with p65 were pre-treated with SB203580, prior to NTHi stimulation. SB203580 decreased CXCL5 expression in p65-transfected cells (Fig. 4g). We further confirmed the requirement of p38 for NF-κ B activation by multiple approaches. Pre-treatment with SB203580 markedly decreased NTHi-induced NF-κ B promoter-driven luciferase activity (Fig. 4h). Consistent with this result, over-expression of either or both p38α and p38β DN forms decreased NTHi-induced NF-κ B promoter activity (Fig. 4i). These results suggest that p38 MAPK also mediates NTHi-induced CXCL5 expression in a p65 dependent mechanism.
Since nuclear translocation of p65 is imperative for NF-κ B-driven gene expression 32 , we determined if p38 up-regulates CXCL5 expression via facilitating nuclear translocation of p65 by immunofluorescence staining. CAPE markedly inhibited nuclear translocation of p65, whereas SB203580 did not show any significant effect (Fig. 4j). Taken together these results suggest that p38 mediates CXCL5 expression via a mechanism independent of the nuclear translocation of p65. Therefore, these results suggest that IKKβ and p38 signaling pathways mediate NTHi-induced CXCL5 up-regulation via activation of p65 in a p65-nuclear translocation-dependent and -independent mechanism, respectively.

Curcumin suppresses NTHi-induced CXCL5 expression in vitro and in vivo.
Having identified the molecular mechanisms underlying NTHi-induced up-regulation of CXCL5, we next sought to explore the translational significance of these findings. Because curcumin, a promising anti-inflammatory agent, has previously been shown to inhibit NF-κ B 33,34 , we first determined if curcumin inhibits NTHi-induced up-regulation of CXCL5. Curcumin pre-treatment inhibited CXCL5 mRNA expression in a dose-dependent manner (Fig. 5a). Curcumin's inhibitory effect on CXCL5 protein levels was also confirmed (Fig. 5b). Additionally, the inhibitory effect of curcumin on CXCL5 mRNA expression was also observed in HMEECs stimulated with other common clinical NTHi strains 2627 and 9274 (Fig. 5c), thereby suggesting the generalizability to more OM-causing NTHi strains. Consistent with in vitro findings, curcumin pre-treatment inhibited CXCL5 mRNA expression in the middle ear of mice inoculated with NTHi (Fig. 5d). These data suggest that curcumin inhibits NTHi-induced CXCL5 expression in middle ear epithelial cells in vitro and in vivo.
Next, we sought to determine the therapeutic relevance of the inhibitory effect of curcumin in NTHi-induced OM model. Thus, we evaluated the effect of administering curcumin post-NTHi infection that resembles a clinically relevant setting. Administering curcumin post-NTHi infection significantly suppressed NTHi-induced CXCL5 mRNA expression in vitro (Fig. 5e) and in vivo (Fig. 5f). Curcumin suppressed CXCL5 expression to the same extent under both pre-NTHi and post-NTHi infection conditions. Since CXCL5 is a neutrophil chemoattractant, we further evaluated the effect of curcumin on PMN recruitment in response to NTHi infection in a mouse model of OM. Consistent with the above results curcumin (pre-NTHi and post-NTHi infection) inhibited PMN infiltration as assessed by PMN staining of middle ear effusion from mice (Fig. 5g). Thus, these data suggest that curcumin is a potential therapeutic for treating NTHi-induced inflammation as seen in OM.
Curcumin suppresses CXCL5 expression via up-regulation of negative regulator MKP-1. MKP-1, a member of a class of dual specificity phosphatases collectively termed MAPK phosphatases, has been shown to be a key negative regulator of inflammatory responses via dephosphorylation and inactivation of MAPKs, including p38 35,36 . Since we identified the requirement of p38 MAPK activation in NTHi-induced CXCL5 expression, we determined the role of MKP-1 in CXCL5 regulation. Overexpression of MKP-1 suppressed NTHi-induced CXCL5 expression (Fig. 7a). Depletion of endogenous MKP-1 with MKP-1 shRNA enhanced CXCL5 mRNA expression (Fig. 7b). MKP-1 shRNA knockdown efficiency was confirmed by Q-PCR. These data suggest that MKP-1 is a negative regulator of NTHi-induced CXCL5 induction.
To further determine if MKP-1 acts as a negative regulator of CXCL5 induction via inactivation of p38, we evaluated the effect of MKP-1 on p38 phosphorylation. Overexpression of MKP-1 reduced NTHi-induced p38 phosphorylation (Fig. 7c). In contrast, depletion of MKP-1 enhanced p38 phosphorylation (Fig. 7d). These data suggest that MKP-1 negatively regulates NTHi-induced CXCL5 expression by targeting p38 MAPK. Having shown that curcumin inhibits NTHi-induced CXCL5 expression via inhibition of p38 and up-regulation of MKP-1 expression, we sought to determine if the inhibitory effect of curcumin on p38 is dependent on the up-regulation of MKP-1. Depletion of MKP-1 with shMKP-1 rendered curcumin treatment ineffective in inhibiting NTHi-induced CXCL5 expression (Fig. 7g). Additionally, curcumin no longer suppressed NTHi-induced p38 phosphorylation in the absence of MKP-1 (Fig. 7h). These data suggest that curcumin inhibits NTHi-induced activation of p38 via up-regulating MKP-1. Thus, our results suggest that curcumin inhibits NTHi-induced CXCL5 expression via MKP-1-dependent inhibition of p38 MAPK.

Discussion
Inflammatory responses are essential for the containment, removal of the invading pathogens and recovery of the host. However, excess inflammation can be detrimental to the host as seen in OM 6-9,37,38 . Therefore, tight regulation of the intensity and duration of inflammatory responses is necessary. In the present study, we show that curcumin inhibits CXCL5 chemokine up-regulation in NTHi-induced OM model, in vitro and in vivo. We found that NTHi up-regulated CXCL5 expression by activating IKKβ -Iκ Bα and p38 MAPK pathways via NF-κ B nuclear translocation-dependent and -independent mechanism. Curcumin not only inhibited the positive IKKβ pathway but also up-regulated the expression of MKP-1, a key negative regulator of p38 MAPK, thereby suppressing CXCL5 expression by dual action. Thus, the current study provides novel insights into the molecular mechanism underlying the tight regulation of neutrophil attractant chemokine CXCL5 in the pathogenesis of NTHi-induced OM and also demonstrates the potential of curcumin as a novel therapeutic for treating OM (Fig. 8).
NF-κ B was found to be the major transcription factor regulating NTHi-induced CXCL5 expression. In our study, both IKKβ -Iκ Bα and p38 MAPK pathways were found to act via p65 subunit to regulate NF-κ B transcriptional activity, albeit through different mechanisms to induce CXCL5 expression. Under resting conditions NF-κ B is present in the cytoplasm bound to Iκ Bα . Upon activation of upstream signaling pathways, NF-κ B dissociates from Iκ Bα and translocates to the nucleus to regulate gene expression. CAPE, an inhibitor of NF-κ B nuclear translocation, inhibited p65 nuclear translocation and suppressed NF-κ B transcriptional activity that in turn suppressed CXCL5 expression. Interestingly p38 MAPK inhibitor SB203580 failed to inhibit p65 nuclear translocation but suppressed NF-κ B transcriptional activity and CXCL5 expression. These findings suggest that p38 MAPK regulates NF-κ B transcriptional activity itself but not p65 nuclear translocation. Previous studies have reported that post-translational modifications such as phosphorylation and acetylation of p65 are critical for promoting its DNA binding and interaction with transcriptional machinery to regulate gene expression. p38 MAPK was found to regulate the acetylation status of p65 but not its phosphorylation. In response to activating stimuli, p38 was found to phosphorylate the transcriptional coactivator p300 (a histone acetyltransferase). Phosphorylated p300 binds to and acetylates K310 residue on p65. Acetylation of K310 was shown to enhance p65 transcriptional activity 39 . Therefore it likely that p38 MAPK mediates NTHi-induced CXCL5 expression in a similar manner by increasing NF-κ B transcriptional activity, independent of p65 nuclear translocation.
OM, a leading cause of conductive hearing loss in children, is caused by NTHi 1 . OM is characterized by the presence of excessive inflammation in the middle ear 3,40 . Current therapies for OM involve the use of analgesics and antipyretics for symptomatic treatment 41 . Though these medications are effective during certain stages of the disease, prolonged usage poses the risk of serious side effects due to unknown "off-targets" and weakened immune system. Decongestants, antihistamines, and corticosteroids have not been effective in treating OM 42 . Prophylactic use of antibiotics has rendered over 80% of the NTHi strains drug-resistant 4,5 . Also, development of vaccines against NTHi remains a challenge due to the high genetic diversity of NTHi strains and high antigenic variability of surface-exposed antigens 43,44 . Thus, there is an urgent need for developing alternative therapeutics for OM with increased efficiency and safety. Therefore, identifying the underlying molecular mechanisms leading to up-regulation of inflammation is critical for the development of novel therapeutic strategies with increased specificity and reduced side effects. Interestingly, in the current study we provide evidence that curcumin inhibits NTHi-induced CXCL5 chemokine expression in middle ear epithelial cells. Interestingly, both pre-infection and post-infection treatment with curcumin not only inhibited NTHi-induced CXCL5 up-regulation but also suppressed PMN infiltration into the middle ear in a mouse model of OM. Curcumin treatment's efficacy in inhibiting CXCL5 expression and PMN recruitment post-NTHi infection is of particular clinical significance. Recently chemokines and chemokine receptors are increasingly considered as targets for developing new drugs to control inflammation 45 . Our finding that curcumin suppresses NTHi-induced CXCL5 chemokine expression is of particular relevance in the current scheme of identifying chemokine-drug combinations to treat inflammation. Thus, curcumin could be repurposed as a new therapeutic for treating OM.
Curcumin is a nutraceutical that has been in use in South Asian countries for many centuries owing to its medicinal properties 20 . Curcumin can interact with a myriad of signaling molecules including transcription factors, protein kinases, growth factors, receptors, adhesion molecules, pro-inflammatory cytokines 46 , thus explaining it pleiotropic therapeutic potential against a wide range of diseases. Curcumin does not present a dose-limiting toxicity, making it suitable for prolonged usage. Completed clinical trials reported usage of curcumin dosage ranging from 0.045 to 8 g/day. Currently, 38 clinical trials evaluating the efficacy of curcumin at a dosage ranging from 0.18 to 8 g/day for treating pathologies such as Alzheimer's disease, diabetes, kidney disease, Crohn's disease, cancer are underway 47 . United States Food and Drug Administration classified curcumin as GRAS (generally recognized as safe), warranting its use as a supplement. In the current study, we identified a novel role of curcumin in suppressing CXCL5 chemokine production. Co-administration of curcumin along with piperine, docetaxel, soy isoflavones, bioperine, lactoferrin, mesalamine in clinical trials [48][49][50][51][52] , suggest the possibility of customizing curcumin-based therapies to maximize its therapeutic efficiency. Since bioavailability of curcumin is a challenge 47 , further studies combining the use of adjuvants, lipids, nanoparticles are needed to elucidate further the potency of curcumin in treating OM.
Another relevant finding of biological significance in the current study is the dual acting mechanism of curcumin in inhibiting CXCL5 expression. Curcumin inhibited NTHi-induced IKKβ phosphorylation, thereby suppressing CXCL5 up-regulation. Moreover, we found that curcumin also inhibits NTHi-induced CXCL5 expression via MKP-1-dependent suppression of p38 MAPK. In the absence of MKP-1, curcumin failed to suppress CXCL5 expression. Due to the importance of p38 in maintaining homeostasis, up-regulation of negative regulator MKP-1 by curcumin could play an important role in controlling the over-active immune responses with minimal side effects. This finding is of particular translational significance due to the attractiveness of targeting overactive inflammation via induction of negative-regulators 19 .
We previously demonstrated that dexamethasone glucocorticoid inhibits p38 MAPK via up-regulation of MKP-1 53 . Glucocorticoids owing to their potent immunosuppressive and anti-inflammatory effects have been in use for treating a gamut of diseases such as asthma, allergies, skin disorders, multiple sclerosis, immune disorders and cancer. However, prolonged usage has been reported to cause severe, sometimes irreversible side effects such as osteoporosis, endocrine and metabolic disorders, behavioral and cognitive changes, gastrointestinal tract complications, uveitis and weakened immune system 54 . Numerous studies over the past years have demonstrated that curcumin's efficacy in resolving pathologies was similar to that of dexamethasone 55,56 . No evidence of side effects with low to moderate consumption of curcumin exists. With a higher curcumin dosage 12 g/day, mild symptoms such as diarrhea, low blood sugar, abdominal pain, and indigestion have been reported 47 . Additionally, curcumin has been reported to aid in overcoming the side effects of glucocorticoid usage 57 . Curcumin is also effective against oxidative stress, characteristic of many inflammatory conditions. Curcumin supplementation could be an effective disease preventive strategy due to its immunomodulatory activity 58 . Thus, curcumin fits the bill for an alternative therapeutic with minimal side effects.
In conclusion, our study demonstrates for the first time that curcumin is a potent inhibitor of CXCL5 chemokine, which could, in turn, suppress inflammation. Further studies promoting curcumin bioavailability may provide means to develop therapies to modulate inflammation more stringently without adverse effects. Development of drug delivery systems in the form of a topical ointment and ear drops could be of clinical significance in treating OM. The findings of this study may have applications in a broader context to other pathologies including chronic obstructive pulmonary disease, tuberculosis, cancer and Alzheimer's disease.
Cell culture. All media described below were supplemented with 10% fetal bovine serum and 100 U/ml penicillin and 100 μ g/ml streptomycin (Gibco). Human middle ear epithelial cells (HMEECs) were maintained in DMEM (Cellgro) supplemented with BEGM SingleQuots (Lonza). Lung epithelial A549 cells were maintained in F-12K medium (Gibco). Human cervical epithelial HeLa cells were maintained in DMEM (Cellgro). Cells were cultured at 37 °C in a humidified 5% CO 2 atmosphere.
Bacterial strains and culture conditions. Clinical isolates of NTHi strains 12, 2627, 9274 were used for this study 28,29 . NTHi was grown on chocolate agar plate in 5% CO 2 atmosphere for 16 h, followed by overnight culture in brain heart infusion (BHI) broth supplemented with 3.5 μ g/ml NAD and 10 μ g/ml hemoglobin (BD Biosciences). Subsequently, bacteria were subcultured in 5 ml fresh BHI broth and the growth was monitored by measurement of optical density (OD). Log phase bacteria were harvested, washed and re-suspended in DMEM for in vitro experiments and isotonic saline for in vivo experiments. For all in vitro experiments the cells were stimulated with NTHi at a multiplicity of infection (MOI) of 50, with an exception for dose-dependent experiment. Cells were stimulated with NTHi for 5 h, or otherwise as indicated. For inhibition study, cells were pretreated with the respective inhibitor for 1 h prior to NTHi stimulation. For post-treatment studies cells were treated with curcumin 1 h after NTHi stimulation.
Enzyme-linked immunosorbent assay (ELISA). Cells were stimulated with NTHi for 12 h. Culture media was harvested and centrifuged at 12,000 × g for 10 min to precipitate cell debris. Culture supernatants were assayed using human ENA78/ CXCL5 ELISA kit (Sigma) according to manufacturer's protocol. OD was measured using Benchmark Plus microplate spectrophotometer. A standard curve showing the relationship between concentration and OD was generated for CXCL5 protein standards. CXCL5 protein concentration in culture supernatants was determined by interpolating from the standard curve.
Western Blot Analysis. Following NTHi stimulation, whole cell extracts were recovered with lysis buffer containing 50 mM Tris-HCl (pH 7.4), 1% Nonidet P-40, 0.25% deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM NaF, supplemented with1 mM PMSF, 1 mM Na 3 VO 4 and protease inhibitor cocktail). Cell extracts were incubated on ice for 30 min and centrifuged at 12,000 × g for 30 min to precipitate cell debris. Supernatants were separated on 10% SDS-PAGE gel, transferred to polyvinylidene difluoride (PVDF) membrane. The membrane was blocked with blocking buffer (TBS containing 0.1% Tween 20 (TBS-T) and 5% nonfat dry milk). After 3 washes with TBS-T, the membrane was incubated overnight with primary antibodies at 1: 1,000-1: 2,000 dilutions in antibody dilution buffer (TBS-T containing 5% BSA) at 4 °C. After 3 washes with TBS-T, the membrane was incubated with corresponding secondary antibody at 1: 5,000 dilution in blocking buffer for 1 h. After 3 washes with TBS-T, the proteins were visualized using Amersham ECL Prime Detection Reagent (GE Healthcare). Images have been cropped for presentation. Full-size images are presented in Supplementary Figs 1 -4. Mice and Animal Experiments. C57BL/6 mice were purchased from Jackson Laboratories. Anesthetized mice were trans-tympanically inoculated with NTHi at a concentration of 5 × 10 7 CFU per mouse. Saline was inoculated as control. For inhibition studies, mice were injected intraperitoneally (i.p) with curcumin (50 mg/kg) 1 h prior or 1 h after NTHi inoculation. For mRNA analysis, mice were sacrificed 6 h post-NTHi inoculation. Total RNA was extracted from the dissected mice middle ear. For PMN analysis, mice were sacrificed 9 h post-NTHi inoculation. Middle ear effusions from mice were harvested with 10 μ l saline (x3). Following cytocentrifugation cells were stained with Diff-Quik stain kit (Siemens) according to manufacturer's protocol. Images were recorded with light microscopy system (AxioVert 40 CFL, AxioCam MRC and AxioVision LE Image system, Carl Zeiss). All animal studies were carried out in accordance with the guidelines of, and were approved by, The Institutional Animal Care and Use Committee at Georgia State University.