TLR9 expression is required for the development of cigarette smoke-induced emphysema in mice

Abstract

The expression of Toll-like receptor (TLR)-9, a pathogen recognition receptor that recognizes unmethylated CpG sequences in microbial DNA molecules, is linked to the pathogenesis of several lung diseases. TLR9 expression and signaling was investigated in animal and cell models of chronic obstructive pulmonary disease (COPD). We observed enhanced TLR9 expression in mouse lungs following exposure to cigarette smoke. Tlr9^−/− mice were resistant to cigarette smoke-induced loss of lung function as determined by mean linear intercept, total lung capacity, lung compliance, and tissue elastance analysis. Tlr9 expression also regulated smoke-mediated immune cell recruitment to the lung; apoptosis; expression of granulocyte-colony stimulating factor (G-CSF), the CXCL5 protein, and matrix metalloproteinase-2 (MMP-2); and protein tyrosine phosphatase 1B (PTP1B) activity in the lung. PTP1B, a phosphatase with anti-inflammatory abilities, was identified as binding to TLR9. In vivo delivery of a TLR9 agonist enhanced TLR9 binding to PTP1B, which inactivated PTP1B. Ptp1b^−/− mice had elevated lung concentrations of G-CSF, CXCL5, and MMP-2, and tissue expression of type-1 interferon following TLR9 agonist administration, compared with wild-type mice. TLR9 responses were further determined in fully differentiated normal human bronchial epithelial (NHBE) cells isolated from nonsmoker, smoker, and COPD donors, and then cultured at air liquid interface. NHBE cells from smokers and patients with COPD expressed more TLR9 and secreted greater levels of G-CSF, IL-6, CXCL5, IL-1β, and MMP-2 upon TLR9 ligand stimulation compared with cells from nonsmoker donors. Although TLR9 combats infection, our results indicate that TLR9 induction can affect lung function by inactivating PTP1B and upregulating expression of proinflammatory cytokines.

in susceptible smokers, lifelong cigarette smoke exposure causes a gradual decline in pulmonary function that leads to chronic obstructive pulmonary disease (COPD) (67). COPD is a complex disease and is currently the third leading cause of death in the United States (46). The disease is characterized microscopically by alterations in airway epithelial cell morphology, peribronchial and perivascular inflammation, and permanent alveolar space enlargement (55). Exposure to cigarette smoke mediates a range of complex immunomodulatory effects, from decreased immune cell activation to depressed phagocytic function (32, 57). Both the innate and adaptive immune responses release inflammatory and protease mediators that contribute to the pathogenesis of COPD (58). Equally, cigarette smoke damages lung residential cells, which heightens inflammation and alters the lung microenvironment. An enhanced understanding of the specific mechanisms by which cigarette smoke alters lung homeostasis and immune responses is needed to prevent the onset and progression of COPD.

Others (14) and we (24) have determined that the expression and functional responses of certain Toll-like receptors (TLR) are altered by cigarette smoke. TLRs recognize a variety of pathogen-associated microbial patterns, but TLR expression and signaling is not restricted to microorganism components, because oxidative stress (22) and neutrophil elastase (26) can trigger a TLR response. Expression of TLRs is linked to lung disease progression, such as idiopathic pulmonary fibrosis (73) and COPD (9, 68). Although several TLRs and MyD88 signaling (14) have been associated with COPD progression, little information is known about whether TLR9 plays a role in COPD progression. TLR9, a pathogen recognition receptor that is primarily triggered by unmethylated CpG sequences in DNA viruses and bacteria, is significantly expressed in multiple cell types throughout the lung, such as the bronchial epithelium, vascular endothelium, alveolar septae cells, and alveolar macrophages (62). Recently, a TLR9 polymorphism [single nucleotide polymorphism (SNP) T1237C] was associated with both lung function decline and dysfunctional antibacterial responses by alveolar macrophages (4). Another TLR9 SNP, T1486C, is frequently observed in patients with COPD (52). Although smoke exposure inhibits TLR3 (71) and TLR7 immune responses (6), TLR9 responses are heightened in T cells (23, 49) and neutrophils (47) from patients with COPD. This is particularly important because DNA viruses such as herpes (79) and Epstein-Barr (EBV) (43) are frequently detected in the lungs of patients with COPD and could contribute to disease exacerbations through modulation of TLR9 signaling.

In view of the potential link between TLR9 and COPD progression, we explored whether cigarette smoke alters TLR9 signaling to exacerbate pulmonary inflammation and impair lung function. Using Tlr9^−/− mice, we determined that TLR9 expression negatively affects cigarette smoke-induced pulmonary physiology. One plausible mechanism for TLR9 expression to alter inflammation is by binding and inactivating protein tyrosine phosphatase 1B (PTP1B), a key anti-inflammatory protein in the lung (18, 25). Smoke exposure desensitizes PTP1B responses, and Ptp1b^−/− mice have heightened airway enlargement in response to smoke (18). Interestingly, TLR9 and PTP1B have opposing effects on granulocyte-colony stimulating factor (G-CSF) levels, immune cell recruitment, apoptosis, and lung physiology (18). Therefore, TLR9 regulation of PTP1B could regulate lung immune responses that alter airway injury and function. TLR9 expression was also significantly increased in normal human bronchial epithelial (NHBE) cells from subjects who smoke and those with COPD compared with cells from nonsmoker donors. Furthermore, this increase in TLR9 expression corresponded with augmented expression of G-CSF, IL-6, CXCL5, IL-1β, matrix metalloproteinase-2 (MMP-2), and MMP-12 following administration of the TLR9 agonist CpG [oligodeoxynucleotide (ODN) 2006]. These findings indicate that chronic cigarette smoke enhances TLR9 signaling to increase inflammation and impair lung function in COPD.

METHODS

Animal models.

Tlr9^−/− (C57BL/6J-Tlr9^M7Btlr/Mmjax) mice, on a C57BL/6J background, were purchased from the Mutant Mouse Resource Center at the Jackson Laboratory (Bar Harbor, ME). All mice were maintained in a specific pathogen-free facility at Mount Sinai Medical Center. Both male and female mice, 8-wk-old, were used at the initiation point for all experiments, and each experimental parameter had at least 10 animals per group. Mice were exposed to cigarette smoke in a chamber (Teague Enterprises, Davis, CA) for 4 h/day, 5 days/wk at a total particulate matter concentration of 80 mg/m³. Smoke exposure was continued for 6 mo. University of Kentucky reference research cigarettes 3R4F (Lexington, KY) were used to generate cigarette smoke. An additional group of wild-type and Ptp1b^−/− (FVB.129S4(B6)-Ptpn1^tm1Bbk/Mmjax) (Mutant Mouse Resource Center at the Jackson Laboratory) mice on an FVB/NJ background were intranasally administered 100 μg of ODN 1826 CpG or GpC (negative control) (both from InvivoGen, San Diego, CA) in endotoxin-free PBS, and euthanized 24 h later. All animal experiments were performed with approval from the Mount Sinai Institutional Animal Care and Use Committee (IACUC). This overall study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and IACUC guidelines.

Forced oscillation and expiratory measurements.

Mice were anesthetized with an injection of ketamine/xylazine hydrochloride solution (100/10 mg/kg ip; Sigma Aldrich, St. Louis, MO). Animals were tracheostomized and connected via an endotracheal cannula to a SCIREQ flexiVent system (SCIREQ, Montreal, QC, Canada). After initiating mechanical ventilation, animals were paralyzed with 1 mg/kg pancuronium bromide (Sigma Aldrich) via ip injection, and several pulmonary function measurements (total lung capacity, lung compliance, and tissue elastance analysis) were determined as previously described (65).

Histology.

Following euthanasia by cervical dislocation, the lungs underwent pressure-fixation and morphometric analysis in accordance with our previously published protocol (17) and in accordance with the American Thoracic Society/European Respiratory Society issue statement on quantitative assessment of lung structure (35). Fixed sections (4 μm) of paraffin-embedded lungs were stained with hematoxylin and eosin. Mean linear intercept analysis was performed as previously described (16). Immunohistochemistry staining was performed for TLR-9 on 4-μm deparaffinized sections of glutaraldehyde-fixed lung tissue using goat anti-TLR9 polyclonal antibodies (sc-13215; Santa Cruz Biotechnology). Donkey anti-goat antibodies were used to detect primary antibodies (Life Technologies, Carlsbad, CA). As negative control, primary antibodies were replaced by normal IgG (Santa Cruz Biotechnology). Visualization of antibody binding by staining with diaminobenzidine (DAB) was performed using an ABC Standard Kit (Vector Laboratories) with DAB/H₂O₂ as substrates following the manufacturer's suggestions. Tissue was counterstained with hematoxylin (Sigma Aldrich).

Lung immune cell measurements.

Bronchoalveolar lavage fluid (BALF) and BALF cells were obtained from animals of each group. Lung cells were analyzed for viability utilizing the LIVE/DEAD cell viability assay from Life Technologies on the Guava EasyCyte flow cytometer (EMD Millipore, Temecula, CA). Apoptotic cells were expressed as a percentage of total lung cells. BALF cells were characterized in neutrophil cell populations by flow cytometry as previously described (12). BALF cells were also cytocentrifuged onto slides to determine macrophage and lymphocyte numbers. Cells were stained with Diff-Quik stain, and at least 200 cells were examined per slide. Lung protein extracts were assayed for myeloperoxidase (MPO) activity using a kit and following the manufacturer's instructions (Cayman Chemical, Ann Arbor, MI).

Cell cultures.

NHBE cells from nonsmokers, smokers, and patients with COPD were isolated from human lungs. Lungs were obtained from organ donors whose lungs were rejected for transplant (see Table 1 for demographics). Consent for research was obtained by the Life Alliance Organ Recovery Agency of the University of Miami. All consents were approved by an institutional review board and conformed to the Declaration of Helsinki. For lungs from patients with disease, the diagnosis of COPD was made by clinical criteria before the death of the patient. All patients with COPD had a significant smoking history and, upon dissection, their lungs had macropathological evidence of emphysema. NHBE cells isolated from lungs of nonsmokers, smokers, and patients with COPD were dedifferentiated through expansion and redifferentiated at an air-liquid interface (ALI) on 24-mm T-clear filters (Costar Corning, Corning, NY) as previously described (50). CpG (1 μM ODN 2006) or GpC dinucleotides (1 μM ODN 2137, negative control) (all from InvivoGen) were added to the apical surface of the cultures and incubated for 2 h at 37°C in 5% CO₂. Subsequently, the apical surface was rinsed five times with PBS, and ALI conditions were restored. Twenty-four hours later, the apical surface was rinsed with 600 μl of PBS, and the rinse was harvested and investigated for cytokine and protease release. The cells were collected for protein and RNA analyses. NHBE cells from nonsmokers were also treated with TLR ligands (tlrl-kit1hw; Human TLR1-9 Agonist kit, InvivoGen). Apical washes were tested for levels of lactate dehydrogenase (LDH) using a commercially available kit (Sigma Aldrich). Fully differentiated NHBE cells from nonsmokers were also exposed to cigarette smoke using a Vitrocell VC-10 smoking robot (Vitrocell Systems, Waldkirch, Germany). Four, eight, or twelve cigarettes were smoked according to ISO standard 3308: six puffs per cigarette with a 35-ml volume per puff and a waiting time between each puff of 60 s. RNA was extracted from the NHBE cells for quantitative PCR (qPCR) analysis.

Table 1.

Demographics of epithelial cell donors

	Nonsmoker	Smoker	COPD
Number	9	6	6
Age, yr	36.3 ± 14.8	36.0 ± 13.2	50 ± 8.3
Gender, male/female	2/7	3/3	3/3
Race, Caucasian/African American	77.8%/22.2%	100%/0%	100%/0%
Pack years	0 ± 0	N/A	51.4 ± 17.2

Values are means ± SD.

COPD, chronic obstructive pulmonary disease; N/A, not available.

Cytokine and protease measurements.

Gene expression was performed by qPCR using TaqMan probes (Life Technologies/Applied Biosystems, Carlsbad, CA). Human and mouse cytokines and MMP-2 were examined in human cell apical washes and media or mouse BALF, respectively, using beads assays (Magnetic Cytokine Bead Panels; Bio-Rad Hercules, CA; and MILLIPLEX MAP MMP Magnetic Bead Panels; EMD Millipore, Billerica, MA) with the Bio-Rad Bio-Plex 200 system. Gelatinase activity was determined by gelatin zymography on apical cell washes (19). BALF neutrophil elastase activity was determined using 50 μM fluorogenic substrate N-(methoxysuccinyl)-Ala-Ala-Pro-Val-7-amino-4-methylcoumarin (Enzo Life Sciences) in 0.1 M Hepes and 0.5 M NaCl, pH 7.5, by excitation at 360 nm and emission at 460 nm. Experiments were performed ± neutrophil elastase inhibitor [N-(methoxysuccinyl)-Ala-Ala-Pro-Val-chloromethyl ketone, 1 mM]. Cathepsin G activity was determined in 50 μl of BALF with a colorimetric cathepsin G activity assay kit (ab126780; Abcam, Cambridge, MA), as described by the manufacturer.

Intracellular signaling.

Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer, centrifuged at 13,000 g for 10 min, and supernatants were collected. Immunoblots were conducted to determine levels of PTP1B (Cell Signaling Technologies, Danvers, MA), TLR9 (Novus Biologicals, Littleton, CO), and β-actin (Cell Signaling Technologies). PTP1B and protein phosphatase 2A (PP2A) activities were determined as previously described (25). PTP1B was immunoprecipitated from lung or cell protein lysates, and immunoblots were performed for TLR9 and PTP1B. Chemiluminescence detection was performed using the Molecular Imager ChemiDoc XRS+ imaging system (Bio-Rad). Densitometry was performed on each target and represented as a ratio of pixel intensity compared with total immunoprecipitated PTP1B, using Image Lab software (version 4.0, build 16; Bio-Rad). Immunofluorescence staining was performed for TLR9 and PTP1B in submerged cultured NHBE cells using goat anti-TLR9 polyclonal antibodies (sc-13215; Santa Cruz Biotechnology) and rabbit anti-PTP1B polyclonal antibodies (5311; Cell Signaling) as primary antibodies. Cells were fixed with acetone following culture on chamber slides (Nunc Lab-Tek, Fisher Scientific). Appropriate secondary controls were used to detect primary antibodies (Alexafluor conjugated; Life Technologies). As negative controls, primary antibodies were replaced with normal IgG from rabbits or goats (Cell Signaling). The nuclei were counterstained with DAPI.

Statistical analyses.

Data are expressed as dot plots with means ± SE highlighted. Differences between two groups were compared by Student's t-test (two-tailed). Experiments with more than two groups were analyzed by two-way ANOVA with Tukey's post hoc test analysis. Significance values were set at P < 0.05 (in figures, * denotes a significant change). All analyses were performed using GraphPad Prism Software (version 6.0h for Mac OS X).

RESULTS

Cigarette smoke exposure enhances TLR9 gene and protein expression in mouse lungs.

To investigate the effect of cigarette smoke on TLR9 expression, C57BL/6J animals were exposed to cigarette smoke daily for 6 mo, and TLR9 expression was determined by qPCR, immunohistochemistry, and immunoblots. Smoke exposure resulted in a significant increase in Tlr9 expression in the lung (Fig. 1A). TLR9 detection by immunohistochemistry demonstrated that TLR9 was expressed throughout the lung but notably in airway cells (Fig. 1B). TLR9 staining intensity was enhanced in animals exposed to cigarette smoke for 6 mo. Western blot analysis confirmed elevated TLR9 protein levels in whole lung homogenates from mice exposed to cigarette smoke for 6 mo (Fig. 1C). These results demonstrate that smoke exposure elevates TLR9 expression levels and that TLR9 is highly expressed in the airways.

Fig. 1.

Smoke exposure enhances Toll-like receptor 9 (TLR9) gene and protein expression in mouse lungs. Enhanced Tlr9 expression is observed in whole lungs of C57BL/6 mice exposed to cigarette smoke for 6 mo. TLR9 was detected in mouse lung tissue following 6 mo exposure to room air and cigarette smoke using quantitative PCR (A), immunohistochemistry (B) (scale bars = 20 μm), and immunoblotting (C). Graphs are represented as means ± SE, with each measurement performed on 3 separate days from at least 6 animals/group. Relative quantification (RQ) was performed compared with room air-treated mice. Immunoblots were quantified by densitometry analysis. Every lane represents an individual mouse. DU, densitometry units. β-Actin immunoblots were used as loading controls. *P < 0.05 when comparing both treatments connected by a line, determined by Student's t-tests.

Tlr9 deficiency prevents smoke-induced loss of lung function in mice.

To determine whether TLR9 expression affected lung function in mice, Tlr9^−/− mice and their wild-type littermates were exposed to cigarette smoke daily for 6 mo. Tlr9^−/− mice did not develop emphysematous changes following exposure to smoke compared with control mice (Fig. 2). Morphometric quantification demonstrated that the loss of Tlr9 expression prevented the increase in mean linear intercept (MLI) (58.1 ± 4 wild-type mice exposed to smoke for 6 mo vs. 50.4 ± 4 for Tlr9^−/− mice exposed to smoke for 6 mo) (Fig. 2A). Forced oscillation and expiratory measurements also determined that loss of Tlr9 expression prevented smoke-induced changes in total lung capacity, lung compliance, and elastance (Fig. 2B). Therefore, Tlr9 expression affects lung function during chronic cigarette smoke exposure.

Fig. 2.

Tlr9 deficiency prevents smoke-induced loss of lung function in mice. Wild-type and Tlr9^−/− mice were exposed to cigarette smoke for 6 mo. A: mean linear intercepts (MLI) (left) were measured in the lungs of the mice to assess air space size and comparative histological images (right) of the four mouse groups are presented (scale bars = 40 μm). B: forced oscillation and expiratory measurements were performed for each mouse to determine total lung capacity, lung compliance, tissue and respiratory system elastance. Graphs are represented as means ± SE, where n = 10 per group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

Loss of Tlr9 expression prevents smoke-induced inflammation in mice.

Several other parameters of COPD were investigated in Tlr9^−/− mice exposed to smoke, such as immune cell infiltration, apoptosis, protease and cytokine production, and phosphatase activity (80). Immune cell infiltration is frequently observed in the lungs of patients with COPD (34). BALF immune cell counts were significantly increased in smoke-exposed mice, which was dependent on Tlr9 expression (Fig. 3A). Neutrophil infiltration and their granule components are implicated in COPD progression (34). Accumulation of neutrophils was observed in BALF following smoke exposure, as defined by flow cytometry for CD11b^highGr-1^high cells (Fig. 3B). Lung neutrophils were reduced in Tlr9^−/− mice following smoke exposure compared with wild-type mice (Fig. 3B). Tlr9 expression also affected smoke-mediated macrophage and lymphocyte lung infiltration (Fig. 3C). BALF levels of MMP-2 were lower in Tlr9^−/− mice (Fig. 3D, left). Increased structural and immune cell apoptosis is observed in COPD lungs (33). Reduced neutrophil counts also coincided with decreased activity of the neutrophil-derived proteases cathepsin G and neutrophil elastase (Fig. 3D, middle and right, respectively). Tissue levels of MPO, one of the principal enzymes released from the azurophilic granules of neutrophils, were diminished in Tlr9^−/− mice (Fig. 3E). Loss of Tlr9 expression was associated with lower numbers of lung cells undergoing apoptosis in cigarette smoke-exposed mice (Fig. 3F).

Fig. 3.

Tlr9 deficiency prevents smoke-induced lung immune cell infiltration, protease release, and apoptosis in mice. Wild-type and Tlr9^−/− mice were exposed to cigarette smoke for 6 mo. A: total bronchoalveolar fluid (BALF) cellularity was determined in each group by flow cytometry. B: a typical representation of neutrophils from wild-type and Tlr9^−/− mice exposed to room air or 6 mo of cigarette smoke, gated as SSC^highCD11b⁺Gr-1⁺. Mean number of neutrophils from whole lung was quantified. C: BALF macrophages and lymphocytes were quantified in each group. D: BALF concentrations of matrix metalloproteinase-2 (MMP-2) (left), and activity of cathepsin G (middle) and neutrophil elastase (right) were determined in each mouse. E: tissue myeloperoxidase (MPO) was quantified in each group. F: lung cells undergoing apoptosis were determined in each group by flow cytometry. Graphs are represented as means ± SE where n = 10 animals per group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

BALF levels of G-CSF and CXCL5 and tissue expression of IFN-β were also lower in Tlr9^−/− mice exposed to smoke compared with wild-type mice (Fig. 4, A and B). Our group previously demonstrated that lung G-CSF expression was regulated by PTP1B (18). Loss of PTP1B enhances lung cell apoptosis and enhances lung remodeling, and chronic cigarette smoke exposure desensitizes the PTP1B anti-inflammation response (18, 25). Because TLR9 and PTP1B have opposing effects on G-CSF levels, immune cell recruitment, apoptosis, and lung function, we investigated whether these regulatory pathways overlap. The loss of Tlr9 expression enhanced lung PTP1B activity in room air and smoke-exposed mice (Fig. 4C). Therefore, TLR9 expression regulates PTP1B phosphatase activity, and this could enhance COPD progression.

Fig. 4.

Tlr9 deficiency prevents smoke-induced lung inflammation in mice. Wild-type and Tlr9^−/− mice were exposed to cigarette smoke for 6 mo. BALF concentrations of granulocyte-colony stimulating factor (G-CSF) and the CXCL5 protein (A) and lung tissue IFN-β gene expression (B) were determined in each mouse. C: enhanced protein tyrosine phosphatase 1B (PTP1B) activity was observed in the lungs of Tlr9^−/− mice compared with wild-type mice, whether exposed to room air or cigarette smoke. Graphs are represented as means ± SE where n = 10 animals per group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

Stimulation of TLR9 signaling blocks PTP1B activity.

To further explore the TLR9 signaling and PTP1B responses, TLR9 activation was triggered in vivo, and interactions between PTP1B and TLR9 were investigated in the lungs (Fig. 5). In vivo lung delivery of the TLR9 agonist (CpG ODN 1826) significantly reduced PTP1B activity in wild-type mice without influencing another prominent phosphatase with anti-inflammatory functions, PP2A (Fig. 5A). Following CpG stimuli, PTP1B bound to TLR9 with a greater affinity than in negative control-treated animals (Fig. 5B). To determine whether TLR9-mediated lung inflammation was dependent on PTP1B expression, Ptp1b^−/− mice were administered ODN 1826 via the nares. BALF cytokine and protease levels were determined. Secretion of G-CSF, CXCL5, and MMP-2, and gene expression of IFN-α and IFN-β were increased in Ptp1b^−/− mice following TLR9 stimulation (Fig. 5, C and D). The loss of PTP1B expression enhances airspace enlargement and inflammation during cigarette smoke exposure (18). Moreover, Ptp1b^−/− mice are more susceptible to viral-induced lung damage (18). Therefore, TLR9 regulation of PTP1B activation can modulate lung inflammation and protease expression.

Fig. 5.

Ptp1b deficiency enhances TLR9 signaling in mice. Methylated (negative control) and unmethylated CpG [oligodeoxynucleotide (ODN) 1826] were intranasally administered, and mice were euthanized 24 h later. A: stimulation of TLR9 inhibits PTP1B activity but not PP2A activity in mouse lungs. B: PTP1B was immunoprecipitated and immunoblots were performed for PTP1B pull-down and total input protein. Every lane represents an individual mouse. Activation of TLR9 enhances TLR9 binding to PTP1B, determined by densitometry analysis of immunoblots. C: intranasal administration of unmethylated CpG to Ptp1b^−/− enhanced BALF concentrations of G-CSF, CXCL5, and MMP-2. D: gene expression of IFN-α and -β genes were determined in each mouse. Graphs are represented as means ± SE, where n = 10 animals per group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

NHBE cells isolated from smokers and patients with COPD express more TLR9 that binds PTP1B with enhanced affinity.

Because TLR9 is highly expressed in the airway epithelium of smoke-exposed mice, we investigated whether this increased expression corresponded with heightened TLR9 signaling in NHBE cells. First, to determine whether cells isolated from the lungs of smokers and individuals with COPD have enhanced TLR9 expression levels, NHBE cells isolated from nonsmokers, smokers, and COPD donors were fully differentiated and cultured at the ALI (Fig. 6A). NHBE cells from smokers and subjects with COPD expressed more TLR9 protein (Fig. 6A) and mRNA (Fig. 6B) than cells from nonsmoker donors. To determine whether TLR9 was the sole TLR to regulate PTP1B activity, ligands for multiple TLRs were administered to NHBE cells from nonsmoker donors, and PTP1B phosphatase activity was determined. Administration of TLR agonists to NHBE cells modulated PTP1B activity without altering total PTP1B protein levels (Fig. 6C). TLR3 (Poly(i:c)) and TLR 6/2 (FSL-1) ligands activated PTP1B; in contrast, TLR9 ligand (CpG ODN 2006) resulted in significantly subdued PTP1B activity in NHBE cells (Fig. 6C). Treatment with CpG ODN 2006 affected PTP1B activity without altering cell viability (Fig. 6D). However, cells from subjects with COPD secreted higher levels of LDH compared with cells from nonsmoker and smoker donors prior to CpG ODN 2006 stimulation (Fig. 6D). Because TLR9 expression and activity influence PTP1B activity, we examined the interaction of these proteins in NHBE cells grown at the ALI. Using immunoprecipitation techniques, we determined that TLR9 binds to PTP1B in NHBE cells with greater affinity in cells isolated from patients with COPD compared with nonsmoker donors (Fig. 6E). TLR9 also trended to bind more to PTP1B in smokers (Fig. 6E). To analyze whether PTP1B colocalizes with TLR9, unstimulated NHBE cells from nonsmokers were stained for TLR9 and PTP1B following submerged culture techniques (Fig. 6F). TLR9 localizes to the endoplasmic reticulum (ER) (40) and endolysosomes (7) in unstimulated cells. Overlapping PTP1B and TLR9 staining occurs mainly in the perinuclear region, suggesting ER localization. However, further studies are required to determine other cellular location of PTP1B and TLR9 binding and how these proteins interact. Overall, these studies show that TLR9 binds and inhibits lung PTP1B tyrosine phosphatase activity.

Fig. 6.

TLR9 binds PTP1B and blocks PTP1B phosphatase activity in human airway epithelial cells. A: schematic of fully differentiated NHBE cells grown at the air-liquid interface (ALI) from nonsmoker, smoker, and donors with chronic obstructive pulmonary disease (COPD) (n = 6 donors per group). Enhanced TLR9 protein (A) and gene expression (B) was observed in NHBE cells from donors with COPD compared with nonsmoker donors, determined by qPCR and immunoblots. C: NHBE cells grown at the ALI from nonsmoker individuals were treated with various TLR ligands for 2 h and washed with PBS. PTP1B activity was recorded in cells 24 h later. PTP1B and β-actin immunoblots show no significant PTP1B protein changes. D: cells from donors with COPD released more lactate dehydrogenase (LDH) onto the apical surface than cells from nonsmoker subjects. Treatment with 1 μM CpG (ODN 2006) or GpG (negative control) did not enhance LDH release in either group. E: PTP1B was immunoprecipitated from normal and COPD cells, and TLR9 and PTP1B immunoblots were performed. NHBE cells from subjects with COPD had enhanced TLR9 binding to PTP1B compared with cells from nonsmoker donors, determined by densitometry analysis of immunoblots. Every lane represents an individual donor cell sample. Graphs are represented as means ± SE where n = 6 donors per group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test. F: immunofluorescence for PTP1B and TLR9 in NHBE cells demonstrates colocalization of both proteins in cells.

NHBE cells from smokers and donors with COPD have enhanced cytokine and protease responses to TLR9 ligand stimulation.

Many studies address the immune responses of infiltrating inflammatory cells following exposure to cigarette smoke or the COPD disease state (3, 23, 45); however, NHBE cell responses from smokers and patients with COPD are rarely investigated. To explore the effect of heightened TLR9 expression in NHBE cells, we exposed the apical surface of cells to CpG (ODN 2006) or negative control for 2 h, after which the cells were returned to ALI conditions for 24 h. CpG stimulation significantly enhanced CXCL5 and G-CSF gene expression in cells from nonsmoker donors (Fig. 7A). NHBE cells from smokers significantly responded to CpG stimulation with increased gene expression of G-CSF, IL-6, and CXCL5 (Fig. 7A). Cells from patients with COPD have higher gene expression of G-CSF, IL-6, CXCL5, and IL-1β compared with cells from nonsmokers prior to stimulation (Fig. 7A). Following CpG stimulation, NHBE cells from patients with COPD exhibited significantly higher G-CSF, IL-6, CXCL5, and IL-1β expression (Fig. 7A). Elevated secretion of IL-6, CXCL5, and IL-1β was observed on the apical surface of NHBE cells from both smokers and patients with COPD compared with nonsmoker donors following CpG stimulation (Fig. 7B). Following CpG administration, COPD cells also had enhanced G-CSF secretion compared with both other cell groups (Fig. 7B). NHBE cells grown at the ALI produce very little type 1 interferon both at baseline and following CpG stimuli (37). Type 1 interferons were not detectable in NHBE cells from any donor group both before and after CpG stimuli (data not shown).

Fig. 7.

NHBE cells from patients with COPD have enhanced cytokine responses to TLR9 ligand stimulation. Fully differentiated NHBE cells grown at the ALI from smokers, nonsmokers, and individuals with COPD (n = 6 donors per group) were treated with CpG (ODN 2006) or GpG (negative control) for 2 h and washed with PBS. Apical surface washes were taken 24 h later. IL-1β, CXCL5, IL-6, and G-CSF gene expression (A) and protein levels (B) in cells or apical washes were determined in NHBE cells from nonsmoker and COPD donors. Graphs represent means ± SE, where each measurement was performed on 3 independent days on 4 donors/group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

MMP expression was investigated under the same conditions. CpG challenge had no effect on MMP-2 and -12 levels in cells from nonsmoker donors. However, NHBE cells from patients with COPD had elevated expressions of MMP-2 and -12, which was further enhanced by CpG (Fig. 8A). NHBE cells from smokers had comparable levels at baseline but MMP-2 and -12 expressions were significantly enhanced in this cohort following CpG treatment (Fig. 8A). Multiplex and zymography analysis demonstrated elevated MMP-2 concentrations and activity on the apical surface of cells from smokers and patients with COPD compared with nonsmoker donors following CpG treatment (Fig. 8, B and C). Secreted MMP-12 protein was undetectable from NHBE cells (data not shown). Therefore, smoke exposure and the COPD disease state predispose NHBE cells to enhanced responses to TLR9 stimuli.

Fig. 8.

NHBE cells from patients with COPD have enhanced protease responses to TLR9 ligand stimulation. Fully differentiated NHBE cells grown at the ALI from smokers, nonsmokers, and individuals with COPD (n = 6 donors per group) were treated with CpG (ODN 2006) or GpG (negative control) for 2 h and washed with PBS. Apical surface washes were taken 24 h later. A: gene expression of MMP-2 (left) and MMP-12 (right) were determined in NHBE cells from nonsmoker and COPD donors. B: MMP-2 levels were detected in apical washes and basal media from each cell group by multiplex analysis and gelatin zymography (C) was performed on the apical washes. Every lane represents an individual donor apical wash sample. MMP-12 protein was undetectable (not shown). Graphs are represented as means ± SE, where each measurement was performed on 3 independent days on 4 donors/group. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

Smoke exposure enhances TLR9 expression in NHBE cells.

NHBE cells isolated from nonsmokers were exposed to smoke using a Vitrocell VC-10 smoking robot, generated from 0, 4, 8, or 12 cigarettes to determine whether smoke alone alters TLR9 expression in NHBE cells. Expressions of TLR9, IL-1β, IL-6, CXCL5, G-CSF, and MMP-2 were determined by qPCR. Smoke enhanced TLR9, IL-1β, IL-6, CXCL5, and G-CSF expression in NHBE cells and induced a trend induction in MMP-2 (Fig. 9). Therefore, NHBE cells are a major source of cytokines and proteases, and these cells have enhanced TLR9 responses following smoke exposure.

Fig. 9.

Smoke exposure enhances TLR9-associated signaling in NHBE cells. Fully differentiated NHBE cells grown at the ALI from nonsmoker individuals (n = 6 donors) were exposed to smoke using a Vitrocell VC-10 smoking robot, generated from 0, 4, 8, or 12 cigarettes, to determine whether smoke alone alters TLR9 signaling in NHBE cells. Expressions of TLR9 (A); IL-1β, IL-6 CXCL5, and G-CSF (B); and MMP-2 (C) were determined by qPCR. Graphs represent means ± SE. *P < 0.05 when comparing both treatments connected by a line, determined by two-way ANOVA with Tukey's post hoc test.

DISCUSSION

In this study we establish that cigarette smoke enhances TLR9 expression, which contributes to COPD through a mechanism that may involve inhibition of PTP1B activity. Tlr9^−/− mice were resistant to cigarette smoke-induced loss of lung function. These TLR9-dependent changes coincided with smoke-induced expression of CXCL5, G-CSF, MMP-2, and IFN-β. In fact, we have identified that NHBE cells from patients with COPD have heightened G-CSF, IL-6, CXCL5, IL-1β, MMP-2, and MMP-12 responses to TLR9 stimulation. Additionally, the binding of TLR9 to PTP1B facilitates regulation of PTP1B activity and represents a novel posttranslational modification that alters immune responses in the COPD lung. Therefore, we propose that TLR9 promotes the loss of lung function in COPD by negating PTP1B responses that counter pulmonary inflammatory responses (Fig. 10).

Fig. 10.

Possible signaling mechanism for TLR9 regulation of PTP1B. Evidence presented in this study indicates that under room air conditions, PTP1B prevents TLR-mediated inflammation (A), but following smoke exposure TLR9 expression is enhanced, and TLR9 binds to PTP1B to prevent the activity of the phosphatase (B). This regulation of PTP1B alters inflammatory responses within the lung.

TLR9 signaling is triggered primarily by unmethylated CpG sequences in DNA viruses and bacteria. There is evidence that unmethylated CpG-rich viruses and bacteria are present in the lungs of patients with COPD. In a recent study, EBV was detected in 48% of patients with COPD undergoing an exacerbation (43). Interestingly, smokers who do not develop COPD rarely have EBV in their sputum (43). Our data suggest that the airway epithelium in COPD has heightened TLR9 responses that induce sensitivity to microbial CpG than in nonsmoker subjects. Higher TLR9 expression (73) and the presence of EBV (38) have also frequently been found in the lungs of patients with pulmonary fibrosis. This enhanced EBV presence was primarily in the airway epithelial cells (70), a site where greater TLR9 responses are observed. Another DNA virus, herpes virus, is also detected more frequently in patients with COPD than in nonsmoker controls (79), and could influence TLR9 signaling on COPD exacerbations. Equally, many bacterial strains rich in unmethylated CpGs, such as Haemophilus influenza (15) and Chlamydia pneumoniae (11), have been detected in patients with COPD. The heightened TLR9 response in the airway mucosa could account for the marked infiltration of cytotoxic CD8+ T-lymphocytes in lungs of individuals with COPD (60). Other researchers have suggested that cigarette smoke induces lung injury via nucleic acid-sensing TLR signaling that is triggered by endogenous nucleic acids (76). Unmethylated CpG motifs are prevalent in bacterial and viral genomes but rarely in vertebrate genomic DNA (74). Vertebrate genomic DNA could plausibly affect TLR9 signaling because mice engineered to express TLR9 on the cell surface experience severe systemic inflammation (48). Whether the trigger for TLR9 is due to endogenous nucleic acids, cell death, microbial influences, or an unknown stimuli has yet to be fully determined in COPD, but the effect of enhanced TLR9 signaling on lung function is evident, independent of the source of the CpG. The change in TLR9 expression may be partially due to epigenetic effects produced by changes in DNA methylation and histone modification following smoke exposure and disease initiation. Additionally, miR7 was recently observed to regulate TLR9 signaling in human lung cancer cells (78). PTP1B expression is also susceptible to micro RNA (miRNA) regulation, with miR-210, miR-122, and miR-744 regulating PTP1B expression (81). Future studies will focus both on the regulation of TLR9 expression and key downstream effectors of TLR9 signaling that affect COPD progression.

One of the TLR9 targets identified in this study, PTP1B, is ubiquitously expressed in eukaryotic cells and primarily localizes to the cytoplasmic face of the ER (20), where it can be relocated to the cytoplasm by a calpain-mediated proteolytic cleavage (21). Our immunofluorescence data suggest that PTP1B and TLR9 interactions may occurs in the perinuclear region, suggesting ER localization. Functionally, PTP1B is mainly known to negatively regulate insulin receptor signaling (8), but PTP1B also regulates several other receptors, including TLRs (44, 77). We previously determined that enhancing PTP1B expression in cigarette smoke conditions could resolve inflammation and protease production in mice (25). Once activated, PTP1B acts as a common negative regulator of numerous inflammatory-associated signaling, such as nuclear factor-κB, AKT, and MAPK activities (72), and PTP1B activity can also protect mice against irradiation injury (72). Enhancing PTP1B levels in patients may not be the best means of treatment because PTP1B negatively regulates insulin signaling (63), and enhancing PTP1B could have several secondary systemic effects in the body. However, targeting specific PTP1B-dependent signaling responses may be a feasible approach and will be a major area for future investigation. We have recently identified that PTP1B negatively regulates S100A9, a damage-associated molecular pattern (DAMP) protein (18). S100A9 levels are elevated in the serum of patients with COPD during an exacerbation (56), and smoke exposure resulted in enhanced S100A9 responses to viral infection (18). TLR9 regulation of similar proteins, via negative regulation of PTP1B, may identify future therapies for treating the acute inflammation observed in the COPD lung.

TLR9 has various roles in airway diseases. In asthma, synthetic TLR9 agonists are beneficial in various rodent and primate models of asthma (31) and have led to the initiation of human clinical trials with conflicting outcomes (1, 5). TLR9 agonists help to protect against seasonal influenza virus infection in mice but at a cost of increased destructive tissue inflammation (75). These enhanced TLR9 responses might be the cause of heightened lung injury in the smoke-exposed animals observed in this study. In contrast to its protective role in influenza, Tlr9 deficiency was recently described to deter Pseudomonas aeruginosa lung infection, with alveolar macrophages from Tlr9^−/− mice clearing bacteria faster than in their wild-type littermates (2). In this study, G-CSF, IL-6, CXCL5, IL-1β, IFN-β, MMP-2, and MMP-12 were all heightened in a TLR9-dependent manner. TLR9 also regulated immune cell infiltration into the lung during smoke exposure, with neutrophils, macrophages, and lymphocytes significantly reduced in the lungs of Tlr9^−/− mice. Reduced neutrophil counts coincided with lower neutrophil elastase and cathepsin G activity and lower MPO levels in the Tlr9^−/− mice following smoke exposure. Neutrophil elastase is elevated in the sputum of smokers and correlates with the severity of airflow obstruction (53). Inhibition of cathepsin G is also an effective means of preventing pulmonary inflammation (42). Mice with knockout of neutrophil serine proteases (proteinase 3, cathepsin G, and neutrophil elastase) are substantially protected against lung tissue destruction after long-term exposure to cigarette smoke (29). Therefore, prevention of TLR9-mediated neutrophil recruitment could reduce lung damage. CXCL5 promotes recruitment of neutrophils to the lung (27) and G-CSF enhances survival of neutrophils (10), which could enhance lung inflammation and damage. TLR9-regulated expression of CXCL5 and G-CSF could contribute to neutrophil recruitment into the lung during smoke exposure. Increased levels of both IL-6 (13) and IL-1β (66) are observed in clinical samples from patients with COPD, and both correlate with disease severity. Ptp1b^−/− mice have elevated IFN-α and -β expression following CpG stimuli compared with wild-type mice. Recently, miR744 has been shown to regulate type 1 IFN signaling via inhibition of PTP1B expression in human renal mesangial cells (81). Therefore, the signals of PTP1B and type 1 IFNs overlap. Tlr9^−/− mice exposed to smoke had reduced IFN-β expression compared with wild-type mice. Therefore, TLR9 and PTP1B interactions mediate key immune responses that change the lung microenvironment.

The role of MMP-2 in COPD is uncertain, with investigators observing heightened expression in alveolar macrophages and airway epithelial cells (64), whereas others found that MMP-2 gene expression levels decreased with increasing stage according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria (28). However, MMP-2 is increased in the lungs of patients with COPD during an exacerbation (54), and elevated MMP-2 could contribute to lung extracellular matrix remodeling. Others have also observed a correlation between TLR9 expression and MMP-2 (41, 59). Expression of MMP-12 is critical for the development of smoke-induced emphysema in mice (30). Hypoxia regulation of TLR9 may play a major role in MMP expression, because TLR9 expression coincides with MMP profiles in hypoxic conditions (61). In fact, hypoxia-sensitive HIF-1α heightens TLR-mediated inflammation in rheumatoid arthritis (36). Regulation of PTP1B expression by miR210 is also mediated by hypoxia-mediated expression of HIF-1α (51). Therefore, TLR9 regulation of G-CSF, IL-6, CXCL5, IL-1β, MMP-2, and MMP-12 could have a major effect on COPD progression and may be driven by hypoxic conditions.

TLR9 and other nucleotide-sensing TLRs such as TLRs 3, 7, 8, and 13, are largely found inside cells, and vast amounts of extracellular host nucleic acids could possibly trigger signaling of these TLRs. Intracellular receptors could easily interact with intracellular proteins, such as the PTP1B and TLR9 interaction, as observed here. These nucleotide-sensing TLRs require UNC93B1 to leave the ER (39). A mutation in UNC93B1 results in signaling defects of TLRs 3, 7, and 9 (69), and prevents the trafficking of TLRs 7 and 9 to the endolysosomes. Mice with a mutation in the Unc93b1 gene are protected against smoke-induced airway remodeling (76). Whether smoke exposure affects UNC93B1 expression to modulate TLR9 signaling remains to be determined. Indeed, the details of the interactions between UNC93B1 and TLRs remain poorly understood, and the role of PTP1B binding to these molecules has yet to be reported. Given our findings, UNC93B1 interaction with TLR9 and possibly PTP1B in COPD lungs merits further investigation.

Together, our data identify the interaction of PTP1B and TLR9, which represents a novel posttranslational modification that occurs following modulation of TLR9 signaling and coincides with enhanced inflammation. Indeed, our work highlights the potential detrimental side effects for the long-term use of TLR9 ligands on lung integrity.

GRANTS

This work was supported by Flight Attendant Medical Research Institute Grant YCSA 113380 to P. Geraghty, National Heart, Lung, and Blood Institute Grant 5R01 HL-098528-05 and Flight Attendant Medical Research Institute CIA Grant 130020 to R.F. Foronjy, and Flight Attendant Medical Research Institute CIA Grant 103027 to M.A. Salathe.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

R.F., M.S., E.E., and P.G. conception and design of research; M.S., A.J.D., N.B., N.C., and P.G. performed experiments; R.F., M.S., and P.G. analyzed data; R.F., M.S., E.E., and P.G. interpreted results of experiments; P.G. prepared figures; P.G. drafted manuscript; R.F., M.S., A.J.D., E.E., and P.G. edited and revised manuscript; R.F., M.S., A.J.D., N.B., N.C., E.E., and P.G. approved final version of manuscript.