Toll-like receptor 4 is not targeted to the lysosome in cystic fibrosis airway epithelial cells

Catriona Kelly, Paul Canning, Paul J. Buchanan, Mark T. Williams, Vanessa Brown, Dieter C. Gruenert, J. Stuart Elborn, Madeleine Ennis, Bettina C. Schock

Abstract

The innate immune response to bacterial infection is mediated through Toll-like receptors (TLRs), which trigger tightly regulated signaling cascades through transcription factors including NF-κB. LPS activation of TLR4 triggers internalization of the receptor-ligand complex which is directed toward lysosomal degradation or endocytic recycling. Cystic fibrosis (CF) patients display a robust and uncontrolled inflammatory response to bacterial infection, suggesting a defect in regulation. This study examined the intracellular trafficking of TLR4 in CF and non-CF airway epithelial cells following stimulation with LPS. We employed cells lines [16hBE14o-, CFBE41o- (CF), and CFTR-complemented CFBE41o-] and confirmed selected experiments in primary nasal epithelial cells from non-CF controls and CF patients (F508del homozygous). In control cells, TLR4 expression (surface and cytoplasmic) was reduced after LPS stimulation but remained unchanged in CF cells and was accompanied by a heightened inflammatory response 24 h after stimulation. All cells expressed markers of the early (EEA1) and late (Rab7b) endosomes at basal levels. However, only CF cells displayed persistent expression of Rab7b following LPS stimulation. Rab7 variants may directly internalize bacteria to the Golgi for recycling or to the lysosome for degradation. TLR4 colocalized with the lysosomal marker LAMP1 in 16 hBE14o- cells, suggesting that TLR4 is targeted for lysosomal degradation in these cells. However, this colocalization was not observed in CFBE41o- cells, where persistent expression of Rab7 and release of proinflammatory cytokines was detected. Consistent with the apparent inability of CF cells to target TLR4 toward the lysosome for degradation, we observed persistent surface and cytoplasmic expression of this pathogen recognition receptor. This defect may account for the prolonged cycle of chronic inflammation associated with CF.

  • inflammation
  • TLR4
  • endosome
  • Golgi
  • Rab GTPases

cystic fibrosis (CF) is a multiorgan disease with obstructive problems in the respiratory, digestive, and reproductive tracts. Inappropriate epithelial function and inflammatory immune responses in the airways play a key role in the development of respiratory failure, the most common cause of mortality in CF patients (34). In CF, mutations in the cystic fibrosis transmembrane conductance regulator gene (CFTR) decrease the ion transport across the pulmonary epithelium and lead to mucus dehydration and a subsequent reduction in muciliary clearance. Infection with gram-negative Pseudomonas aeruginosa is an important cause of pulmonary inflammation in CF lung disease and is associated with an accelerated decline in lung function, morbidity, and mortality. The innate inflammatory response to such infection is mediated through Toll-like receptors (TLRs), which trigger tightly regulated signaling cascades through transcription factors including nuclear factor-κB (NF-κB) (37).

Intracellular membrane trafficking and routing of the receptor-ligand complex is essential to achieve a controlled inflammatory response and a timely and regulated termination of inflammatory signaling. Infection with P. aeruginosa and subsequent detection of bacterial wall components such as lipopolysaccharides (LPS) induces TLR4 activation via dimerization within surface membrane lipid rafts (20, 31, 33, 40, 42, 44). This triggers the activation of MyD88-dependent and independent signal transduction pathways (39, 40). The TLR4-LPS complex is then rapidly internalized and TLR-induced inflammatory signaling is stopped by targeting the complex for degradation (22, 40, 41). Endosomes play a vital role in this process, serving as sorting facilities for the biosynthetic and endocytic pathways. Acidification of the endosome leads to dissociation of the TLR dimer and its association with LPS, thereby initiating the first step in shutting off proinflammatory signaling (29). The TLR4-LPS complex may be directed from the early endosomes toward endosomal-lysosomal degradation or toward the Golgi-apparatus for recycling. Several chaperons and adaptor molecules play an important role in directing the internalized complex into the appropriate compartment (38). Rab proteins are members of the Rho family of GTPases (a subgroup of the Ras superfamily of GTPases), which regulate endocytic sorting. Rab7 is associated with late endosomes and traffics to the lysosome. However, an alternative isoform of Rab7, Rab7b has been implicated in trafficking toward the Golgi despite sharing significant homology with Rab7 (6). Two further members of the Rho GTPases, Rab10 and Rab11a, have been shown to interact with TLR4 and to be important in the trafficking of internalized TLR4 to the Golgi for recycling (Rab10) (43) or to bacteria containing phagosomes (Rab11a) (21). Functional CFTR alters the internal compartmentalization of TLR4 in peripheral blood macrophages (5). Here we sought to investigate the potential role of these GTPases in directing TLR4 compartmentalization in the CF lung.

CF patients display signs of ongoing and uncontrolled inflammation (30), suggesting that inflammatory regulation is defective. Failure to regulate inflammatory signaling contributes to the cycle of chronic infection and inflammation and may further enhance the decline in lung function. Prolonged intracellular signaling has been shown for epidermal growth factor receptor (EGFR), when the receptor failed to enter the endosome but remained in a perinuclear compartment (28). Although the mechanism by which intracellular signaling is initiated is independent of the subsequent trafficking route (40, 41), inappropriate trafficking may prolong inflammatory signaling. Additionally, several studies have shown altered TLR4 expression in CF airways cells, contributing to an increased inflammatory response (17, 19, 26). Our study aimed to investigate the role of endolysosomal degradation in prolonged TLR4-dependent NF-κB signaling in CF epithelium. Here, we report deviations from the normal intracellular trafficking route of TLR4 in LPS-stimulated CF epithelial cells.

MATERIALS AND METHODS

Cell culture and stimulations.

The bronchial epithelial cell lines 16hBE14o-, CFBE41o- (homozygous for the F508del mutation), and CFBE41o- comp [CFBE41o- cells complemented with wild-type (wt) CFTR] were obtained from Dr. D. Gruenert (UCSF). Cells were maintained in flasks coated with 1% PureCol (Advanced Biomatrix) in water, and grown in MEM containing glutamine (PAA Laboratories) at 37°C and in 5% CO2 in air. Culture medium was supplemented with 10% heat-inactivated fetal bovine serum (FBS), and 100 U/ml penicillin and 100 μg/ml streptomycin, and cells were routinely passaged in 10% trypsin (all from PAA). Care was taken to passage cells when 70–80% confluent and all stimulations and experiments were conducted at this density. Cells from passages 8–14 were used for these experiments.

16hBE14o- cells are SV40 large T-antigen immortalized bronchial epithelial cells that have been described as being derived from a 1-yr-old male heart-lung transplant patient (11). This cell line was first described about 30 years ago (9). However, subsequent analysis has indicated that 16hBE14o- cells might be female, although a loss of all or part of the Y chromosome could also be the reason for a female designation (D. Gruenert, personal communication). 16hBE14o- have a cobblestone appearance in culture, retain the differentiated epithelial morphology and functions of the epithelial cells, express cilia (in air liquid interface cultures) and tight junctions, and display vectoral ion transport. They express CFTR protein in both mature and immature forms in the cytoplasm and in the mature form only in the membrane (13).

CFBE41o- cells are also SV40 immortalized bronchial epithelial cells (1st bifurcation). They derived from a female CF patient homozygous for the F508del mutation (D. Gruenert, personal communication) (4, 23). CFBE41o- cells appear morphologically larger than 16hBE14o- cells. They express functional tight junctions in immersed cultures (16). Forskolin has no effect on Cl ion transport in CFBE41o- cells (12, 18). Cl transport in complemented CF bronchial epithelial cells correlates with CFTR mRNA expression levels (23), which is the same for cultured CF primary epithelial cells (16, 23).

CFBE41o- cells have been complemented with wt CFTR cDNA. These cell models have been used as isogenically matched controls for the parental F508del/F508del CFBE41o- cell line to study the CF inflammatory response (10). They appear to express the transgene CFTR and have restored chloride transport (cAMP dependent) (26).

Primary nasal epithelial cells (NECs) were obtained from CF patients homozygous for F508del and age- and sex-matched control volunteers as previously described (25, 26). The study was approved by the Research Ethics Committee of Northern Ireland and all participants provided informed consent (Ethics Number 07/NIR02/23). NECs were grown in monolayers and maintained in supplemented Airway Epithelial Growth Medium (AEGM, PromoCell, Houston, TX) until cells were 70–80% confluent. Cells were passaged twice with trypsin and seeded onto 12-mm Transwell inserts, which were coated with a 1:10 solution of PureCol-sterile H2O, air-dried and exposed to UV light to allow type I collagen fibers to cross-link prior to use. Cells were maintained in air-liquid interface (ALI) medium (AEGM with extensive supplements, PromoCell) and fed basolaterally and apically until fully confluent (typically 5 days). The formation of tight junctions was confirmed by measuring transepithelial electrical resistance values with an EVOM meter (World Precision Instruments). Once confluent, medium was removed from the apical surface and cultures maintained at ALI for 21 days prior to use in experiments. All cells were treated with P. aeruginosa-derived lipopolysaccharide (LPS; Sigma-Aldrich). Primary cells and cell lines were treated with 50 μg/ml LPS for 0 to 24 h as indicated in the figures. A range of LPS concentrations was assessed and 50 μg/ml was selected as the concentration that produced a robust inflammatory response (p65 expression and/or IL-8 release), while not inducing in significant change in cellular viability (as determined by MTT and LDH assays; data not shown).

Surface and intracellular TLR4.

Surface and intracellular protein expression of TLR4 was quantified by flow cytometry in all cell lines. Cells (2.5 × 105 cells) were detached with enzyme-free cell dissociation buffer (PBS based; Sigma-Aldrich) and washed twice with PBS. For intracellular/cytoplasmic expression cells were additionally fixed with 4% paraformaldehyde (Sigma-Aldrich), washed with PBS and permeabilized for 15 min in 0.1% saponin (with 0.5% BSA) at room temperature. For cell surface expression, washed cells were directly labeled with antibodies. The cells were incubated with a primary FITC-conjugated antibody against TLR4 (Imogen) diluted in 0.5% BSA or in 0.5% BSA containing 0.1% saponin for 30 min. Data acquisition and analyses were performed immediately on a flow cytometer (Epics XL, Beckman Coulter). For each sample, cells were gated in the forward angle and side scatter to exclude dead cells or aggregates and a minimum of 10,000 events were collected. The threshold of TLR4-positive cells (expressed as % positive) was set using cells stained with isotype-matched, directly labeled nonspecific antibodies in identical concentrations and labeled with the same fluorochrome (BD Pharmingen). Data are expressed as mean fluorescence intensity (MFI) to evaluate the shift in fluorescence intensity of TLR4-positive cells compared with isotype controls.

Determination of IL-8 and IL-10 concentrations.

The concentration of IL-8 and IL-10 in cell line supernatants was measured by commercially available ELISA (PeproTech EC) according to the manufacturer's instructions.

Quantification of NF-κB expression.

The percentage of epithelial cells positive for the nuclear p65 subunit of NF-κB was determined by an adapted flow cytometric method (35). Nuclei were isolated using the CycleTest PLUS DNA Reagent Kit (BD Biosciences) and stained with anti-human p65 antibody (Santa Cruz Biotechnology) or isotype-matched control (SouthernBiotech), followed by anti-rabbit FITC conjugate (Sigma-Aldrich) and propidium iodide (PI) (14). Localization of nuclei (forward scatter/side scatter plot) was determined by use of previously separated epithelial cell nuclei. Acquisition of stained nuclei (minimum 5,000 events) was carried out on an Epics XL flow cytometer. Single nuclei were gated on the basis of PI staining (FL-3, at 630 nm), after doublet elimination by FL-3 peak vs. integral. Positive p65 expression (in FL-1 plot) was expressed as % of epithelial cell nuclei. MFI was subsequently calculated to evaluate the shift in fluorescence intensity of nuclear p65-positive cells compared with isotype controls.

Quantitative PCR.

Total RNA was extracted from cells by using an RNeasy Micro kit (Qiagen) according to the manufacturer's protocol. RNA samples were quantified by using a NanoDrop spectrophotometer (Thermo Scientific) and equal amounts of RNA (1 μg for cell lines and 500 ng for primary cells) reverse transcribed into cDNA by use of a Sensiscript Reverse Transcription Kit (Qiagen). Only RNA with 260/280 and 260/230 ratios between 1.8 and 2.2 was used for these experiments. Primers were designed by using gene accession numbers and Primer3 open-source PCR primer design software and obtained from Invitrogen (Paisley, UK). Quantitative PCR was performed with a LightCycler rapid thermal cycler system (Roche Diagnostics). Multiple housekeeping genes were tested and β-actin was chosen for consistency within cycles and between different samples. An internal calibrator (Jurkat cells) was used to monitor consistency and reaction efficiency between cycles. Primer sequences, gene accession numbers, and product sizes are given in Table 1. Relative expression to β-actin was calculated by the ΔΔCt method, and Jurkat cells served as an internal calibrator for all experiments.

View this table:
Table 1.

Primer sequences for quantitative PCR

Western immunoblotting.

Protein was extracted with RIPA buffer and quantified via a BCA protein assay (Thermo Scientific). Cell lysates were diluted in nuclease-free water prior to adding Laemmli loading buffer containing 5% β-mercaptoethanol. The samples were boiled for 10 min, loaded onto 10% Tris·HCl polyacrylamide gels (Thermo Scientific), and separated by SDS-PAGE. Proteins were transferred to a PVDF membrane, and blocked with 5% Blotto (Santa Cruz Biotechnology) in Tris-buffered saline with 0.1% Tween-20. Membranes were incubated with specific antibodies (all Abcam) against EEA1 (no. ab2900), Rab7 (no. ab58029), 58K Golgi protein (no. ab27043), and LAMP1 (no. ab24170) and a polyclonal GAPDH loading control (no. ab9485) overnight at 4°C. All antibodies were used at 1 μg/ml. After washing with PBS with Tween 20, horseradish peroxidase-conjugated secondary antibodies (Cell Signaling Technology) were applied, and the blots were visualized on a Bio-Rad Chemi Doc XRS system (Bio-Rad).

Immunocytochemistry.

When significant differences in the expression of markers of the endolysosomal pathway were observed by Western blotting, we further investigated the relationship of these markers with TLR4 by immunocytochemistry (ICC). Cell lines were seeded onto collagen-coated glass coverslips (13-mm diameter) by using 12-well cell culture plates and were stimulated with LPS as described above. At selected time points cells were fixed with 1% paraformaldehyde (10 min) and washed with PBS. After blocking and permeabilization with 10% normal goat serum in PBS containing 0.5% Triton X-100, cells were stained overnight for LAMP1 (1:200, after Ag retrieval) or 58K Golgi protein (1:200) and subsequently stained with TLR4 (E-bioscience, no. 14.9920.83, 1:200). Visualization was performed with goat anti-mouse or goat anti-rabbit Alexa Fluor 568 (red) for LAMP1 or 58K and goat anti-rabbit Alexa Fluor 488 (green) for TLR4 (all 1:1,000, 1 h). Coverslips were mounted onto slides with Vectashield mounting medium containing DAPI, prior to being imaged on a Leica DM5500 fluorescence microscope or a Nikon C1 confocal microscope. Images were collected by using identical settings for the respective fluorescence channels, and areas of colocalization appear yellow/orange.

Statistical analysis.

All data are presented as means ± SE. Differences between groups were analyzed by the Kruskal-Wallis nonparametric ANOVA with Dunn's posttest and were considered to be significant if P < 0.05. GraphPad Prism was used to plot graphs and to analyze the data.

RESULTS

In the following sections, significant changes in gene or protein expression are shown relative to the corresponding untreated control. Given the limited amount of material available from primary ALI cultures, all primary NEC investigations were performed at the mRNA level. To account for interpatient variation, the expression at 24 h is given relative to the untreated control for each individual sample.

CF cells display sustained expression of TLR4 after LPS stimulation.

Surface and intracellular expression of TLR4 was determined by flow cytometry. Following LPS stimulation, 16hBE14o- cells displayed significant reductions (P < 0.01–0.001) in surface TLR4 expression from 4 h poststimulation onward (Fig. 1A). Similarly cytoplasmic TLR4 expression was reduced in a time-dependent manner in 16hBE14o- cells with significant differences (P < 0.05) observed 12 h poststimulation (Fig. 1B). Surface and intracellular TLR4 expression in the CFBE41o- comp cell line followed a similar pattern to that of 16 hBE14o- cells. However, the reduction in TLR4 expression in CFBE41o- comp cells took longer to initiate and was not as pronounced as in 16hBE14o- cells (Figs. 1, A and B). Conversely, CFBE41o- cells showed sustained surface expression of TLR4 that did not alter with LPS stimulation (Fig. 1A) and increased (P < 0.01) intracellular expression of TLR4 (Fig. 1B) 24 h after stimulation.

Fig. 1.

Surface and intracellular Toll-like receptor 4 (TLR4) expression. 16hBE14o-, CFBE41o- comp [CFBE41o- cells complemented with wild-type (wt) CFTR], and CFBE41o- cells were exposed to LPS for 0–12 h as indicated in the figure. The percentage of cells positive for surface expression (A) and intracellular expression (B) of TLR4 was determined by flow cytometry. Data are presented as means ± SE with n = 3–5 for all experiments and with *P < 0.05, **P < 0.01, ***P < 0.001.

TLR4 expression is accompanied by persistent nuclear expression of p65.

We have previously reported increased nuclear expression of p65 in CFBE41o- cells and primary CF NECs, which was accompanied by significant release of IL-8 (3). Here, we have extended this study to CFBE41o- comp cells to determine the role of CFTR in sustained p65 expression. For comparative purposes, these data are shown in Fig. 2, alongside some of our previously published observations. The nuclear expression of p65 subunit was significantly upregulated in 16hBE14o- cells 1 and 4 h poststimulation but returned to basal levels 12 h poststimulation (Fig. 2A). CFBE41o- comp cells did not show significant upregulation of p65 following LPS stimulation. In contrast, LPS stimulation of CFBE41o- cells provoked significant increases (P < 0.001) in the nuclear expression of p65, which remained significantly upregulated (P < 0.001) up to 12 h poststimulation (Fig. 2A).

Fig. 2.

Nuclear NF-κB (p65) expression and release of IL-8 and IL-10. 16hBE14o-, and CFBE41o- cells were exposed to LPS for 0, 1, 4, and 12 h and the percentage of cells positive for the nuclear expression of the p65 (A) subunit of NF-κB was determined by using an adapted published flow cytometric method (26) with n = 3–5. The expression of p65 was found to be persistently upregulated in CFBE41o- cells following stimulation (A). Heightened mRNA expression of p65 was also observed in primary nasal epithelial cells (NECs) from patients homozygous for F508del following LPS stimulation (n = 4) (B). Data are presented as means ± SE with ***P < 0.001. The release of IL-8 (B) and IL-10 (C) was measured from cell-free supernatants by ELISA. Samples were assayed in duplicate and data are presented as means ± SE with n = 4 for all cell line experiments. Cytokine release for each cell line has been compared with the corresponding untreated control (0 h) with ***P < 0.001.

Altered release of pro- and anti-inflammatory cytokines in CF epithelial cells.

Following LPS stimulation, the release of IL-8 increased in a time-dependent manner in CFBE41o- cells with significant differences (P < 0.01–0.001) observed from 4 h onward (Fig. 2B). A significant increase in LPS-medicated IL-8 release was not observed in control cell lines (16hBE14o- or CFBE41o- comp cells, Fig. 2B). The release of the anti-inflammatory cytokine, IL-10 (Fig. 2C), increased in a time-dependent manner in 16hBE14o- and CFBE41o- comp cells, with a significant increase (P < 0.01) observed 24 h after LPS stimulation. However, the opposite effect was observed in CFBE41o- cells, where a significant reduction (P < 0.001) in IL-10 release was observed 24 h after LPS stimulation (Fig. 2C).

CF cells display prolonged expression of endosomal markers.

The expression of markers of the early (EEA1) and late (Rab7) endosome were investigated in cell lines by Western blot. 16hBE14o- and CFBE41o- comp cells treated with LPS displayed peak expression of EEA1 (Figs. 3, A and B) 1 h after LPS stimulation and peak expression of Rab7 (Figs. 4, A and B) 1–8 h postsimulation. Conversely, CFBE41o- cells showed sustained expression of both markers up to 24 h poststimulation with LPS (Figs. 3C and 4C). The mRNA expression of EEA1 and Rab7 was also investigated in primary NECs from CF patients and age-matched controls (Figs. 3D and 4D). Rab7b, transcript variant 1 was the only Rab7 isoform detectable in NECs. Although mean expression levels of EEA1 and Rab7b were increased in CF NECs compared with control NECs, this did not reach statistical significance (Figs. 3D and 4D).

Fig. 3.

Expression of the early endosomal marker EEA1. The expression of early endosome antigen1 (EEA1) was assessed in 16hBE14o- (A), CFBE41o- comp (B), and CFBE41o- (C) cells by Western blot, and in primary NECs (D) by quantitative PCR (qPCR). All cells were stimulated with LPS as indicated. Western blot images were analyzed via the Bio-Rad Chemi Doc XRS system, standardized to a GAPDH loading control, and presented as percentage change from untreated control (0 h) for each cell line. Expression of EEA1 was confirmed in NECs by qPCR. mRNA expression after 24-h LPS stimulation was expressed relative to the untreated control (0 h) for each sample. Rab7b was the only isoform detectable in NECs. Data are presented as means ± SE with n = 4 for all experiments. Representative Western blot images are shown.

Fig. 4.

Expression of the late endosomal marker Rab7. The expression of the small GTPase member of the RAS oncogene family, Rab7 was assessed in 16hBE14o- (A), CFBE41o- comp (B), and CFBE41o- (C) cells by Western blot, and in primary NECs (D) by qPCR. All cells were stimulated with LPS as indicated. Western blot images were analyzed by using the Bio-Rad Chemi Doc XRS system, standardized to a GAPDH loading control, and presented as percentage change from untreated control (0 h) for each cell line. Expression of Rab7 was confirmed in NECs by qPCR. mRNA expression after 24-h LPS stimulation was expressed relative to the untreated control (0 h) for each sample. Rab7b was the only isoform detectable in NECs. Data are presented as means ± SE with n = 4 for all experiments. Representative Western blot images are shown.

Expression of the 58K Golgi protein is only detected in CF cells.

Expression of the 58K Golgi protein was not detected in 16hBE14o- or CFBE41o- comp cells by Western blot (Fig. 5, A and B). However, CFBE41o- cells showed persistent expression of 58K that was not significantly altered by LPS stimulation (Fig. 5C). Consistently, control NECs showed a significant reduction in the expression of 58K mRNA 24 h after LPS stimulation whereas no significant change was seen in CF NECs (Fig. 5D). To investigate potential colocalizations cells were stained by ICC for TLR4 (488 nm, green) and 58K protein (568 nm, red). However, despite persistent expression of 58K in CFBE14o- cells, colocalization with TLR4 was not observed (data not shown).

Fig. 5.

Expression of the 58K Golgi protein. The expression of the 58K Golgi protein was assessed in 16hBE14o-, CFBE41o- comp, and CFBE41o- cells by Western blot, and in primary NECs by qPCR (A). All cells were stimulated with LPS as indicated. Western blot images were analyzed by using the Bio-Rad Chemi Doc XRS system, standardized to a GAPDH loading control, and presented as percentage change from untreated control (0 h) for each cell line. Expression of 58K was confirmed in NECs by qPCR (A). mRNA expression after 24-h LPS stimulation was expressed relative to the untreated control (0 h) for each sample. Data are presented as means ± SE with n = 4 for all experiments. Representative Western blot images are shown.

TLR4 is not targeted toward the lysosome in the CF epithelium.

We next investigated the expression of the lysosomal marker LAMP1 in both cell lines and primary NECs. Peak LAMP1 expression was observed 1 h poststimulation in 16hBE14o- cells (Fig. 6A) and 4 h poststimulation in CFBE41o- comp cells (Fig. 6B). After these time points, expression returned to or below basal levels in both cell lines. However, LAMP1 was barely detectable in CFBE41o- cells at basal levels and 1 h after LPS stimulation. Thereafter, LAMP1 expression was not observed in CFBE41o- cells (Fig. 6C). The mRNA expression of LAMP1 in primary NECs followed a similar pattern, with CF NECs showing a significant reduction in LAMP1 expression after 24-h LPS stimulation. Consistent with results in the cell lines, control NECs displayed no change in LAMP1 mRNA expression 24 h after LPS stimulation (Fig. 6D). To further investigate this observation, we stained 16hBE14o- and CFBE41o- cells for TLR4 (488 nm, green) and LAMP1 (568 nm, red) by ICC. Considerable colocalization (yellow/orange) of TLR4 and LAMP1 was observed 1 h after LPS stimulation in 16hBE14o- cells, whereas CFBE41o- cells showed little colocalization (Fig. 6E).

Fig. 6.

Expression of the lysosomal marker LAMP1. The expression of the lysosomal marker LAMP1 was assessed in 16hBE14o-, CFBE41o- comp, and CFBE41o- cells by Western blot, and in primary NECs by qPCR (A). All cells were stimulated with LPS as indicated. Western blot images were analyzed by using the Bio-Rad Chemi Doc XRS system, standardized to a GAPDH loading control, and presented as percentage change from untreated control (0 h) for each cell line. Expression of LAMP1 was confirmed in NECs by qPCR (A). mRNA expression after 24-h LPS stimulation was expressed relative to the untreated control (0 h) for each sample. Data are presented as means ± SE with n = 4 for all experiments. Representative Western blot images are shown. Colocalization of TLR4 with LAMP1 was assessed by immunocytochemistry (B). Nuclei are stained blue with DAPI, TLR4 is stained green (488 nm), and LAMP1 is stained red (568 nm) with areas of colocalization following 1 h LPS stimulation shown in yellow/orange. Representative images of 2 independent experiments are shown. Scale bars = 10 μm.

The CF epithelium shows sustained expression of the small GTPase Rab10.

The small GTPases Rab10 and Rab11a have been reported to colocalize with TLR4 and regulate intracellular trafficking of TLR4 to the Golgi (27) or phagosome (43) respectively. Here, we examined the mRNA expression of Rab10 and Rab11a at basal levels and following LPS stimulation. A full time course (0–24 h; data not shown) was examined in cell lines, with notable differences in expression only observed 24 h after LPS stimulation. Rab10 expression was reduced in 16hBE14o- and primary NECs after 24-h LPS stimulation but remained unaltered in CFBE41o- comp, CFBE41o-, and primary NECs from F508del homozygotes (Fig. 7A). Rab11a expression was reduced in CFBE41o- comp cells only (Fig. 7B).

Fig. 7.

mRNA expression of the small GTPases Rab10 and Rab11a. The small GTPases Rab10 and Rab11a regulate endolysosomal trafficking toward the Golgi and phagosome, respectively. mRNA expression of Rab10 (A) and Rab11a (B) was examined in 16 hBE14o-, CFBE41o- comp, and CFBE41o- cells and in primary NECs by qPCR following treatment with LPS for 24 h. Data are presented as means ± SE with n = 3 for all experiments and *P < 0.05.

DISCUSSION

The internalization and targeted degradation of activated TLRs is an essential component in terminating inflammatory signaling. Patients with CF display chronic uncontrolled activation of NF-κB and higher release of inflammatory mediators than the general population. In this study we show that LPS stimulation triggers persistent intracellular TLR4 signaling in CF, which is not terminated in a timely manner. The cell surface expression of TLR4 in CF cells has proved contentious in the literature. Reports have shown that it is both elevated (21) and reduced following activation of the receptor with appropriate bacterial derived stimuli. In this study we report that surface and intracellular TLR4 expression remains largely unchanged in airway epithelial cells with a nonsignificant trend for increased intracellular TLR4 expression following LPS stimulation. This is consistent with recent findings in CF macrophages (5). In contrast, TLR4 expression was reduced in control cells stimulated with LPS in a time-dependent manner. The discrepancies as to whether TLR4 is increased or reduced in CF cells are complicated by the range of different stimuli, bacteria, bacterial products, and cell lines used for each study. However, the clear distinction is that the expression profile differs greatly between CF and non-CF cells and that altered TLR4 expression is linked with chronic inflammation in CF airway cells. The persistent expression of TLR4 (both surface and intracellular) in CF cells reported here would indicate that TLR4 does not undergo the same degradative processes in the CF airway epithelium.

The activation of TLR4 by LPS was accompanied by an inflammatory response in CF cells that was characterized by persistent expression of p65 and a time-dependent increase in IL-8 and reduction in IL-10 release. Previously, we have shown that p65 selectively regulates IL-8 processing in the CFBE41o- cell line (5). There is disagreement about the secretion of IL-10 from epithelial cells in vitro in the literature. However, Bonfield et al. (2) showed that bronchial epithelial cells in culture constitutively produce anti-inflammatory IL-10. Furthermore, her work also shows that IL-10 secretion is significantly reduced in cultured epithelial cells from patients with cystic fibrosis, which is consistent with our findings. The inability of mutant cells to release the anti-inflammatory cytokine IL-10 may well exacerbate the inflammatory response to LPS challenge. This increase in inflammation was associated with prolonged expression of markers of the endosome (EEA1, Rab7b) and Golgi (58K) and a relative lack of the lysosomal marker LAMP1 in CF epithelial cells only. Bruscia et al. (5) have recently reported increased and prolonged localization of TLR4 in the early endosome in peripheral blood derived macrophages from CF patients. Here, we investigated the expression of the early endosomal marker EEA1 in the lung using bronchial epithelial cell lines and primary NECs in culture. Peak EEA1 expression was observed in non-CF cell lines 1 h after LPS stimulation and was barely detectable thereafter. EEA1 expression was increased in the CF cell line following LPS stimulation at all time points examined and remained above basal levels 24 h after stimulation. These trends were reflected in NECs from control participants and CF patients, although the difference between control and CF NECs did not reach significance. The work of Bruscia and colleagues (5) together with the findings of this study would imply that in pulmonary and immune cells from CF patients TLR4 spends an extended period in the early endosomal compartment, and this may contribute to the enhanced inflammatory signaling we observed in CF epithelial cells. Prolonged intracellular signaling has been described when the receptor involved remains in a perinuclear compartment (5) and does not enter the endolysosomal path.

To increase our understanding of the potential role of the endosome in airway inflammation in CF, we next examined markers of the late endosome. Rab7 is a small GTPase and member of the Ras superfamily of GTPase proteins which is responsible for trafficking to the late endosome/lysosomes. Our initial work in cell lines used an antibody against total Rab7 (including the Rab7b isoform) and found that persistent expression of the protein was observed in the CF but not in the control cell line. However, only Rab7b, which shares ∼68% similarity with Rab7 (28), was detectable in primary NECs. Rab7b has also been shown to localize on the late endosomes and lysosomes (45) and to play an important role in trafficking lysosomal enzymes to the late endosome/lysosome (45). Furthermore, Rab7b directs TLR4 toward the lysosome for degradation, resulting in the negative regulation of TLR4 protein levels on the cell membrane (36). Conversely, the activated, GTP-bound form of Rab7b (Q67L) has been found to traffic almost exclusively to the Golgi apparatus and the trans-Golgi network, revealing an additional but essential role for Rab7b in endosome-to-Golgi transport of TLRs (44).

The 58K Golgi protein is a peripheral membrane-associated protein exposed to the cytoplasmic side of the Golgi apparatus (6). When stimulated with LPS, the control cell line failed to express 58K, whereas control NECs showed significant reductions in the mRNA expression of 58K from basal unstimulated levels. However, CF cells (cell line and primary NECs) displayed persistent expression of 58K at all time points examined. Given its role in trafficking toward the Golgi and the heightened expression of 58K observed in CF cells, we hypothesized that Rab7b may play an important role in directing TLR4 toward the Golgi in the CF epithelium. Rapid recycling of TLR4 complexes between the Golgi and the plasma membrane have been reported following LPS stimulation (1, 15). In line with this theory, we report persistent expression of TLR4 on the cell surface in CF. However, we were unable to detect colocalization of TLR4 and 58K in the CF cell line at the times examined in this study.

Jarry and Cheung (24) examined the internalization of live Staphylococcus aureus in CF and non-CF epithelial cell lines. They reported equal acquisition of the lysosomal marker LAMP1 to S. aureus-containing vesicles immediately after infection. However, 4 h postinvasion, the percentage of S. aureus associated with LAMP1 decreased significantly in CF cells compared with control cells, where the association remained at ∼100% (31). This suggests a fundamental disparity in the fate of bacteria in CF cells. Similarly, we examined the expression of LAMP1 in CF and non-CF cell lines following LPS stimulation. Consistent with the observations of Jarry and Cheung (24), we found little if any LAMP1 in the CF cell line by 4 h poststimulation, while LAMP1 was highly expressed in the control cell line at all time points examined. These findings suggest that in CF, bacteria and bacterial products (LPS) are not targeted to the lysosome for degradation. Consistent with this, we observed colocalization of TLR4 and LAMP1 in the control cell line, but not in CF cells.

Continuous replenishment of TLR4 from Golgi to plasma membrane may be regulated by the small GTPase Rab10, which colocalizes with TLR4 and is predominantly found in 58K and EEA1-positive subcellular compartments (24). Here, we show that LPS stimulation causes a reduction in Rab10 mRNA expression in 16 hBE14o-, but not in CFBE41o- comp. However, similar to our findings on Rab7b expression, Rab10 expression does not change from basal levels following LPS stimulation in CF cells. The persistent expression of Rab10 may promote TLR4 recycling to the cell membrane. In line with this, we report sustained cell surface expression of TLR4 in response to LPS in CF cells only. However, the finding that the pattern of Rab10 expression in CFBE41o- comp cells was similar to that of CFBE41o- cells suggests that this process may not be entirely dependent on CFTR and may be influenced by other factors such as the chronic inflammatory state.

Rab11a is involved in the recycling of previously internalized endosomal membranes to the cell surface (43) and controls fusion with the plasma membrane at sites of phagocytosis. Recently, Husebye et al. (8) reported that TLR4 colocalizes with Rab11a-positive recycling endosomes and is specifically directed toward phagosomes containing gram-negative bacteria. We hypothesized that a reduction in Rab11a in CF cells may account for the continued intracellular expression of TLR4 observed here. Conversely, a difference in the expression of Rab11a between control and CF cells was not observed, suggesting that Rab11a is unlikely to play a significant role in intracellular trafficking of TLR4 in LPS-stimulated CF epithelium.

The use of immortalized cell lines could be seen as a limitation of our study. To avoid drawing direct conclusions from the comparison of nonisogenically matched cell lines, we have also included CFBE41o- cells complemented with wt CFTR in our analyses. Furthermore, we have concentrated on differences in expression profiles within each cell line over time. Although isogenically matched, the use of the CFTR-complemented CFBE41o- cell line comes with the additional caveat that the existing parental F508del/F508del CFTR may still be expressed in low levels additionally to the wt CFTR. Overall, our data suggest that complementation of CFBE41o- with wt CFTR may help to normalize the endolysosomal response of CFBE41o- epithelial cells to an inflammatory stimulus. However, it is possible that CFTR is not solely responsible for endolysosomal trafficking within the cell and that additional factors may also have a role to play.

In summary, our work suggests a significant dysfunction of the endolysosomal process in LPS-induced inflammation in CF airway epithelial cells. Targeting of the TLR4-LPS complex toward the lysosome is an important step in terminating the inflammatory response. However, alternative mechanisms including incomplete processing and degradation of TLR4 in the ER (see Ref. 21) and perturbation of cholesterol trafficking leading to TLR4 accumulation in endosomal compartments (38) may also need to be taken into account. Therefore, the apparent discrepancy in endosomal sorting of TLR4 toward the lysosome reported in this study is likely to be a contributing factor but may not in itself fully account for the chronic inflammatory state in CF. Clearly, there is an additional possibility that TLR4 may be mutated in CF [as previously suggested by Greene et al. (9)] or has undergone a conformational change that may account for differential intracellular trafficking. Recently TLR4 genotype has been shown to affect the survival of CF mice (32). However, the observation that CF primary cells express Rab7b specifically and show sustained expression of the 58K Golgi marker while control cells do not would indicate a difference in endosomal sorting that is inherent in CF.

A proposed model of endolysosomal dysfunction in CF is given in Fig. 8. Our findings suggest that, in all cells, LPS stimulation triggers internalization of the TLR4 complex, which enters the endosome. However, CF cells display persistent expression of Rab7b and we hypothesize that the TLR4-LPS complex is targeted toward the Golgi, where it may undergo recycling toward the cell membrane. Future studies will investigate this theory by examining the turnover, recycling, and half-life of surface TLR4 in CF epithelial cells and by further examining the important of the Golgi in this process. The clear distinction between CF and non-CF cells is that TLR4 is directed toward the lysosomes for degradation in non-CF cells and the inflammatory response is terminated (Fig. 8).

Fig. 8.

Proposed model of endolysosomal signaling in healthy and cystic fibrosis airways. LPS stimulation triggers internalization of the TLR4 complex which enters the endosome in both CF and healthy control cells. However, CF cells display persistent expression of Rab7b and the 58K Golgi protein. We hypothesize that TLR4 is directed toward the Golgi, where it may undergo recycling toward the cell membrane. In contrast, however, TLR4 is directed toward the lysosomes for degradation in healthy control cells and the inflammatory response is terminated.

The present study examines endolysosomal trafficking in human CF airways epithelial cells (both primary cells and cell lines) for the first time and builds on recent observations that abnormal TLR trafficking underlies inflammation in peripheral blood macrophages from CF patients (7). However, a thorough understanding of the role of Rab GTPases in this process is likely to greatly enhance our understanding of this complex area of cellular signaling and may bring new opportunities to target defects in signaling in CF. Here, a comprehensive investigation of the role of Rab7b in chronic inflammation is required to investigate whether Rab7b could be a novel target for therapeutic interventions in CF airway inflammation.

GRANTS

This work was funded by a grant (to B. C. Schock) from the Cystic Fibrosis Trust UK (PJ541).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

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

ACKNOWLEDGMENTS

We are grateful to Rebecca Cromie, Queen's University of Belfast, for assistance with ELISAs.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
View Abstract