IL-4-induced caveolin-1-containing lipid rafts aggregation contributes to MUC5AC synthesis in bronchial epithelial cells

BECs were isolated from large bronchial airway of rat as described in the literature [26]. The experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee (IACUC) of the Tongji Medical College, Huazhong University of Science and Technology. In brief, airway was isolated from rat under sterile conditions, rinsed with ice D-Hanks twice, then digested with 1% Pronase E in DMEM/F-12 at 4°C for 14 h. Then BECs were harvested with ice D-hanks contain 5% newborn bovine serum (NBS) and centrifuged at 1000 rpm for 5 min. The spun down cells were resuspended with epithelial cell medium (Cell Biologics, Chicago, IL, USA) and incubated in a 5% CO₂ incubator (Thermo Fisher Forma, Waltham, MA, USA) at 37 °C. These cells were used in experiments without subculture. In most experiments, BECs were treated by the indicated factors in serum-free medium when the cells were 60–70% confluence. For MUC5AC immunostaining, the isolated cells digested with protease were cultured on cover slip with serum-containing medium for 24 h in the Petri dish. Some digested cells were presented as small clusters, but not completely separated cells. IL-4 containing serum-free medium was administered when the cells reached about 30% confluence.

BECs [Ca²⁺]_i were measured after loading with 10 μM of the acetoxymethyl ester form of Fura-2 for 30 min at room temperature. Then, the fluorescence of Fura-2 was recorded with a stimulation of 20 ng/ml IL-4, in the presence or absence of 10 mM M-βCD or 50 μM 2-APB after excitation at 340 ± 10 and 380 ± 10 nm using a xenon short-arc lamp (Ushio). Bandpass interference filters (Omega Optical, Brattleboro, VT 05301) selected wavelength bands of emitted fluorescence at 510 ± 10 nm. In the process of [Ca²⁺]_i measurements, we firstly got stable [Ca²⁺]_i signal, 250 s later, we treated cells with 2-APB or IL-4. M-βCD was used before [Ca²⁺]_i measurements. In some experiments, we firstly treated cells with 2-APB or M-βCD, then added IL-4 to the cells at 1800s. Emitted Fura-2 fluorescence was collected and measured using a spectrofluorometer (PTI, Deltascan).

To acquire the final [Ca²⁺]_i with the ratios from Fura-2 fluorescence, we had detected the minimum and maximum Fura-2 fluorescence ratio (R_min and R_max) with Ca²⁺-free buffer without calcium (8.182 g NaCl, 0.335 g KCl, 0.385 g EGTA, 2 g Glucose, 1.2 g HEPES, 0.2 g MgCl₂, 1 L H₂O, PH7.4) and Ca²⁺-high buffer with high concentration of Ca²⁺ (8.182 g NaCl, 0.335 g KCl, 0.555 g CaCl₂, 2 g Glucose, 1.2 g HEPES, 0.2 g MgCl₂, 1 L H₂O, PH7.4). The R_min was determined using Ca²⁺-free buffer and stimulated by 3 μM ionomycin (Sigma). The same to R_min, the R_max was measured with Ca²⁺-high buffer and stimulated by 3 μM ionomycin. The final [Ca²⁺]_i was calculated with R_min and R_max follow this equation: [Ca²⁺]_i = Kd*β*(R – R_min)/(R_max – R), Kd is a constant which is 224 for BECs, and the β values is the fluorescence intensity emitted by 380 ± 10 nm within Ca²⁺-free buffer. Ca²⁺ influx duration was calculated from the beginning of Ca²⁺ level rise to the time of Ca²⁺ level return to the base.

An ELISA-based assay was performed to measure endogenous NF-κB activities as described previously [27]. Briefly, after BECs were treated with 20 ng/ml IL-4 for 24 h in the presence or absence of M-βCD or 2-APB (M-βCD or 2-APB was administrated half an hour before IL-4 using), the cells were lysed with lysis buffer (20 mM HEPES pH 7.5, 0.35 M NaCl, 20% glycerol, 1% NP-40, 1 mM MgCl₂·6H₂O, 0.5 mM EDTA, 0.1 mM EGTA) containing a protease inhibitor cocktail (Calbiochem, La Jolla, CA, USA) on ice for 10 min. The supernatant obtained after centrifugation at 14,000 rpm for 30 min at 4 °C was recovered. The double-stranded probe with single-stranded-linker was generated by 1:1 mix of the following two oligonucleotide with the sequences of 5′-AGTTGAGGGGACTTTCCCAGGCC-(C)34-C-3′, the 3′ end biotinylated and 5′-GCCTGGGAAAGTCCCCTCAACT-3′, respectively. The probe was denatured at 94 °C for 10 min, annealed at room temperature overnight and then linked to streptavidin-coated 96-well plates (Roche, Mannheim, Germany) by incubating 2 pmol of probe per well for 1 h at 37 °C in 50 μl PBS. After wash, 20 μl of cell extract were mixed with 30 μl of binding buffer (4 mM HEPES pH 7.5, 100 mM KCl, 8% glycerol, 5 mM DTT, 0.2% BSA, 40 μg/ml salmon sperm DNA) in the above microwells incubated at room temperature with mild agitation (200 rpm) for 1 h. After wash, mouse anti-NF-κB p65 monoclonal antibody (1:1000 diluted) was incubated for 1 h at room temperature. After wash, peroxidase-conjugated goat anti-mouse IgG were incubated at room temperature for 1 h. After wash, 100 μl tetramethylbenzidine was incubated at room temperature for 10 min before adding 100 μl of stopping solution (2 M H₂SO₄). Optical density was then read at 450 nm under a microplate reader (Biotek Instruments, Winooski, VT, USA) using a 655-nm reference wavelength. Backgrounds are determined in lysis buffer and subtracted before data analysis.

Total RNA was extracted from BECs using Trizol Reagent. cDNAs were synthesized from total RNA by reverse transcription according to the manufacturer’s instruction. Then cDNAs were used for amplification by qRT-PCR in a 25 μl reaction using SYBR® Premix EX Taq™ II (TaKaRa) with a Mini Opticon Real-time PCR Systerm (BIORAD). MUC5AC mRNA expression was normalized with β-actin. The MUC5AC primers were F: 3-GCTCATCCTAA GCGACGTCT-5, R: 3-GGGGGCATAACTTCTCTTGG-5, and the β-actin primers were F: 3-CGGCATTGTCACCAACTG-5, R: 3-CGCTCGGTCAGGATCTTC-5.

Intracellular MUC5AC protein levels were measure using ELISA kits following the manufacturer’s instructions (Cloud-Clone Corp, Wuhan, China).

To determine the protein level of MUC5AC, BECs were cultured on cover slips, after pretreated with EGTA (0.5 mM, 2 mM or 5 mM), AS1517499 or LKK16 for 30 min, cells were treated with IL-4 (20 ng/ml). 24 h later, BECs were stained with rabbit anti-MUC5AC antibody (1:100 dilution) at 4 °C overnight, and then incubated with tetramethyl rhodamin isothiocyanate (TRITC)-conjugated goat anti-rabbit antibody at room temperature for 60 min. The nuclei were stained for DAPI for 10 min in dark. The fluorescence-labeled cells were examined using a Zeiss-LSM780 Confocal laser scanning microscope (Oberkochen, German).

To determine the protein level of caveolin-1 and TRPC1, as well as lipid rafts aggregation, BECs were cultured on cover slips, after pretreated with 10 mM M-βCD or 100 μM 2-APB for 30 min, cells were treated with 20 ng/ml IL-4 for 15 min in the presence or absence of M-βCD or 2-APB, BECs were stained with a mouse monoclonal antibody specific to caveolin-1 and rabbit anti-TRPC1 polyclonal antibody. Lipid rafts were visualized by staining with 2 μg/ml cholera toxin subunit B- Alexa Fluor 488 for 20 min.

To isolate lipid raft fractions from the cell membrane, BECs were lysed in 1.5 ml buffer containing 10 mM Tris·HCl, 150 mM NaCl, 5 mM EDTA, 1 mMPMSF, 3 mM Na₃VO₄, protease inhibitors cocktail and 1% Triton X-100 (pH 7.4). Cell extracts were homogenized with five passages through a 25-gauge needle. Homogenates were adjusted with 60% sucrose density gradient medium to 40% and overlaid with 2 ml 90% sucrose, 4 ml 35% sucrose and 4 ml 5% sucrose Density Gradient medium gently.

Samples were centrifuged at 39000 rpm and 4°C for 18 h using a SW32.1 rotor. Fractions were collected from top to bottom, each sample had 11 fractions. For immunoblot analysis of lipid raft-associated proteins, these fractions were precipitated by mixing with equal volume of 100% trichloroacetic acid and 30 min of incubation on ice, and then samples were centrifuged at 15000 g and 4°C for 15 min, protein sediments were washed with ice acetone twice, air dried, and then resuspended in 1 M Tris·HCl (pH 8.0), which was ready for immunoblot analysis. Western blots were performed to detect protein levels of caveolin-1 and TRPC1 in each fraction.

Results are shown as the mean ± SD for n experiments, n means cells from n rats. Differences between groups were analyzed using unpaired t tests or two-way analysis of variance. A P value less than 0.05 was considered to be statistically significant.

To study the effect of IL-4 on [Ca²⁺]_i in BECs, we treated BECs with IL-4 in the presence or absence of 2-APB, an inhibitor of Ca²⁺ release. As shown in Fig. 1a-d, IL-4 significantly increased the level of [Ca²⁺]_i, but this was abolished by 2-APB. After careful analysis of Ca²⁺ influx duration and amplitude, we found both Ca²⁺ influx duration and amplitude were increased with the treatment of IL-4 in BECs (Fig. 1e, f). These data suggested that IL-4 induced Ca²⁺ influx could be prevented by 2-APB in BECs.

To investigate the role of Ca²⁺ influx in MUC5AC synthesis, we detected MUC5AC mRNA and protein levels in BECs treated by IL-4. As shown in Fig. 2a and b, IL-4 increased the levels of MUC5AC mRNA and protein expression, and these were prevented by 2-APB (Fig. 2a, b). Because 2-ABP is nonspecific and blocks both TRP channels and intracellular inositol triphosphate receptor (IP3R), to distinguish between extracellular and intracellular Ca²⁺ source, we blocked extracellular Ca²⁺ using EGTA. As shown in Fig. 2c and d, IL-4 induced increases in MUC5AC expression, but this was attenuated by EGTA at a dose-dependent manner. These data suggested extracellular Ca²⁺ influx played a role in IL-4 induced MUC5AC expression.

To study downstream signals of IL-4 in BECs, we investigate STAT6 and NF-κB pathways. As shown in Fig. 3a and b, IL-4 induced time-dependent phosphorylation of STAT6 and P65 which suggested both signals were activated in IL-4 treated BECs. To further distinguish the role of STAT6 and NF-kB pathway in IL-4-induced MUC5AC expression, we used specific STAT6 inhibitor AS1517499 and a selective IκB kinase (IKK) inhibitor IKK 16 to treat cells. As shown in Fig. 3c, IKK 16 induced significant down-regulation of MUC5AC protein, but AS1517499 had no effect on IL-4 induced up-regulation of MUC5AC protein. These data suggested that NF-κB rather than STAT6 signal pathway mediated MUC5AC up-regulation induced by IL-4. To further dissect the role of NF-κB as well as the relationship with Ca²⁺ influx, we detected NF-κB activity. As shown in Fig. 3d, IL-4 enhanced NF-κB activity, and this was blocked by 2-APB. These data suggested extracellular Ca²⁺ influx played a role in IL-4 induced NF-κB activation and further MUC5AC expression.

To explore the potential mechanism underlying extracellular Ca²⁺ influx induced by IL-4, we next focused on the lipid rafts clustering in BECs. We detected the colocalization and clustering of caveolin-1 and ganglioside GM1 enriched lipid rafts. Cells were first incubated with an anti-calveolin-1 antibody, followed by staining with cholera toxin subunit B-Alexa Fluor 488, which is a specific clustering agent for GM1 enriched lipid rafts. As shown in Fig. 4a and b, IL-4 induced clustering and colocalization of GM1 enriched lipid rafts and caveolin-1. Moreover, when the cells were treated with the inducer of lipid rafts disruption (M-βCD), IL-4 failed to induce clustering of lipid rafts. By using sucrose gradient ultracentrifugation and immunoblotting, lysates were separated into lipid raft and non-lipid raft fractions. As shown in Fig. 4c, caveolin-1 was mainly found in non-lipid raft fractions in the control. However, the treatment of IL-4 led to significantly increased caveolin-1 in lipid raft fractions, and this was prevented by M-βCD (Fig. 4c, d). These data suggested that IL-4 induced caveolin-1 containing lipid rafts aggregation.

TRPC activation is necessary in extracellular Ca²⁺ influx when IP3 binds with IP3R, so to further uncover the relationship between extracellular Ca²⁺ influx and caveolin-1-containing lipid rafts aggregation, we studied the colocalization of TRPC1 and caveolin-1. As shown in Fig. 4e and f, IL-4 induced clustering and colocalization of caveolin-1 and TRPC1. After sucrose gradient ultracentrifugation, the immunoblotting images revealed that TRPC1 was found in non-lipid raft fractions in normal control, but IL-4 caused significant increase of TRPC1 in lipid raft fractions (Fig. 4g, h). These data suggested that IL-4 induced colocalization of TRPC1 in caveolin-1-containing lipid rafts aggregation.

To further confirm the colocalization of TRPC1 in caveolin-1-containing lipid rafts aggregation, we detected [Ca²⁺]_i levels. BECs were exposed to IL-4 in the presence or absence of M-βCD. As shown in Fig. 5, IL-4 triggered increases in [Ca²⁺]_i including amplitude and duration, but after disruption of lipid raft clustering with M-βCD, IL-4 did not induce any change in [Ca²⁺]_i signal. These data suggested that IL-4 induced intracellular calcium response is mediated by lipid rafts aggregation.

To study the role of IL-4-induced caveolin-1-containing lipid rafts aggregation in MUC5AC synthesis in BECs, we used M-βCD to disrupt lipid rafts aggregation, and then detected changes in NF-κB activity and MUC5AC levels. As shown in Fig. 6a, IL-4 induced increases in NF-κB activity, and this was blocked by M-βCD. IL-4 significantly increased the intracellular MUC5AC mRNA and protein levels (Fig. 6b, c), disruption of lipid raft by M-βCD attenuated IL-4-induced increases of MUC5AC. These data suggest that IL-4 induced caveolin-1-containing lipid rafts aggregation which contributed to MUC5AC synthesis.