Repression of Connexin26 hemichannel activity protects the barrier function of respiratory airway epithelial cells against LPS-induced alteration

Abstract

In respiratory airway epithelial cells, lipopolysaccharide (LPS) treatment induced an enhancement of connexin 26 (Cx26) hemichannel activity shown by dye uptake experiments after siRNA-mediated knock-down of Cx26. This effect was already observed at infection relevant concentrations (≤ 10 ng/mL LPS) and involved tumor necrosis factor alpha (TNF-α)- and Ca²⁺-dependent signaling. High concentrations (1 µg/mL LPS) reduced the transepithelial electrical resistance (TEER) of Calu-3 cells by 35% within an application time of 3 h followed by a recovery. Parallel to barrier alteration, a reduced tight junction organization rate (TiJOR) of claudin-4 (CLDN4) by 75% was observed within an application time of 3 h. After TEER recovery, CLDN4 TiJOR stayed reduced. Low concentrations (10 ng/mL LPS) required three times repeated application for barrier reduction and CLDN4 TiJOR reduction by 30%. The small molecule CVB4-57, newly published as a potential inhibitor of Cx26 hemichannels, mitigated the effects of LPS on the epithelial barrier function. Molecular docking studies revealed a potential interaction between CVB4-57 and Cx26 thereby reducing its hemichannel activity. We conclude that LPS-related enhancement of Cx26 hemichannel activity acts like a "molecular scar" that weakens the lung epithelium, which could be attenuated by agents targeting Cx26 hemichannels.

Keywords: Airway epithelium; Barrier function; Calu-3 cells; Connexin channels; lipopolysaccharide; Cytokine; PCLS; Primary cells.

Fig. 1

LPS enhances the EtdBr dye uptake in airway epithelial cells. a-b EtdBr dye uptake rates (fluorescence intensities (AU)/min) under different perfusion conditions (exemplary experiment see Fig. S1) in Calu-3 cells (a) and PBEPCs (b) cultivated under control conditions or with 1 ng/mL LPS for 24 h. Perfusion conditions: 2 mM [Ca²⁺]_ex, 0 mM [Ca²⁺]_ex or 0 mM [Ca²⁺]_ex + 1 mM La³⁺ (n = cell patches for Calu-3 cells, single cells for PBPECs). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.05 *, p < 0.001 *** vs. different perfusion conditions; p < 0.05 #, p < 0.001 ### vs. control). c Real-time qRT-PCR for different Cx isoforms in PBEPCs. d Immunofluorescence staining against Cx26 or Cx43 (yellow) in PBEPCs. Scale bar = 20 μm. e Exemplary immunofluorescence staining against Cx26 (yellow) in the airways of PCLS cultivated under control conditions or with 1 µg/mL LPS for 3 h. Scale bar = 10 μm. f Cx26 immunofluorescence signal (particle count/area, relative to control) in PCLS after treatment with 1 µg/mL LPS for 3 h (n = analyzed PCLS from 3 donors, 2 PCLS/donor with mean of 3 airway areas/PCLS). Different symbol shapes visualize different donors. Unpaired two-tailed Student’s t-test (p < 0.05 * vs. control). g EtdBr dye uptake rates in absence of [Ca²⁺]_ex relative to the rates obtained in presence of [Ca²⁺]_ex in Calu-3 cells in which Cx26 or Cx43 expression was suppressed using respective siRNA and which were treated with 1 ng/mL LPS or vehicle for 24 h (n = measured cell patches). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.01 **, p < 0.001 *** vs. negative siRNA + vehicle; p < 0.05 #, p < 0.001 ### vs. respective siRNA + vehicle; p < 0.001 +++ vs. LPS). h EtdBr dye uptake rates in absence of [Ca²⁺]_ex ± 5 µM CVB4-57 relative to the rates obtained in presence of [Ca²⁺]_ex in Calu-3 cells treated with vehicle or 1 ng/mL LPS for 24 h (n = cell patches). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.001 *** vs. vehicle; p < 0.001 ### vs. LPS)

Fig. 2

Molecular docking of CVB4-57 in the structure of Cx26 hemichannels. a-b Putative binding modes of CVB4-57 on the Cx26 7QEV structure (clipped molecular surface) based on the distribution of best docked poses in site-01 (a) and site-02 (b). NTHs, relative to site-01 cavity (left NTH in brown) and site-02 cavity (between NTHs) are shown in brown and gold (tubes, ribbons), respectively. Some tagged residues (navy blue) are shown below the molecular surface (transparency). For the molecular docking of CVB4-57 in site-01 (a), three pathways could be recorded: linear along the cavity (pink pose); constrained (green poses); extended towards site-02 cavity (purple). For the molecular docking in site-02 (b), a main linear fluctuation was recorded near the NTH (right, pink-purple). Poses were found in the central sub-cavity (teal) and in site-01 (light green). c-d Since the Cx26 structure 7QEV consists of only two chains (a, b, tan color), three 7QEV structures (tan, teal, blue) were aligned on 7QEW [35] (Cx26 type, hexamer) to reconstruct a hexamer. c Clipped hexamer (orthogonal, view from cytoplasmic side) showing best docked poses in site-01 (left, under brown ribbons) and site-02 (right, near gold ribbons). One linear pose of CVB4-57 per site was highlighted by molecular surface (site-01: purple, site-02: dark grey). d Laterally clipped hexamer showing the pore and molecular surfaces (transparency) of highlighted docked poses. The position of Ca²⁺ atoms (green spheres) is also shown after alignment with 5ER7 [54] (Cx26 type, dodecamer split as hexamer)

Fig. 3

Involvement of TLR4 and TNF-α signaling in the LPS-induced enhancement of Cx26 hemichannel activity. a Real-time qRT-PCR experiments for TLR4 mRNA expression in Calu-3 cells and PBEPCs (n = 3 biological replicates or donors, respectively). b Immunofluorescence staining against TLR4 (yellow) in Calu-3 cells and PBEPCs. Scale bar = 20 μm. c EtdBr dye uptake rates in absence of [Ca²⁺]_ex relative to the rates obtained in presence of [Ca²⁺]_ex in Calu-3 cells treated for 24 h with vehicle, 1 ng/mL LPS ± preincubation (0.5 h) with 20 µM C34 or C34 alone (n = cell patches). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.001 *** vs. vehicle, p < 0.001 ### vs. LPS). d Real-time qRT-PCR for TNF-α mRNA amount in Calu-3 cells treated for 3 h with 1 µg/mL LPS (n = 3). Unpaired two-tailed Student’s t-test. e EtdBr dye uptake rates in absence of [Ca²⁺]_ex relative to the rates obtained in presence of [Ca²⁺]_ex in Calu-3 cells treated for 24 h with vehicle, 1 ng/mL LPS ± preincubation (0.5 h) with 10 µM SPD-304 or 10 µM Marimastat, with SPD-304/Marimastat alone or for 1 h with 10 ng/mL TNF-α (n = cell patches). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.05 *, p < 0.001 *** vs. vehicle; p < 0.001 ### vs. LPS; p < 0.001 +++ vs. TNF-α). f EtdBr dye uptake rates in absence of [Ca²⁺]_ex relative to the rates obtained in presence of [Ca²⁺]_ex in Calu-3 cells in which Cx26 expression was suppressed using respective siRNA and which were treated with 10 ng/mL TNF-α or vehicle for 1 h (n = cell patches). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.01 **, p < 0.001 *** vs. negative siRNA + vehicle; p < 0.05 #, p < 0.001 ### vs. respective siRNA + vehicle; p < 0.001 +++ vs. TNF-α)

Fig. 4

Involvement of intracellular Ca²⁺ in the LPS- and TNF-α-induced enhanced dye uptake rate. a Ca²⁺ imaging using Fura-2-loaded Calu-3 cells to estimate baseline [Ca²⁺]_i levels (mean F_{360 nm}/F_{380 nm} of 5 min measurement) in cells treated with 1 ng/mL LPS for 24 h or 10 ng/mL TNF-α for 1 h (n = single cells). One-way ANOVA with Sidak’s multiple comparison test (p < 0.05 * vs. vehicle). b EtdBr dye uptake rates in absence of [Ca²⁺]_ex relative to the rates obtained in presence of [Ca²⁺]_ex in Calu-3 cells treated for 24 h with vehicle, 1 ng/mL LPS ± preincubation (0.5 h) with 10 µM BAPTA, with BAPTA alone or for 1 h with 10 ng/mL TNF-α ± preincubation (0.5 h) with 10 µM BAPTA (n = cell patches). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.001 *** vs. vehicle; p < 0.001 ### vs. LPS; p < 0.001 +++ vs. TNF-α)

Fig. 5

LPS modulates the tight junction barrier in Calu-3 cells, which can be attenuated by CVB4-57. a Relative TEER of cells cultivated on transwell inserts and treated for 3.5 h with 1 ng/mL, 10 ng/mL, 100 ng/mL or 1 µg/mL LPS compared to control conditions (n = transwell inserts). One-way ANOVA with Dunnet’s multiple comparison test (p < 0.05 *, p < 0.01 ** vs. control). b Exemplary TEER Measurement over 24 h of cells cultivated on transwell inserts treated with 1 µg/mL LPS ± 5 µM CVB4-57 or CVB4-57 alone. Red arrow: analyzed time point shown in c. c Relative TEER of cells cultivated on transwell inserts 3 h after treatment with 1 µg/mL ± LPS 5 µM CVB4-57 or CVB4-57 alone (n = transwell inserts). One-way ANOVA with Sidak’s multiple comparison test (p < 0.001 *** vs. control; p < 0.001 ### vs. LPS). d Exemplary TEER measurement of repetitive treatment with 10 ng/mL LPS ± 5 µM CVB4-57 every 24 h for a total time of 72 h (1st, 2nd, 3rd application). Red arrow: analyzed time point shown in e. e Relative TEER of cells cultivated on transwell inserts 24 h after the 3rd treatment with 1 µg/mL LPS ± 5 µM CVB4-57 or CVB4-57 alone (n = transwell inserts). One-way ANOVA with Sidak’s multiple comparison test (p < 0.05 * vs. control)

Fig. 6

LPS leads to CLDN4 remodeling in epithelial cells of the airways, which can be attenuated by CVB4-57. a Real-time qRT-PCR for relative mRNA amounts of CLDN1, CLDN3, CLDN4, CLDN7 and ZO-1 in Calu-3 cells cultivated on transwell inserts after treatment for 3 h and 24 h with 1 µg/mL LPS, respectively (n = 3). One-way ANOVA with Dunnet’s multiple comparison test. b Exemplary immunofluorescence staining against ZO-1 (magenta) and CLDN4 (yellow) in Calu-3 cells cultivated on transwell inserts treated with 1 µg/mL LPS ± 5 µM CVB4-57 for 3 h and 24 h. Scale bar = 20 μm. c Tight junction organization rate (TiJOR, intersections/µm) of CLDN4 after application of 1 µg/mL LPS ± 5 µM CVB4-57 for 3 h or 24 h (n = transwell inserts). Kruskal-Wallis test with Dunn’s multiple comparison test (p < 0.01 ** vs. control). d Exemplary immunofluorescence staining against CLDN4 (yellow) in human PCLS cultivated for 24 h with 1 µg/mL LPS or culture medium (control). Upper images: 5 × 3 tile scan, scale bar = 200 μm. Lower images: z-stack of 30 μm with one image every 3 μm, scale bar = 10 μm). A similar result was obtained in PCLS of a second donor. e Exemplary immunofluorescence staining against ZO-1 (magenta) and CLDN4 (yellow) in Calu-3 cells cultivated on transwell inserts 24 h after 3rd application (application every 24 h) of 10 ng/mL LPS ± 5 µM CVB4-57. Scale bar = 20 μm. f TiJOR (intersection/µm) of CLDN4 24 h after the 3rd application of 10 ng/mL LPS ± 5 µM CVB4-57 (n = transwell inserts). One-way ANOVA with Sidak’s multiple comparison test (p < 0.05 * vs. control; p < 0.001 ### vs. LPS)