Cigarette smoke disrupts the integrity of airway adherens junctions through the aberrant interaction of p120-catenin with the cytoplasmic tail of MUC1

Abstract

Adherens junctions (AJs) containing epithelial cadherin (E-cad) bound to p120-catenin (p120ctn) and β-catenin (β-ctn) play a crucial role in regulating cell-cell adhesion. Cigarette smoke abrogates cell-cell adhesion between epithelial cells by disrupting E-cad, a hallmark of epithelial-mesenchymal transition (EMT), yet the underlying mechanism remains unknown. We used an organotypic culture of primary human bronchial epithelial (HBE) cells treated with smoke-concentrated medium (Smk) to establish an essential role for the interaction between p120ctn and the cytoplasmic tail of MUC1 (MUC1-CT) in regulating E-cad disruption. Within the first 4 h of smoke exposure, apical MUC1-CT repositioned to the basolateral membrane of pseudo-stratified HBE cells, where it interacted with p120ctn. A time-dependent increase in MUC1-CT/p120ctn complexes occurred in conjunction with a time-dependent dissociation of p120ctn/E-cad/β-ctn complexes, as well as the coordinated degradation of p120ctn and E-cad. Interestingly, Smk induced a similar interaction between MUC1-CT and β-ctn, but this occurred 44 h after MUC1-CT's initial interaction with p120ctn, and well after the AJs were destroyed. Blocking MUC1-CT's interaction with p120ctn using a MUC1-CT dominant-negative peptide, PMIP, successfully abolished Smk's disruptive effects on AJs and recovered apical-basolateral polarity of HBE cells. The MUC1-CT/p120ctn interaction was highly dependent on EGFR/Src/Jnk-mediated tyrosine phosphorylation (TyrP) of MUC1-CT. Accordingly, EGFR, Src or Jnk inhibitors (AG1478, PP2, SP600125, respectively) abrogated Smk-induced MUC1-CT-TyrP, MUC1-CT/p120ctn interaction, AJ disruption, and loss of cellular polarity. Our work identified MUC1-CT and p120ctn as important regulators of epithelial polarity and cell-cell adhesion during a smoke-induced EMT-like process. Novel therapeutics designed to inhibit MUC1-CT/p120ctn complex formation may prevent EMT in the smoker's airway.

Figure 1

Smoke disrupts cell polarity by dissociating E-cadherin, p120-catenin, and β-catenin in adherens junctions (AJs) of polarized HBE cells. (A) AJs in pseudo-stratified HBE cells are stained by immunofluorescence after exposure to smoke (Smk) and smoke-free control medium (Ctrl) for 4 h. Antibodies directed against p120-catenin (p120ctn, green, top panels), β-catenin (β-ctn, green, middle panels), E-cadherin (E-cad, green, lower panels), and MUC1-cytoplasmic tail (MUC1-CT, red, all panels) are shown. Nuclei were visualized with DAPI (blue). In Ctrl-exposed cells (left panels), white boxes indicate AJs where p120ctn, β-ctn, and E-cad are intact. In Smk-exposed cells (right panels), white boxes indicate intercellular areas where p120ctn, β-ctn, and E-cad are lost. Yellow arrowheads point to areas on the apical surface where MUC1-CT is lost. Scale bar = 50 μm. (B) Membranous and non-membranous fractions of pseudo-stratified HBE cell lysates exposed to Smk or Ctrl medium for 4 h were analysed by western blots probed with p120ctn, β-ctn, and E-cad antibodies. Fraction purity and loading were determined by immunoblotting for BAP31. Densitometric analysis of the above proteins in smoke-treated cells was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold changes (^*p < 0.05; ^**p < 0.01). Black columns denote membranous fractions, and grey columns denote total proteins (the sum of membranous and non-membranous fractions). (C, D) Pseudo-stratified HBE cells were incubated with Smk or Ctrl medium and harvested at the indicated time points. (C) P120ctn immunoprecipitates (IP) were immunoblotted for p120ctn, β-ctn, and E-cad. Equal loading was revealed by immunoglobulin light chain (IgL). Densitometric quantitation of β-ctn and E-cad immunoprecipitated by p120ctn after a 4 h Smk treatment was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold changes (^*p < 0.05). (D) β-ctn immunoprecipitates (IP) were immunoblotted with β-ctn and E-cad. Equal loading was revealed by immunoglobulin heavy chain (IgH). Densitometric quantitation of E-cad immunoprecipitated by β-ctn after a 4 h Smk treatment was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold changes (^*p < 0.05).

Figure 2

Smoke-mediated MUC1-CT/p120ctn complex formation occurred in conjunction with dissociation of the E-cad/p120ctn/β-ctn complexes at AJs. (A, B) Pseudo-stratified HBE cells exposed to smoke (Smk) and smoke-free medium (Ctrl) were harvested at the indicated time points. Immunoprecipitates of MUC1-CT were probed with antibodies directed against p120ctn, β-ctn, E-cad, and MUC1-CT. Equal loading was confirmed by IgH. (C) To adjust for observed fluctuations in baseline levels of the MUC1-CT/p120ctn interaction in primary cells, densitometric quantitation of p120ctn (black columns) and β-ctn (grey columns) immunoprecipitated by MUC1-CT following 4, 24, and 48 h of Smk treatment was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold change of three independent experiments (^**p < 0.01). (D) Pseudo-stratified HBE cells are stained by immunofluorescence after exposure to Smk and Ctrl medium for 4 h. Staining with antibodies directed against p120ctn (green, top panels) and MUC1-CT (red, middle panels) is shown. Cell nuclei were visualized with DAPI (blue). White arrowheads indicate cytoplasmic translocation of p120ctn and MUC1-CT from the adherens junction and the apical membrane, respectively. Merged p120ctn and MUC1-CT images (generating a yellow signal, bottom panels) demonstrate co-localization of p120ctn and MUC1-CT in Smk-exposed cells (white arrowheads). Scale bar = 50 μm.

Figure 3

Disruptive effects of smoke on AJs of polarized HBE cells depend on p120ctn/MUC1-CT interaction. (A–D) Polarized HBE cells were preincubated overnight with 50 μM PMIP (a MUC1 inhibitory peptide conjugated to a protein transduction domain) [33] or vehicle control (water). The cells were exposed to smoke (Smk) or smoke-free medium (Ctrl) in the presence of 50 μM PMIP or water for 4 h. (A) Immunofluorescent staining of polarized HBE cells for p120ctn (green, top panels), β-ctn (green, middle panels), E-cad (green, lower panels), and MUC1-CT (red, all panels). Cellular nuclei were counterstained with DAPI (blue). Scale bar = 50 μm. (B) MUC1-CT immunoprecipitates (IP) were probed for p120ctn, β-ctn, and MUC1-CT. Equal loading was confirmed by IgH. To adjust for observed fluctuations in baseline levels of the MUC1-CT/p120ctn interaction in primary cells, densitometric quantitation of p120ctn immunoprecipitated by MUC1-CT in Smk- and/or PMIP-treated cells was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold change of three independent experiments. (C) P120ctn immunoprecipitates (IP) were immunoblotted for p120ctn, β-ctn, and MUC1-CT. Equal loading was revealed with IgH. Densitometric quantitation of β-ctn immunoprecipitated by p120ctn in Smk- and/or PMIP-treated cells was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold changes. (D) Immunoblotting of cell lysates for E-cad. GAPDH was used as a loading control. Densitometric quantitation of E-cad in Smk- and/or PMIP-treated cells was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold changes. (B–D) ^*p < 0.05, ^**p < 0.01, Smk- and/or PMIP-treated cells are compared with untreated Ctrl. ^*p < 0.05, ^**p < 0.01, PMIP-treated cells are compared with vehicle control.

Figure 4

Smoke activates intracellular signalling in pseudo-stratified HBE cells. (A) TJ protein zona occludens-1 (ZO-1, green) in pseudo-stratified HBE cells is stained by immunofluorescence after exposure to smoke (Smk) and smoke-free control medium (Ctrl) for 1 h. Nuclei were visualized with DAPI (blue). In Ctrl cells (upper panel), white arrowheads in yellow box-a indicate intact ZO-1. In Smk-exposed cells (lower panel), yellow box-b indicates complete loss of ZO-1 in an apical area. Scale bar = 50 μm. (B) Smoke promotes tyrosine phosphorylation (TyrP) of EGFR. Polarized HBE cells exposed to Smk and Ctrl medium were harvested at the indicated time points. Cell lysates were analysed by western blots probed with phosphotyrosine (4G10) antibody. The blots were stripped and reprobed with anti-EGFR antibody. The 180 kDa band recognized by 4G10 completely overlapped with EGFR. Quantitation of EGFR-TyrP after 1 h Smk was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold change (^*p < 0.05). (C) Smoke activates Src and Jnk while suppressing Akt signalling. Polarized HBE cells exposed to Smk or Ctrl medium were harvested at the indicated time points. Cell lysates were analysed by western blots probed with Tyr⁴¹⁶-phosphorylated Src (Src-P), Thr¹⁸³/Tyr¹⁸⁵-phosphorylated SAPK/Jnk (Jnk-P), and Ser⁴⁷³-phosphorylated Akt (Akt-P). Side-by-side controls for each time point produced equivalent signals up to 8 h and thus only 1 h Ctrl is shown. Equal loading was confirmed using total Src, Jnk, and Akt. Levels of phospho-proteins following 4 h Smk treatment were normalized to their corresponding Ctrl (designated as 1-fold) and graphed as mean ± SEM fold change (^**p < 0.01). (D) Smoke increases MUC1-CT tyrosine phosphorylation (TyrP) but has no effect on serine phosphorylation (SerP). Polarized HBE cells exposed to Smk or Ctrl medium were harvested at the indicated time points. Cell lysates were analysed by western blots probed with phosphotyrosine (PY20+99 and 4G10) and phosphoserine (Z-PS1) antibodies. Blots were stripped and reprobed with MUC1-CT antibody. Bands recognized by PY20+99 and 4G10 overlapped with each other and completely overlapped with MUC1-CT. Side-by-side controls for each time point revealed no change through 8 h; thus, only 1 h Ctrl is shown. Equal loading was confirmed with actin. Densitometric quantitation of MUC1-CT-TyrP (PY20+99), MUC1-CT-TyrP (4G10), and MUC1-CT-SerP (Z-PS1) after 4 h Smk treatment was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM (^*p < 0.05).

Figure 5

Smoke-induced p120ctn/MUC1-CT interaction in polarized HBE cells depends on Jnk signalling. (A–C) Polarized HBE cells were preincubated with SP (10 μM SP600125) or vehicle control (DMSO) for 1 h. The cells were subsequently exposed to smoke (Smk) and smoke-free (Ctrl) medium for 4 h in the presence of SP or DMSO. (A) Immunofluorescent staining of Thr¹⁸³/Tyr¹⁸⁵-phosphorylated SAPK/Jnk (Jnk-P, green) in pseudo-stratified HBE cells. Nuclei were visualized with DAPI (blue). In Smk-exposed cells (Smk panel), Jnk-P was localized on apical membranes (yellow arrowheads), basolateral membranes (white box-a), and junctional areas (white box-b). Scale bar = 50 μm. (B) Cell lysates were analysed by western blot for Thr¹⁸³/Tyr¹⁸⁵-phosphorylated SAPK/Jnk (Jnk-P) and phosphotyrosine (PY20+99). Bands recognized by PY20+99 overlapped with MUC1-CT (MUC1-CT-TyrP). Equal loading was confirmed with Jnk and MUC1-CT. Densitometric analysis of Jnk-P (black columns) and MUC1-CT-TyrP (grey columns) in Smk- and/or SP-treated cells was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold changes. (C) MUC1-CT immunoprecipitates (IP) were immunoblotted for p120ctn, β-ctn, E-cad, and MUC1-CT. Equal loading was shown by IgH. To adjust for the influence of variations in basal p120ctn/MUC1-CT interaction in primary HBE cells, densitometric analysis of p120ctn immunoprecipitated by MUC1-CT in Smk- and/or SP-treated cells was normalized to untreated Ctrl (designated as 1-fold) and reported as mean ± SEM fold change of three independent experiments. (B, C) ^**p < 0.01, Smk- and/or SP-treated cells versus untreated Ctrl. ^**p < 0.01, SP-treated cells versus vehicle control.

Figure 6

Smoke-induced p120ctn/MUC1-CT interaction in polarized HBE cells depends on EGFR/Src signalling. (A–F) Polarized HBE cells were preincubated with AG (1 μM AG1478), PP (12.5 μM PP2), SP (10 μM SP600125) or vehicle control (DMSO) for 1 h. The cells were subsequently exposed to Smk and Ctrl medium for 4 h in the presence of AG, PP, SP or DMSO. (A–C) Cell lysates analysed by western blot were probed for phosphotyrosine (PY20+99 and 4G10), Tyr⁴¹⁶-phosphorylated Src (Src-P), and Thr¹⁸³/Tyr¹⁸⁵-phosphorylated SAPK/Jnk (Jnk-P). The 180 kDa bands recognized by PY20+99 and 4G10 overlapped with each other and completely overlapped with EGFR (EGFR-TyrP). Equal loading was confirmed with EGFR, Src, and Jnk. (D, E) MUC1-CT immunoprecipitates (IP) were immunoblotted for p120ctn and MUC1-CT. Equal loading was shown by IgH. (F) Immunofluorescent staining for p120ctn (green) and MUC1-CT (red) in polarized HBE cells exposed to Ctrl and Smk medium for 4 h in the presence and absence of inhibitors (SP, AG or PP). Cell nuclei were counterstained with DAPI (blue). Scale bar = 50 μm.

Figure 7

Schematic representation of the MUC1-CT/p120ctn interaction in regulating smoke-induced disruption of adherens junctions (AJs). In polarized HBE cells, MUC1 is localized on the apical plasma membrane (PM) of normal HBE cells and spatially segregated from AJs on the basolateral PM. Cigarette smoke (Smk) disrupts tight junctions (TJs) to increase paracellular permeability and thereby activate EGFR localized on the basolateral PM. The cytoplasmic tail of MUC1 (MUC1-CT) is tyrosine-phosphorylated (Y-P) by Smk-activated EGFR/Src/Jnk signalling but is not serine-phosphorylated (S). Activation of MUC1-CT through Y-P promotes its interaction with p120ctn. Through this interaction, MUC1-CT competes with E-cad for p120ctn binding and junctional p120ctn/E-cad/β-ctn complexes are disrupted. Within 4 h of Smk-induced AJ disruption, E-cad is internalized and degraded. Within approximately 48 h of smoke exposure, cytoplasmic MUC1-CT/β-ctn complexes are formed and subsequently shuttled into the nucleus. Blocking the interaction between MUC1-CT and p120ctn using a MUC1 inhibitory peptide (PMIP) or inhibitors directed against EGFR (AG), Src (PP) or Jnk (SP) signalling prevents the loss of E-cad and restores AJ integrity. Thus, blocking the interaction between MUC1-CT and p120ctn may provide a novel therapeutic approach to maintain AJ integrity and prevent the initiation of an EMT-like process in airway epithelial cells exposed to smoke.