CAR regulates epithelial cell junction stability through control of E-cadherin trafficking

OPEN SUBJECT AREAS: ADHERENS JUNCTIONS ENDOCYTOSIS CAR regulates epithelial cell junction stability through control of E-cadherin trafficking Penny E. Morton1,2, Alexander Hicks1, Theodoros Nastos1, George Santis1* & Maddy Parsons2* Received 31 July 2013 Accepted 20 September 2013 Published 7 October 2013 Correspondence and requests for materials should be addressed to M.P. (maddy. ) * These authors contributed equally to this work. 1 Division of Asthma, Allergy & Lung Biology, King’s College London, 5th Floor Tower Wing, Guy’s Hospital Campus, London, United Kingdom, 2Randall Division of Cell and Molecular Biophysics, King’s College London, New Hunt’s House, Guys Campus, London, United Kingdom. CAR (Coxsackie and Adenovirus Receptor) is the primary docking receptor for typeB coxsackie viruses and subgroup C adenoviruses. CAR is a member of the JAM family of adhesion receptors and is located to both tight and adherens junctions between epithelial cells where it can assemble adhesive contacts through homodimerisation in trans. However, the role of CAR in controlling epithelial junction dynamics remains poorly understood. Here we demonstrate that levels of CAR in human epithelial cells play a key role in determining epithelial cell adhesion through control of E-cadherin stability at cell-cell junctions. Mechanistically, we show that CAR is phosphorylated within the C-terminus by PKCd and that this in turn controls Src-dependent endocytosis of E-cadherin at cell junctions. This data demonstrates a novel role for CAR in regulating epithelial homeostasis. D ynamic control of epithelial cell-cell junctions is vital to many biological processes including embryonic development, tissue homeostasis, wound healing and inflammation. Incorrect control of epithelial cell-cell junctions has been implicated in many disease processes, including cancer. The integrity of epithelial junctions is maintained by families of transmembrane receptors whose inter-cellular homo- or heterodimerisation form and maintains cell-cell contacts. Control of these adhesions is then mediated by extracellular factors, endocytosis/exocytosis of the receptors and interactions with cytoplasmic binding partners including the cytoskeleton1. One such cell adhesion protein is the coxsackie and adenovirus receptor (CAR) which is a member of the Junctional Adhesion Molecule (JAM) family of junctional adhesion molecules, as well as a viral receptor for type B coxsackievirus and subgroup C adenovirus2. The role of CAR in adenovirus infection has been well studied; however its role in cell adhesion and tissue homeostasis is not well understood. CAR has been shown to localise to both tight and adherens junctions in different epithelial cell types, and to co-precipitate with tight junction proteins in MDCK cells3 and endothelial cells4 suggesting that it has a role in formation and/or maintenance of tight junctions. However recent studies have highlighted tissue specific roles as CAR knockout mice have defects in vascular permeability in the lymphatic system5 and in heart muscle6, but tight junctions in the intestine are unaffected by the loss of CAR6. The role of CAR in other epithelial types, for example respiratory epithelium, has not been well studied despite the gross changes in lung morphology observed in CAR knockout mice6. Our own previous work has shown that CAR expression levels affect recruitment of adherens junction proteins in MCF-7 human breast carcinoma cells without affecting basal paracellular permeability7. Moreover, CAR expression has been shown to correlate with tumour progression in some types of epithelial-derived cancer including lung cancers and in some cases to specifically correlate with reduced E-cadherin expression, a marker for EMT8–11. These data suggest CAR can play important tissue specific roles in the control of cell-cell adhesion. Our previous studies have shown that high levels of CAR can affect recruitment of E-cadherin to cell-cell contacts in MCF-7 cells7. However how CAR controls localisation of E-cadherin, whether this effect is specific to E-cadherin and the functional consequences of this are unknown. E-cadherin localisation to junctions can be controlled through a number of different mechanisms12. At cell-cell contacts E-cadherin forms calcium dependent homodimers in trans with opposing E-cadherin molecules on neighbouring cells. Stabilisation of E-cadherin is in part controlled by interaction to cytoplasmic adaptor proteins; b-catenin and p120-catenin. b-catenin binds to E-cadherin soon after transcription and facilitates trafficking of E-cadherin to the cell membrane. Whereas p120-catenin is thought to join this complex at or adjacent to the membrane and inhibit endocytosis, in addition SCIENTIFIC REPORTS | 3 : 2889 | DOI: 10.1038/srep02889 1 www.nature.com/scientificreports Figure 1 | CAR mediates E-cadherin localisation to epithelial cell junctions and mediates junction stability. (A) Confocal microscopy of E-cadherin localisation in a 50550 mix of WT and CARRFP HBEC. Arrows highlight loss of E-cadherin at CARRFP positive junctions (left), quantification of Ecadherin intensity in monolayers of WT or CARGFP HBEC by wide-field microscopy, with and without calcium (right). (B) Confocal microscopy of Ecadherin localisation in a 50550 mix of WT and CARGFP HBEC, in untreated, buffer alone control and Ad5FK treated cells. Colocalisation of E-cadherin and CARGFP in the presence of Ad5FK is pseudo-coloured yellow. (C) Western blot analysis of wild-type and CAR-GFP HBEC in the presence or absence of calcium probed for E-cadherin and HSC-70. (D) Confocal microscopy of E-cadherin localisation in WT, control shRNA expressing, CAR shRNA expressing HBEC and CAR shRNA HBEC expressing sh-resistant CAR-RFP(arrow highlights and sh-resistant CAR-RFP expressing cell-cell junctions showing reduced E-cadherin). Western blot showing CAR and E-cadherin expression in WT HBEC or HBEC expressing control shRNA or shRNA directed at CAR (right). (E) Quantification of FRAP recovery data of E-cadherin-GFP expressed in wild-type or CAR-RFP HBEC. Histogram shows t1/2 recovery time for E-cadherin-GFP at junctions in wild-type HBEC (n 5 18) and CAR-RFP HBEC (n 5 15). (F) Dissociation of cell-cell contacts in wildtype and CAR GFP HBEC cells upon removal of calcium. Images show phase contrast of wild-type or CAR-GFP HBEC grown in calcium containing media, before and after the media was replaced with calcium free media (for times indicated). Graph shows analysis of junction dissolution quantified as the average time taken for individual cell-cell junctions to dissociate. Data is the mean of at least 100 junctions per data set. Error bars are SEM. * 5 p , 0.05, ** 5 p , 0.01 *** 5 p , 0.005. Scale bars correspond to 10 mm. SCIENTIFIC REPORTS | 3 : 2889 | DOI: 10.1038/srep02889 2 www.nature.com/scientificreports to promoting exocytosis of E-cadherin (reviewed in13,14. Recent studies using fluores (...truncated)