Late domain dependent E-cadherin recruitment into extracellular vesicles

Abstract

E-cadherin, a transmembrane protein involved in epithelial cell-cell adhesion and signaling, is found in exosomal fractions isolated from human body fluids. A cellular mechanism for recruitment of E-cadherin into extracellular vesicles (EVs) has not yet been defined. Here, we show that E-cadherin is incorporated into the membrane of EVs with the extracellular domain exposed at the vesicle surface. This recruitment depends on the endosomal sorting complex required for transport I (ESCRT-I) component Tsg101 and a highly conserved tetrapeptide P(S/T)AP late domain motif in the cytoplasmic tail of E-cadherin that mediates interaction with Tsg101. Mutation of this motif results in a loss of interaction and a dramatic decrease in exosomal E-cadherin secretion. We conclude, that the process of late domain mediated exosomal recruitment is exerted by this endogenous non-ESCRT transmembrane protein.

Keywords: E-cadherin; ESCRT (endosomal sorting complex required for transport); exosomes; extracellular vesicles; late domain; multivesicular bodies (MVB).

FIGURE 1

The cytoplasmic tail of E-cadherin comprises a highly conserved PTAP-late domain motif. (A) Pie chart showing 242 primary polypeptide sequences of the UniProtKB/Swiss-Prot database that contain a P(S/T)AP late domain motif and are listed in ExoCarta (see also Supplementary Table). 624 late domain containing sequences from the UniProtKB/Swiss-Prot database are not listed in ExoCarta. (B) Domain structure of E-cadherin. S, signaling sequence; pro, pro-domain; EC, E-cadherin domain; TM, transmembrane domain; CD, cytoplasmic domain. The location of the PTAP late domain motif is indicated. (C) Alignment of E-cadherin late domain-like motifs found in 32 vertebrates was used to generate a sequence logo of the decapeptide (765)DTPTAPPYD(773).

FIGURE 2

E-cadherin colocalizes with Hrs in MDCK cells. Confocal codistribution analysis of immunostained E-cadherin with the MVE-protein Hrs (A) or mitochondrial outer membrane protein TOM20 (B). MDCK cells were cultivated for 2 days, fixed and stained by immunofluorescence with mAb anti-E-cadherin/Alexa Fluor 555 and pAb anti-Hrs/Alexa Fluor 647 or pAb anti-TOM20/Alexa Fluor 647. Nuclei are depicted in blue. Co-stained vesicular structures are indicated by arrows. Arrowheads point at punctate E-cadherin-positive structures that are not co-stained with Hrs or TOM20. (C) Manders’ correlation coefficient was used for quantification of experiments. Means ± s.e.m., 15–20 cells per experiment, n = 3 independent experiments. Nuclei were excluded from quantification. Statistical analysis in this figure: Student’s unpaired t-test, ***p < 0.001.

FIGURE 3

Identification of E-cadherin on isolated EVs. (A) Western blotting analysis of EV fractions and cell lysates. Representative results, n = 3 independent experiments. EVs and the corresponding cell lysates were analyzed by immunoblot with antibodies directed against E-cadherin and Tsg101. Antibodies directed against giantin (Golgi), PDI (ER) and TOM20 (mitochondria) were used as negative controls to validate the purity of EV-isolation. EVs isolated from MDCK_CD63-GFP cells were immunoblotted with anti-GFP antibodies and used as positive controls. (B) Proteinase protection assay. Antibodies directed against the cytoplasmic tail of E-cadherin were used for E-cadherin staining. Antibodies directed against Tsg101 and GFP were used as indicated. Representative results, n = 3 independent experiments. (C) Schematic drawing of the orientation of E-cadherin in the exosomal membrane. Schematic drawing of the orientation of E-cadherin in the membrane of an extracellular vesicle (EV). The extracellular part of the protein is shown in blue, and the part directed to the interior of the vesicle is shown in yellow.

FIGURE 4

Tsg-101 dependent EV recruitment of E-cadherin. (A) Co-immunoprecipitation of E-cadherin-GFP or Tsg101-GFP from MDCK cell lysates. E-cadherin-GFP or Tsg101-GFP fusion proteins and their binding partners were immunoprecipitated with GFP-nanobody beads and detected by immunoblot using antibodies directed against E-cadherin, GFP and Tsg101. Non-specific precipitation was monitored by using control beads without nanobodies (neg. control) and GFP from MDCK_GFP cells (GFP-contr.). Antibodies used for western blots (WB) are indicated. Representative results, n = 3 independent experiments. (B) Tsg101 knockdown in MDCK cells and the corresponding EVs isolated from cell culture media. EVs and the corresponding cell lysates were analyzed by immunoblot using antibodies directed against E-cadherin and Tsg101. Antibodies directed against PDI were used as negative controls to validate the purity of EV-isolation. Actin was monitored as EV cargo molecule. (C) Efficient knockdown through Tsg101 siRNA administration was verified by quantification of Tsg101 in cell lysates from 3 independent experiments. Tsg101 quantities were normalized to GAPDH. Means ± s.e.m. (D) Quantification of the immunoblot analysis of the exosomal fraction after Tsg101 knockdown in MDCK cells as in (C). E-cadherin was significantly reduced in Tsg101 siRNA-treated cells. Normalized to the E-cadherin quantities in the respective cell lysates. Means ± s.e.m., n = 3 independent experiments. Statistical analysis in this figure: Student’s unpaired t-test, ***p < 0.001; **p < 0.005; *p < 0.01.

FIGURE 5

Subcellular localization and EV recruitment of E-cadherin variants. (A) Schematic drawing of the two E-cadherin deletion constructs. S, signaling sequence; pro, pro-domain; EC, E-cadherin domain; TM, transmembrane domain; CD, cytoplasmic domain. The location of the PTAP late domain motif and the mutagenized ASAA stretch are indicated. (B) Confocal codistribution analysis of MDCK cells transiently transfected with E-cadherinΔE2-5GFP or E-cadherinΔE2-5GFP_ASAA with the MVE-proteins Hrs and Tsg101. The cells were cultivated for 2 days post transfection, fixed and stained by immunofluorescence with mAb anti-Tsg101/Alexa Fluor 555 and pAb anti-Hrs/Alexa Fluor 647. Enlarged views of areas encircled by dotted lines are depicted on the right. Nuclei are depicted in cyan. Vesicular structures co-stained for E-cadherin, Hrs, and Tsg101 are indicated by arrows. (C) Colocalization of the two E-cadherinΔE2-5GFP variants and Tsg101 was estimated from at least 15 cells per experiment via the Manders’ colocalization coefficient. Data are represented as means ± s.e.m., significance was tested with Student’s t-test (***p < 0.001), n = 3 independent experiments. (D) EVs isolated from MDCK_{E-cadherinΔE2-5GFPPTAP} (PTAP) and MDCK_{E-cadherinΔE2-5GFPASAA} (ASAA) culture media and the corresponding cell lysates were analyzed by immunoblot using antibodies directed against E-cadherin and Tsg101. Antibodies directed against PDI and TOM20 were used as negative controls to validate the purity of EV-isolation. (E) Co-immunoprecipitation of E-cadherin variants with Tsg101 from MDCK_{E-cadherinΔE2-5GFPPTAP} (PTAP) and MDCK_{E-cadherinΔE2-5GFPASAA} (ASAA) cell lysates. E-cadherinΔE2-5GFP or E-cadherinΔE2-5GFP_ASAA and their binding partners were immunoprecipitated with GFP-nanobody beads and detected by immunoblot using antibodies directed against GFP or Tsg101. Non-specific precipitation was monitored by using control beads without nanobodies (neg. control) and GFP from MDCK_GFP cells (GFP-contr.).

FIGURE 6

Single-particle phenotyping of EVs derived from MDCK_Gal3-GFP cells. Representative plots of galectin-3-GFP and E-cadherin (Ecad) expression on single EVs using the GFP signal and PE-conjugated antibodies by nano-flow cytometry (nFCM).Bivariate dot-plots of indicated fluorescence versus side scatter (SS-A). In addition, EVs harboring a CD63-GFP or fluorescently labeled with FITC-conjugated antibodies specific to CD9 were stained with PE-conjugated antibodies specific to Ecad. For E-cadherin-detection on vesicle surfaces, the Genetex/Biozol antibody GTX134997 was used. Double positives for CD9/Ecad and CD63/Ecad are depicted. Fluorescently labeled IgG isotypes were used as a control. Numbers indicate events detected in the corresponding gate in percent of total events.