First body of evidence suggesting a role of a tankyrase-binding motif (TBM) of vinculin (VCL) in epithelial cells

Abstract

Background

Adherens junctions (AJ) are involved in cancer, infections and neurodegeneration. Still, their composition has not been completely disclosed. Poly(ADP-ribose) polymerases (PARPs) catalyze the synthesis of poly(ADP-ribose) (PAR) as a posttranslational modification. Four PARPs synthesize PAR, namely PARP-1/2 and Tankyrase-1/2 (TNKS). In the epithelial belt, AJ are accompanied by a PAR belt and a subcortical F-actin ring. F-actin depolymerization alters the AJ and PAR belts while PARP inhibitors prevent the assembly of the AJ belt and cortical actin. We wondered which PARP synthesizes the belt and which is the PARylation target protein. Vinculin (VCL) participates in the anchorage of F-actin to the AJ, regulating its functions, and colocalized with the PAR belt. TNKS has been formerly involved in the assembly of epithelial cell junctions.

Hypothesis

TNKS poly(ADP-ribosylates) (PARylates) epithelial belt VCL, affecting its functions in AJ, including cell shape maintenance.

Materials and Methods

Tankyrase-binding motif (TBM) sequences in hVCL gene were identified and VCL sequences from various vertebrates, Drosophila melanogaster and Caenorhabditis elegans were aligned and compared. Plasma membrane-associated PAR was tested by immunocytofluorescence (ICF) and subcellular fractionation in Vero cells while TNKS role in this structure and cell junction assembly was evaluated using specific inhibitors. The identity of the PARylated proteins was tested by affinity precipitation with PAR-binding reagent followed by western blots. Finally, MCF-7 human breast cancer epithelial cells were subjected to transfection with Tol2-plasmids, carrying a dicistronic expression sequence including Gallus gallus wt VCL (Tol-2-GgVCL), or the same VCL gene with a point mutation in TBM-II (Tol2-GgVCL/*TBM) under the control of a β-actin promoter, plus green fluorescent protein following an internal ribosome entry site (IRES-GFP) to allow the identification of transfected cells without modifying the transfected protein of interest.

Results and discussion

In this work, some of the hypothesis predictions have been tested. We have demonstrated that: (1) VCL TBMs were conserved in vertebrate evolution while absent in C. elegans; (2) TNKS inhibitors disrupted the PAR belt synthesis, while PAR and an endogenous TNKS pool were associated to the plasma membrane; (3) a VCL pool was covalently PARylated; (4) transfection of MCF-7 cells leading to overexpression of Gg-VCL/*TBM induced mesenchymal-like cell shape changes. This last point deserves further investigation, bypassing the limits of our transient transfection and overexpression system. In fact, a 5 ^th testable prediction would be that a single point mutation in VCL TBM-II under endogenous expression control would induce an epithelial to mesenchymal transition (EMT). To check this, a CRISPR/Cas9 substitution approach followed by migration, invasion, gene expression and chemo-resistance assays should be performed.