Studies on CAR and CLMP : two proteins of epithelial tight junctions

Abstract

Tight junctions are important structures for the function of epithelial cells. They are composed of multi-protein complexes that connect the plasma membranes of two adjacent epithelial cells to each other and function as paracellular barriers, which regulate flux of ions, solutes, proteins and cells across the epithelium. Classical transmembrane components of tight junctions include the tetra- membrane- spann i ng proteins claudins and occludin, which are involved in the formation of these junction complexes.

In this thesis, a newly identified group of tight junction proteins is described. These proteins belong to the cortical thymocyte marker in Xenopus (CTX)-subfamily, which belongs to the larger immunoglobulin superfamily of cell adhesion molecules. All CTX-like proteins are type-1 single spanning transmembrane proteins that are composed of two extracellular immunoglobulin loops and an intracellular tail containing a PDZ-domain binding motif, which is used for interactions with PDZ-domain containing proteins that serve as scaffolds, anchoring the transmembrane components of tight junctions to the cytoskeleton.

CTX-like proteins mediate cell-cell adhesion and have the capacity to increase transepithelial electrical resistance (TER) across epithelial cells, which suggests a role in barrier function. One member of the CTX-family has been of particular interest: the coxsackie and adenovirus receptor (CAR). CAR was originally identified as a virus receptor and has, as such, been in focus in various adenovirus-based gene transfer and gene therapy studies. Its in vitro properties as being a cell adhesion molecule, expressed at epithelial tight junctions, have also been analyzed. The physiological role of CAR in vivo, however, is less clear.

The major aims of these studies were to determine the function of CAR and a newly identified member of the CTX-family, CAR-like membrane protein (CLMP), in adult tissues and during embryonic development. The ligand-of-numb protein-X (LNX) was identified as a novel interacting partner of CAR and co-localized with CAR at cell-cell contacts in epithelial cells. Moreover, CAR was essential for the recruitment of LNX to this sub-cellular localization. In the adult mouse, CAR was predominantly expressed in epithelial cells, where it localized to tight junctions. The intracellular tail of CAR was essential for correct tight junction localization in vivo.

A positive correlation between CAR expression and tight junction maturity was found in epithelial cells lining different segments of the tubules in the kidneys. This suggested a role for CAR in regulating permeability and tissue homeostasis, which was further supported by results from studies of the effect of CAR knockdown during zebrafish development. Fish embryos lacking CAR expression in all cells and tissues developed abnormalities that were associated with kidney failure, such as pericardial and body edema and formation of renal cysts. These studies showed that CAR is essential for proper kidney development and function and thus uncovered a novel function for CAR, which had not been detected in mouse embryos lacking CAR expression.

CLMP, which was identified through computer-based database searches, was similarly to CAR, expressed in epithelial cells, where it localized to tight junctions. Functional studies showed that CLMP could mediate cell-cell adhesion and increase TER across epithelial cells in culture.

In conclusion, the findings presented in this thesis expand our knowledge of the physiological function of CAR and give new insights into the complexity of the CTX-subfamily group of proteins.