Abstract
Tissue factor pathway inhibitor (TFPI) is the major endogenous inhibitor of tissue factor initiated blood coagulation and a is key regulator of the development of intravascular thrombosis. TFPI indirectly binds to the endothelial surface through tight association with a GPI-anchored co-receptor. The location of recombinant TFPI expression in mammalian cells varies depending on the cell line used. In cell lines that do not produce endogenous TFPI, CHO and HEK293 cells, recombinant TFPI is secreted into the culture media. However, in a cell line that produces endogenous surface TFPI, EaHy926 cells, recombinant TFPI is expressed on the cell surface. These data suggest that TFPI is expressed on the surface of cells that produce the GPI-anchored co-receptor and is secreted by cells that do not. To further investigate the function of the co-receptor in TFPI cellular trafficking we developed aerolysin resistant ECV304 and EaHy926 cells lines. Both of these cell lines produce endogenous, surface associated TFPI. The cell lines were mutated with ethyl methanesulfonate and selected with aerolysin. Mutant cells lacking surface GPI-anchored proteins are resistant to the toxic effects of aerolysin and survive. The morphology and growth rate of the two aerolysin resistant cell lines are identical to that of the wild-type cells. They were first characterized to rule out the presence of a mutation that could directly alter cellular metabolism of TFPI. Sequencing of TFPI cDNA indicates that no mutations are present in the TFPI exons. Analysis of mRNA by real time PCR demonstrates that the aerolysin resistant cells make similar amounts of TFPI mRNA as their wild-type counterparts. Thus, transcription and translation of TFPI appear identical to the wild-type cells. In addition, the two independently derived cell lines have very similar phenotypes, as described below, indicating that the aerolysin resistant cell lines have a defect in GPI-anchor biosynthesis but not additional random mutations that could alter cellular processing of TFPI. Characterization of protein expression by flow cytometry indicates that the aerolysin resistant cell lines do not express GPI-anchored proteins (CD59, uPAR) or TFPI on the cell surface but do have wild-type surface expression of transmembrane proteins (CD9, tissue factor). Interestingly, instead of being secreted, western blot analysis of cellular lysates indicates that TFPI is degraded within the aerolysin resistant cells in a manner similar to that observed for GPI-anchored proteins. Intracellular degradation of TFPI is prevented by brefeldin A indicating that degradation takes place in a post endoplasmic reticulum compartment. Pepstatin A, but not MG-132, also prevents degradation, indicating that degradation is lysosomal rather than proteosomal. It appears that binding of TFPI to its co-receptor occurs early in cellular processing, likely within the endoplasmic reticulum. Cellular trafficking of TFPI is controlled by its co-receptor, which has not yet been identified. The co-receptor directs TFPI to the cell surface in wild-type endothelial cells or to be degraded in aerolysin resistant cells. In the absence if its co-receptor TFPI is secreted. Therefore, regulation of co-receptor expression provides a mechanism for the production of cell associated TFPI, as occurs in endothelial cells, versus soluble TFPI, as may occur in megakaryocytes/platelets.
Author notes
Corresponding author