TFPI-Beta Is An Effective Inhibitor of TF-FVIIa Mediated Coagulation and Tumor Cell Invasion

Abstract 347

Tissue Factor Pathway Inhibitor (TFPI) is the primary physiological inhibitor of TF-fVIIa, the in vivo activator of blood coagulation. TFPI is an alternatively spliced protein with two major isoforms, TFPI-alpha and TFPI-beta. These isoforms differ in their C-terminal domain structure and their mechanism for cell surface attachment with TFPI-alpha indirectly associating with the endothelial surface through binding to a GPI-floated protein while TFPI-beta is directly attached to a GPI-float. In addition, the isoforms are differentially expressed in mouse tissues, with TFPI-alpha present in placenta, embryo and platelets, while TFPI-beta is the predominant isoform in adult vascular beds. However, TFPI-beta has ∼20-fold decreased activity when compared to TFPI-alpha in solution phase plasma clotting assays. We hypothesized that TFPI-beta will have more physiologically relevant activity when associated with a cell surface instead of in solution phase. To test this hypothesis, a CHO cell system that allows measurement of the anti-TF activity of different cell surface forms of TFPI in both in vitro and in vivo assays was developed. CHO cells were stably transfected with TF creating CHO-TF cells. In contrast to wild type CHO cells, CHO-TF cells 1) generate fXa in vitro in the presence of fVIIa, fX and calcium ions, 2) migrate through matrigel in transwell assays, and 3) produce tumors in the lungs of SCID mice following tail vein injection. The ability of solution phase TFPI-alpha and TFPI-beta to inhibit TF-fVIIa mediated fXa generation on the surface of the CHO-TF cells was examined with amidolytic assays. The Ki(final) for TFPI-alpha and TFPI-beta were 5.81 nM and 21.6 nM, respectively, confirming that solution phase TFPI-beta has decreased inhibitory activity. To examine the activity of cell surface associated TFPI-beta, CHO-TF cells were co-transfected with either TFPI-beta or equal amounts of an altered form of TFPI, called K1K2K3-GPI that is similar to TFPI-alpha but lacks its basic C-terminal region. Forms of TFPI containing the basic C-terminal region with a GPI-float could not be studied because they were not expressed by the CHO cells. CHO-TF cells expressing TFPI-beta had equal inhibitory activity to that observed in CHO-TF cells with K1K2K3-GPI suggesting that TFPI-beta is a potent inhibitor of TF activity when associated with the cell surface. The activity of cell surface associated TFPI-beta was further examined in transwell migration assays. The number of cells migrating through matrigel in multiple 20X fields was averaged. In this assay, cell surface associated TFPI-beta was a potent inhibitor of TF activity. The results are as follows: wild type CHO cells—6.6+/−4.1; CHO-TF cells—98.0+/−32.5; CHO-TF cells with TFPI-beta—9.05+/−7.0; CHO-TF cells with K1K2K3-GPI—43.5+/−21.5. Migration of CHO-TF cells was blocked using argatroban, an active site directed inhibitor of thrombin (15.0+/−4.5 CHO-TF cells migrated in the presence of 100 micromolar argatroban), demonstrating that the likely mechanism for the TF-mediated cell migration is generation of thrombin with cellular activation through cleavage of protease activated receptors. In the SCID tumor model, the severity of lung tumor burden was graded as 1–4 by a pathologist blinded to the cell type injected, n=5-8 per group. Average scores were: wild type CHO cells—1.17; CHO-TF cells 2.80; CHO-TF cells with TFPI-beta—2.14; demonstrating that cell surface associated TFPI-beta is an effective inhibitor of TF activity in vivo. Thus, while TFPI-beta has limited anticoagulant activity when examined in solution phase assays in vitro, evaluation of cell surface associated TFPI-beta reveals that it is a highly effective inhibitor of TF-fVIIa activity both in vitro and in vivo.