Characterization of TFPI and TFPI-beta in Mouse Tissues.

Tissue factor pathway inhibitor (TFPI) is a trivalent Kunitz-type serine protease inhibitor. The first two Kunitz domains efficiently inhibit factor VIIa and factor Xa and thereby abrogate tissue factor initiated blood coagulation. Two isoforms of TFPI have been identified. Full-length TFPI is thought to indirectly associate with endothelium via binding to a GPI-anchored co-receptor. TFPI-beta is produced by alternative splicing of full-length TFPI. It consists of the first two Kunitz domains followed by a unique C-terminal region with a direct GPI-anchor attachment site. The physiological significance of TFPI-beta is not known. We performed 3′RACE of mouse placental cDNA to determine if additional isoforms of TFPI exist. These studies identified only TFPI and TFPI-beta. Of importance, a full-length form of TFPI containing a direct GPI-anchor attachment site was not identified. This finding supports the hypothesis that full-length TFPI binds to endothelium indirectly through association with a GPI-anchored co-receptor. In order to identify potential physiological roles for TFPI and TFPI-beta, we analyzed mRNA expression and protein production in mouse tissues. Message for both TFPI and TFPI-beta was present in all tissues examined. Real time PCR showed TFPI message 2.5- to 65-fold higher than TFPI-beta message in the different tissues (heart (2.5)>lung (5.1)>testicle (5.8)>spleen (9.3)>thymus (10)>uterus (11)>kidney (26)> liver (35)> placenta (37)>brain (65)). The high relative amount of TFPI-beta in heart, lung and testicle suggest that it may have an important physiological role in these tissues. In situ hybridization studies did not identify vascular bed specific expression of TFPI or TFPI-beta. These studies demonstrated that TFPI and TFPI-beta localize predominantly within endothelial cells with identical staining patterns in all tissues studied. Western blot analysis for TFPI and TFPI-beta in detergent lysates of the different tissues demonstrated the presence of full-length TFPI in all tissues in amounts proportional to the vascularization of the tissue (placenta>>heart>lung>spleen>kidney>liver>>>brain).

TFPI-beta was not observed by western blot analysis in any of the tissues indicating that TFPI-beta message may be transcriptionally repressed. Brain contains much less TFPI than other tissues, yet has high amounts of tissue factor. The low TFPI in the brain could be due to either low numbers of endothelial cells within the brain or to low expression of TFPI by brain endothelial cells. We used real time PCR to compare the level of TFPI and TFPI-beta mRNA expression to that of the endothelial specific proteins VE-cadherin and PECAM. When compared to these markers, the relative amount of TFPI mRNA brain is similar to that present in the heart and lung, indicating that TFPI likely has an important role in prevention of thrombosis within the cerebral vasculature.