Expression of Alternatively-Spliced Human Tissue Factor, but Not Full-Length Tissue Factor, Enhances Tumor Growth and Tumor-Associated Vasculature.

It is well-established that cancer is associated with activation of the blood coagulation system, with associated thrombosis as a major cause of morbidity and mortality. Increased expression of Tissue Factor (TF) by cancer cells correlates with a more aggressive grade and clinical course. It is widely presumed that activation of coagulation facilitates cancer growth, and in mouse models, anticoagulation can reduce development of lung metastases. Yet primary tumors are not reduced in a fibrinogen knock-out mouse host, and most importantly, anticoagulation has not been shown to reduce tumor growth in cancer patients. We therefore studied the effect of expression of full-length Tissue Factor (FLTF) and alternatively-spliced human Tissue Factor (asHTF) in a mouse model of human pancreatic cancer. Due to the loss of exon 5, asHTF has a truncated extracellular domain with incomplete procoagulant activity. And due to a frame shift, exon 6 does not code for the transmembrane domain and cytoplasmic tail of FLTF, but codes for a novel peptide sequence. asHTF is soluble and of unknown function. We show that 5 of 6 human pancreatic cancer cell lines tested expressed both FLTF as well as asHTF. The MiaPaca-2 line did not express detectable mRNA or protein of either TF isoform. We generated mammalian expression vectors for both FLTF and asHTF, and established Miapaca-2 clones, stably expressing FLTF, asHTF, or control clones with an empty vector. As anticipated, conditioned media from all FLTF clones shortened the whole blood clotting times by approximately 75%. Conditioned media from control cells and asHTF expressing cells had no effect on clotting times. To evaluate the effect of the TF isoforms on primary tumor growth, 5 X 10⁶ cells from three independent clones of stably transfected clones of FLTF, asHTF, or control clones were injected into the flanks of nude mice (4 mice per clone). At 31 days, the mice were sacrificed and tumor mass measured. Tumors grew in 10 of 12 control mice, but were small (mean tumors 90 mg, SEM 21 mg). Interestingly, FLTF was associated with reduced primary tumor growth; only 4 of 12 developed measurable tumors (mean tumors 10 mg, SEM 4 mg, p = 0.002). In contrast, asHTF expression was associated with enhanced tumor growth; 12 of 12 animals developed tumors (mean tumors 390 mg, SEM 102 mg, p=0.018). In animals with asHTF expressing tumors, circulating asHTF protein was observed in the plasma. The asHTF tumors had increased vascular density compared with controls, suggesting a role of asHTF promoting angiogenesis. In contrast to the prevailing paradigm, our data suggest that FLTF, with procoagulant activity, not only fails to promote primary tumor growth, but may actually inhibit tumor growth. In contrast, asHTF, may be the more important TF isoform in the enhancement of tumor growth.