Abstract
Although it has been generally accepted that hypercoagulability contributes to enhancing tumor growth via generation of thrombin (Cancer Cell10:355, 2006), it has not been rigorously proven, nor has the mechanism been established at the cell cycle level. Previous studies have employed thrombin-treated tumor cell lines in vitro and in vivo. In vitro studies were performed in the presence of serum which contains a panoply of growth factors. In vivo studies have used huge non-pathologic concentrations of tumor cells injected into the flank, organ or blood of a mouse. In these situations, tumor growth could be a result of thrombin-induced angiogenesis. We therefore employed a transgenic mouse prostate cancer model (TRAMP) programmed to develop prostate CA over a period of 140–175 days. We treated these animals with thrombin to induce hypercoagulability or hirudin to inhibit endogenous thrombin production, to determine whether thrombin regulates this process independent of angiogenesis. Repetitive thrombin injection enhanced prostate tumor volume 6–8 fold (p<0.04). Repetitive hirudin decreased tumor volume 13–24 fold (p<0.04) via its effect on generated endogenous thrombin, n=6. Thrombin enhanced the production of several vascular growth factors and receptors 2.5 – 3 fold in the liver (VEGF, KDR, ANG-2, Tie2, GRO-1, CD31) and enhanced angiogenesis in the liver, n=3–4. Thrombin had no effect on tumor angiogenesis. Thus, the thrombin-induced spontaneous tumor growth was independent of angiogenesis. We next turned our attention to cell cycle regulators in serum-starved (72 hr) Go-synchronized LNcap prostate CA cells, employing Brdu and Propidium iodide staining. Addition of thrombin (0.5 u/ml) or its PAR-1 receptor agonist, TFLLRN (100 uM) had the same effect as androgen containing serum, inducing cells to leave Go, enter G1 and progress to S-phase. At 8 hrs the number of S-phase cells increased dramatically for both the serum (29 fold) as well as thrombin-treated cells (48 fold), n=3. Similar observations were noted in a Glioblastoma cell line, T98G. We further analyzed the effect of thrombin by performing immunoblots on cell cycle components mediated during cell growth and proliferation. In synchronized Go cells, levels of p27Kip1, a cyclin-dependent kinase inhibitor are high, while levels of cyclins D1 and A, the activation subunits for cyclin-dependent kinases are low. Both thrombin or serum addition led to down-regulation of p27Kip1 with concomitant induction of Skp2, the E3 ubiquitin ligase for p27Kip1. Cyclins D1 and A are induced by similar kinetics, indicating entry into S-phase by 8 hrs. Since p27Kip1 appears to be a rate-limiting down-regulator of the cell cycle (absent with high tumor grade and predicts poor prognosis), we confirmed its role by testing the effect of thrombin or TFLLRN by transfecting p27Kip1 in LNcap cells. This transfection completely prevented the cell cycle stimulation induced by these agonists. A similar approach was used with Skp2 knock down (KD), a negative down-regulator of p27Kip1. KD of Skp2 (over expressed in numerous cancers) completely prevented cell cycle progression induced by thrombin/TFLLRN. MiRNA 222 (upregulated in many cancers) is another down-regulator of p27Kip1. Further analysis following thrombin treatment revealed a robust upregulation at 4 and 8 hrs, providing further proof for the role of thrombin in down-regulating p27Kip1 and stimulating tumor cell entrance into S-phase. Thus, 1) Thrombin enhances spontaneous prostate cell growth in vivo in the absence of enhanced angiogenesis; 2) Thrombin activates the tumor cell cycle by stimulating the down-regulation of p27Kip1 through the upregulation of Skp2 and MiRNA 222.
Tumor quiescence: elevating SOX2 in diverse tumor cell types downregulates a broad spectrum of the cell cycle machinery and inhibits tumor growth
