Increased Thrombosis in Vivo and Altered Coagulation in Vitro in STAT5AB-Deficient Mice

The risk of thrombosis differs between men and women; however, the mechanisms underlying this sex difference are unknown. We recently reported differences in thrombosis that are dependent on sexually-dimorphic patterns of growth hormone (GH) secretion and its subsequent effects on the expression of coagulation-related genes in the liver. Mice that are deficient in STAT5A and STAT5B are largely insensitive to growth hormone and demonstrate loss or reversal of sex-specific patterns of liver gene expression. Thus, genetic manipulation of STAT5 factors provides a means of probing the mechanism of GH-dependent sex differences in thrombosis. We analyzed mice with global (STAT5g) and liver-specific (STAT5c) disruptions of the Stat5a and Stat5b genes, and determined their susceptibility to thrombosis in vivo and coagulation in vitro. In an in vivo model of pulmonary embolism, control female mice were relatively protected from thrombosis compared to control males, demonstrating percent survival of 61.3% and 50.0% and median survival times of 1200 and 972 seconds, respectively. In contrast, STAT5g males and females showed similar susceptibility to pulmonary embolism and had significantly reduced survival compared to matched controls, with percentages of 12.5% and 11.1% and median survival times of 468 and 275 seconds, respectively. As with the global loss of STAT5AB, liver-specific disruption of STAT5AB resulted in reduced survival compared to control mice. In vitro coagulation studies of whole blood triggered with dilute tissue factor (TF 1:1000) revealed significantly longer clot times in control females compared to control males. In contrast, mice with either global or liver-specific disruption of STAT5AB showed similar whole blood clot times between males and females, but had significantly shorter clot times as compared to controls. Similar results were obtained from coagulation studies in platelet poor plasma, indicating that platelets or white blood cells do not play a major role in this phenotype. No significant differences were found between control and STAT5AB-deficient mice in the expression or activity of a panel of pro- or anti-coagulant factors. APC sensitivity ratios were also similar between control and STAT5g males and females. An in vitro clotting model was employed to measure thrombin-antithrombin (TAT) complexes, fibrinogen, and D-dimer levels over time. Clotting was triggered with a 1:10,000 dilution of TF and analyzed at two minute intervals for ten minutes. Levels of TAT and fibrinogen did not differ between genotypes throughout the time course. In contrast, D-dimer was decreased in STAT5AB-deficient mice at each interval with a 52% reduction present at four minutes and a 46% reduction remaining at ten minutes. These studies have shown that global and liver-specific loss of STAT5AB results in reduced survival in a pulmonary embolism model, in vivo and shorter clotting times, in vitro. Furthermore, STAT5AB -deficiency eliminated the relative protection from thrombosis afforded to control female mice. This observed increase in thrombosis was not due to significant changes in coagulation factor activity, APC resistance, fibrinogen levels, or thrombin generation. However, a major decrease in D-dimer levels suggests a defect in the fibrinolytic system. While future studies are aimed at further defining the mechanistic regulation of the fibrinolytic system by growth hormone and downstream mediators, the current findings demonstrate a novel role for STAT5AB as a significant regulator of sexdependent thrombosis.