Lessons from Transgenic Mice Expressing Supra-Physiological Levels of Activated Murine Factor VII.

Abstract

A gene transfer approach using an engineered activated murine FVII (mVIIa) has previously shown efficacy in correcting the bleeding diathesis of hemophilia B mice. To assess the effects of long-term expression, we generated transgenic mice expressing supra-physiological concentrations of mVIIa (TgVIIa), and divided the mice into high and low expressors (3.1±0.8 and 1.4±0.6 microg/ml), respectively. Both high and low mVIIa levels shortened the aPTT in TgVIIa-HB mice when compared to their HB littermates (18.1±1.6 and 32.4±5.2 vs. 56.6±11.2 sec, P<0.004 and P<0.005, respectively). In the first high expressor transgenic line, all mice died with lifespans of 6 mons, 2 mons and 2 wks for F0, F1 and F2 mice, respectively. This is in contrast to the low expressor mice, which had normal survival based on a Kaplan-Mier curve with a follow up of 15 mons. The early mortality seen in high expressors was accompanied by fibrin deposition and thrombi formation in the hearts and lungs of the high TgVIIa, in addition to intimal thickening in several pulmonary veins. In order to rescue the high expressor mice and to investigate the cause of their early mortality, we established new lines with high levels of mVIIa and crossbred them with our recently generated mice expressing low murine factor X activity (Lo FX). This resulted in an F2 generation of high TgVIIa with factor X activity in the range of 1–5% of normal (TgVIIa-LoFX). The VIIa levels shortened the prolonged PT in TgVIIa-LoFX compared to LoFX littermates (71.8±6.5 vs. 132.6±5.1 sec, P<0.005) and thus far, the TgVIIa-LoFX mice have longer survival compared to hemostatically normal high VIIa expressor littermates. While TgVIIa and TgVIIa-LoFX mice were comparable in terms of mVIIa levels (4.8±1.6 vs. 3.6±1.6 microg/ml, P>0.3) and PT (10.1±0.6 vs. 9.8±0.9 sec, P>0.6) respectively, thrombin-anti thrombin (TAT) levels were significantly lower in TgVIIa mice with low than with normal FX activity (5.7±4.1 vs. 71.8±32.2 ng/ml, P<0.004, respectively), with a dose dependent decrease of TAT levels parallel to the decreased FX activity, and a trend of decreased D-Dimer levels in TgVIIa-LoFX compared to TgVIIa (11.3±3.2 vs. 18.1±12.9, P>0.4, respectively). The TgVIIa mice were also crossed to low tissue factor mice, kindly provided by Nigel Mackman, and demonstrated a trend of decreased TAT levels in the F1 generation of TgVIIa with ~50% of tissue factor (TF) activity compared to their TgVIIa-normal TF littermates (21.4±19.9 vs. 51.0±44.3 ng/ml, P>0.2, respectively). Interestingly, TAT levels were not less in TgVIIa-HB or TgVIIa-Hemophilia A mice than in hemostatically normal TgVIIa (41.8±17.1 and 52.4±19.9 vs. 51.1± 20.4, P>0.4 and P>0.9, respectively). This suggests an efficient thrombin generation with high VIIa levels in spite of the defective intrinsic system in hemophilic A and B mice, whereas LoFX and LoTF hindered the effect of VIIa on thrombin generation. Correspondingly, the litter sizes were increased when breeding the TgVIIa males with LoFX but not with HB females, both compared to breeding with wt females (9.3±1.3 and 4.3±1.3 vs. 4.5±0.7 pups/litter, with P<0.002 and P>0.9, respectively). This observation is relevant to human disease, as a recent report documented elevated VIIa levels in women with recurrent fetal loss. In conclusion, this mouse model demonstrates pathological consequences, including fetal loss and intimal thickening in the pulmonary vasculature as a consequence of overexpressing VIIa, and provides a valuable tool to determine both safe and effective levels of VIIa in hemophilic mice.