APC is a potent mediator of anticoagulant and anti-inflammatory properties. Clinical studies showed that APC reduces mortality in severe sepsis patients. Animal studies are essential for providing a framework to the assessment of mechanism of complex diseases and to test therapeutic agents. Studies of APC function in vivo models has been limited by its short half-life, lack of data on plasma levels and therapeutic effect, species specific effects of the protein. Here we report novel models to assess APC function in vivo. Both murine (m) and human (h) protein C was first engineered by introducing an extra cleavage site for the intracellular protease-cleavage enzyme (PACE/furin), which results in secretion of active forms of PC. Secondly, to overcome the time-consuming and expensive transgenic technology, we use adeno-associated viral (AAV) to ensure continuous expression of APC forms. Gene transfer was hepatocyte-restricted by using a liver-specific promoter. Following portal vein delivery of vectors, APC were monitored using aPTT-based and species-specific ELISA capture assays. Six weeks post-injection, three cohorts of mice (n=19) expressed hAPC at levels of 80±7, 160±27 and 260±23ng/ml, in a dose-dependent manner, and their respective aPTT values were 24±0.8, 27±1.5 and 31±1.8sec. These aPTT were higher than baseline values (p<0.05). At doses tested, mice injected with mAPC showed functional levels of 7±0.7, 14±2 and 83±2ng/ml, which resulted in aPTT values of 27±1.3, 30±0.2 and 36±1.4sec, respectively and differ from baseline data (p<0.01). Thus, mAPC at levels comparable to the hAPC (80ng/ml) showed prolonged clotting times (36 vs. 24sec, p<0.02). Only by increasing hAPC by 3-fold (260ng/ml), the aPTT values reached those of mAPC. Mice were further challenged by tail clipping assay. Blood loss was increased 2-fold in mice with highest APC levels (260ng/ml, p<0.03) compared to controls. Similar blood loss was observed in mice expressing 3-fold lower mAPC levels (83ng/ml). Anti-thrombotic properties of APC were first tested at microcirculation level by real-time imaging of thrombus formation upon laser-induced injury. In hAPC-treated mice at levels of 80ng/ml, we observed 19 thrombi/29 injury sites (66%), whereas among those with APC levels of 260ng/ml thrombus formation was detected in only 8% (2/24 sites). Interestingly, expression of mAPC at levels of 83ng/ml resulted in substantial reduction of clotting formation (5 thrombi/45 sites, 11%). Exposure of carotid artery to FeCl3 failed to induce vascular occlusion in all hAPC-injected mice (n=9). However, 1/6 mouse expressing mAPC (83ng/ml) showed vessel occlusion, but the remained mice presented either transient clot or no occlusion. In conclusion, both murine and human APC can be efficiently expressed at a range of levels in murine models. Murine APC showed a higher anticoagulant active than human APC at both in vitro and in vivo assays. However, we also demonstrate species-related differences in the protective antithrombotic effects. Murine APC is more effective in the laser-induced injury, characterized by heat damage to a limited region, an injury model likely induced by inflammation. In this latter model, it is likely that mAPC by providing both anticoagulant and antiinflammatory effects have enhanced in vivo functions. Thus, we demonstrate the feasibility and the potential of these models in assessing in vivo effects of APC.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author