Generation of Transgenic Mice for Platelet-Specific Expression of Human GPVI From GPVI-Knockout Mice

Abstract 1139

Transgenic mice with human gene knock-in (KI) in platelets can be very useful tools for the evaluation of anti-thrombotic therapeutics in vivo, as many murine thrombosis models have been established and well-characterized. Previously we have successfully generated the GPVI-knockout (KO) mice (S Lockyer et al. Thromb. Res. 2006, 118: 371–80), and used the KO mice for the development of potent anti-human GPVI (hGPVI)-specific monoclonal antibodies (mAb) (Y Matsumoto et al. Thromb. Res. 2007, 119: 319–29), one of which (OM-2) was humanized for further clinical development. As the mAbs only recognize human and monkey GPVI but not murine counterpart, there is a need to develop hGPVI-expressing murine model for the evaluation of these mAbs in vivo. In the present study, we further introduced the hGPVI gene under the control of platelet-specific promoters into the fully back-crossed GPVI-KO C57BL/6 mice, by perivitelline space microinjection with lentivirus. In order to achieve platelet-specific expression, we cloned the promoters from human GPIIb, rat platelet factor 4 and murine GPIbα, confirmed their specific promoter activity in megakaryocytes and used in lentiviral vector construction for the generation of KI mice. Full length cDNA encoding human GPVI was cloned from platelet and put downstream of these promoters in the lentivirus. One-day old fertilized eggs from GPVI-KO mice were isolated. Concentrated lentiviruses were injected into the eggs with an automated Eppendorf micro-injection apparatus. Next day, embryos at two-cell stage were implanted into the uterus of pseudo-pregnant CD-1 mice. Six weeks after delivery, blood was collected for the examination of hGPVI expression in platelets and other cells with flow cyotmetry after immunofluorescent staining with FITC-conjugated OM-2. Results: We demonstrated the expression of hGPVI on the surface of platelets from the hGPVI-KI transgenic mice by immunofluorescent staining with OM-2 and with GPVI-specific FITC-convulxin. No significant difference was observed in platelet counts and other blood cells in these mice. It was further shown that the platelets from hGPVI-KI mice aggregated in platelet-rich plasma (PRP) and whole blood with collagen stimulation concentration-dependently and was inhibited by OM-2 Fab. The platelets from hGPVI-KI mice also adhered to fibrillar collagen-coated plate under both static and flow condition, similar to wild-type platelets, and was inhibited by OM-2 Fab. Pulmonary thromboembolism induced by co-injection of collagen and epinephrine was tested with the mice. GPVI-KO mice survived the pulmonary thromboembolism challenge due to lack of GPVI in platelets, but 9 out of 10 hGPVI-KI mice died after collagen/epinephrine injection. However the death of the KI mice (10 out of 10) were prevented by OM-2 Fab (0.2 mg/kg body weight) pre-injection. These data indicate that the expressed hGPVI in the platelets of the transgenic mice is fully functional, and the platelets activate to collagen similar to the platelets from wild-type mice. Using FITC-conjugated convulxin, we found the expression levels of hGPVI in the platelets from the hGPVI-KI transgenic mice were comparable to wild-type platelets. It is therefore concluded that GPVI-mediated platelet function is rescued in these hGPVI-KI transgenic mice from the GPVI-KO mice which manifest the lack of platelet response to collagen in aggregation and adhesion studies. These hGPVI-KI transgenic mice have been further used successfully in the evaluation of the role of GPVI in the development of rheumatoid arthritis (RA) and of the therapeutic effect of OM-2 in RA, and would be very useful and serve as cheap animal models for the evaluation of other anti-hGPVI therapeutics (such as for the reduction of cardiac infarct size after ischemic injury and reperfusion and prevention of cancer cell metastasis), and for pre-clinical toxicity and dose-finding studies. To our best knowledge, this is the first time of reported success of functional replacement of murine GPVI with its human homologue in mice.