Protein S (PS) homozygous deficiency causes purpura fulminans (PF) thought to result from the imbalance between procoagulant and anticoagulant factors.PS shares similarities with Growth Arrest Specific gene 6 (Gas6) but Gas6-/- mice display an antithrombotic phenotype whereas complete PS deficient (ProS1-/-) mice have a thrombotic phenotype. Beyond their role in coagulation both Gas6 and PS have cellular functions through tyrosine kinase receptors TAM. Gas6 plays also a role in erythropoiesis and previous data suggest that ProS1-/- embryos are anemic.

To mimic severe acquired PS deficiency and thus monitor PF development in adult mice, we induced Pros1 gene silencing via poly I:C-inducible Mx1-Cre+ mice. Once induced, ProS1lox/loxMx1 Cre+ and ProS1lox/- Mx1Cre+ mice showed a marked reduction of plasma PS levels compared with ProS1lox/lox mice (47.7±6% ProS1lox/loxMx1 Cre+; 14.5±7% ProS1lox/-Mx1 Cre+, p<0.001). Although spontaneous thromboses were found in lungs and liver sections from both ProS1lox/loxMx1 Cre+ and ProS1lox/- Mx1Cre+ mice, no PF lesions were observed within a 2-month observation period, suggesting that circulating PS level was still too high to induce PF.

As a second strategy to reduce PS, we treated ProS1+/- and ProS1+/+ adult mice with warfarin (0.8mg/day, during 5 days), and the treatment induced 75% mortality in ProS1+/- and 5% in WT mice (p<0.001). Interestingly, only a few ProS1+/- mice developed lesions compatible with PF. Early lesions showed vascular engorgement, intradermal edema and rare intravascular thrombosis. In advanced lesions, we noticed massive red blood cell (RBC) extravasation, several intra-epidermal hemorrhagic blisters and necrotic areas.

Because Gas6-/- mice are protected against thrombosis, our third strategy was to generate a mouse model combining ProS1 and Gas6 deficiency to overcome the ProS1-/- lethal phenotype. Surprisingly, embryonic mortality, due to massive thrombosis with fibrin deposition, occurred earlier in ProS1-/- Gas6-/- embryosthan in ProS1-/- embryos.

RBC extravasation observed during warfarin treatment pointed to a vascular wall involvement during PF. Consequently, we investigated the vasculature using E15 embryonic dorsal skin. Staining with anti-VE-Cadherin and anti-Ter119 antibodies (Ab) confirmed RBC extravasation while anti-CD31 Ab showed areas with underdeveloped and less dense vascular networks, collapsed vessels and less vessels branch points in ProS1-/- andProS1-/-Gas6-/- than in WT mice. Massively enlarged lymphatic vessels and increased macrophages infiltration were observed in ProS1-/- andProS1-/-Gas6 -/- embryos by anti-Lyve1 and F4/80 Ab.

Histology of blood vessels and liver from ProS1-/- and ProS1-/- Gas6-/- embryos revealed a high number of circulating immature RBC compatible with increased erythropoiesis because of severe bleeding due to consumption coagulopathy and vascular damage. In addition, numerous isolated erythroblast nuclei were found in these embryos indicating altered phagocytosis. Moreover, preliminary data showed higher cytokine levels in ProS1-/- than in WT embryos. Exclusively in ProS1-/- Gas6-/-, we found a very high level of hypochromic erythrocytes suggesting troubled iron metabolism linked to inflammation and iron recycling.

To investigate potential erythropoietic defects, fetal liver (E14.5) single-cell suspensions were cultured for colony forming assays. No difference was found in BFU-E colonies from ProS1-/- and WT embryos. In contrast, BFU-E colonies were 2-fold reduced in ProS1-/-Gas6-/- compared toProS1+/+ Gas6-/- embryos. These data were consistent with our previous observations on the role of Gas6 in erythropoiesis. Both ProS1-/- and ProS1-/-Gas6-/- displayed 50% less CFU-E colonies than WT embryos, pointing to a blunted response to erythropoietin (EPO) in the context of inflammation promoted by PF. These results were confirmed by flow cytometry.

In conclusion, with 14.5% plasmaticPS, no PF lesions were observed in adult mice. Gas6 deficiency did not overcome lethality in mice homozygous for PS deficiency. PF developed during warfarin treatment and in ProS1-/- andProS1-/-Gas6-/- embryos which showed thrombosis occurring together with vascular wall damage and inflammatory processes that appeared to impact iron metabolism and response to EPO.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution