A Syngeneic Murine Model of Antiphospholipid Syndrome (APS) Recapitulates Aspects of Human APS

Background

Antiphospholipid syndrome (APS) is characterized by arterial or venous thrombosis, or recurrent pregnancy loss, in the presence of persistently positive tests for antiphospholipid antibodies (aPL). aPL include lupus anticoagulants (LAC), anti-β2-glycoprotein 1 (β2GPI) and anti-cardiolipin antibodies (aCL). Among them, antibodies directed toward β2GPI have been best characterized and shown in most studies to have a significant association with thrombosis. Most murine models of APS are generated by infusion of either polyclonal IgG isolated from APS patients, or monoclonal IgG directed against human or murine β2GPI. However, one or more of the following limitations may limit their clinical relevance: 1) Unlike chronically elevated autologous aPL in APS patients, cross-species IgG infused into mice circulate for only limited periods of time and may provoke allogeneic responses; 2) Monoclonal human or murine anti-β2GPI IgG, unlike polyclonal aPL from patients, tend to be high affinity, monospecific, and may recognize epitopes different than those recognized by clinical aPL. Hence, we sought to develop a relevant syngeneic murine model of APS that employed polyclonal murine anti-β2GPI IgG that would circulate in mice for extended periods, and not induce allogeneic inflammatory responses.

Methods

β2GPI-deficient (apoh^-/-) mice were previously developed in our lab on a C57/BL6 background using CRISPR/cas9. IgG isolated from apoh^-/- mice immunized with human plasma β2GPI was used as a source of polyclonal murine aPL IgG, while IgG isolated from apoh^-/- mice treated with adjuvant alone served as control IgG in all experiments. Wild-type (WT) C57/BL6 mice (two groups of 8 mice each) were injected intravenously with 20 µg of either murine polyclonal aPL or control IgG, and after 72 hours were evaluated for venous thrombosis using the IVC stasis model, with thrombi harvested and measured 48 h after IVC ligation. In parallel studies, we injected mice with aPL or control IgG and evaluated 1) activation of washed platelets, 2) aortic wall gene expression, and 3) levels of circulating neutrophil extracellular traps (NETs) using blood and tissue samples isolated at identical time points as IVC ligation studies (mice in these studies did not undergo IVC ligation). Platelet activation was assessed by measuring the binding of JON/A and expression of platelet P-selectin by flow-cytometry. Total RNA isolated from aortic walls of 3 mice each treated with murine aPL or control IgG was analyzed by RNA sequencing.

Results

WT C57/BL6 mice that received 20 µg of polyclonal murine aPL sustained plasma anti-β2GPI IgG levels >500 ng/mL for more than 120 h after IgG injection. In the IVC stasis model, six out of eight mice receiving aPL IgG and four out of eight mice receiving control IgG developed thrombi (lack of thrombi in all mice was likely due to bleeding of extensive IVC back branches in C57/BL6 mice). Thrombi weight in aPL-treated mice were significantly larger than those from mice that received control IgG (19.7 ± 1.7 mg vs 13.5 ± 1.0 mg, P = 0.015). Thrombin (0.1 U/ml)-induced platelet activation was significantly greater for platelets isolated from aPL-administered mice compared to those from control mice 72h after IgG injection as determined by binding of JON/A (1246.7 ± 55.3 MFI vs 879 ± 37.9 MFI, P = 0.005) and expression of P-selectin (5729.7 ± 553.9 MFI vs 3777.3 ± 402.5 MFI, P=0.046). Gene enrichment analysis of mRNA from aortas demonstrated significant upregulation of gene pathways related to complement and coagulation cascades, blood coagulation, and oxidoreductase activity in the aPL-treated mice. Individually upregulated genes included C8, C9, fga and fgb. Elevated levels of circulating nucleosomes were also observed in aPL treated mice but did not reach significance due to small numbers.

Conclusion

These results demonstrate that anti-β2GPI polyclonal aPL raised in apoh-/- mice circulate for extended periods when reinfused into WT mice, enhance thrombosis, induce platelet hyperreactivity, and cause changes in mRNA expression in the vascular wall. Further characterization of anti-β2GPI IgG for LAC activity, anti-PS/PT reactivity and β2GPI domain 1 reactivity is in progress. While all animal models of APS have weaknesses, this syngeneic model provides an additional approach that recapitulates many aspects of vascular inflammation seen in human APS.