Antiphospholipid/Anti-β₂gpi Antibodies Induce Endothelial Cell Activation through Formation of a Multi-Protein Signaling Complex Consisting of Annexin A2, Toll-Like Receptor 4, Calreticulin and Nucleolin

Abstract 151

Antiphospholipid syndrome (APS) is a multisystem autoimmune disorder characterized by an increased risk of arterial and/or venous thrombosis and recurrent fetal loss in the presence of elevated levels of antiphospholipid antibodies (APLA). Most pathologic APLA actually recognize phospholipid-bound proteins, the most common of which is beta-2 glycoprotein I (β₂GPI). Previous work from our laboratory and others have demonstrated that anti-β₂GPI antibodies activate endothelial cells in the presence of β₂GPI through cross-linking or clustering of endothelial cell surface Annexin A2, to which β₂GPI binds with high affinity. Endothelial cells activated by β₂GPI/anti-β₂GPI antibodies increase their expression of cell surface adhesion molecules including ICAM-1, VCAM-1, and E-selectin. Since annexin A2 is not a transmembrane protein, APLA-induced endothelial cell activation may occur through a TLR4-MyD88-dependent pathway leading to NFκB activation. However, the nature of the interactions between annexin A2, β₂GPI and TLR4 have not been well defined and whether additional proteins may contribute to the formation of a signaling complex remains unknown. To address this issue, we used a cell surface ELISA for E-selectin to measure endothelial cell activation in response to β₂GPI/anti-β₂GPI antibodies. We observed that inhibition of TLR4 expression by siRNA caused a 90% inhibition of E-selectin expression following exposure of cells to β₂GPI/anti-β₂GPI antibodies, while TLR2 siRNA had no effect. siRNA-induced inhibition of apolipoprotein E receptor 2 (ApoER2) expression also did not affect the activation of endothelial cells by β₂GPI/anti-β₂GPI antibodies. To determine whether annexin A2 binds directly to TLR4, we determined whether TLR4 could be affinity purified from APLA-induced endothelial cell extracts using immobilized annexin A2. These studies led to the isolation of not only TLR4, but two additional proteins, calreticulin and nucleolin. The interaction between Annexin A2 and TLR4 was also demonstrated using Annexin A2 coupled to Affigel-HZ. To explore the role of calreticulin and nucleolin in APLA-induced endothelial cell activation, we measured APLA/anti-β₂GPI antibody-induced E-selectin expression by endothelial cells after siRNA mediated inhibition of calreticulin and nucleolin expression, observing that siRNAs against either of these proteins significantly inhibited endothelial cell activation. Additional studies confirmed that siRNA-mediated knockdown of Annexin A2, TLR4, calreticulin and/or nucleolin inhibited the cell surface expression of not only E-selectin, but ICAM-I and VCAM-I. Interestingly, endothelial cell activation caused by APLA/anti-β₂GPI antibodies was induced increased expression of mRNA encoding annexin A2, TLR4, MD2 (a TLR4 co-receptor involved in dimerization), MyD88, and S100A10 (a component of the endothelial cell annexin A2-S100A10 heterotetramer) (12, 3, 12, 5 and 6-fold, respectively). Taken together, these studies suggest that a complex consisting of annexin A2, TLR4, calreticulin and nucleolin may mediate endothelial cell activation by APLA/anti-β₂GPI antibodies. We hypothesize these proteins may form a multi-component signaling complex, perhaps in lipid rafts, that culminates in the activation of NF-κB.