The antiphospholipid syndrome (APS) is characterized by predisposition to thrombosis. The cause for this pathology is poorly understood but is likely multifactorial, involving activation of blood cells and vasculature. The role that anti-β2-GPI antibodies play in von Willebrand factor (VWF) release from endothelial cells, VWF-platelet binding, and VWF cleavage by ADAMTS13 has not been well characterized in APS. We decided to study the effect of these antibodies on expressed ultra large VWF strings (ULVWF strings) that bind platelets (VWF-PLT strings) under flow to better understand platelet–VWF binding and ADAMTS13 regulation in APS.


We hypothesized that Anti-β2-GPI antibodies could induce VWF release from endothelial cells and modulate VWF’s prothombotic effect through alterations in VWF-Platelet binding and VWF cleavage by ADAMTS13.


Human umbilical vein endothelial cells were seeded in 96-well plates/flow chambers prepared with a collagen Type I substrate for static/flow experiments, respectively.

Static assays: Cells were incubated for 1 hr with Anti-β2-GPI or control antibodies and the conditioned media was assayed for VWF by ELISA, normalized to normal plasma.

Flow Assay Analysis: After stimulation with agonist and perfusion with a platelet suspension, platelets bound to ULVWF in a string pattern were quantified via brightfield microscopy. Images of chambers were captured and VWF-PLT string-units (defined as a string length of 25μM) per slide were quantified. To minimize bias, image acquisition was standardized and the investigator was blinded at time of image acquisition/analysis.

β2-GPI Flow assays: Endothelial cells in flow chambers were stimulated with 50ng/mL of phorbol myristate acetate (PMA), and a solution of fixed platelets with β2-GPI or β2-GPI+Anti-β2-GPI were perfused prior to image acquisition.

ADAMTS13 assays: After stimulation with 25ng/mL PMA and perfusion with fixed platelets, images were acquired. Then control/patient plasma was perfused over formed strings. Images taken after plasma perfusion were quantified and compared to images prior to plasma perfusion.

Data are shown as mean +/- SEM, and significance was determined as p<0.05 by student’s t-test or Mann-Whitney U Test, when appropriate.


Static Assays: Compared to control human IgG (8.28 +/- 3.34 mU/mL), VWF release was increased in the presence of two patient-derived Anti-β2-GPI antibodies, APS25-6 Anti-β2-GPI, 35.73 +/- 7.83 mU/mL (P = 0.008) and APS203-2 Anti-β2-GPI, 34.08 +/- 7.119 mU/mL (P = 0.039). As compared to control rabbit IgG (15.80 +/- 7.12 mU/mL), a rabbit polyclonal Anti-β2-GPI antibody, R24-6, also demonstrated increased soluble VWF (43.16 +/- 9.60 mU/mL, P = 0.013) release.

β2GPI Flow Assays:The presence of β2GPI (2µM) reduced String-unit formation from 50.10 +/-5.57 Sting-units/image to 20.98 +/- 2.05 String Units/image (P < 0.0001) as compared to buffer. Addition of goat Anti-β2-GPI antibody (1µM) increased the VWF-PLT string observed as compared to β2GPI (2µM), 30.09 +/- 1.83 String Units to 20.98 +/- 2.05 String Units (P = 0.012) indicating that an Anti-β2-GPI antibody partially reverses the effect of β2GPI on reducing VWF-PLT string formation.

ADAMTS13 Assay:Compared to pooled normal plasma (ADAMTS13 Activity 100%) (4.57 +/- 0.60 String Units/image cleaved), there was a significant decrease in the amount of string units/image cleaved in two APS plasmas with Anti-β2-GPI antibodies, APS232-9 (-0.23 +/- 0.98, P = 0.0003) and APS227-9 (2.23 +/- 0.73, P = 0.0009). ADAMTS13 Activity of patient plasma was 98.37% and 83.97%, respectively. These results suggest an inhibitory role of APS plasma on the cleavage of ULVWF strings.


Anti-β2-GPI antibodies and antiphospholipid syndrome plasma may contribute to the prothrombotic phenotype observed in APS by three mechanisms: 1) the increased release of VWF from endothelial cells after incubation with Anti-β2-GPI, 2) increased platelet binding to ULVWF strings likely mediated by interfering with β2GPI’s known inhibition of Gp1bα VWF-platelet binding, and 3) a reduced ability to cleave VWF-PLT strings by APS plasma, suggestive of ADAMTS13 inhibition that does not correlate with ADAMTS13 activity. Taken together, our results suggest that VWF and its modulation may contribute to the prothrombotic phenotype observed in the antiphospholipid syndrome.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.