Plasmin-cleaved von Willebrand factor as a biomarker for microvascular thrombosis

Abstract

von Willebrand factor (VWF) is an essential contributor to microvascular thrombosis. Physiological cleavage by ADAMTS13 (a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13) limits its prothrombotic properties, explaining why ADAMTS13 deficiency leads to attacks of microthrombosis in patients with thrombotic thrombocytopenic purpura (TTP). We previously reported that plasminogen activation takes place during TTP attacks in these patients. Furthermore, stimulation of plasminogen activation attenuates pathogenesis in preclinical TTP models in vivo. This suggests that plasmin is an endogenous regulator of VWF thrombogenicity, in particular when ADAMTS13 falls short to prevent microvascular occlusions. VWF cleavage by plasmin is biochemically distinct from cleavage by ADAMTS13. We hypothesized that plasmin-cleaved VWF (cVWF) holds value as a biomarker of microvascular thrombosis. Here, we describe the development of a variable domain of heavy-chain-only antibody (VHH)-based bioassay that can distinguish cVWF from intact and ADAMTS13-cleaved VWF in plasma. We validate this assay by tracking cVWF release during degradation of microthombi in vitro. We demonstrate that endogenous cVWF formation takes place in patients with TTP during acute attacks of thrombotic microangiopathy but not in those in remission. Finally, we show that therapeutic plasminogen activation in a mouse model of TTP amplifies cVWF formation, which is accompanied by VWF clearance. Our combined findings indicate that cVWF is released from microthrombi in the context of microvascular occlusion.

Graphical abstract

Figure 1.

Time-dependent VWF cleavage by plasmin during microthrombus breakdown. (A) Time course of sampling during microthrombus breakdown by plasmin (75 μg/mL). Plasmin activity was blocked with PPACK, cells removed by centrifugation, and supernatants analyzed by western blotting. (B) Western blot under nonreducing conditions; (C) western blot under reducing conditions. Data represent 3 independently executed experiments. Plm, plasmin.

Figure 2.

cVWF assay. (A) cVWF in buffer or (B) citrated NPP in the absence or presence of ristocetin (600 μg/mL). (C) cVWF or ADAMTS13-cleaved VWF in buffer or normal pooled citrated plasma. Data are expressed as means ± standard deviation (SD) for 3 independently executed experiments.

Figure 3.

Characterization of a binding site in cVWF that enables V_HH capture. Conditioned cell culture supernatants containing C-terminally truncated VWF variants were incubated with plasmin (150 μg/mL) or vehicle and analyzed by (A) western blotting and by ELISA. (B) V_HH G5 was immobilized, and captured VWF variants were detected with a polyclonal antibody. (C) Western blot of captured VWF products, spiked in buffer or NPP. Data represent 3 independently executed experiments. Bar graphs show means ± SD. M, marker; Neg. Sup, negative control supernatant; Plm, plasmin.

Figure 4.

cVWF release from microthrombi. Microthrombi were exposed to plasmin (75 μg/mL) or buffer. (A-C) Platelet-free supernatants were collected and analyzed by ELISA. Next, preformed microthrombi were separated from soluble VWF by centrifugation and subjected to either degradation by plasmin (75 μg/mL) or buffer. Platelet complexes were removed by a second centrifugation step and supernatants analyzed by western blotting (D) and ELISA (E-G). Data represent 3 independently executed experiments. Bar graphs show means ± SD. No Plt, no platelets; Plm, plasmin.

Figure 5.

cVWF formation under flow. (A-C) Washed platelets were perfused over histamine-stimulated HUVECs. After being stable for 10 minutes, HUVEC-bound platelet-covered VWF strings were exposed to plasmin (50 μg/mL) or vehicle control under flow. Samples were collected into PPACK and analyzed by ELISA for VWF antigen levels, cVWF levels, and fraction of VWF in a plasmin-cleaved state. Data represent 5 independently executed experiments, and are shown as means ± SD.

Figure 6.

cVWF formation during TMA attacks in patients with TTP. (A-C) VWF antigen levels, cVWF levels, and fraction of VWF in a plasmin-cleaved state in healthy controls, patients with TTP in remission, and during acute attacks. Data represent 3 independently executed experiments and are shown as scatter plots with medians. Data were analyzed with the Kruskal-Wallis test followed by the Dunn multiple comparisons test. ∗P < .05; ∗∗∗∗P < .0001; ns, nonsignificant. (D-F) Correlations between PAP complex levels and platelet counts (D), cVWF levels and platelet counts (E), and cVWF levels and PAP complex levels (F). (G-I) Correlations between VWF antigen and cVWF levels (G), VWF antigen levels and platelet counts (H), and VWF antigen levels and PAP complex levels (I). Correlations were computed by Pearson correlation coefficients. Healthy, healthy controls.

Figure 7.

Therapeutic plasminogen activation accelerates cVWF formation. Levels of VWF antigen (A), cVWF (B), and fraction of circulating VWF that exists in a plasmin-cleaved state (C) after Microlyse treatment (blue squares) or saline treatment (red circles) in Adamts13^−/− mice, challenged by administration of rhVWF. Data are displayed as scatter plots with medians, and represent 3 independently executed experiments. Results were analyzed by 2-way analysis of variance followed by the Šidák’s post-hoc test. ∗∗P < .005; ∗∗∗P < .005; ∗∗∗∗P < .0001; ns, nonsignificant.