Abstract

Thrombospondin-1 (TSP-1) is a large homotrimeric glycoprotein of approximatively 450 kDa, synthetised by several cell types, including platelets where it is stored in α-granules. Its multidomain structure enables TSP-1 to interact with many cell-adhesive receptors, integrins, heparan sulfate, as well as with other adhesive glycoproteins, including VWF. In a mouse model, soluble or local platelet released TSP-1 can protect unfolded endothelium-bound and subendothelial VWF from degradation by its plasma protease ADAMTS13. TSP-1 has also been shown to bind to the A3 domain of VWF suggesting a role for competitive inhibition of ADAMTS13-mediated VWF proteolysis. With this background we chose to investigate the inhibition effect of TSP-1 on ADAMTS13 activity by measuring the cleavage of recombinant (r)VWF multimers using rADAMTS13 in absence or presence of TSP-1 at different concentrations. We also assessed whether TSP-1 competes with ADAMTS13 for interaction with the A3 domain of VWF, by comparing cleavage of VWF-A1A2A3 or A1A2 or A2A3 peptides by ADAMTS13 in the presence or absence of TSP-1 by using HPLC analysis.Significant inhibition by TSP1 on ADAMTS13 activity was observed only with TSP-1 plasma concentrations reached after platelet activation (24–160 nM). However, this result was not confirmed at the low concentrations physiologically present in plasma (0.3–0.7 nM). The relative percentage of high molecular weight multimers was proportionally correlated to the TSP-1 concentration. These findings were confirmed using the collagen binding assay to assess VWF activity relative to antigen concentrations. The mean ratio of CBA/Ag was 0.22 in absence of TSP-1 compared to 0.5 and 0.78 in presence of 80 and 160nM of TSP-1. The HPLC analysis of VWF peptides cleavage by ADAMTS13 in presence of increasing concentrations of TSP-1 (0–225 nM) showed that TSP-1 acts as a competitive inhibitor of ADAMTS13 in binding both A1A2A3 and A1A2 peptides. The A1A2A3 peptide demonstrated an apparent affinity for ADAMTS-13 about three-fold higher than that of A1A2 construct (IC50=36.7±4nM and IC50=100.4±25nM, respectively). We also observed a TSP-1 inhibitory effect of ADAMTS13 activity with the A2A3 peptide (IC50=120 nM). These results indicate that the entire A1A2A3 region is necessary to achieve an optimal inhibition of the ADAMTS-13/VWF interaction by TSP-1. The A3 domain plays a central role in this inhibition, although its effect is dependent on the allosteric role played by the A1 domain. Thus, the lower inhibition observed with A2A3 peptide than with the A1A2A3 peptide, could be explained by increased accessibility of ADAMTS13 to A2 domain when the inhibition effect of A1 domain is lacking, as previously described by

Nishio et al (
PNAS
2004
;
101
:
10578
–83).

Author notes

Disclosure: No relevant conflicts of interest to declare.