Characterization of the interactions of ADAMTS13 CUB1 domain to WT- and GOF-Spacer domain by molecular dynamics simulation

https://doi.org/10.1016/j.jmgm.2021.108029Get rights and content

Highlights

  • Exosite-3, exosite-4, and a novel G607-S610 region in the Spacer domain stabilized the Spacer-CUB1 complex.

  • GOF Spacer attenuated the exosite-3 and the G607-610 region but enhanced the exosite-4 to bind to the CUB1 domain.

  • Residues D1259, L1258, T1261, E1231, and R1251 might play a critical role in the Spacer-CUB1 interaction.

Abstract

Metalloprotease ADAMTS13 specifically cleaves VWF (von Willebrand Factor) to prevent excessive platelet aggregation and thrombus formation at the sites of vascular injury. To avoid non-specific cleavage, ADAMTS13 has the auto-inhibition effect in which the Spacer domain in N-terminal interacts with the CUB1 domain in C-terminal, resulting in decreased proteolytic activity. Previous studies reported that exosite-3 in the Spacer domain was a key binding site in the Spacer-CUB1 interaction. When exosite-3 was mutated (R660K/F592Y/R568K/Y661F/Y665F, GOF), the auto-inhibition of ADAMTS13 was disrupted and the enzymatic activity was markedly increased. However, the characteristics of the Spacer-CUB1 interaction is not fully understood. Here, we constructed the model of Spacer-CUB1 complex by homologous modeling and molecular docking to characterize the Spacer-CUB1 binding and predict key amino acid residues via molecular dynamics simulation. Our data showed that G607-S610 was a non-reported potential binding site in the Spacer domain; GOF mutation attenuated the formation of hydrogen bond between exosite-3 and the CUB1 domain; Residues E1231, R1251, L1258, D1259 and T1261 in the CUB1 domain might play an important role in the Spacer-CUB1 interaction. Our study advances the understanding of the structural basis of the auto-inhibition of ADAMTS13 and provides information about the key residues in the binding interface.

Introduction

ADAMTS13 (A Disintegrin and Metalloprotease with Thrombospondin type 1 repeats, member 13) specifically cleaves the Tyr1605-Met1606 scissile bond in the A2 domain of VWF (von Willebrand Factor) to regular the size and activity of VWF multimers, preventing the excessive platelet aggregation and thrombus formation at the sites of vascular injury [1]. A severe deficiency in ADAMTS13 activity leads to a life-threatening disease thrombotic thrombocytopenic purpura (TTP), which could be classified as acquired TTP and congenital TTP. The former is mainly caused by the production of various auto-antibodies against ADAMTS13 and the latter is due to the mutation of the ADAMTS13 gene [2]. Recombinant human ADAMTS13 has completed the phase Ⅰ clinical trial and has been proved to be safe and effective in the treatment of congenital TTP [3].

ADAMTS13 has the auto-inhibition effect in which the Spacer domain at the amino-terminal interacts with the TSP8 (Thrombospondin type-1 repeats 8) domain and two CUB (Complement components C1r and C1s, sea urchin protein Uegf, and Bone morphogenetic protein 1) domains at the carboxyl-terminal [4,5]. The auto-inhibition blocks the critical regions of the Spacer domain that bind to VWF A2, resulting in the weakened ability of ADAMTS13 to recognize and bind to the substrate, and thus the decreased proteolytic activity [4,6,7]. Besides, ADAMTS13 has the mechanism of conformational activation. VWF D4-CK domains could bind to ADAMTS13 TSP8-CUB domains, disrupting the auto-inhibition of ADAMTS13. This disruption induces the conformation of ADAMTS13 from a “close” to “open” state, exposing the key residues in the Spacer domain to promote substrate binding and increasing the proteolytic activity [4].

The Spacer domain is required for the cleavage of VWF [8]. Previous studies reported that Y659–Y665 residues in the Spacer domain were critical to the recognition and interaction with the VWF A2 domain [9,10]. In addition, Casina et al. reported that the E634-R639 region in the Spacer domain, called exosite-4, was also a key binding site of Spacer-A2 interaction, in which D635 and R636 residues played a critical role in substrate recognition [11]. The Spacer domain is also a primary epitope for autoantibodies against ADAMTS13 in TTP patients. These autoantibodies block the binding of the Spacer domain and the VWF A2 domain and thus attenuate enzyme digestion. The external epitope composes of R568, F592, R660, Y661, and Y665, called exosite-3, the key target of anti-Spacer autoantibodies in TTP patients [9,12]. Cui et al. mutated the exosite-3 (R660K/F592Y/R568K/Y661F/Y665F) and obtained a gain-of-function ADAMTS13 (GOF ADAMTS13), which effectively resisted the bindings of autoantibodies, providing a new strategy for the treatment of acquired TTP [13]. Compared to wild-type, GOF ADAMTS13 has higher activity and “open” conformation because the auto-inhibition is disrupted [7,14], suggesting that exosite-3 is a key binding site in the auto-inhibition. However, a previous study reported that GOF ADAMTS13 nonspecifically cleaved the Aα chain of human fibrinogen [15].

The crystal structure of the Spacer domain was resolved a decade ago [16,17], and the one of the CUB1 domain was resolved recently [18]. Kim et al. investigated the interaction of the CUB1-2 domain and the Spacer domain by docking simulation in combination with functional assays, revealing that the CUB2 domain but not the CUB1 domain bound to exosite-3, while W1245, W1250, and K1252 in the CUB1 domain and R1326, E1387, and E1389 in the CUB2 domain were critical in Spacer-CUB1-2 binding [18]. However, surface Plasmon Resonance (SPR) data demonstrated that isolated CUB1 domain could bind to wild-type Spacer but not GOF Spacer, indicating that exosite-3 was the main binding site of the Spacer-CUB1 interaction [4]. In addition, we previously revealed that 11.78% and 75.36% of GOF ADAMTS13 molecules retained the “close” and intermediate state respectively by scanning the WT and GOF ADAMTS13 molecules with atomic force microscope [14]. This finding implied that the C-terminal of ADAMTS13 could still bind to the Spacer domain in GOF ADAMTS13. In order to characterize the binding of CUB1 to WT and GOF Spacer, we herein used computational approaches to obtain a model of the Spacer-CUB1 complex, analyzed the binding information of the complex at the atomic level via molecular dynamics simulation, and predicted the key amino acid residues in the Spacer-CUB1 interaction.

Section snippets

Construction of the Spacer-CUB1 complex

The structure of the Spacer domain was selected from the crystal structure of truncated ADAMTS13 DTCS (PDB ID: 3GHM) from the protein data bank (PDB) and the amino acid sequence was S556–P682. The CUB1 structures were obtained by homologous modeling via PHYRE2 server [19]. The structure with the highest score was selected, which was built upon the crystal structure of the acidic seminal fluid protein (PDB ID: 1SFP), the highest scoring template with 94.3% confidence for modeling the CUB1

A reliable and stable structure of the Spacer-CUB1 complex was obtained

To obtain the poses of the wild-type Spacer-CUB1 complex, the Spacer domain was docked with the CUB1 domain via HADDOCK 2.2 server. Three clusters in which the exosite-3 was located in the binding interface were screened out from the total ten clusters. In order to find out the most stable pose, the highest scoring pose of each selected cluster: poses 1-1, 2–1, and 8–1 (Fig. 2A), were performed 20 ns equilibrium (EQU). RMSD-Cα, Buried SASA, distribution of H-bond, and binding energy were used

Discussion

The auto-inhibition of ADAMTS13 is of great biological significance. This auto-inhibition mechanism could prevent ADAMTS13 non-specific proteolysis against fibrinogen [15] and protect epitopes in the Spacer domain from anti-ADAMTS13 autoantibodies [33]. It is believed that the auto-inhibition is governed by the interaction of the Spacer domain and the C-terminal TSP8-CUB2 domains [5]. Although the crystal structure of MDTCS, an N-terminal truncated ADAMTS13, and the CUB1-2 domains have been

Author contributions

J.L. designed this research; J.Y., Z.W., and X.X. performed the research and analyzed data; J.W., Y.F. and G.L. provided a server for MD simulations. J.Y., G.L., Y.F. J.W., and J.L. wrote this paper; All authors contributed to the article and approved the published version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China, 31771012 and 32071303 (to J.L.), 11672109 (to Y·F.), and 11432006 (to J.W.).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (34)

  • G.C.P. Van Zundert et al.

    The HADDOCK2.2 web server: user-friendly integrative modeling of biomolecular complexes

    J. Mol. Biol.

    (2016)
  • Y. Yamada et al.

    Carbon materials with controlled edge structures

    Carbon N. Y.

    (2017)
  • G. Liu et al.

    A mechanism for localized dynamics-driven affinity regulation of the binding of von willebrand factor to platelet glycoprotein Ibα

    J. Biol. Chem.

    (2013)
  • E. Roose et al.

    Open ADAMTS13, induced by antibodies, is a biomarker for subclinical immune-mediated thrombotic thrombocytopenic purpura

    Blood

    (2020)
  • B. Ercig et al.

    N-glycan–mediated shielding of ADAMTS13 prevents binding of pathogenic autoantibodies in immune-mediated TTP

    Blood

    (2021)
  • J. Muia et al.

    Allosteric activation of ADAMTS13 by von Willebrand factor

    Proc. Natl. Acad. Sci. Unit. States Am.

    (2014)
  • K. South et al.

    Conformational activation of ADAMTS13

    Proc. Natl. Acad. Sci. Unit. States Am.

    (2014)
  • Cited by (2)

    View full text