Arginine Methylation by Protein Arginine Methyltransferase 1 (PRMT1) Affacts ADAMTS13 Secretion and Enzymatic Activity

Abstract

Background:

ADAMTS13, primarily synthesized in hepatic stellate and endothelial cells, plays a pivotal role in regulation of hemostasis by proteolytic cleavage of von Willebrand factor. Severe deficiency of plasma ADAMTS13 activity may result in thrombotic thrombocytopenic purpura (TTP), a potentially fatal blood disorder. ADAMTS13 undergoes posttranslational modifications including glycosylation, citrullination, and oxidation, etc. The present study determines the impact of arginine methylation by protein arginine methyltransferase 1 (PRMT1) on ADAMTS13 secretion and function.

Methods:

Cell culture, recombinant protein, biochemical analysis, site-directed mutagenesis, and animal models were utilized.

Results:

An inhibition of arginine methylation by a PRMTs inhibitor (MS023) in HEK293 cells expressing recombinant ADAMTS13 and mice results in a significant reduction of ADAMTS13 secretion but the secreted ADAMTS13 shows an increased specific activity; conversely, an overexpression of PRMT1 in HEK293 cells and mice results in an increase of ADAMTS13 secretion but the secreted ADAMTS13 exhibits a significantly reduced proteolytic activity. The altered ADAMTS13 activity appeared to be related to its conformational changes. LC-MS/MS identified >100 arginine methylation events on recombinant ADAMTS13. Site-directed mutagenesis of 5 highly conserved methylation sites (R193, R498, R692, R1123, and R1206) identifies the critical role of R1206 in ADAMTS13 function. The R1206K variant exhibits 4~5-fold increase of ADAMTS13 activity, likely resulting from an alleviation of allosteric inhibition.

Conclusions:

These results demonstrate for the first time the crucial role of arginine methylation on ADAMTS13 secretion and activity. Our findings may shed new light on the mechanism of allosteric regulation of ADAMTS13, which may have a therapeutic implication.

Graphical Abstract

Introduction

Thrombotic thrombocytopenic purpura (TTP), a potentially fatal thrombotic microangiopathy, is characterized by severe thrombocytopenia, microangiopathic hemolytic anemia, and various degrees of organ dysfunctions^1–3. Central to pathophysiology of TTP lies in severe deficiency of plasma ADAMTS13 (A Disintegrin And Metalloprotease with ThromboSpondin type 1 repeats, 13) activity⁴. In most cases, the severe deficiency is resulted from acquired autoantibodies that inhibit ADAMTS13 activity and/or enhance its clearance^5–7.

ADAMTS13, primarily synthesized and secreted in hepatic stellate cells⁸ and endothelial cells⁹, is a key plasma metalloprotease that cleaves ultra large forms of von Willebrand factor (VWF), derived from endothelium¹⁰, megakaryocytes¹¹, and platelets¹². This proteolytic cleavage is crucial for preventing the formation of spontaneous VWF-platelet agglutinates and occlusive thrombi in small arterioles and capillaries. An inability to cleave ultra large VWF, resulting from severe deficiency of plasma ADAMTS13 activity leads to the formation of disseminated microvascular thrombosis¹³, a pathognomonic feature of TTP. Additionally, moderate or relative deficiency of plasma ADAMTS13 activity is common, which may be associated with an increased risk for myocardial infarction¹⁴, ischemic stroke¹⁵, trauma-¹⁶ or COVID19-associated coagulopathy¹⁷. ADAMTS13 also exhibits an anti-inflammatory activity by inhibiting the VWF-platelet mediated accumulation of monocytes/neutrophils^18,19 and neutrophil extracellular traps (NETs) such as in the cases of COVID-19²⁰ or immune TTP²¹. Moreover, low levels of plasma ADAMTS13 activity are associated with an unfavorable outcome in hospitalized patients with a suspected or a confirmed heparin-induced thrombocytopenia and sepsis^22,23.

How ADAMTS13 synthesis, secretion, and proteolytic activity are regulated is not fully understood. Previous studies have demonstrated that inflammatory cytokines such as interleukin 6, tumor necrosis factor (TNF)-α, and interferon-γ suppress ADAMTS13 synthesis at transcriptional levels in hepatic stellate cells and endothelial cells in culture⁹. Additionally, oxidation^24,25, glycosylation^26,27, and citrullination^28,29 are shown to affect ADAMTS13 function. Human neutrophil peptide-1 (HNP1), released from degranulation of neutrophils, may inhibit or accelerate the VWF cleavage by ADAMTS13 depending on whether fluidic shear is absent or present³⁰. It has not yet been reported if ADAMTS13 is post-translationally modified by protein arginine methylation, which is one of the most prevalent post-translational modifications in eukaryotic organisms^31,32.

Protein arginine methyltransferase 1 (PRMT1), responsible for more than 80% of total protein arginine methylation in mammals^33,34, is significantly increased in its expression during many disease processes including liver cirrhosis, pulmonary hypertension, and obesity³⁵. This suggests a broad involvement of arginine methylation in the pathogenesis of immune, thrombotic, and inflammatory disorders.

The present study aims to: 1) determine if ADAMTS13 is post-translationally modified by arginine methylation; 2) determine how arginine methylation on ADAMTS13 affects its secretion and proteolytic activity; 3) understand the molecular mechanism underlying an altered ADAMTS13 function with hypo- or hyper-methylation and its physiological relevance. The findings may shed new light on ADAMTS13 regulation and its pathogenetic mechanism of immune TTP, and perhaps other inflammatory thrombotic disorders.

Materials and Methods

Materials:

All data, analytic methods, and study materials will be available to other researchers at the request to the corresponding authors. Human recombinant VWF73 was expressed, labeled with 5’-maleimide fluorescein (Thermo Fisher, Rockford, lL), and purified to homogeneity as described previously^36,37. The final product, designated as FRETS-rVWF73, was stored in 0.5% trifluoroacetic acid³⁸ (Sigma, St. Louis, MO) at −80 °C. A cattle FRETS-rVWF71 was kindly provided by Dr. Joshua Muia at Oklahoma State University, Stillwater, OK^39,40. Human VWF was purified from cryoprecipitate as described in detail in our recent publication⁴¹. Recombinant human ADAMTS13 was expressed in stably transfected HEK293 cells and purified using three-step column chromatography according to our published procedures^42,43. Protein concentrations were determined by using the following molecular weights and extinction coefficient (E₂₈₀, 1 mg/ml): ADAMTS13, 195 kDa, 0.68 and VWF, 250 kDa per monomer, 1.0). Human PRMT1 cDNA in pcDNA3 vector and PRMTs inhibitor (MS023) were previously described⁴⁴.

Treatment of HEK293 cells with a PRMTs inhibitor.

Human embryonic kidney cells (HEK293) stably expressing a His-V5 tagged recombinant ADAMTS13 were treated with a a potent, selective, and cell-active inhibitor of human type I protein arginine methyltransferases (PRMTs) inhibitor (MS023) at various concentrations (0–1.0 μM) or a fixed concentration of 0.5 μM for 48 hours. The serum-free conditioned medium was collected and concentrated 20-fold prior to analysis. Cells were scraped out of culture dishes and lysed with a lysis buffer consisting of 20 mM HEPES, pH 7.9, 150 mM NaCl, 1mM MgCl₂, 1% NP40, 10 mM NaF, 0.2 mM NaVO₄, and 10 mM β-glycerol phosphate in the presence of a proteinase inhibitor cocktail (Roche, Indianapolis, IN).

Overexpression of PRMT1 in HEK293 cells expressing recombinant ADAMTS13.

HEK293 cells expressing a His-V5 tagged recombinant ADAMTS13 were transiently transfected with a pcDNA3-Flag-PRMT1 using PolyJet^™ In Vitro DNA Transfection Reagent (SignaGen, Frederick, MD). Conditioned medium and cell lysate were collected 48 hours following transfection for analysis.

Western blotting and immunoprecipitation:

Stably transfected HEK293 cells from 10-cm dish were lysed with 1 ml of lysis buffer (20 mM HEPES, pH 7.9, 150 mM NaCl, 1mM MgCl₂, 1% NP40, 10 mM NaF, 0.2 mM NaVO₄, 10 mM β-glycerol phosphate) in the presence of a proteinase inhibitor cocktail (Roche, Indianapolis, IN). The cell lysates were incubated for 30 min on ice, and the extracts were cleared by centrifugation at 12,000xg for 15 min at 4°C. The conditioned medium or cell lysates were used directly for Western blotting analysis or following immunoprecipitations. The immunoprecipitation was performed at 4°C in the lysis buffer in the presence of RNase A (10 μg/ml) using an indicated antibody, and 50 μl of 50% (v/v) slurry of Protein A agarose or anti-Flag (for PRMT1) M2 agarose for 4 hr. The precipitates were extensively washed and resuspended in 2x SDS-PAGE sample buffer directly. The immunopurified proteins and cell lysates were resolved by 10% SDS-PAGE and transferred to PVDF membranes for Western blotting analysis. The proteins were visualized by the Immobilon Western Chemiluminescent horseradish peroxidase substrate detection kit (Millipore). The antibodies used were commercially available, including anti-Flag M2 antibody (Sigma-Aldrich, St. Louis, MO) or anti-PRMT1 antibody (Millipore, Burlington, MA) or anti-His antibody (Abcam, Waltham, Boston, MA), or anti- anti-asymmetric dimethyl arginine antibody (ThermoFisher, Rockford, lL) or anti-V5 antibody (Invitrogen, Carlsbad, California), or beta-actin monoclonal antibody (Sigma- Aldrich). Polyclonal anti-ADAMTS13 IgG was produced with a mixture (8:2) of human and mouse recombinant ADAMTS13⁴⁵ and monoclonal anti-ADAMTS13 IgG were produced from Green Mountain antibodies, Burlington, VT)⁴⁵.

Immunofluorescent microscopic imaging:

Cells were spread on 3.5 mm imaging cell culture dish and then fixed with ice-cold 4% paraformaldehyde in PBS for 15 minutes. Cell were permeabilized with 0.1% Triton X-100 for 10 minutes and incubated with a mouse anti-V5 IgG (1.0 μg/mL) (Invitrogen) and rabbit anti-Calnexin (0.8 μg/mL) (Cell signaling, Danvers, MA) overnight at 4 °C. After washing 3 times with PBS with 0.1% Tween-20, cells were incubated with anti-mouse Alexa 488 or anti-rabbit Alexa 594. The cell nucleus was stained with Hoechst 33342. The fluorescent images were obtained using a Nikon A1-SHR confocal fluorescence microscope (Belmont, CA).

Glycosidase digestion.

To better understand the intracellular trafficking of recombinant ADAMTS13, we assessed the glycosylation status of recombinant ADAMTS13 inside cells and in the conditioned medium. In these experiments, 10 μl of cell lysate or conditioned medium was heated at 100 °C for 5 minutes in a total volume of 30 μl, containing 20 mM sodium phosphate (pH 7.5), 50 mM EDTA-Na, 0.4% SDS, 4% β-mercaptoethanol, and 0.02% sodium azide. The denatured and reduced proteins were then incubated at 37 °C for 2–4 hours, either in the absence (−) of enzyme or presence (+) of 1 milliunit of endoglycosidase H (Millipore, Burlington, MA) or 0.5 units of peptide-N-glycosidase F (New England Biolabs, Ipswich, MA) in a total reaction volume of 30 μl according to manufacturer’s recommendation. Subsequently, 10 μl of the digested samples was subjected to Western blot analysis.

Animal studies.

All animal study protocols are approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Kansas Medical Center. Mice were all fed with diet 5053, a PicoLab Rodent Diet 20 (formulated with 20% protein diet and 4.5% fat) from Richmond, Indiana. Both male and female mice at age 4–5 weeks were used. The data from males and females were combined, and sex difference and interaction were not determined in the study. Wild type (WT) in C57/BL6 background or transgenic mice expressing human PRMT1 under platelet factor 4 promoter which overexpress PRMT1 in megakaryocytes and platelets as previously described⁴⁴. WT mice were treated via an intraperitoneal cavity with a PRMTs inhibitor (MS023) at 80 mg/kg body weight (n=5) or a same volume of DMSO (n=5) every other day for 14 days. Plasma was obtained prior to sacrifice of the mice for ADAMTS13 antigen and proteolytic activity.

Capillary-based Western blotting analysis for detecting plasma ADAMTS13 in mice.

Murine plasma (10 μL) diluted with sample buffer and denatured. The samples were loaded onto a Protein Simple capillary electrophoresis system for separation (Biotechne, Minneapolis, MN)⁴⁶ and detected with polyclonal anti-human and mouse ADAMTS13 IgG as previously described⁷.

ADAMTS13 activity in the conditioned medium or cell lysate or plasma by FRETS-VWF73 or cattle VWF71:

The proteolytic activity of huma recombinant ADAMTS13 and its variants was assessed through the cleavage of FRETS-rVWF73 as previously described^37,47. The reaction system utilized a buffer containing 5 mM Bis-Tris, 25 mM CaCl₂, 0.005% Tween 20, and 1 mM Pefabloc (Sigma-Aldrich) at pH 6. Murine plasma ADAMTS13 activity was determined by a cattle FRETS-rVWF71 substrate, which was performed in a buffer consisting of 50 mM HEPES, 150 mM NaCl, 10 mM CaCl2, 0.05% Tween-20, and 1 mg/mL bovine serum albumin³⁹. The reaction was performed in a 96-well opaque white Nunc plate (ThermoFisher, Norristown, PA) at 25°C and the rate of fluorescence generation was detected using a SpectraMax Gemini XPS plate reader (Molecular Devices, San Jose, CA).

Cleavage of multimeric VWF by ADAMTS13 under denaturing conditions:

Purified plasma VWF (~150 nM) was incubated with ADAMTS13 proteases in various conditioned media at 37°C for 4 hours on a dialysis membrane (0.25 μm pore size) floating over 50 mL of dialysis buffer (10 mM Tris-HCl, pH 8.0, containing 1.5 M urea) in a conical tube. The digested material was collected from the top of membrane and denatured with a sample buffer (70 mM Tris-HCl, pH 6.5, 2.4% SDS, 0.67 M urea, and 4 mM EDTA) at 60°C for 20 minutes. The denatured VWF was then fractionated using 1% (w/v) SeaKem HGT agarose gel (Cambrex, East Rutherford, NJ) and 5% SDS-PAGE), respectively. The proteins were transferred onto a nitrocellulose membrane, blocked with 5% nonfat milk in TBST, and incubated with anti-VWF IgG (0.2 μg/mL) (Dako North America, Carpinteria CA) for overnight. After three washes with TBST, the bound primary antibody was detected using an IRDye 800CW-labeled goat anti-rabbit IgG (0.1 μg/mL) (LI-COR, Lincoln, NE) in TBST. The fluorescent signal was detected using the Odyssey imaging system (LI-COR), as previously described⁴⁸.

Site-directed mutagenesis.

ADAMTS13 variants (e.g., R193K, R498K, R692K, R1123K and R1206K) were generated using the Q5 site-directed mutagenesis kit (New England Biolabs, Ipswich, MA). All primers are listed in supplemental Table S1. All plasmids were sequenced to confirm the presence of mutations using Sanger sequencing.

Conformational index by ELISAs:

Total rADAMTS13 antigen in the conditioned medium was determined using two different enzyme-linked immunosorbent assays (ELISAs). The rADAMTS13 was captured by anti-metalloprotease (3H9) and detected by a biotinylated anti-TSP1–8 or anti-Spacer (15D1), respectively. This avoids the loss of a antibody binding site resulting from site-directed mutagenesis. The conformations of ADAMTS13 mutants were assessed using the method previously described^49–51. A microtiter plate was coated with antibody (1C4) (5 μg/mL in a bi-carbonate buffer), which specifically recognizes a cryptic epitope in the spacer domain that is only accessible when ADAMTS13 is in its “open” conformation. Serially diluted conditioned medium containing wild type and ADAMTS13 mutants was incubated with immobilized 1C4 at 37°C. As mutations in ADAMTS13 might affect the binding of the anti-CUB1 antibody 17G2, the conditioned medium was preincubated with our anti-TSP1 antibody (18H10) that was shown to induce an “open” conformation of ADAMTS13. Captured “open” ADAMTS13 was detected using a biotinylated anti-MP antibody (3H9) that binds metalloprotease domains, and a HRP-labeled high sensitivity streptavidin (ThermoFisher Scientific, USA). Colorimetric reactions were performed for ten minutes using o-phenylenediamine (OPD) and H₂O₂, and stopped by adding 4 M H₂SO₄. Optical density was measured at 492 nm and relative binding to 1C4 was reported considering “open” ADAMTS13 binding in presence of 18H10 as a reference set as one. Bars indicate mean relative binding (n=3) with error bars showing one standard deviation.

LC-MS/MS:

Recombinant ADAMTS13 protein affinity-purified from HEK293 cells treated with a PRMTs inhibitor MS023 at 0.5 μM for 48 hours or from HEK193 stably overexpressing both recombinant ADAMTS13 and PRMT1 was resolved in 4–20% SDS-PAGE). The specific bands corresponding to the full-length and degraded recombinant ADAMTS13 proteins were cut out from stained gel and digested with trypsin and other proteases for analysis with LC-MS/MS according to the method previously described⁵². The spectrum counts for each peptide with a mono- or a di-methyl group attached were performed and expressed as a function of arginine residues.

Statistical analysis:

All graphs and statistical analyes were performed with Prism 9.0 software (GraphPad, Boston, MA). Student t-test or Mann-Whitney test was used to compare the difference between two groups and and ANOVA one-way variant or Krustal-Wallis test was used for comparing three or more groups. P<0.05 and <0.01 are considered to be statistically significant and highly significant, respectively.

Results

Effect of a PRMT inhibitor on ADAMTS13 secretion and function.

Using the PRmePRed tool (https://bioinfo.icgeb.res.in/PRmePRed/), we predicted based on the RG or an RGG motif that approximately 35 mono-methylated arginine (MMA) residues and 35 asymmetric dimethylated arginine (ADMA) residues may be present on human ADAMTS13 protein (Fig. S1). The arginine residues may be modified by one or two methyl groups (-CH3) using S-Adenosyl methionine (SAM) as a co-substrate, resulting in the generation of MMA and ADMA (Fig. 1A).

Fig. 1. Effects of arginine methylation on ADAMTS13 secretion and proteolytic activity.

A. Depiction of the chemical reaction of arginine methylation. Type I, II, and III PRMTs catalyze a transfer of methyl groups (-CH3) to the guanidino nitrogen moieties in arginine residues of the substrate protein to form monomethyl arginine (MMA), followed by a reaction medidated by type I PRMTs to form asymmetric dimethyl arginine (ADMA). B. ADAMTS13 activity by FRETS-VWF73 in the conditioned medium of stably transfected HEK293 cells treated with a PRMTs inhibitor (MS023) at 0–1.0 μM or a vehicle alone (Ctr). Student t-test was performed to compare the difference between vehicle-treated (Ctrl) and MS023-treated groups. C and D. Western blotting images and densitometric quantification, respectively, of rADAMTS13 secreted in the conditioned medium and in the cell lysate, respectively, of stable HEK293 cells treated with a vehicle control (Ctrl) or MS023 (0.5 μM) for 48 hours. Student t-test was performed to compare the difference between Ctrl and MS023-treated groups in the conditioned medium or in the cell lysate. E and F. Proteolytic activity of the rADAMTS13 (rA13), assessed by the cleavage of a FRETS-VWF73 and a multimeric VWF assay, respectively, in the conditioned medium of stable HEK293 cells treated with a vehicle control (Ctrl) or MS023 (0.5 μM) for 48 hours. Student t-test was performed for statistical analysis. The + and - in panel F indicate the presence and absence, respectively, of 10 mM EDTA. UL and P indicate the uncleaved ultra large VWF multimers and the proteolytic cleavage product (~350 kDa) of VWF, respectively. G. Western blotting demonstrates the rADAMTS13 in the conditioned medium and in the cell lysate following a treatment for 48 hours without (−) or with (+) MS023 (0.5 μM), followed by a washout for 0 and 48-hours. A beta-actin was used for a protein loading control for the cell lysate.

To determine if arginine methylation plays a role in regulating ADAMTS13 secretion and function, we first treated a stably transfected HEK293 cell line expressing a full-length human recombinant ADAMTS13 (rADAMTS13) with an increasing concentration (0–1.0 μM) of a PRMT inhibitor (MS023) for 48 hours. The specific proteolytic activity of the secreted rADAMTS13 in the conditioned medium from MS023-treated cells increased in a concentration-dependent manner compared with that in the vehicle-treated controls (Fig. 1B). These preliminary results suggest that an inhibition of protein arginine methylation on rADAMTS13 may affect its function.

To further determine if the inhibition of arginine methylation affects ADAMTS13 function and protein secretion, stably transfected HEK293 cells expressing rADAMTS13 were treated with a fixed concentration (0.5 μM) of MS023 for 48 hours. ADAMTS13 antigen and proteolytic activity in the conditioned medium and cell lysates were determined with FRETS-VWF73 and Western blotting, respectively. Compared with the vehicle control, MS023 treatment resulted in significantly reduced secretion of rADAMTS13 into the conditioned medium with an accumulation of intracellular pool of rADAMTS13 protein (Fig. 1C & Fig. 1D). Consistent with the early preliminary results, rADAMTS13 secreted into the conditioned medium from MS023-treated cells has a significantly increased specific proteolytic activity towards FRETS-VWF73 (Fig. 1E) and multimeric VWF under denaturing conditions (Fig. 1F) compared with that in the vehicle-treated controls. The inhibition of rADAMTS13 secretion from HEK293 cells was reversible, as the removal of MS023 for 48 hours (washout) restored the secretion of rADAMTS13 into the conditioned medium and normalized the intracellular pool of rADAMTS13 protein (Fig. 1G). These results confirmed that the inhibition of protein arginine methylation with a PRMT inhibitor may affect rADAMTS13 secretion, but the secreted rADAMTS13 in the conditioned medium may have enhanced specific proteolytic activity towards varous VWF substrates.

Effect of overexpression of PRMT1 on ADAMTS13 secretion and its proteolytic activity.

PRMT1 is responsible for more than 80% of protein arginine methylation in mammals^33,34. To assess the impact of PRMT1-mediated arginine methylation on ADAMTS13 secretion and function, we overexpressed a human PRMT1 in stably transfected HEK293 cells expressing rADAMTS13 via a transient or stable transfection of a pcDNA3.0 containing a Flag-PRMT1 or a vector alone. Our result demonstrated that the overexpression of PRMT1 in HEK293 cells resulted in significantly increased secretion of rADAMTS13 with concomitant reduction of an intracellular pool of rADAMTS13 as compared with that in the vehicle-transfected controls (Fig. 2A & 2B). Conversely, the secreted rADAMTS13 in the conditioned medium exhibited a significantly reduced proteolytic activity towards a FRETS-VWF73 substrate (Fig. 2C) but only a slightly reduced activity towars a multimeric VWF under denaturing conditions (Fig. 2D). These results indicate that PRMT1-mediated protein arginine methylation may have a significant impact on rADAMTS13 secretion and proteolytic activity.

Fig. 2. Effect of overexpression of PRMT1 on rADAMTS13 secretion and proteolytic activity.

A and B. Western blotting images and densitometric quantification, respectively, of rADAMTS13 in the conditioned medium and in the cell lysate of stable HEK293 cells overexpressing a Flag-PRMT1 (PRMT1) or a vehicle control (pcDNA). A β-actin protein serves as the protein extraction and loading control of the cell lysate. C and D. Proteolytic activity of rADAMTS13 by a FRETS-VWF73 and a multimeric VWF assay, respectively, in the conditioned medium of stable HEK293 cells overexpressing a Flag-PRMT1 (PRMT1) or a vehicle alone (pcDNA). The – and + signs indicate the absence and presence of 10 mM EDTA, respectively, in the cleavage reaction. Here, UL and P indicate uncleaved ultra large VWF multimers and the proteolytic cleavage product, respectively. Student t-test was performed to compare every two groups either in the conditioned medium or in the cell lysate.

Conformational changes of rADAMTS13 with different methylation status.

To determine if methylation affects rADAMTS13 conformation to expose or hide the cryptic site that is important for substrate recognition, we determined binding of an anti-spacer IgG (1C4) that binds only the “open” rADAMTS13, but not the “closed” one⁴⁹. Additionally, the “closed” to “open” conformation of rADAMTS13 could be induced by binding of an antibody against the first TSP1 repeat (18H10) or anti-CUB antibody such as scFv3-3 as illustrated in Fig. 3A. The results showed that an addition of scFv3-3 to rADAMTS13 in the conditioned medium containing WT, pcDNA vector-transfected, and PRMT1-overexpressed, but not MS023-treated cells resulted in significantly increased proteolytic activity towards the FRETS-VWF73 substrate (Fig. 3B), suggesting that the conformational change can be induced by scFv3-3 for WT and hyper-methylated ADAMTS13.

Fig. 3. Antibody-induced activation or conformational changes of rADAMTS13.

A. Schematic illustration of conformational changes induced by two different antibodies (scFv3.3 and 18H10), which were detected by an ELISA method using anti-S (1C4) that binds cryptic epitope in the spacer domain. B. Relative proteolytic activity of rADAMTS13 of different sources (e.g., NHP, normal human plasma; WT-1 and WT-2, two different preparations of wild type rADAMTS13; PRMT1, rADAMTS13 from HEK293 cells overexpressing Flag-PRMT1; MS023, rADAMTS13 from HEK293 cells treated with a PRMTs inhibitor MS023) in the absence (−) or presence (+) of scFv3-3, an anti-CUB antibody. C. Relative binding index of 1C4 to rADAMTS13 of various sources in the absence (−) or presence (+) of 18H10, an anti-TSP1. Student t-test was performed for comparing the difference between each two groups. All experiments were performed three times. Unpaired student t-test was performed to determine the statistical significance between each two groups without (−) and with (+) scFv3.3 in Panel B or 18H10 in Panel C.

To further confirm the conformational changes of rADAMTS13 with varous methylation status, an anti-TSP1–1 IgG (18H10) was added to native ADAMTS13 in normal human plasma (NHP) and rADAMTS13 in the conditioned medium from WT, PRMT1-overexpressed, and MS023-treated cells. The results showed that 18H10 treatment dramatically enhanced the binding of anti-Spacer (1C4) (Fig. 3C). These results indicate that hypomethylation may allievate the allosteric inhibtion by the distal CUB domain, but not by the adjacent TSP1 repeats.

The effects of inhibition of type I PRMTs or overexpression of PRMT1 on intracellular maturation of rADAMTS13.

To assess if arginine methylation affects intracellular trafficking and posttranslational modifications such as N-linked glycosylation, we performed immunofluorescence, confocal mcroscopic analysis, and glycosidase digestion experiments. We showed that regardless of treatment rADAMTS13 in all HEK293 cells was predominantly localized to endoplasmic reticulum (ER), Golgi aparatus, and trans-Golgi-network (TGN) (Fig. 4A), consistent with the normal intracellular distribution of a secretory protein.

Fig. 4. Intracellular distribution of rADAMTS13 with different methylation states and interaction between ADAMTS13 and PRMT1.

A. Immunofluorescent staining for rADAMTS13 with anti-V5, endoplsmic reticulum calreticulin with anti-calr, and nucleus with DAPI in stable HEK293 cells expressing rADAMTS13 without (WT) or with a PRMTs inhibitor (MS023) or an overexpression of PRMT1 (PRMT1). Fluorescent images were obtained with Niko A1-SHR confocal microscope. B. The rADAMTS13 in the cell lysates or in the conditioned medium of HEK293 cells treated without (WT) or with MS023 or overexpressing PRMT1 as indicated was treated without (−) or with (+) either endo H or PNGase F for 2–4 hours, followed by Western blotting with anti-V5. Arrowheads indicate uncleaved rADAMTS13 (u) and cleaved rADAMTS13 (c). C. Western blotting for rADAMTS13 (rA13) with anti-V5, PRMT1 with anti-Flag, and β-actin with anti-actin, respectively, in the lysate of stable HEK293 without (−) or with (+) overexpression of Flag-PRMT1. D. Western blotting analysis of the anti-V5 pull down products for rADAMTS13 (rA13) with anti-V5, methylated arginines with anit-Methyl-R, total PRMT1 with anti-PRMT1, and overexpressed PRMT1 with anti-Flag, respectively, as indicated, in the cell lysate of stable HEK293 cells expressing rADAMTS13 without (−) or with (+) an overexpression of a Flag-tagged PRMT1.

Endoglycosidase digestion showed that rADAMTS13 in all cell lysates was sensitive to endoglycosidase H (Endo H), suggesting the presence of only high mannose glycan, which occurs in the ER. However, rADAMTS13 in the conditioned media was resistant to Endo H, but sensitive to peptide N-glycosidase F (PNGase F), consistent with the addition of more complex GlcNAc carbohydrates, which occurs in the Golgi and TGN (Fig. 4B). These results demonstrate that either an inhibition or an augmentation of arginine methylation of rADAMTS13 does not significantly affect its posttranslational glycosylation and intracellular trafficking. Other mechanism resulting in reduced rADAMTS13 secretion should be sought in our future study.

Intracellular interaction between PRMT1 and rADAMTS13.

To determine if PRMT1 directly interacts with rADAMTS13, stable HEK293 cells expressing human rADAMTS13 without or with an overexpression of Flag-PRMT1 were lysed and subjected for immunoprecipitation followed by Western blotting. The expression of the Flag-PRMT1 in rADAMTS13-producing cells was confirmed by Western blotting in the cell lysate of HEK293 with or without transfection of a pcDNA3.1 Flag-PRMT1. As expected, the Flag-PRMT1 was not detected in HEK293 cell lysate that did not overexpress an exogenous PRMT1 but detected in the cell lysate overexpressing Flag-PRMT1 (Fig. 4C). These cell lysates were then subjected for immunoprecipitation with anti-V5 resin. The pulldown products were probed with anti-V5, anti-methyl R, anti-PRMT1, and anti-Flag, respectively. As shown, significantly more PRMT1 was detected in the pulldown product with anti-V5 in cells overexpressing the Flag-PRMT1 than that in the vehicle control. Consistently, a stronger methyl-R signal was detected in the anti-V5 pulldown products from cell lysate with than those without an overexpression of the Flag-PRMT1 (Fig. 4D). These results demonstrate the direct physical interaction between PRMT1 and recombinant ADAMTS13 that may explain the function of PRMT1-mediated arginine methylation on ADAMTS13.

Effects of a PRMTs inhibitor or an overexpression of PRMT1 on plasma ADAMTS13 antigen and proteolytic activity in mice.

PRMT1 expression increases following chronic inflammation, immune responses, and cancer^35,53,54. To explore the physiological relevance of PRMT1-mediated arginine methylation on ADAMTS13 secretion and function, we treated wild-type mice (C57/BL6) with MS023 (80 mg/kg body weight) (i.p.) or a vechile, every other day for 14 days to inhibit arginine methylation. Whole blood samples were collected from vehicle- and MS023-treated mice and anticoagulated with sodium citrate (Fig. 5A). Plasma was prepared for analysis of ADAMTS13 antigen by the capillary-based Western blotting⁷ and proteolytic activity by a cattle FRETS-VWF71 assay⁵. In agreement with the results of in vitro cultured cells, plasma levels of ADAMTS13 antigen in MS023-treated mice were significantly reduced (Fig. 5B & 5C), but plasma ADAMTS13 specific activity was significantly increased (p<0.05) compared with those in the vehicle-treated controls (Fig. 5D), suggesting that an inhibition of arginine methylation via a type I PRMT inhibitor resduces ADAMTS13 secretion, but the secreted ADAMTS13 has a relatively increased specific activity in vivo.

Fig. 5. Plasma levels of ADAMTS13 antigen and activity in mice.

A. schematic representation of treatment with MS023 or vehicle control (Ctrl) in 5 WT-mice in each group and their bleeding protocols. B. Blood sample collection from WT and PRMT-1 transgenic mice for analysis. C. Capillary-based Western blotting analysis (WES) for detection of murine plasma Adamts13 antigen with a polyclonal anti-human/murine ADAMTS13 from the mice treated according to the Panel A schedule. D and E. Quantification of relative plasma Adamts13 antigen by WES and proteolytic activity by the cleavage of a cattle-FRETS-VWF71, respectively, in the vehicle controls (Ctrl) and MS023-treated mice. F. Also, WES detection of plasma Adamts13 antigen in WT control (Ctrl.) and PRMT1-transgenic mice. G and H. Quantification of plasma Adamts13 antigen by WES and activity by the cleavage of a cattle-FRETS-VWF71, respectively, in WT and PRMT1-transgenic mice. Here, M1 to M4/5 indicate 4/5 different mice from each group. The data are presented as invidual values, median, and interqurtile range (IQR). Unpaired t-test was performed to determine the statistical significance between the two groups.

Additionally, whole blood samples were collected from wild type (WT) and transgenic mice (Prmt^TG) ectopically expressing a human PRMT1 in megakaryocytes and platelets(Fig. 5E)⁵⁵ for analysis of ADAMTS13 antigen and activity by the methods described in the previous section. The results showed that plasma levels of ADAMTS13 antigen in the Prmt1^TG mice were significantly increased (Fig. 5F & 5G), but plasma ADAMTS13 activity was significantly reduced compared with those in the WT controls (Fig. 5H). These results demonstrate that hypermethatlation of arginine mediated by PRMT1 may signficantly increase ADAMTS13 secretion but reduce its proteolytic activity in vivo.

LC-MS/MS identified arginine methylation sites on ADAMTS13.

To identify the arginine sites on ADAMTS13 that may be modified by a methyl group, a novel LC-MS/MS spectrometric analysis was developed and performed on affinity-purified human rADAMTS13 from the conditioned medium of stably transfected cells treated without or with MS023 (0.5 μM) for 48 hours. The purified human rADAMTS13 was subjected to enzyme digestion and proteomic analysis as described in the Experimental Procedures. Peptides obtained from LC-MS/MS analysis covered nearly 95% of the rADAMTS13 protein sequence. More than 100 arginine sites were identified that contained mono- or di-, or tri-methylated arginine residues indicated in Fig. 6A. The rADAMTS13 purified from the conditioned medium of MS023-treated HEK293 cells exhibited significantly reduced number of arginine methylated peptides in most areas of the protein (Fig. 6B). Examples demonstrating how a methylated arginine in a specific peptide was identified and the relative abundancy of each peptide that is modified by arginine methylation are listed in Fig. 6C and in the supplemental materials (Fig. S2–S5). These results demonstrate the success of unequivocal identification and abundance of mono-, di-, and tri-methyl group on each arginine of human rADAMTS13 via LC-MS/MS.

Fig. 6. Identification of the sites of arginine methylation on ADAMTS13 by LC-MS/MS.

A. Spectral counts of total arginines, mono-, di-, and tri-methylated peptides derived from recombinant ADAMTS13 purified from the conditioned medium of HEK293 cells. Schematic representation of ADAMTS13 domains is shown underneath panel A. B. Spectral counts of arginine-methylated peptides derived from recombinant ADAMTS13 purified from the conditioned medium of HEK293 cells being treated without (Ctrl.) and with MS023 (0.5 μM) for 48 hours. Underneath the panel B is the full-length ADAMTS13 with methylated arginine peptides idenfied (green). C. One example of mass spectrometric identification of methylated arginine such as R193 from overlapping peptides.

The role of several highly conserved methylated arginine residues on ADAMTS13 in its secretion and proteolytic activity.

The RG motif is highly conserved among animal species from zebrafish to mammals. Alignment of ADAMTS13 protein sequences of various animal species allowed us to identify 5 highly conserved arginine residues (e.g., R193, R498, R692, R1123, and R1206) that are potentially methylated (Fig. 7A & 7B) but later confirmed by LC-MS/MS (not shown). To assess their relative contribution to rADAMTS13 secretion and proteolytic activity, site-directed mutagenesis was performed to replace each of these arginine residues with a lysine residue. The resulting rADAMTS13 variants were transiently expressed in HEK293 cells, and the conditioned medium and cell lysate were collected 48 hours following the transient transfection for analysis. Western blotting analysis detected all recombinant ADAMTS13 variants except for R1123K in the conditioned medium and all recombinant ADAMTS13 variants in the cell lysate (Fig. 7C). The proteolytic activity of all ADAMTS13 variants except for R193K in the cell lysates was significantly reduced when compared with that in the WT controls (Fig. 7D). Also, the proteolytic activity of all ADAMTS13 variants except for R193K was reduced in the conditioned medium. Interestingly, the proteolytic activity of the R1206K variant was ~4-fold higher than that of the WT control (Fig. 7E). These results demonstrated that a specific arginine residue or alteration of the methylation status on certain residues of ADAMTS13 may be crucial for its secretion and proteolytic function.

Fig. 7. Identification of several conserved methylated arginine residues on ADAMTS13 among various species and their role in ADAMTS13 secretion and activity.

A. Schematic domain representation of full-length human ADAMTS13 protein, consisting of metalloprotease domain (MP), disintegrin (Dis), the first thrombospondin type 1 (TSP1) repeat (1), Cysteine-rich (Cys), and spacer domain. The distal C-terminus contains 7 more TSP1 repeats (2–8) and CUB domains. B. Comparison of 5 highly conserved consensus RG motifs (highlighted in red) from zebrafish to mammals that are confirmed by LC/MS/MS to be methylated. C. Western blotting analysis of the expression WT and five variants of recombinant ADAMTS13 (rA13) in the conditioned medium and cell lysate of transfected HEK293 cells. The β-actin is used for protein extraction and loading controls of cell lysates. D and E. Relative proteolytic activity by FRETS-VWF73, normalized to the ADAMTS13 antigen of each construct, in the cell lysates and the conditioned medium, respectively, of HEK293 cells expressing WT and ADAMTS13 variants (e.g., R193K, R498K, R1123K, and R1206K). The data shown are the mean ± standard error of the mean (SEM). ANOVA one-way variant analysis was performed to determine the statistical significance.

Potential conformational changes of rADAMTS13 variants may explain the increased specific activity.

Under physiological conditions, ADAMTS13 primarily exhibits a “closed” or “latent” conformation, which can be activated (or “opened”) by binding to an antibody against the distal C-terminal domains (Fig. 3A)⁵⁶ or by lowering pH in the assay conditions (Fig. 8A)⁵⁷. To better understand how rADAMTS13 variant (R1206K) exhibited an enhaced activity, we determined the conformational changes of WT and R1206K by lowering pH to 6.0, or by an addition of anti-CUB antibody (scFv3-3), isolated from a patient with iTTP^56,58. The concentration of rADAMTS13 protein in the conditioned medium of various constructs was quantified with two different ELISA methods illustrated in the supplmental Figure S6A. No rADAMTS13 was detected in the conditioned medium following transfection with R1123K construct, suggesting lack of protein secretion. Lower rADAMTS13 concentration for R1206K constructu by method 1 than method 2, suggesting R1206K mutation affects anti-TSP1–8, but not anti-Spacer binding (supplemental Figure S6B). The R1206K variant exhibited 3~4-fold increased specific activity towards a FRETS-VWF-73 substrate compared with the WT control; lowering pH from 7.0 to 6.0 significantly increased the proteolytic activity of WT but not that of the R1206K variant (Fig. 8B). Also, an addition of scFv3.3 also signicantly increased the proteolytic activity of WT but not that of the R1206K variant (Fig. 8C). Moreover, MS023 treatment of HEK293 cells expressing the WT construct resulted in a signicantly increased proteolytic activity in the conditioned medium, but such a treatment had no effect on the R1206K variant (Fig. 8D). Conversely, an overexpression of PRMT1 signicantly reduced the proteolytic activity of WT, but not had no effect on the R1206K variant (Fig. 8E). Together, these results suggest that the R1206K variant may exhibit an “open” conformation that may expose the spacer domain where the primary substrate recognition site is located.

Fig. 8. Conformational changes of WT and ADAMTS13 variants induced under various conditions.

A. Schematic representation how low pH (6.0) or anti-CUB (scFv3-3), or a mutation in the distal C-terminal domain of ADAMTS13 (e.g., R1206K) may affect ADAMTS13 conformations. B and C. The effect of lowering pH (6.0) and binding of scFv3-3, respectively, on the proteolytic activity of WT and ADAMTS13 variant (R1206K). D and E. Proteolytic activity of WT and R1206K in the conditioned medium of HEK293 cells being treated a vehicle (Ctrl) or MS023 at final concentration of 0.5 μM and in the conditioned medium of cells overexpressing a Flag-PRMT1 or a vector (pcDNA) alone, respectively. The data are shown as the mean ± standard errors of the mean (SEM). Student t-test was performed to determine the statistical significance between each two groups. F. ELISA detects the binding of native ADAMTS13 in normal human plasma (NHP), wild type rADAMTS13 (WT), and ADAMTS13 variants to anti-S (1C4) in the absence (open bars) or presence (closed bars) of anti-T1 (18H10) antibody. Student t-test was performed to determine the statistical significance of the differences between ADAMTS13 activity without (−) and with (+)18H10.

To confirm the exposure of cryptic sapcer domain in R1206K variant, a binding index was determined using an ELISA-based method described^49,51. The results showed that in the presence of anti-TSP1–1 (18H10), binding of anti-Spacer (1C4) to native ADAMTS13 in normal human plasma, WT-rADAMTS13, and several rADAMTS13 variants (e.g., R193K, R498K, and R692K) in the conditioned medium was significantly increased. Interestingly, even in the absence of 18H10 the binding of 1C4 to rADAMTS13 (R1206K) variant in the conditioned medium was better than that of WT and other variants. However, addition of 18H10 resulted in only modest but statistically significant increase of binding of R1206K to 1C4 (Fig. 8F). These results indicate that the R1206K variant appears to be in its partially “open” conformation, likely resulting from the disruption of CUB and spacer interaction, in which the cryptic spacer domain is at least partially exposed, explaining the increased proteolytic activity towards the VWF substrates.

Discussion

Arginine methylation by PRMT1 affects protein functions in diverse biological processes, including cell cycle progression⁵⁹, transcriptional regulation⁶⁰, and cellular differentiation⁶¹, as well as protein trafficking and secretion⁶². The present study demonstrates for the first time that ADAMTS13 is post-translationally modified by arginine methylation. The arginine methylation on ADAMTS13 by PRMT1 and other PRMTs may play a critical role in regulation of ADAMTS13 secretion and its enzymatic activity under various pathophysiological conditions.

First, using a stably transfected cell line, we show that significantly reduced secretion of rADAMTS13 following the treatment with a PRMTs inhibitor (MS023) in a concentration-dependnet manner. Then, we show that an overexpression of PRMT1 that is responsible for 80% of arginine methylation in mammalian cells results in significantly increased secretion of rADAMTS13 in stably transfected cells. However, the mechanism underlying the altered secretion of rADAMTS13 is not fully understood, which may be rather complicated. MS023 or PRMT1 may alter the methylation status not only on rADATMS13 but also on many other intracellular proteins including synthesis, trafficking, and degradation machineries. Our preliminary results also demonstrate that overexpression of PRMT1 upregulates rADAMTS13 mRNA expression (not shown). Thus, further investigations into specific cellular components or pathoways affected by the PRMTs inhibitors or PRMT1 may help elucidate the mechanism underlying altered secretion of rADAMTS13 under these conditions.

Interestingly, the secreted rADAMTS13 into the conditioned medium of stable cell lines following the treatment with a PRMTs inhibitor (MS023) exhibits a significantly increased specific activity towards peptidyl and native VWF substrates. These results suggest that arginine methylation on rADAMTS13 may activate the protease although the molecular mechanism underlying such an enhanced specific activity is not fully understood.

Native ADAMTS13 ro full-length rADAMTS13 in solution may exist in multiple conformations from a more “closed” (or latent) to a more “open” (or activated) conformation⁵⁷. The various conformations are allosterically controlled by the interactions between the proximal domains (e.g., spacer) and the distal domains (e.g., distal TSP1 or CUB) of ADAMTS13^49,63–65. Previous studies demonstrate that lowering pH to 6.0^47,57, or binding of a VWF-D4 domain to the distal C-terminal domains of ADAMTS13⁶⁶, or binding of an anti-CUB antibody^56,57,65 or truncation of the distal C-terminal domains^65,67 appears to cause conformational changes that activate ADAMTS13⁶⁸. In our study, a hypometylated rADAMTS13 (with MS023) may exhibit a more open conformation that binding of scFv3-3 does not increase its proteolytic activity compaed with the WT, and hypermethylated rADAMTS13 (with overexpression of PRMT1). Interesting, both native and recombinant ADAMTS13 from untreated (WT), MS023-treated, and PRMT1 overexpressed HEK293 cells are all sensitive to anti-TSP1–1 that results in exposure of cryptic epitope in the spacer domain, suggesting anti-TSP1–1 induced local conformational change. These results indicate that arginine methylation may regulate the interaction between the proximal and distal domains, rather than the domain-domain interaction withint the proximal regions of ADAMTS13. Further investigation into the conformational changes of purified rADAMTS13 using hydrogen-deuterium exchange plus mass spectrometry (HX-MS) or cryo-electron microscopy may help elucidate the precise structural alterations associated with hypo- or hypermethylation of the arginine residues in rADAMTS13 protein.

Arginine methylation based on purified rADAMTS13 appears to be common in cultured mammalian cells. Over 100 arginine residues are identified to be mono-, di-, and tri-methylated. The methylated arginine residues span the entire rADAMTS13 molecule. Such a widespread distribution of methylated arginine residues may suggest the potential diverse functions of ADAMTS13 methylation including its synthesis, secretion, degradation, trafficking, ligand or substrate binding, and autoregulation of its conformation and activity. The arginine methylation sites (e.g., R193, R498, R692, R1123, R1206) on ADAMTS13 are highly conserved among various species from zebrafish to mammals. Alteration in some of these highly conserved arginine residues on ADAMTS13 may affect its secretion and enzymatic function. Notably, recombinant ADAMTS13 R1206K demonstrate a reduced secretion but significantly increased specific activity. The increased specific activity of R1206K may be associated with change from a closed to more open conformations, demonstrated by the increased binding to anti-S IgG. In patients with congenital TTP, mutations such as R193W in the metalloprotease domain, R692C in the TSP-1 domain, R1123C in the TSP-8 domain, and R1206strop in the CUB1 domain of ADAMTS13 result in severe deficiency of ADAMTS13 activity^69–72, although changing from arginine to cysteine residue or introduction of a stop may have a profound change in ADAMTS13 mRNA stability and protein structure due to aberrant disulfide bond formation or non-sense-mediated ADAMTS13 mRNA decay.

To assess if arginine methylation on ADAMTS13 is biologically important, we determined plasma Adamts13 antigen and activity in WT mice treated with MS023 or in transgenic mice ectopically overexpressing human PRMT1⁵⁵. We found significant reduced plasma Adamts13 antigen but increased its proteolytic activity in MS023-treated mice or significantly increased plasma Adamts13 antigen, but reduced its proteotic activity in the transgenic mice compared with those in the WT controls. These results suggest that arginine methylation by PRMT1 may modulate ADAMTS13 secretion and function under various (patho)physiological conditions.

Plasma ADAMTS13 is primarily synthesized and secreted from hepatic stellate cells^8,73 and endothelial cells^74,75 and little if any in megakaryocytes and platelets^76,77, in which PRMT1 is overexpressed under the control of PF4 promoter. It is not known at the present time how the overexpressed PRMT1 in one cell type (e.g., megakaryocytes and platelets) would affect ADAMTS13 secretion and function in another cell type (e.g., hepatic stellate and endothelial cells). One possibility is that endothelial cells or hepatic stellate cells uptake the PRMT1-bearing platelets or microparticles via a receptor-mediated endocytosis. Such a phenotypic modification in brain endothelial cells via endocytosis of platelet-derived microparticles has been reported⁷⁸.

Regulation of ADAMTS13 secretion and function may have a profound implication in hemostasis and inflammation. ADAMTS13 is of both antithrombotic and anti-inflammatory function by proteolytic cleavage of VWF and inhibition of platelet adhesion and aggregation on activated or injured endothelium^79,80. Severe deficiency of plasma ADAMTS13 activity, resulting from mutations in ADAMTS13 or acquired autoantibodies against ADAMTS13, leads to thrombotic thrombocytopenic purpura (TTP), a potentially fatal blood disorder^81,82. Mild to moderate deficiency of plasma ADAMTS13 activity is a risk factor for the development of many other inflammatory and thrombotic disorders including myocardial infarction^83,84, stroke^15,85, preeclampsia^86,87, and trauma-associated coagulopathy¹⁶, etc. Low levels of plasma ADAMTS13 and high levels of plasma VWF, resulting in reduced ratios of ADAMTS13/VWF, predict mortality in patients with traumatic brain injury^14,88,89, severe COVID-19^90,91, and suspected heparin-induced thrombocytopenia⁹². Thus, understanding the regulation of ADAMTS13 synthesis, secretion, and function is of clinical importance for diagnosis and management of patients with inflammatory and thrombotic disorders.

Conclusions

We conclude that our results unveil an intricate and novel relationship between PRMT1-mediated arginine methylation and ADAMTS13 secretion and function. The impact of arginine methylation on ADAMTS13 secretion and enzymatic activity may have profound clinical relevance for better understanding of normal hemostasis, inflammation, and thrombosis under various pathological conditions. Moreover, an inhibition of arginine methylation by PRMTs inhibitors or a modification of certain conserved arginine residues on ADAMTS13 through site-directed mutagenesis may allow us to identify potential gain-of-functional ADAMTS13 variants for therapeutic purpose.

Highlights.

• ADAMTS13 is a critical enzyme required to cleave von Willebrand factor and regulate hemostasis.

• ADAMTS13 undergoes a posttranslational modification by arginine methylation, which affects ADAMTS13 secretion and proteolytic activity.

• Inhibition of arginine methylation on ADAMTS13 or change of arginine residue to a lysine at the position of 1206 significantly enhances its activity.

• These resutls may shed new light on how a posttranslational modification affects ADAMTS13 biology.

Abbreviation

ADAMTS13

A disintegrin and metalloprotease with thrombosponin type 1 repeats, 13

TTP

thrombotic thrombocytopenic pupura

VWF

von Willebrand factor

COVID-19

Coronavirus disease 2019

NETs

neutrophil extracellular traps

TNF

tissue necrosis factor

HNP1

human neutrophil peptide-1

PRMT1

protein arginine methyltransferase 1

VWF73 or VWF71

a 73 or 71 amino acid peptide derived from central A2 domain of VWF

FRETS

fluorescence resonance energy transfers

SDS-PAGE

sodium dodecyl sulfate polyacrylamide gel electrophoresis

ELISA

enzyme-linked immunosorbent assay

MS023

a type I methyl transferase inhibitor

PVDF

polyvinylidene fluoride

OPD

o-phenylenediamine

DMSO

dimethyl sulfoxide

MMA

mono-methylated arginine

ADMA

asymmetric dimethylated arginine

DMA

di-methylated arginine

Data availability

All data generated or analyzed during this study are included in this article or in the supplementary information files.