Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Feb 1.
Published in final edited form as: J Thromb Haemost. 2018 Jan 8;16(2):330–341. doi: 10.1111/jth.13920

Expression and Functional Characterization of Two Natural Heparin-binding Site Variants of Antithrombin

Peyman Dinarvand †,1, Likui Yang ¶,2, Bruno O Villoutreix ‡, Alireza R Rezaie ¶,*
PMCID: PMC5809256  NIHMSID: NIHMS923212  PMID: 29215785

Summary

Background

Several heparin-binding site (HBS) variants of antithrombin (AT) have been identified which predispose carriers to higher incidence of thrombosis. The cause of thrombosis in carriers of HBS variants has been primarily attributed to a loss in their heparin-dependent anticoagulant function.

Objective

The objective of this study was to determine whether HSB mutations affect the anti-inflammatory functions of variants.

Methods

Two HBS variants of AT (AT-I7N and AT-L99F), which are known to be associated with higher incidence of thrombosis, were expressed in mammalian cells and purified to homogeneity. These variants were characterized by kinetic assays followed by analysis of their activities in established cellular and/or in vivo inflammatory models. The possible effects of mutations on the AT structure was also evaluated by molecular modeling.

Results

Results indicate that while progressive inhibitory activities of variants have been minimally affected, their heparin affinity and inhibitory activity in the presence of heparin has been markedly decreased. Unlike wild-type, neither AT variant was capable of inhibiting activation of NF-κB or downregulation of expression of cell adhesion molecules in response to LPS. Similarly, neither variant elicited barrier protective activity in response to LPS. Structural analysis suggests that the L99F substitution locally destabilizes the AT structure.

Conclusions

It is concluded that the L99F mutation of AT is associated with destabilization of the serpin structure and that the loss of anti-inflammatory signaling function of the HBS variants may also contribute to enhanced thrombosis in carriers of HSB mutations.

Keywords: antithrombin, HSPG, heparin-binding, signaling, inflammation

Introduction

Antithrombin (AT) is a plasma inhibitor of the serpin superfamily, which regulates the proteolytic activity of coagulation proteases of the clotting cascade both in intrinsic and extrinsic pathways (1–4). It circulates in plasma with a concentration of ~0.125 mg/ml (~2.5 μM) and its hereditary or acquired deficiency is associated with an increased risk of thrombosis (5,6). Complete AT deficiency in mice results in embryonic lethality (7). AT is a heparin-binding serpin possessing a basic D-helix and the binding of a distinct pentasaccharide (H5) sequence of heparin containing 3-O-sulfate (3-OS) modification to this helical motif induces a conformational change that activates the serpin, thereby improving its reactivity with vitamin K-dependent coagulation proteases by ~100–500-fold (1,8,9). Interestingly, however, the activation of AT by H5 plays a minimal role in promoting the AT inhibition of thrombin and longer chain heparins are required to enhance the protease inhibition by a bridging (template) mechanism (10). This mechanism of heparin-mediated AT inhibition of coagulation proteases has been extensively studied by in vitro assays. It has been traditionally assumed that this model of heparin-mediated coagulation protease inhibition by AT can also occur in vivo by AT binding to specific vascular heparan sulfate proteoglycans (HSPGs) lining the vessel wall. In support of this hypothesis, it has been shown that a small population of vascular HSPGs (1–5%) has glycosaminoglycans (GAGs) containing H5 sequences with the same characteristic 3-OS modification (11,12). Nevertheless, other than in the therapeutic setting, there is hardly any in vivo data in the literature to firmly support this mechanism of vascular GAG-mediated AT inhibition of coagulation proteases under different pathophysiological conditions.

In addition to its anticoagulant function through direct inhibition of procoagulant proteases, AT also possesses potent anti-inflammatory signaling activities (13–16). The anti-inflammatory function of AT is mediated through the same D-helix-dependent interaction of the serpin with 3-OS containing vascular HSPGs (13–17). In contrast to its poorly understood physiological role in the anticoagulant pathway, the critical role of D-helix-dependent interaction of AT with 3-OS containing vascular GAGs in the signaling function of the serpin has been relatively well established. Thus, results from numerous studies have indicated that the binding of AT to vascular GAGs elicits potent anti-inflammatory responses by inducing prostacyclin (PGI2) production that culminates in inhibition of NF-κB activation, down-regulation of the expression of vascular cell adhesion molecules and inhibition of synthesis of pro-inflammatory cytokines (13–18). At least 12 distinct heparin-binding site (HBS) mutations of AT (type II deficiency) which are associated with higher incidence of venous and/or arterial thrombosis have been thus far identified (5). It has been demonstrated that progressive inhibitory activities of HBS mutants, purified from patients’ plasmas, have been minimally affected, thus the underlying basis of thrombosis in carriers of these mutations has been primarily attributed to the loss of their heparin cofactor-dependent protease inhibitory function. The possible effect of such mutations on the anti-inflammatory function of HBS variants has not been investigated.

To address this question, we expressed two natural HBS variants of AT which are associated with higher incidence of venous/arterial thrombosis in heterozygous and homozygous carriers (5,19). The two AT mutants are Ile-7 to Asn (I7N, Rouen III) and Leu-99 to Phe (L99F, Budapest III) substitutions neither of which is actually located on D-helix, nevertheless, both variants exhibit low affinities for heparin (19–21). Characterization of these mutants revealed that both mutants have lost their anti-inflammatory signaling activities in response to pro-inflammatory stimuli as determined by several established assay systems. Computational and kinetic analysis predict a destabilizing effect for the L99F mutation on the structure of the serpin. Results of this proof-of-concept study support for the hypothesis that the loss of D-helix-dependent signaling function of HBS variants may also contribute to enhanced thrombosis in carriers under certain pathophysiological conditions.

Materials and Methods

Expression and purification of AT variants

Recombinant human AT (AT-WT) was expressed in HEK-293 cells using the RSV-PL4 expression/purification vector system as described (22). The same vector system was used to prepare two AT variants in which residues Ile-7 and Leu-99 of the serpin were replaced with either Asn (AT-I7N) or Phe (AT-L99F) in two separate constructs. AT derivatives were purified to homogeneity from 20-liters of cell culture supernatants by immunoaffinity chromatography using the HPC4 monoclonal antibody linked to Affi-gel 10 (Bio-Rad) as described (22). Concentrations of AT derivatives were determined from the absorbance at 280 nm using a molar absorption coefficient of 37,700 M−1 cm−1 and by stoichiometric titration of serpins with calibrated concentrations of thrombin as described (22,23). Complete list of other reagents and details of experimental methods have been presented as Supplementary Materials.

Inhibition assays

Rate of inactivation of proteases by AT derivatives in both the absence and presence of pentasaccharide (H5) and heparin was measured under pseudo-first-order rate conditions by a discontinuous assay method and observed pseudo-first-order (kobs) and second-order association rate constants (k2(app)) for uncatalyzed and catalyzed reactions were calculated as described (22,23). This assay was also used to evaluate thermal stability of AT derivatives (1 μM) after their incubation at 56 °C for 0–60 min.

Permeability assay

The EA.hy926 endothelial cell permeability in response to LPS (15 ng/mL) in the absence or presence of increasing concentrations of AT derivatives (0.67–5.0 μM for 4h) was assessed by spectrophotometric measurement of the flux of Evans blue-bound albumin across functional cell monolayers by a modified two-compartment chamber model as described (24).

Analysis of expression of cell adhesion molecules

The cell surface expression levels of ICAM-1 and VCAM-1 on endothelial cells were measured by a whole-cell based ELISA as described (25).

NF-κB assay

The activation of NF-κB in nuclear lysates of LPS-treated endothelial cells with or without pre-treatment with AT derivatives (0.67–2.5 μM, 4h) was measured using an ELISA kit (Cell Signaling Technology, Inc, Danvers, MA) according to manufacturers’ protocol.

In vivo permeability and leukocyte migration assays

Each group of mice (N=10) was injected intraperitoneally with 200 μL of AT derivatives (0.04 mg/g body weight) or with 200 μL normal saline as a control. After 3h, 1% BSA-bound Evans blue dye in normal saline was injected intravenously followed by immediate intraperitoneal injection of LPS (2 mg/kg bodyweight). Thirty minutes later, mice were sacrificed, peritoneal exudates were collected and vascular permeability was measured from leakage of dye (650 nm) into peritoneal cavity (26). For analysis of leukocyte migration, 20 μL of peritoneal fluid was mixed with 0.38 mL of Turk’s solution (0.01% crystal violet in 3% acetic acid) and number of leukocytes was counted under a light microscope (26).

Statistical analysis

Data were expressed as means ± S.D. from at least three independent experiments. Group data were compared using analysis of variance (ANOVA) followed by Bonferroni post-hoc tests (for comparisons among three or more groups) or Student’s t-test, as appropriate. p values of <0.05 were considered statistically significant.

Structural analysis and estimation of ΔΔG values

The two amino acid substitutions were investigated using x-ray crystal structures of both native (PDB entry 2beh) (27) and pentasaccharide-activated AT (PDB entry 1azx) (28) and of pentasaccharide-AT in complex with FXa-S195A (PDB entry 2gd4) (29) using PyMOL molecular graphics system (Schrödinger, LLC). Residue interaction networks (30) were computed with RING (31). Stability predictions (ΔΔG values) upon mutations were estimated using three different and complementary approaches. Structure-based prediction engine DUET (32), which combines two different methods named mCSM and SDM (see Suppl. Materials) was used for analysis. Computations were also performed with structure-based PoPMuSiC method (33), and with the sequence and structure-based INPS-MD approach (34).

Results

Expression and characterization of AT derivatives by kinetic assays

The AT derivatives were expressed in HEK-293 cells and purified to homogeneity by a combination of immunoaffinity and heparin-Sepharose chromatography (22). In contrast to AT-WT, which eluted at ~1 M NaCl from the heparin column, AT-I7N and AT-L99F were eluted at lower NaCl concentrations of ~0.3 M and ~0.5 M, respectively. These results are in agreement with previous observations that these mutants exhibit much weaker affinity for heparin (5,19–21). Both mutants reacted with near normal or slightly lower rate constants with both FXa and thrombin in the absence of a cofactor (Table 1). Thus, the concentration-dependence of AT inhibition of FXa yielded similar k(obs) (Fig. 1A) and k2(app) values for inhibition of the protease by AT derivatives (Table 1). However, both AT variants exhibited significantly lower rate constants for FXa in the presence of H5 (Fig. 1B), thus yielding k2(app) values which were ~8-fold (I7N) and ~47-fold (L99F) lower than the rate of protease inhibition by AT-WT (Table 1). Heparin concentration-dependence of inhibition of proteases by AT mutants (Fig. 1C,D) indicated the reactivity of mutants with both FXa and thrombin has been similarly impaired (Table 1). Thus, at optimal concentrations of heparin (Fig. 1C,D), the reactivity of AT-I7N with FXa and thrombin was impaired ~13-fold and ~20-fold, respectively (Table 1). The same values for the AT-L99F inactivation of both FXa and thrombin at optimal concentrations of heparin were lower than AT-WT by ~40-fold (Table 1).

Table 1.

Second-order rate (k2(app)) and equilibrium dissociation (KD) constants for the inhibition of FXa and thrombin by AT derivatives in the absence and presence of heparins.

FXa −Hep (103 M−1s−1) +H5 (105 M−1s−1) +Hep (107 M−1s−1) KD (Hep) nM KD (H5) nM
AT-WT 2.5 ± 0.2 6.6 ± 0.3 8.7 ± 0.5 12.5 ± 1.8 19.1 ± 2.1
AT-I7N 2.3 ± 0.1 0.83 ± 0.05 0.65 ± 0.04 113 ± 15 118 ± 17
AT-L99F 2.4 ± 0.2 0.14 ± 0.01 0.22 ± 0.02 ND ND
Thrombin
AT-WT 7.5 ± 0.4 - 8.3 ± 0.3 - -
AT-I7N 6.2 ± 0.3 - 0.40 ± 0.06 - -
AT-L99F 5.7 ± 0.5 - 0.21 ± 0.01 - -
Open in a new tab

All k2 values for the AT inhibition of FXa and thrombin were determined by a discontinuous assay and KD values were determined from changes in the intrinsic protein fluorescence upon titration with heparins as described under “Materials and Methods”. All values are averages of three independent measurements ± S.D. ND, not determined; H5, pentasaccharide; Hep, heparin.

Figure 1.

Figure 1

Open in a new tab

Analysis of the reactivity of AT vraints with FXa and thrombin in the absence and presence of heparins. (A) The pseudo-first order rate constants (kobs) for the inhibition of FXa by AT-WT (○), AT-I7N (●) and AT-L99F (□) were determined from the time-dependent inhibition of the protease by different concentrations of AT derivatives in TBS/Ca2+ as described under “Materials and Methods”. (B) The same as panel A except that the rates were measured in the presence of a saturating concentration of pentasaccharide (2 μM H5). Solid lines are best fit of kinetic data to a linear equation. (C) Heparin concentration dependence of FXa (●) and thrombin (○) inhibition by AT-I7N was monitored and k2 values were calculated as described under “Materials and Methods”. (D) The same as panel C except that the heparin-concentration dependence of proteases by AT-L99F was monitored. k2(app) values from three independent measurements are presented in Table 1.

The binding of heparin to D-helix of AT is associated with an enhancement in the intrinsic fluorescence intensity of the serpin (35). Analysis of the heparin-dependent intrinsic protein fluorescence changes in AT-WT and AT-I7N yielded equilibrium dissociation constants (KD) of 12.5 nM and 113 nM, respectively (Fig. 2A and Table 1). Thus, consistent with kinetic data, the affinity of the AT mutant for interaction with heparin was decreased by ~an order of magnitude. Same results were obtained if KD values were measured for H5 binding to serpins, thus, AT-I7N exhibited ~6-fold lower affinity for the pentasaccharide (Table 1). The same analysis for AT-L99F was not successful since repeated experiments in TBS (ionic strength 0.1) with this mutant did not result in any significant intrinsic protein fluorescence change upon binding heparin. These results suggested that L99F mutation might be associated with dramatic structural changes in the mutant serpin. In agreement with this hypothesis, when measuring the inhibition rate constants, we noted that the stability of AT-L99F has been negatively impacted. This was evidenced by the mutant showing lower and variable activity from day to day and upon freeze thawing of the serpin. Moreover, less than 10% of the protein, recovered from the affinity column, exhibited activity even though the protein was essentially homogenous on SDS-PAGE (data not shown). To investigate this question further, we compared the thermal stability of the L99F mutant by a kinetic assay by incubating the serpin at 56 °C and comparing its protease inhibitory activity to AT-WT. Analysis of results supported our hypothesis that AT-L99F is highly unstable since unlike AT-WT, most of the inhibitory activity of the L99F mutant was abolished by 60 min (Fig. 2B,C). It appears that the rate of conversion of AT-L99F to the latent form has been markedly enhanced in the mutant serpin. These results suggested that the L99F mutation is associated with dramatic structural changes in the mutant serpin. Further analysis revealed that the apparent stoichiometry of inhibition (SI) by L99F has also been elevated ~2–3-fold for thrombin and >5-fold for FXa, suggesting that the reactivity of both proteases with the mutant serpin in the substrate pathway may have been increased. However, an accurate estimation of SI values proved to be challenging since the inhibitory activity of the mutant was not stable during the incubation time and if heparin was included in the assay the rate of the reaction in the substrate pathway for FXa was also increased. The I7N mutation did not exhibit an increase in the SI values with either protease (data not shown).

Figure 2.

Figure 2

Open in a new tab

Binding to heparin and analysis of the thermal stability of recombinant AT derivatives. (A) The spectral changes were monitored by addition of 1–2 μL of a concentrated stock solution of heparin to 50 nM AT in TBS (pH 7.5) and dissociation constants (KD) were calculated from the changes of the intrinsic protein fluorescence as described under “Materials and Methods”. The symbols are: (○) AT-WT and (●) AT-I7N. (B) The thermal stability of the AT derivatives (panel A for AT-WT and panel B for L99F) were assessed from their ability to inhibit the amidolytic activity of thrombin after incubation at 56 °C for difference time (x-axis) as described under “Methods”.

Structural analysis and estimation of ΔΔG values

The residue Ile-7 is located in a turn in the N-terminal heparin-binding region of the AT (Suppl. Figs. S1 and S2). The side chain of this residue is essentially solvent-exposed in the native AT structure while it is more buried in the heparin-activated AT form (Suppl. Fig. S1). This residue is next to the C8–C128 disulfide bond (Fig 3A). This region is partially flexible and structural changes upon heparin binding were noticed (Suppl. Fig. S2). There is space for accommodating an Asn side chain in this environment and in silico mutagenesis experiments performed with PyMOL showed that most of the known rotamers for Asn could fit in the native AT structure without creating severe atomic clashes (Fig 3A). The residue interaction networks (Suppl. Figs. S3 and S4) show that Ile-7 has a few non-covalent contacts with hydrophobic residues and is near several positively charged amino acids present in this region of AT. The node degree (non-covalent interactions such as hydrogen bonds and van-der-Waals contacts) for Ile-7 is small, having a value of 2. Several residues surrounding Ile-7 also have a low node degree. In general, substitution of a residue with a small node degree in such an environment is predicted to be structurally tolerated. The predicted ΔΔG stability changes for the I7N substitution were: DUET = −0.91 kcal/mol (mCSM = −0.8 kcal/mol; SDM = −0.90 kcal/mol); PoPMuSiC = +0.7 kcal/mol (positive ΔΔG sign with this method indicates destabilization); and INPS-MD = −0.5 kcal/mol. Thus, all methods suggest that the I7N substitution should only be moderately destabilizing (Fig. 3A).

Figure 3.

Figure 3

Figure 3

Open in a new tab

Overview of the AT structures- (A) Structural analysis of AT (PDB entry 2beh) with selected side chains in the vicinity of Ile-7 are labeled. The image is centered on Ile-7 (magenta). Ile-7 was mutated to Asn in silico (lower corner). The side of mutant residue is readily accommodated in the structure. The estimated stability changes upon mutation also suggest that the substitution could be tolerated and the scores predict that the amino acid change is only slightly destabilizing. Right panel. A sequence logo was generated using WebLogo sequence logo generator (44). The AT sequences from 5 different species (human, mouse, bovine, sheep, monkey) are shown. The residue Ile-7 is not fully conserved but is generally replaced by a hydrophobic residue. (B) Structural analysis of Leu-99 (in magenta). The side chains of selected residues surrounding Leu-99 are shown in cyan while a glycan side chain grafted on Asn-96 is colored in grey. Substitution of Leu-99 by a Phe was carried out on in silico and for all the available Phe rotamers, important steric clashes were noticed. These clashes are visualized by small red disks and only the side chains of Leu-99 (magenta) and of Phe (white) and of the disulfide bon C21–C95 are shown to simplify the figure. Right panel. The sequence logo shows that C95, N96, T98, L99, Q101, L102, L103 and E104 are strictly conserved in the AT sequences. The estimated stability changes computed suggest that the substitution should be destabilizing.

Leu-99 is located on the N-terminus of the C-helix, at some distance from the heparin-binding D-helix of the serpin (Suppl. Fig. S1). Leu-99 in all three different AT structures is buried (relative solvent accessibility ~3.6%) and located in a primarily hydrophobic environment (Suppl. Figs. S1 and S5). This region is stabilized by the C21–C95 disulfide bond and the nearby Asn-96 residue is glycosylated (Fig. 3B). The b-factors on the C-helix are relatively low as compared to other regions of the protein suggesting that the area is relatively rigid. There is little free space around position 99 to accommodate a larger side chain such as Phe. Thus, a Leu-99 to Phe in silico mutagenesis performed with PyMOL, with or without energy minimization, shows severe steric clashes for all Phe rotamers available, not only with some surrounding side chains but also and mainly with backbone atoms (Fig. 3B). Analysis of the residue interaction networks (Suppl. Figs. S3 and S4) shows that Leu-99 is in part surrounded by hydrophobic residues and that it is involved in 8 non-covalent interactions (node degree) with surrounding residues. Residues near Leu-99 also tend to have relatively high node degrees. Substitution of such a residue is in general destabilizing and can induce some severe local structural changes. This expected destabilization is further supported by the predicted ΔΔG stability changes as the L99F substitution led to a destabilizing score of −1.7 kcal/mol with the DUET method (mCSM = −1.5 kcal/mol; SDM = −0.84 kcal/mol) while for PoPMuSiC it was +0.9 kcal/mol (positive ΔΔG with this method indicates destabilization), and −0.97 kcal/mol for INPS-MD (Fig. 3B).

Analysis of the anti-inflammatory activity of serpin variants

Next, the anti-inflammatory activity of serpin variants was evaluated by several established cell-based assays. LPS is known to induce activation of the nuclear transcription factor, NF-κB, and disrupt the barrier permeability function of endothelial cells. We have demonstrated that AT elicits a protective effect in both assays, thus reversing LPS-mediated pro-inflammatory processes by a concentration dependent manner (36). Interestingly, unlike AT-WT, neither AT-L99F (Fig. 4A) nor AT-I7N (Fig. 4B) exhibited any modulatory effect on the NF-κB pathway in response to LPS. Similarly, neither AT-L99F (Fig. 5A) nor AT-I7N (Fig. 5B) exhibited any barrier protective activity in the LPS-mediated barrier hyper-permeability assay. Analysis of LPS-mediated cell surface expression of cell adhesion molecules on endothelial cells pretreated with AT derivatives indicated that unlike AT-WT, which inhibited the expression of both ICAM-1 and VCAM-1 in response to LPS, neither AT-L99F nor AT-I7N exhibited protective activity in these cellular assays (Fig. 6).

Figure 4.

Figure 4

Open in a new tab

Analysis of NF-B activation in LPS-stimulated endothelial cells pretreated with different variants of AT. NF-B activation in LPS-stimulated endothelial cells pretreated with AT-wild type and AT-L99F B) Same as A except that NF-B activation was measured on cells pretreated with either AT-wild type or AT-I7N. All results are shown as means ± SD of three different experiments. Control, cells incubated with growth media only. * p < 0.05, ** p < 0.01.

Figure 5.

Figure 5

Open in a new tab

Effects of serpin derivatives on the barrier permeability function of endothelial cells in response to proinflammatory stimuli (LPS). A) Cell permeability in response to LPS was measured by spectrophotometric measurement of the flux of Evans blue-bound albumin across functional endothelial cell monolayer in the absence and presence of increasing concentrations of AT variants (WT or L99F) as described in Materials and methods. B) Same as A except that cell permeability in response to LPS was measured in the presence of increasing concentrations of either AT-Wild-type or AT-I7N. All results are shown as means ± SD of three different experiments. Control, cells incubated with growth media only. * p < 0.05, ** p < 0.01.

Figure 6.

Figure 6

Open in a new tab

LPS-mediated expression of ICAM1 and VCAM1 in HUVECs in the absence and presence of treatment by AT variants. A) Confluent endothelial cells were incubated with AT variants (WT and L99F) followed by LPS treatment. Surface expression of ICAM-1 is measured by a cell-based enzyme-linked immunosorbent assay (ELISA). B) The same as (A), except that AT variants (WT and I7N) were used for endothelial cell activation. C and D) The same as panels A and B except that measurement were con ducted for VCAM1. All results are shown as mean ± standard deviation of 3 different experiments. Control, cells incubated with growth media only. * p < 0.05.

AT-WT but not AT-I7N inhibits the pro-inflammatory effect of LPS in vivo

A potent anti-inflammatory effect for AT has been observed in several in vivo inflammatory models (14,16,37,38). Here we set up an LPS-mediated vascular permeability model by intraperitoneal injection of LPS into mice, followed by measuring vascular permeability based on the extravasation of the BSA-bound Evans blue dye from plasma into the peritoneal cavity (26). LPS dramatically increased endothelium leakiness, similar to the effect of using a solution of 0.7% acetic acid as a positive control (Fig. 7). While consistent with the literature (14,37,38), AT-WT exhibited a therapeutic effect and inhibited LPS-mediated vascular permeability, AT-I7N exhibited no barrier protective activity in this model (Fig. 7A). Similarly, AT-WT, but not AT-I7N, markedly inhibited LPS-mediated binding of leukocytes to the vascular endothelium and their subsequent migration to the peritoneal cavity (Fig. 7B). A 1.5% carboxymethylcellulose-sodium (CMC-Na) solution was used as a positive control for leukocyte migration in this model (26). The in vivo permeability study could not be conducted for AT-L99F due to inability to prepare a sufficient quantity of active mutant since, as described above, only a small fraction of the expressed protein exhibited activity, which was highly unstable and temperature sensitive.

Figure 7.

Figure 7

Open in a new tab

In vivo analysis of protective effects of serpins on HMGB1-mediated vascular leakage and leukocyte infiltration. A) Each group of mice (n=10 for each group) were intraperitoneally injected with AT-Wild type and AT-I7N (0.04 mg/g body weight). After 3 h, mice were intravenously injected with 1% BSA-bound Evans blue dye followed by an immediate intraperitoneal injection of LPS (2 mg/kg) or 0.7% acetic acid as a positive control and normal saline (as a negative control). Vascular permeability was determined from the extent of extravasation of Evans blue to the peritoneal cavity. B) Inhibitory effect of AT variants (0.04 mg/g body weight) on the migration of leukocytes to peritoneal cavity in response to LPS was analyzed. All results are shown as means ± SD. * p < 0.05, ** p < 0.01.

Discussion

AT deficiency has been phenotypically divided into type I deficiency (quantitative), which is characterized by equally low antigen and activity levels, and type II deficiency (qualitative), which is characterized by a lower activity level for the serpin (5). The type II deficiency may be caused by a lower RCL-dependent inhibitory function of AT toward coagulation proteases and/or by loss of the heparin cofactor function due to mutations in the heparin-binding site (HBS) of the serpin which weakens its affinity for heparin (5). The homozygous AT deficiency of both types are not compatible with life, with the exception of two HBS mutations (L99F and R47C) which predispose the carriers to higher incidence of thrombosis (5,19,20,39,40). A recent clinical and laboratory study of a relatively large number of families carrying the L99F mutation has established a founder effect for this mutation (19). It has been found that both heterozygous and homozygous carriers of the L99F mutation experience higher incidence of thrombosis, though the frequency and severity of the homozygous carriers were significantly higher (19). Results from several laboratories using patients’ plasma-derived AT have established that the progressive inhibitory activity of AT-L99F with coagulation proteases has been minimally affected but the capacity of the AT mutant to inactivate these proteases in the presence of heparin has been diminished due to its lower affinity for heparin. The same results have been obtained with several other HBS mutants of AT (at least 12 have been identified) including a heterozygous patient who has been found to carry an Ile-7 to Asn substitution (AT-I7N) (21). The underlying cause of thrombosis in such patients has been primarily attributed to inability of these mutants to effectively regulate the proteolytic activity of procoagulant proteases of the clotting cascade due to their inability to interact with vascular GAGs, thereby presumably not being able to optimally inhibit coagulation proteases.

Since all of the studies with the two HBS variants have been conducted with limited quantities of patients’ plasma-derived AT, we initiated this study and prepared sufficient amounts of recombinant serpin variants for a detailed characterization of their activities in different assay systems. Our results suggest that while consistent with the published data the progressive inhibitory activities of the serpin variants have not been significantly affected, nevertheless, the protective signaling function of both variants has been essentially eliminated. Thus, neither HBS variant under study exhibited a protective anti-inflammatory activity in any one of the established cellular assays described above. Unlike AT-WT, neither one of the AT variants was capable of inhibiting NF-κB activation or expression of cell adhesion molecules in endothelial cells in response to LPS. Neither mutant exhibited barrier-protective activity in response to LPS. Moreover, the I7N variant did not exhibit anti-inflammatory activity in vivo in response to LPS-mediated barrier permeability dysfunction and inflammatory leukocyte infiltration. The in vivo experiments with L99F mutant was not feasible due to a low recovery of the active form of this variant as well its very poor stability. We found that the AT-L99F mutation results in destabilization, temperature sensitivity and an elevated rate of recognition of the serpin as a substrate by the coagulation proteases. Thus, L99F mutation not only negatively affects the interaction of D-helix with 3-OS containing vascular GAGs, but it also induces dramatic structural changes in the mutant serpin (see below). The 3-OS moiety of GAGs is added by the gene encoding heparan sulfate 3-O-sulfotransferase-1 (3-OST-1) (41). Interestingly, it has been demonstrated that 3-OST-1 deficient mice have normal hemostasis and exhibit comparable phenotypes as in wild-type mice in response to prothrombotic challenges (17,42). However, 3-OST-1 deficient mice exhibit severe pro-inflammatory phenotypes in response to LPS challenge that is ameliorated by exogenous AT, providing support for the hypothesis that interaction of AT with 3-OS containing vascular GAGs may actually be responsible for the anti-inflammatory rather than the protease inhibitory function of AT (17,42). We also demonstrated that the siRNA knockdown of 3-OST-1 inhibits the protective signaling function of AT in cellular assays (16). Recent targeting of the type II HBS variant, Arg-47 to Cys mutation, to the mouse genome resulted in life-threatening spontaneous thrombosis with only a fraction of mice surviving to adulthood (43). Interestingly, the thrombin-AT level was similar to that of wild-type in the transgenic mice, possibly indicating that the inhibition of thrombin may not require interaction with GAGs. Thus, the traditional view, which is mostly based on in vitro and therapeutic heparin data, that AT binding to vascular GAGs is required for the AT regulation of coagulation proteases may require further investigation and validation since we have not been able to find conclusive in vivo evidence in the literature that the vascular GAG-binding function of AT is required for its coagulation protease inhibitory function. In light of these results and the established concept that D-helix-dependent interaction of AT with 3-OS containing GAGs elicits a potent barrier protective effect in vascular endothelial cells that might be essential for maintaining the vascular tone, the results of this study suggest that the higher incidence of thrombosis, caused by HBS mutations, may also be contributed to by the loss of the anti-inflammatory signaling function of AT under certain pathophysiological conditions (i.e., acute infection). Thus, further investigation will be needed to determine the extent to which the conformational activation and/or a bridging effect of vascular GAG binding contributes to the protease inhibitory and anti-inflammatory signaling functions of AT. We are in the process of developing appropriate model systems to investigate this important question.

Finally, the structural analysis of the effects of mutations by several types of computational approaches suggest that the I7N substitution could be structurally tolerated and that the defect is only limited to the glycosylation of the mutant Asn residue (21), which impedes the interaction of the serpin variant with heparin. By contrast, the structural analysis of the L99F mutation indicated that the mutation is highly destabilizing. Residue Leu-99 is buried in a relatively rigid environment on the C-helix, which also harbors the glycan-linked Asn-96 and is stabilized by the C21–C95 disulfide bond (Fig. 3B). While Leu-99, C21–C95 disulfide bond and Asn-96 are not conserved in the serpin family, they are nevertheless strictly conserved in the AT sequences from different species (Fig. 3B) highlighting their importance in the AT structure/function. There is little free space around position 99 to accommodate a larger side chain such as Phe. A Phe at this position would not only clash against nearby side chains but also with backbone atoms, most likely perturbing the folding of this area. It appears that this structural perturbation is allosterically transmitting the negative conformational effect of the mutation to the basic residues of the heparin-binding D-helix. Whether the L99F mutation alters the formation of the nearby disulfide bond and/or interferes with the linkage of a glycan side chain on Asn-96 needs further investigation. The destabilizing effect predicted by this interactive structural analysis is further supported by the investigation of the residue interaction networks (Suppl. Figs. S3 and S4) and the ΔΔG computational approaches, all of which suggest that the L99F mutation should be destabilizing. Thus, in addition to affecting the heparin affinity of the AT variant and eliminating its anti-inflammatory signaling properties, the L99F mutation also decreases the thermal stability of the mutant serpin.

Supplementary Material

Supplementary Materials

Essentials.

  1. Heparin-binding site (HBS) variants of AT are associated with a higher incidence of thrombosis.

  2. HSB variants have in general normal progressive inhibitory activity but reduced heparin affinity.

  3. Thrombosis in HSB carriers has been primarily attributed to the loss of heparin cofactor activity.

  4. Results here demonstrate that HSB variants of AT also lack anti-inflammatory signaling functions.

Acknowledgments

We thank continuous financial supports from the Inserm and University Paris Diderot to BOV. The authors also thank Audrey Rezaie for editorial work on the manuscript.

Funding Sources

This research was supported by grants awarded by the National Heart, Lung, and Blood Institute of the National Institutes of Health HL 101917 and HL 62565 to A.R.R.

Footnotes

Addendum

P.D. conducted signaling assays and animal experiments; L.Y. constructed, expressed and characterized the AT mutants; B.O.V. performed molecular modeling and A.R.R. designed experiments, analyzed data and wrote the paper. All authors approved the final version of the manuscript.

Disclosure of Conflict of Interests

The authors declare no conflict of interests.

Supplementary Materials

1. List of reagents and methods for fluorescence measurements and in silico computations of stability changes.

2. Supplementary figures S1–S5 for structural analysis of AT.

References

  • 1.Olson ST, Richard B, Izaguirre G, Schedin-Weiss S, Gettins PG. Molecular mechanism of antithrombin regulation of blood clotting proteinases. A paradigm for understanding proteinase regulation by serpin family protein proteinase inhibitors. Biochimie. 2010;92:1587–96. doi: 10.1016/j.biochi.2010.05.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Gettins PGW. Serpins structure, mechanism, and function. Chem Rev. 2002;102:4751–803. doi: 10.1021/cr010170+. [DOI] [PubMed] [Google Scholar]
  • 3.Damus PS, Hicks M, Rosenberg RD. Anticoagulant action of heparin. Nature. 1973;246:355–7. doi: 10.1038/246355a0. [DOI] [PubMed] [Google Scholar]
  • 4.Carrell RW, Skinner R, Jin L, Abrahams JP. Structural Mobility of Antithrombin and its Modulation by Heparin. Thromb Haemost. 1997;78:516–9. [PubMed] [Google Scholar]
  • 5.Lane DA, Bayston T, Olds RJ, Fitches AC, Cooper DN, Millar DS, Jochmans K, Perry DJ, Okajima K, Thein SL, Emmerich J. Antithrombin mutation database: 2nd (1997) update. For the Plasma Coagulation Inhibitors Subcommittee of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Thromb Haemost. 1997;77:197–211. [PubMed] [Google Scholar]
  • 6.Lane DA, Olds RR, Thein SL. Antithrombin and its deficiency states. Blood Coagul Fibrinolysis. 1992;3:315–41. doi: 10.1097/00001721-199206000-00012. [DOI] [PubMed] [Google Scholar]
  • 7.Ishiguro K, Kojima T, Kadomatsu K, Nakayama Y, Takagi A, Suzuki M, Takeda N, Ito M, Yamamoto K, Matsusita T, Kusugami K, Muramatsu T, Saito H. Complete antithrombin deficiency in mice results in embryonic lethality. J Clin Invest. 2000;106:873–8. doi: 10.1172/JCI10489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Huntington JA, McCoy A, Belzar KJ, Pei XY, Gettins PWG, Carrell RW. The conformational activation of antithrombin. A 2. 85-A structure of a fluorescein derivative reveals an electrostatic link between the hinge and heparin binding regions. J Biol Chem. 2000;275:15377–83. doi: 10.1074/jbc.275.20.15377. [DOI] [PubMed] [Google Scholar]
  • 9.Belzar KJ, Zhou A, Carrell RW, Gettins PGW, Huntington JA. Helix D elongation and allosteric activation of antithrombin. J Biol Chem. 2002;277:8551–8. doi: 10.1074/jbc.M110807200. [DOI] [PubMed] [Google Scholar]
  • 10.Danielsson A, Raub E, Lindahl U, Björk I. Role of ternary complexes, in which heparin binds both antithrombin and proteinase, in the acceleration of the reactions between antithrombin and thrombin or factor Xa. J Biol Chem. 1986;261:15467–73. [PubMed] [Google Scholar]
  • 11.Marcum JA, Rosenberg RD. Anticoagulantly active heparin-like molecules from the vascular tissue. Biochemistry. 1984;23:1730–7. doi: 10.1021/bi00303a023. [DOI] [PubMed] [Google Scholar]
  • 12.de Agostini AI, Watkins SC, Slayter HS, Youssoufian H, Rosenberg RD. Localization of anticoagulantly active heparan sulfate proteoglycans in vascular endothelium: antithrombin binding on cultured endothelial cells and perfused rat aorta. J Cell Biol. 1990;111:1293–304. doi: 10.1083/jcb.111.3.1293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Oelschläger C, Römisch J, Staubitz A, Stauss H, Leithäuser B, Tillmanns H, Hölschermann H. Antithrombin III inhibits nuclear factor kappaB activation in human monocytes and vascular endothelial cells. Blood. 2002;99:4015–20. doi: 10.1182/blood.v99.11.4015. [DOI] [PubMed] [Google Scholar]
  • 14.Harada N, Okajima K, Uchiba M, Kushimoto S, Isobe H. Antithrombin reduces ischemia/reperfusion-induced liver injury in rats by activation of cyclooxygenase-1. Thromb Haemost. 2004;92:550–8. doi: 10.1160/TH03-07-0460. [DOI] [PubMed] [Google Scholar]
  • 15.Bae JS, Rezaie AR. Mutagenesis studies toward understanding the intracellular signaling mechanism of antithrombin. J Thromb Haemost. 2009;7:803–10. doi: 10.1111/j.1538-7836.2009.03337.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Ma Y, Wang J, Gao J, Yang H, Wang Y, Manithody C, Li J, Rezaie AR. Antithrombin up-regulates AMP-activated protein kinase signaling during myocardial ischaemia/reperfusion injury. Thromb Haemost. 2015;113:338–49. doi: 10.1160/TH14-04-0360. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Shworak NW, Kobayashi T, De Agostini A, Smits NC. Anticoagulant heparan sulfate: To not clot-or not? Prog Mol Biol Transl Sci. 2010;93:153–78. doi: 10.1016/S1877-1173(10)93008-1. [DOI] [PubMed] [Google Scholar]
  • 18.Opal SM. Therapeutic rationale for antithrombin III in sepsis. Crit Care Med. 2000;28:S34–7. doi: 10.1097/00003246-200009001-00008. [DOI] [PubMed] [Google Scholar]
  • 19.Gindele R, Oláh Z, Ilonczai P, Speker M, Udvari Á, Selmeczi A, Pfliegler G, Marján E, Kovács B, Boda Z, Muszbek L, Bereczky Z. Founder effect is responsible for the p. Leu131Phe heparin-binding-site antithrombin mutation common in Hungary: phenotype analysis in a large cohort. J Thromb Haemost. 2016;14:704–15. doi: 10.1111/jth.13252. [DOI] [PubMed] [Google Scholar]
  • 20.Olds RJ, Lane DA, Boisclair M, Sas G, Bock SC, Thein SL. Antithrombin Budapest 3. An antithrombin variant with reduced heparin affinity resulting from the substitution L99F. FEBS Lett. 1992;300:241–6. doi: 10.1016/0014-5793(92)80854-a. [DOI] [PubMed] [Google Scholar]
  • 21.Brennan SO, Borg JY, George PM, Soria C, Soria J, Caen J, Carrell RW. New carbohydrate site in mutant antithrombin (7 Ile----Asn) with decreased heparin affinity. FEBS Lett. 1988;237:118–22. doi: 10.1016/0014-5793(88)80183-2. [DOI] [PubMed] [Google Scholar]
  • 22.Rezaie AR, Yang L. Probing the molecular basis of factor Xa specificity by mutagenesis of the serpin, antithrombin. Biochim Biophys Acta. 2001;1528:167–76. doi: 10.1016/s0304-4165(01)00189-1. [DOI] [PubMed] [Google Scholar]
  • 23.Olson ST, Björk I, Shore JD. Kinetic characterization of heparin-catalyzed and uncatalyzed inhibition of blood coagulation proteinases by antithrombin. Methods Enzymol. 1993;222:525–60. doi: 10.1016/0076-6879(93)22033-c. [DOI] [PubMed] [Google Scholar]
  • 24.Bae JS, Yang L, Manithody C, Rezaie AR. The ligand occupancy of endothelial protein C receptor switches the protease-activated receptor 1-dependent signaling specificity of thrombin from a permeability-enhancing to a barrier-protective response in endothelial cells. Blood. 2007;110:3909–16. doi: 10.1182/blood-2007-06-096651. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Hassanian SM, Dinarvand P, Rezaie AR. Adenosine regulates the proinflammatory signaling function of thrombin in endothelial cells. J Cell Physiol. 2014;229:1292–300. doi: 10.1002/jcp.24568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Dinarvand P, Hassanian SM, Qureshi SH, Manithody C, Eissenberg JC, Yang L, Rezaie AR. Polyphosphate amplifies proinflammatory responses of nuclear proteins through interaction with receptor for advanced glycation end products and P2Y1 purinergic receptor. Blood. 2014;123:935–45. doi: 10.1182/blood-2013-09-529602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Johnson DJ, Langdown J, Li W, Luis SA, Baglin TP, Huntington JA. Crystal structure of monomeric native antithrombin reveals a novel reactive center loop conformation. J Biol Chem. 2006;281:35478–86. doi: 10.1074/jbc.M607204200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Jin L, Abrahams JP, Skinner R, Petitou M, Pike RN, Carrell RW. The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA. 1997;94:14683–8. doi: 10.1073/pnas.94.26.14683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Johnson DJ, Li W, Adams TE, Huntington JA. Antithrombin-S195A factor Xa-heparin structure reveals the allosteric mechanism of antithrombin activation. EMBO J. 2006;25:2029–37. doi: 10.1038/sj.emboj.7601089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Doncheva NT, Klein K, Morris JH, Wybrow M, Domingues FS, Albrecht M. Integrative visual analysis of protein sequence mutations. BMC Proc. 2014;8:S2. doi: 10.1186/1753-6561-8-S2-S2. eCollection 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Piovesan D, Minervini G, Tosatto SCE. The RING 2. 0 web server for high quality residue interaction networks. Nucleic Acids Res. 2016;44:W367–74. doi: 10.1093/nar/gkw315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Pires DE, Ascher DB, Blundell TL. DUET: a server for predicting effects of mutations on protein stability using an integrated computational approach. Nucleic Acids Res. 2014;42:W314–9. doi: 10.1093/nar/gku411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Dehouck Y, Kwasigroch JM, Gilis D, Rooman M. PoPMuSiC 2. 1: a web server for the estimation of protein stability changes upon mutation and sequence optimality. BMC Bioinformatics. 2011;12:151. doi: 10.1186/1471-2105-12-151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Savojardo C, Fariselli P, Martelli PL, Casadio R. INPS-MD: a web server to predict stability of protein variants from sequence and structure. Bioinformatics. 2016;32:2542–4. doi: 10.1093/bioinformatics/btw192. [DOI] [PubMed] [Google Scholar]
  • 35.Meagher JL, Beechem JM, Olson ST, Gettins PG. Deconvolution of the fluorescence emission spectrum of human antithrombin and identification of the tryptophan residues that are responsive to heparin binding. J Biol Chem. 1998;273:23283–9. doi: 10.1074/jbc.273.36.23283. [DOI] [PubMed] [Google Scholar]
  • 36.Bae JS, Rezaie AR. Mutagenesis studies toward understanding the intracellular signaling mechanism of antithrombin. J Thromb Haemost. 2009;7:803–10. doi: 10.1111/j.1538-7836.2009.03337.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Minnema MC, Chang AC, Jansen PM, Lubbers YTP, Pratt BM, Whittaker BG, Taylor FB, Hack EC, Friedman B. Recombinant human antithrombin III improves survival and attenuates inflammatory responses in baboons lethally challenged with Escherichia coli. Blood. 2000;95:1117–23. [PubMed] [Google Scholar]
  • 38.Hagiwara S, Iwasaka H, Matsumoto S, Noguchi T. High dose antithrombin III inhibits HMGB1 and improves endotoxin-induced acute lung injury in rats. Intensive Care Med. 2008;34:361–7. doi: 10.1007/s00134-007-0887-5. [DOI] [PubMed] [Google Scholar]
  • 39.Koide T, Odani S, Takahashi K, Ono T, Sakuragawa N. Antithrombin III Toyama: replacement of arginine-47 by cysteine in hereditary abnormal antithrombin III that lacks heparin-binding ability. Proc Natl Acad Sci USA. 1984;81:289–93. doi: 10.1073/pnas.81.2.289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kuhle S, Lane DA, Jochmanns K, Male C, Quehenberger P, Lechner K, Pabinger I. Homozygous antithrombin deficiency type II (99 Leu to Phe mutation) and childhood thromboembolism. Thromb Haemost. 2001;86:1007–11. [PubMed] [Google Scholar]
  • 41.Shworak NW, Liu J, Fritze LM, Schwartz JJ, Zhang L, Logeart D, Rosenberg RD. Molecular cloning and expression of mouse and human cDNAs encoding heparan sulfate D-glucosaminyl 3-O-sulfotransferase. J Biol Chem. 1997;272:28008–19. doi: 10.1074/jbc.272.44.28008. [DOI] [PubMed] [Google Scholar]
  • 42.HajMohammadi S, Enjyoji K, Princivalle M, Christi P, Lech M, Beeler D, Rayburn H, Schwartz JJ, Barzegar S, de Agostini AI, Post MJ, Rosenberg RD, Shworak NW. Normal levels of anticoagulant heparan sulfate are not essential for normal hemostasis. J Clin Invest. 2003;111:989–99. doi: 10.1172/JCI15809. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Dewerchin M, Hérault JP, Wallays G, Petitou M, Schaeffer P, Millet L, Weitz JI, Moons L, Collen D, Carmeliet P, Herbert JM. Life-threatening thrombosis in mice with targeted Arg48-to-Cys mutation of the heparin-binding domain of antithrombin. Circ Res. 2003;93:1120–6. doi: 10.1161/01.RES.0000103634.69868.4F. [DOI] [PubMed] [Google Scholar]
  • 44.Crooks GE, Hon G, Chandonia JM, Brenner SE. WebLogo: a sequence logo generator. Genome Res. 2004;14:1188–90. doi: 10.1101/gr.849004. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Materials
NIHMS923212-supplement-Supplementary_Materials.pdf (802.9KB, pdf)

RESOURCES