Skeletal muscle myosin and cardiac myosin attenuate heparin’s antithrombin-dependent anticoagulant activity

Abstract

Background:

Heparin enhances the ability of the plasma protease inhibitor, antithrombin, to neutralize coagulation factor Xa and thrombin. Skeletal muscle myosin binds unfractionated heparin.

Objectives:

The aim of this study was to investigate the influence of myosin binding to heparin on antithrombin’s anticoagulant activity.

Methods:

Inhibition of factor Xa and thrombin by antithrombin in the presence of different heparins and skeletal muscle myosin or cardiac myosin was studied by measuring inhibition of each enzyme’s chromogenic substrate hydrolysis.

Results and Conclusions:

Skeletal muscle myosin and cardiac myosin neutralized unfractionated heparin’s enhancement of antithrombin’s inhibition of purified factor Xa and thrombin. Skeletal muscle myosin also reduced the inhibition of factor Xa and thrombin by antithrombin in the presence of heparan sulfate. These two myosins did not protect factor Xa from antithrombin inhibition when tested in the presence of smaller heparins, e.g., low molecular weight heparin or heparin pentasaccharide. This chain length dependence for skeletal muscle myosin’s ability to reduce heparin’s anticoagulant activity might have potential implications for therapy for patients who experience increases in plasma myosin levels, e.g., acute trauma patients. In addition to the chain length, the type and extent of sulfation of glycosaminoglycans influenced the ability of skeletal muscle myosin to neutralize the polysaccharide’s ability to enhance antithrombin’s activity. In summary, these studies show that skeletal muscle myosin and cardiac myosin can influence antithrombin’s anticoagulant activity against factor Xa and thrombin, implying that they may significantly influence the hemostatic balance involving bleeding versus clotting.

1. INTRODUCTION

Research on the role of some striated muscle myosins, namely skeletal muscle myosin (SkM) and cardiac myosin (CM), in the pathophysiology of blood coagulation is a promising new area [1–8]. Structurally, SkM and CM are similar and exist as dimers of heterotrimers and consist of six polypeptide chains: two heavy chains (HC), and two pairs of two light chains (essential light chain (ELC) and regulatory light chain (RLC)) [3, 8–10].

Recently, SkM rare exomic variations were identified as associated with increased venous thrombosis risk through a rare exomic variant genotyping study [7, 8]. SkM and CM were shown to exert procoagulant effects by binding factor (F) Xa and FVa, thereby forming a surface for the assembly of the prothrombinase complex [1, 6]. In addition, SkM was found to enable negative feedback downregulation of thrombin generation by acting as a cofactor for activated protein C’s proteolytic inactivation of FVa [4]. SkM can bind to von Willebrand factor, suggesting another possible procoagulant role by localizing FVIII wherever von Willebrand factor is localized, e.g., on activated platelets [5]. Myosin from rabbit muscle binds heparin with a K_d of 0.3 μM [11]. Specifically, the sub-fragment 1 (S1) of SkM interacts with heparin [12]. However, these reports have had no recent studies, and the effects of SkM’s interactions with heparin have not been assessed for any mechanistic effects on heparin’s antithrombin (AT)-dependent anticoagulant properties.

Heparin is a naturally occurring heterogeneous sulfated glycosaminoglycan with an average chain length of 50 saccharide units, and unfractionated heparin (UFH) is used clinically as an anticoagulant both for prophylaxis and to prevent the extension of existing thrombi in venous thrombosis, pulmonary embolism, myocardial infarction, unstable angina, hemofiltration, and pre- and post-surgeries [13–17]. Although UFH is a very potent anticoagulant, it has several limitations, including wide variability among patients’ response leading to the need for continued monitoring, poor pharmacokinetic properties, and the risk of developing severe side-effects such as heparin-induced thrombocytopenia [14, 15]. This led to the development of heparin derivatives, low molecular weight heparin (LMWH, average chain length of 20 saccharide units) and a synthetic heparin pentasaccharide (H5) [14].

UFH exerts its anticoagulant activity by binding and producing a conformational change in the serpin AT, thus converting AT from a slow and progressive anticoagulant inhibitor of procoagulant proteases to a rapid inhibitor, especially of thrombin and FXa [16]. This binding of heparin to AT requires a unique pentasaccharide sequence (H5), which is only present in one-third of all UFH molecules [18]. To efficiently inhibit thrombin, heparin must bind both thrombin and AT, forming a ternary complex, thus requiring a minimum chain length of 18 saccharide units [15, 16]. This is why LMWH and H5 do not enhance AT’s anticoagulant effect towards thrombin. In contrast, the binding of heparin to FXa is believed not to be essential for its inhibition by the AT-heparin complex [16]. Each form of heparin, notably even H5, binds AT and alters its conformation to make it a more efficient inhibitor of FXa [18]. In the presence of Ca⁺⁺, however, the binding of heparin to FXa may enhance the anticoagulant effect of AT in more physiological settings [19]. When inhibition of FXa and thrombin by AT are compared, UFH has an anti-FXa to anti-thrombin ratio of 1:1, while LMWHs have a ratio of 2:1 to 4:1, depending on the type of LMWH preparation [16, 19]. Notably, because of its short chain length, H5 does not substantially promote inhibition of thrombin by AT but enhances only the inhibition of FXa [14].

Here, we report that both SkM and CM can prevent UFH-assisted AT inhibition of FXa and thrombin. Myosin–heparin interactions appear to be dependent on the type, sulfation, and chain length of the glycosaminoglycan. SkM also prevents inhibition of FXa and thrombin by AT in the presence of heparan sulfate (HS), presenting a possible novel procoagulant role for myosin.

2. METHODS

2.1. Materials

Human FXa, thrombin, AT, and bovine lactadherin were purchased from Hematologic Technologies (Essex Junction, VT). FXa chromogenic substrate (equal parts of Bz-Ile-Glu(OCH₃)-Gly-Arg-pNA.HCl and Bz-Ile-Glu(OH)Gly-Arg-pNA.HCl) was from Aniara Diagnostica (West Chester, OH) and thrombin chromogenic substrate (H-D-Phe-Pip-Arg-pNA.2HCl) from Molecular Innovations (Novi, MI). Rabbit SkM, unfractionated heparin (UFH), low molecular weight heparin (LMWH, enoxaparin), N-acetyl-de-O-sulfated heparin (NAc-deO-Hep), and fatty acid-free and protease-free bovine serum albumin (BSA) were from Sigma Aldrich (St. Louis, MO). Bovine CM was from Cytoskeleton, Inc. (Denver, CO), heparin pentasaccharide (H5, fondaparinux) from Fisher Scientific (Hampton, NH), heparan sulfate (HS) and chondroitin sulfate (CS) from MedChemExpress (Princeton, NJ). Phosphatidylserine (PS) and phosphatidylcholine (PC) were purchased from Avanti Polar Lipids (Alabaster, AL), and recombinant human annexin V from Biovision (Milpitas, CA).

2.2. Phospholipid vesicles

Phospholipid vesicles containing 80% PC and 20% PS were prepared as previously described [20].

2.3. Serine protease inhibition assays

FXa (5 nM, final) or thrombin (5 nM, final) was incubated with AT (25 nM, final) and UFH (0.02 μg/mL, final) or LMWH (0.1 μg/mL and 0.5 μg/mL, final) or H5 (10 nM and 50 nM, final) or HS (0.1 μg/mL with FXa and 1 μg/mL with thrombin, final) in Tris-buffered saline (TBS) containing 0.5% BSA and 5 mM Ca⁺², pH 7.4 at room temperature in the presence of varying concentrations of SkM or CM or PC/PS. After 10 minutes, residual serine protease activity was measured by monitoring the rate of chromogenic substrate hydrolysis. The concentrations of UFH, LMWH and H5 were chosen to obtain an approximate 50% and 90% inhibition of the proteases.

For experiments without heparins, higher concentrations of AT (500 nM for FXa inactivation and 300 nM for thrombin inactivation, final) were used and the reaction mixture incubated for 30 minutes at room temperature.

For competition experiments with other carbohydrate polymers, FXa (5 nM, final) was incubated with AT (25 nM, final), UFH (0.02 μg/mL, final) and CS (0.2–20 μg/mL, final) or NAc-deO-Hep (0.2–20 μg/mL, final) in TBS containing 0.5% BSA and 5 mM Ca⁺² at room temperature in the presence of varying concentrations of SkM for 10 minutes. FXa residual activity was measured by monitoring the rate of chromogenic substrate hydrolysis

2.4. Statistical analysis

SD and EC₅₀ values were determined using Prism™ 8.2.1 (GraphPad Software Inc., San Diego, CA).

3. RESULTS

3.1. SkM reduces UFH-dependent stimulation of inhibition of FXa and thrombin by AT

To assess the effect of heparin’s binding to SkM on its anticoagulant properties, FXa and thrombin inhibition assays by AT in the presence of UFH were carried out. SkM substantially reduced the inhibition of both FXa and thrombin by AT in the presence of UFH in a dose-dependent manner (Figure 1A). The EC₅₀ values for the protective effect of SkM were 65 ± 6 nM for FXa and 54 ± 7 nM for thrombin (Table 1). In control experiments lacking AT, the amidolytic activities of FXa and thrombin were not influenced by SkM alone or CM alone in the presence or absence of UFH (data not shown).

Figure 1: SkM prevents the inhibition of FXa and thrombin by AT in the presence, but not in the absence, of UFH.

FXa or thrombin was incubated with AT in the presence of SkM either (A) with or (B) without UFH in Tris-buffered saline with 0.5% BSA, 5 mM CaCl₂, pH 7.4 at room temperature. Residual protease activity was determined using an appropriate chromogenic substrate. Each value represents the mean ± SD of at least three individual determinations.

Table 1:

Kinetics for SkM’s ability to protect FXa and thrombin from inhibition by AT-UFH and AT-HS.

SkM		EC50 (nM)	ΔY (%)
AT-UFH	FXa	65 ± 6	61 ± 5
	Thrombin	54 ± 7	58 ± 6
AT-HS	FXa	9 ± 3	38 ± 3
	Thrombin	26 ± 8	33 ± 5

EC₅₀ (concentration for 50% inhibition) and ΔY (maximum reduction of inhibition) refer to potency and efficacy, respectively. Error refers to ± 1 SD.

Since SkM was reported to bind to both heparin [11] and FXa [1], it was unclear whether the protective effects seen (in Figure 1A) were because of its interactions with UFH or FXa or both. To address this, the effect of SkM on the inhibition of FXa and thrombin by AT in the absence of UFH was studied. In the absence of UFH, SkM did not prevent the inhibition of FXa and thrombin by AT (Figure 1B), indicating that the protective effects seen in the presence of UFH were very likely due to SkM’s interactions with UFH and not FXa.

Notably, our previous LCMS reports indicated the presence of trace amounts of phosphatidylserine (PS) in the commercial SkM preparations [4, 8]. Walker and Esmon previously showed that phospholipids protect FXa from inhibition by AT when tested in the absence of heparins [21]. Although SkM did not protect FXa from AT inhibition in the absence of heparin, it was imperative to determine if the ability of SkM to inhibit UFH’s activity was due to the protein SkM or the trace amounts of PS in the SkM preparation. Phospholipid vesicles containing 80% phosphatidylcholine (PC) and 20% PS were tested for their effect on AT inhibition of FXa and thrombin, in the presence of UFH. PC/PS vesicles, unlike SkM, did not reduce UFH’s enhancement of AT inhibition of either serine protease (Supplemental Figure S1A). Further, two different PS binding proteins, annexin V and lactadherin, did not reduce the ability of SkM to inhibit UFH’s anticoagulant activity (Supplemental Figure S1B and S1C). Thus, these results confirm the role of SkM in preventing UFH-assisted AT inhibition of FXa and thrombin and indicate that this property of SkM is independent of any minor level of PS in the SkM preparation.

3.2. SkM neutralizes heparan sulfate’s anticoagulant properties

SkM reduced the inhibition of both FXa and thrombin by AT when tested in the presence of heparan sulfate (HS) (Supplemental Figure S2) which by itself enhances AT activity against FXa and thrombin (Figure 2A and Table 1). This may imply a physiological role for SkM in neutralizing HS’s anticoagulant properties.

Figure 2:

(A) SkM reduced the inhibition of FXa and thrombin by AT in the presence of stimulatory HS. (B, C) Chondroitin Sulfate (CS) (B) and NAc-deO-Hep (C) had no effect on SkM’s ability to neutralize UFH’s enhancement of AT’s inhibition of FXa. All the experiments were performed in Tris-buffered saline with 0.5% BSA, 5 mM CaCl₂, pH 7.4 at room temperature. Each value represents the mean ± SD of at least three individual experiments.

3.3. SkM’s interactions with UFH are not solely based on charge–charge interactions

To study the specificity of SkM–heparin interactions, chondroitin sulfate (CS) (Supplemental Figure S2), which does not enhance AT’s inhibitory actions, was included in assays as a decoy. CS failed to reduce SkM’s ability to neutralize UFH’s enhancement of AT activity (Figure 2B), implying that simply the presence of abundant negative charges on a carbohydrate polymer is not enough to explain SkM’s effects on UFH. These findings agree with the previous report which noted that CS was unable to displace heparin bound to myosin [11].

Since a highly negatively charged polymer like CS did not influence myosin–heparin interactions, the importance of negative charges on heparin for its interactions with SkM was studied using N-acetyl-de-O-sulfated heparin (NAc-deO-Hep) in the assays along with UFH. NAc-deO-Hep is produced from heparin by acetylation of the amine groups and desulfation of the hydroxy groups (Supplemental Figure S2), resulting in the loss of anticoagulant properties [22]. Like CS, NAc-deO-Hep failed to reduce SkM’s ability to reduce UFH’s enhancement of AT activity (Figure 2C). Thus, although the mere presence of negative charges on a carbohydrate polymer is not enough for its interactions with SkM, they are likely necessary.

3.4. SkM requires longer chain heparins for its protective effect

To study the influence of heparin’s chain length on its functional interactions with SkM, the inhibition of FXa by AT in the presence of SkM plus low molecular weight heparin (LMWH) or heparin pentasaccharide (H5) was determined. In the presence of these potent AT-cofactors, SkM did not protect FXa from inhibition by AT (Figure 3), indicating that functionally relevant SkM–heparin interactions require the presence of long-chain saccharides.

Figure 3: SkM does not prevent the inhibition of FXa by AT in the presence of LMWH and H5.

FXa (5 nM, final) was incubated with AT (25 nM, final) in the presence of different concentrations of (A) LMWH or (B) H5 and with SkM (20 nM and 100 nM, final) in Tris-buffered saline with 0.5% BSA, 5 mM CaCl₂, pH 7.4 at room temperature for 10 minutes. Residual FXa activity was determined using chromogenic substrate (0.4 mM, final). Each value represents the mean ± SD of at least three individual experiments.

3.5. CM inhibits UFH’s anticoagulant activity

Because CM is structurally very similar to SkM and showed procoagulant properties by binding FXa and enhancing prothrombinase activity [2, 6], the ability of CM to inhibit UFH’s AT-dependent anticoagulant properties was tested. Indeed, like SkM, CM protected FXa and thrombin from inhibition by AT in the presence but not in the absence of UFH (Figures 4A and 4B). Moreover, CM, like SkM, showed chain length dependence on its interactions with heparin as CM did not protect FXa from inhibition by LMWH-AT or H5-AT (Figures 4C and 4D).

Figure 4: CM inhibits AT’s anticoagulant action only in the presence of (A) UFH, but not in the (B) absence of heparins or presence of (C) LMWH or (D) H5.

All experiments were performed in Tris-buffered saline with 0.5% BSA, 5 mM CaCl₂, pH 7.4 at room temperature. Residual protease activity was determined using an appropriate chromogenic substrate. Error values represents the SD of at least three individual determinations.

4. DISCUSSION

Studies conducted to understand the influence of myosin–heparin interactions on heparin’s anticoagulant properties established that SkM prevents the efficient inhibition of the procoagulant serine proteases, FXa and thrombin, by AT in the presence of UFH. SkM also prevents the efficient inhibition of FXa and thrombin by AT in the presence of another AT physiologic cofactor, HS. However, CS, another negatively charged glycosaminoglycan, did not influence SkM’s ability to neutralize UFH’s anticoagulant properties, indicating a degree of specificity in SkM–heparin interactions. Thus, these findings put forth a novel procoagulant mechanism for SkM: neutralization of heparin/HS-assisted anticoagulant properties of AT. This procoagulant nature of SkM, as a result of its interactions with heparin/HS, might lead to thrombotic complications in conditions with increased plasma SkM concentrations, such as rhabdomyolysis, polymyositis and dermatomyositis [23–26]. In fact, people with polymyositis and dermatomyositis are at an increased risk of developing venous thromboembolism [27–32]. However, these hypotheses need further investigation using clinical patient samples, and much work is needed to understand better the pathophysiologic implications of SkM–heparin interactions that affect AT activity.

CM belongs to the family of conventional striated muscle myosins and shares ~80% amino acid sequence identity to SkM [8–10]. Although the current studies used rabbit SkM and bovine CM, instead of human myosins, both SkM and CM amino acid sequences are highly homologous among mammalian species (>95% identity) [33–36]. Similar to SkM, CM reduced FXa inhibition by AT when tested in the presence but not the absence of UFH.

Neither SkM nor CM protected FXa from AT inhibition that was stimulated by shorter heparins, such as LMWH and H5. Since LMWH and H5 each binds directly to AT to exert their anticoagulant activity, SkM and CM appear not to interfere with interactions between AT and UFH. The longer UFH binds not only to AT, as do LMWH and H5, but also to FXa and thrombin. UFH-thrombin interactions are required for AT’s efficient anticoagulant activity. In contrast to thrombin, UFH-FXa interactions are not required for the AT’s anticoagulant activity, but they do enhance AT’s activity against FXa in the presence of Ca⁺⁺ ions [19]. Thus, one attractive speculation is that myosin’s binding to heparins appears to block UFH-FXa and UFH-thrombin interactions, but not UFH-AT interactions, resulting in a reduction in AT’s anticoagulant activity. The observed chain length dependence for SkM’s ability to reduce heparin’s anticoagulant activity might have implications for the clinic. Multiple studies have reported the superior effect of LMWH over UFH in preventing the development of venous thromboembolism in patients following major trauma [37, 38], and another study found that LMWH is more effective in preventing pulmonary embolism following major trauma when compared to UFH [39]. No such differences between the use of LMWH and UFH were noted in preventing venous thromboembolism or pulmonary embolism in patients without trauma. As our previous studies suggest an increase in the concentration of SkM in the plasma of acute trauma patients compared to healthy individuals [1, 2, 8], it is possible that the reduced effect of UFH compared to LMWH in preventing venous thromboembolism and pulmonary embolism following major trauma may be, in part, related to the reduced anticoagulant potency of UFH in the presence of SkM.

Inhibition of AT’s anticoagulant activity can have a variety of effects. Multiple studies have shown the promising effect of inhibiting AT’s anticoagulant pathway as a potential treatment for hemophilia [40, 41], with an RNAi therapy currently being investigated in phase III clinical trials [42, 43]. Since SkM and CM blunt UFH’s and HS’s enhancement of AT activity, it is possible that a better understanding of the multiple molecular interactions responsible for these effects might point towards the development of novel compounds which, like myosins, might also reduce AT potency by binding endogenous heparin/HS, thus affecting hemophilia-related bleeding.

In summary, SkM and CM significantly impede UFH’s and HS’s AT-dependent anticoagulant properties and protect FXa and thrombin from inhibition by AT. The ability of these myosins to accomplish this depends on the type, sulfation, and chain length of the glycosaminoglycans.

Essentials.

• Skeletal muscle myosin (SkM) binds unfractionated heparin (UFH)

• SkM reduces UFH-assisted inhibition of factor Xa and thrombin by antithrombin

• Cardiac myosin reduces UFH-assisted antithrombin inhibition of factor Xa and thrombin

• Heparin chain length influences myosin’s ability to neutralize UFH-antithrombin activity

ABBREVIATIONS AND NOMENCLATURE

SkM

skeletal muscle myosin

CM

cardiac myosin

AT

antithrombin

UFH

unfractionated heparin

LMWH

low molecular weight heparin

H5

heparin pentasaccharide

HS

heparan sulfate

CS

chondroitin sulfate

NAc-deO-Hep

N-acetyl-de-O-sulfated heparin

PC

phosphatidylcholine

PS

phosphatidylserine

S1

sub-fragment 1

RLC

regulatory light chain

ELC

essential light chain

HC

heavy chain