Effects of Enoxaparin on Intravascular Sclerostin Release in Healthy Men

Abstract

Sclerostin (Scl) is implicated in vascular calcification and angiogenesis and localizes within vasculature. Its molecule incorporates a heparin-binding site that implies also binding to endothelial glycocalyx. We preliminary tested whether intravenous (IV) low-molecular-weight heparin enoxaparin can stimulate intravascular release of this calcification inhibitor in humans. Sixteen male volunteers were injected with a bolus of 1 mg/kg body weight of enoxaparin. After 10 minutes, plasma immunoreactive Scl levels increased uniformly by a mean of 184% versus baseline level of 0.56 ± 0.17 ng/mL (P = .0004). Plasma Scl levels were found still elevated after 2 and 6 hours (with a median of 20.9% and 8.69%, respectively) and became normal after 24 hours. The percentage of increase (Δ) in plasma Scl after 10 minutes was directly correlated with enoxaparin dose per kg/m² of body mass index (ρ = 0.587, P = .017) and strongly inversely correlated with the preinjection Scl levels (ρ = −0.747, P = .0008). A robust negative association between the ΔScl increase after 10 minutes and the ΔScl decrease after 2 hours versus 10 minutes was observed (ρ = −0.835, P < .0001). Complementary in vitro spiking experiment showed no effects of enoxaparin addition and whole blood incubation on plasma Scl levels when measured with the immunoassay. This study shows that enoxaparin has a stimulating effect on the intravascular release of calcification inhibitor Scl in healthy men. This novel pharmacological action of the popular anticoagulant drug seems important in cardiovascular medicine.

Introduction

Sclerostin (Scl) is a member of the Cerberus/DAN family of glycoproteins and a bone morphogenetic protein antagonist. It is produced mainly, but not exclusively, by osteocytes and generally acts as an inhibitor of bone formation by repressing differentiation and proliferation of osteoblasts and enhancing their apoptosis. ^1,2 Recently, Scl was found to be also involved in vascular calcification of both intimal (atherosclerotic plaques) and medial layer (Mönckeberg sclerosis) of arterial blood vessels, as well as in calcifying aortic valve disease. ^{2 –4} Vascular calcification is particularly prevalent and severe in patients with type 2 diabetes mellitus ^3,5 and in those with advanced chronic kidney disease. ^6,7 Sclerostin is expressed in calcifying vasculature during the highly regulated process of vascular smooth muscle cell (VSMC) reprogramming and differentiation to mineralizing osteoblast-like cells. ^{2 –4,7} Thus, Scl may be regarded as part of a local counterregulatory mechanism to suppress vascular calcification. ^{2,6 –8}

Characterization of the molecular structure of a 190-amino acid glycoprotein Scl showed that it contains a C-terminal cysteine knot-like domain and 3 loops. Interestingly, a potential binding site for heparin, suggestive of a functional role in localizing Scl to the surface of target cells, was found within the molecule. ^9,10 This implies that Scl also belong to the growing family of heparin-binding proteins. The proteins bind to heparan sulfate glycosaminoglycans (GAGs) present on the luminal surface of vascular endothelial cells (glycocalyx) and in the subendothelial matrix, from where they may be competitively released by endo- or exogenous heparin. ^11,12

Although uncovered for almost a decade, ^9,10 the tempting hypothesis that also Scl may be strongly regulated by heparin has not been challenged so far. The aim of the present pilot study was to investigate for the first time the effects of intravenous (IV) low-molecular-weight heparin (LMWH) enoxaparin administration on plasma immunoreactive Scl levels in healthy male volunteers.

Participants and Methods

Participants

The study protocol was in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration. Sixteen generally healthy male volunteers aged 46.7 ± 11.3 years, with a mean body weight (bw) of 85.5 ± 8.62 kg, height of 1.79 ± 0.07 m, and a slightly increased mean Quételet body mass index (BMI) of 26.7 ± 3.11 kg/m², gave their informed consent and were enrolled. They were no drug or alcohol abusers; 3 of them were previous smokers and 4 had arterial hypertension (without evidence of target organ damage) well controlled by monotherapy (perindopril, nifedipine, or bisoprolol). Their health status was ascertained by history, physical examination, and selected laboratory tests. Special attention was given to mental status, actual arterial blood pressure, history of drug hypersensitivity, bleeding and thrombotic disorders, gastroduodenal disease, complete blood counts, urinalysis, serum C-reactive protein, creatinine, liver enzymes, viral hepatitis B/C and HIV infection, blood glucose, and routine hemostatic tests. None of the participants experienced major trauma, surgery, or infection and was not receiving any additional medications (particularly heparin or heparinoid sulodexide) in the preceding month.

Enoxaparin Administration and Blood Sampling

Fasted blood samples (T₀) were obtained from the antecubital vein (punctured with an 18-gauge needle with minimal stasis) into EDTA-coated Monovette vacutainer. Then LMWH enoxaparin sodium (Clexane; Sanofi-Aventis, Paris, France) was injected IV at a mean dose of 82.5 ± 6.83 mg (a content of a prefilled syringe containing 80 mg or 100 mg of the drug; 10 mg = 1000 IU anti-factor Xa activity). The dose of enoxaparin for the study was chosen to approach that of 1 mg per kg bw, and it actually amounted to 0.97 ± 0.06 mg/kg bw. In relation to BMI, the enoxaparin dose was 3.12 ± 0.32 mg per kg/m².

Following the drug injection, the patients remained ambulatory within the clinic and were allowed habitual water intake; the consecutive blood samples were drawn after 10 minutes (T_10min), 120 minutes (T_2h) and 6 hours (T_6h). Then the participants presented after 24 hours from the first sampling, the fasting blood was drawn (T_24h), and the study was terminated. The follow-up period was uneventful; no adverse effects of enoxaparin administration were noted on medical examination performed 7 to 10 days after study termination.

Ex Vivo Testing for Scl–Enoxaparin Interference

To test for any ex vivo interference, the Scl concentrations were measured in plasma samples spiked with enoxaparin at various (half dose, normal dose, and double dose) concentrations after 10 minutes, 60 minutes, and 2 hours of incubation (Table 1).

Table 1.

Sclerostin in Plasma of Whole Blood Spiked and Incubated With Enoxaparin.

Incubation Time	Plasma Sclerostin, ng/mL, Participant 1			Plasma Sclerostin, ng/mL, Participant 2			Plasma Sclerostin, ng/mL, Participant 3
Zero	0.54			0.71			0.91
	Enoxaparin dose			Enoxaparin dose			Enoxaparin dose
	Half	Normal	Double	Half	Normal	Double	Half	Normal	Double
10 minutes	0.49	0.55	0.44	0.65	0.64	0.70	0.92	0.89	0.84
60 minutes	0.47	0.43	0.42	0.62	0.64	0.53	0.88	0.91	0.90
2 hours	0.52	0.54	0.55	0.68	0.69	0.71	0.88	0.87	0.86

Fasting venous blood was drawn into EDTA-coated vacutainer (20 mL, then aliquoted into 2 mL plastic tubes) from 3 men recruited from the previous group (aged 35, 68, and 55 years; bw of 80, 81 and 87 kg; height of 1.88, 1.73, 1.78 m; BMI of 22.6, 27.1, 27.8 kg/m², respectively). Assuming that their blood volume was approximately 6000 mL, ¹³ they would have purportedly received 80 mg of IV enoxaparin. Thus, to mimic initial blood concentration of the supposedly injected enoxaparin, approximately 0.027 mg of the drug was added to a 2 mL sample of whole blood. The enoxaparin solution for spiking experiments was prepared by dissolving 20 mg of this LMWH (a content of 0.2 mL prefilled syringe) in 100 mL of sterile 0.9% saline that resulted in 0.027 mg of enoxaparin in 135 μL of the solution.

For each of the participants, 10 whole blood samples were prepared: (1) 1 sample with no enoxaparin added, which was centrifuged at 3000g for 20 minutes to obtain platelet-poor plasma; (2) 3 samples spiked with 50 μL (half-dose) of the enoxaparin solution; (3) 3 samples spiked with 100 μL (normal dose); and (4) 3 samples spiked with 200 μL (double dose) of the enoxaparin solution. The enoxaparin-spiked samples were mixed thoroughly by vortexing and incubated at 37°C for 10 minutes, 60 minutes, or 2 hours. Following the incubation periods, plasma samples were prepared, aliquoted, and stored at −70°C until time of Scl assay.

Plasma Scl Measurements

Plasma for Scl quantification in the human study was also prepared as described above, aliquoted, and stored at −70°C until batch analysis. Human Sclerostin High Sensitivity EIA kits from Tecomedical AG (Sissach, Switzerland, cat. no. TE1023-HS) were used. The measurements were performed according to the manufacturer’s protocol, in duplicate, using a 400 SFC microplate reader (SLT-Labinstruments, Gröding/Salzburg, Austria) and calibrated using provided recombinant human reference samples and standards. For calculations of the results, a computer and a curve-fitting program were used. The within- and between-assay coefficients of variations were <8%.

Statistical Analysis

Shapiro-Wilk W test of normality was used for data distribution analysis. The normally distributed data were presented as mean ± 1 standard deviation, and the skewed data as median (minimum-maximum). Comparisons were performed with nonparametric Friedman analysis of variance (ANOVA) by ranks and Wilcoxon tests and bivariate correlations assessed with Spearman regression analysis. All tests were 2 sided, and statistical significance was set at P value <.05. Analyses were conducted using Statistica 12.5 PL for Windows (StatSoft, Tulsa, Oklahoma).

Results

Effects of IV Enoxaparin on Plasma Scl Levels

The Scl levels significantly changed following IV enoxaparin administration (χ² ANOVA = 50.7, df = 4, P < .0001; Figure 1). They increased consistently (in each of the participants) by a mean of 184 ± 97.9% from 0.56 ± 0.17 ng/mL at T₀ to a median of 1.36 (1.08-1.97) ng/mL at T_10min (P = .0004). At T_2h, Scl levels were 0.71 ± 0.19 ng/mL, lower than those at T_10min (P = .004), but still elevated by a median of 20.9% compared with the baseline values (P = .0005). At T_6h, Scl levels were 0.61 ± 0.16 ng/mL and still statistically higher by a median of 8.69% versus baseline (P = .017). At T_24h, they were found normal (0.56 ± 0.16 ng/mL).

Figure 1.

Case profiles of plasma sclerostin levels following enoxaparin administration in 16 healthy men.

Variables Associated With Plasma Scl Increments

At baseline, plasma Scl levels were not significantly associated with age (ρ = 0.395, P = .130), bw (ρ = 0.110, P = .684), or BMI (ρ = 0.374, P = 0.164). The percentage of increase (Δ) in plasma Scl at T_10min versus baseline tended to be directly associated with the dose of enoxaparin per kg bw (ρ = 0.430, P = .096; Figure 2A). It was significantly and directly correlated with enoxaparin dose per kg/m² of BMI (ρ = 0.587, P = .017; Figure 2B). The ΔScl increments at the subsequent time points were also direct but not significant (data not shown).

Figure 2.

Variables influencing sclerostin (Scl) release by enoxaparin in healthy men. The Δ symbol denotes the percentage change in plasma Scl levels at 10 minutes (T_10min) after intravenous enoxaparin administration versus baseline Scl concentrations (T_0min) or at 120 minutes (T_120min) versus T_10min. The regression lines represent the subjective best-fit model for the nonlinear Spearman correlation analysis. BMI indicates body mass index.

The ΔScl increase at T_10min strongly and inversely correlated with the baseline (preinjection) levels of plasma Scl (ρ = −0.747, P = .0008; Figure 2C). The above relation was also inverse but not significant for ΔScl increase at both T_2h (ρ = −0.350, P = .183) and T_6h (ρ = −0.208, P = .438). A strong negative association between ΔScl increment at T_10min versus T₀ and the ΔScl decrease at T_2h versus T_10min was observed (ρ = −0.835, P < .0001; Figure 2D).

Sclerostin in Plasma of Enoxaparin-Spiked Whole Blood

Plasma Scl levels measured before enoxaparin addition to EDTA-anticoagulated venous whole blood samples from 3 participants and in those obtained following enoxaparin spiking and incubation are shown in Table 1. Neither enoxaparin addition nor whole blood incubation for up to 2 hours had any apparent effect on plasma Scl measured with ELISA from Tecomedical AG.

Zero: Scl levels in nonheparinized EDTA plasma (0 minutes; obtained before whole blood spiking with enoxaparin). Normal dose: amount of enoxaparin added to the incubation tube (∼0.027 mg per 2 mL of EDTA-anticoagulated blood) to mimic the likely initial drug concentration achieved by IV injection of 80 mg of enoxaparin to an 80-kg man. Plasma Scl measured by ELISA from Tecomedical AG.

Discussion

In this pilot open-label study, we have suggested for the first time that IV injection of LMWH enoxaparin induces an intense, dose-dependent, and consistent increase in plasma immunoreactive Scl levels in healthy humans. This may provide a completely novel effect of the popular medication and challenge novel insights into the regulation of Scl metabolism and the process of tissue calcification.

Sclerostin has been traditionally regarded as a bone formation inhibitor. It principally reduces osteoblastic activity as an endogenous antagonist of canonical Wnt signaling — by binding to and modulating function of lipoprotein receptor-related protein 5/6, bone morphogenetic proteins, and other minor receptors. ^1,2 Thought-provokingly, immunohistochemical studies showed that Scl is also present within vasculature as evidenced by its expression in calcifying murine VSMCs, ⁷ subcutaneous tissue of human calcific uremic arteriolopathy lesions, ¹⁴ and in the radial artery media layer in patients with end-stage kidney disease. ¹⁵ The important role of Scl in vascular structure and function was further supported by the recent comprehensive report of its potential to drive blood vessel formation. ¹⁶ Sclerostin was found to increase proliferation of human umbilical vein endothelial cells — principally by binding to its operational lipoprotein receptor-related protein 6 receptor and further by upregulating its expression and through the production of vascular endothelial growth factor and placental growth factor. Remarkably, the proangiogenic potential of Scl was comparable to that of the master vascular endothelial growth factor. ¹⁶ On the other hand, the conceptually interconnected experimental study showed that the artificial material involved in Scl binding could also be GAGs — a mixture of hyaluronan sulfate and chondroitin sulfate. ¹⁷ In humans, the GAGs are abundant within the glycocalyx on the endothelial surface and in the extracellular matrix of medial vascular layer, while the GAG-bound molecules (numerous growth factors, cytokines, enzymes, etc) are easily and competitively displaced by exogenous heparin into the circulation. ¹¹ In an original paper describing for the first time Scl molecular structure, besides identifying the heparin-binding site between Scl loops 1 and 3, the authors found a meaningful striking and dose-dependent increase in the amount of Scl detected in the supernatant of culture cells following heparin addition. ⁹ Altogether, congregated evidence indicates that Scl is implicated also in vascular morphology and function, including common vascular calcification, and suggests that heparin may influence Scl metabolism also in a human body.

In the present pilot study testing the above hypothesis, we found the rapid (evidentas soon as after 10 minutes), uniform (existing in all participants), substantial (almost 2-fold), sustained (lasting for at least 6 hours) and dose-dependent (particularly in relation to BMI) rise in plasma Scl following IV enoxaparin injection. Moreover, the magnitude of the Scl increase after 10 minutes was strongly, highly significantly, and remarkably inversely correlated with the baseline (preinjection) plasma Scl levels. From a pharmacokinetic point of view, this regression pattern plausibly indicates the specific and reciprocal equilibrium between the circulating and easily releasable Scl pools. Even more statistically robust and negative association found between the extent of Scl release, and the ability to clear it from the circulation may point to saturation and blocking of the Scl-binding sites by exogenous enoxaparin lodged and embedded in, supposedly, the endothelial layer. ¹⁷ The findings may thus provide indirect but suggestive evidence that Scl rise in plasma originates from its enoxaparin accessible endothelial stores and, in later stages, from the subendothelial extracellular matrix stores. The former of the 2 above meaningful regression patterns is in line with those revealed in our previous studies on other molecules released from its vascular deposits by enoxaparin — activin A, type 1 tissue factor pathway inhibitor, transforming growth factor-β1, and myeloperoxidase, ^{18 –21} or in those liberated by the heparinoid drug sulodexide such as hepatocyte growth factor and again type 1 tissue factor pathway inhibitor. ^22,23 Thus, our present study seems to further add to the emerging concept that vascular GAGs and glycocalyx form a specific and vital reservoir and that heparins are powerful and pluripotent drugs, being far beyond simple blood anticoagulants. ^11,12,18,24

Since the introduction of commercial ELISAs for Scl measurement, it has been known that plasma samples yield up to 1.5-fold higher Scl concentrations than serum samples. It was speculated that Scl circulates in serum as a complex bound to various proteins and that antibodies used in the assays may differently recognize both bound and free forms. ²⁵ The fact that Scl contains a heparin-binding site was appreciated at that time and ex vivo antigenic modification or dissolution of Scl complexes by heparin from prefilled syringes (approximately 50 IU of lithium heparin per 1 mL syringe) was regarded as an alternative explanation for higher Scl concentrations detected in plasma. ²⁵ Nowadays, it is apparent that both heparin and EDTA plasma yield comparable Scl values and the in vitro heparin interference (even in the above relatively high dose) is unlikely. ²⁶ In our study, we used EDTA-coated syringes to collect the already systemically heparinized blood. Thus, our samples supposedly contained minor amounts of active enoxaparin, especially in view of its preceding dissolution in circulating blood, rapid clearance due to binding to vascular endothelium, and then elimination by the kidneys. ^11,25 Our enoxaparin-spiking experiment also did not point to any apparent in vitro effect of enoxaparin on plasma Scl measurements. Therefore, the observed rapid rise in plasma Scl seems to originate from the liberation of this calcification inhibitor from its vascular stores rather than from methodological reasons.

These new effects of enoxaparin on Scl need verification and, if confirmed, they deserve further investigations. The issue may be particularly important in patients with chronic kidney disease stage 5 on maintenance hemodialysis who are at the highest possible risk for cardiovascular morbidity and mortality, have particularly prevalent and severe vascular calcification, and highest serum Scl levels of so far vague predictive value. ^6,7,27,28 Noteworthy, these unusual patients also commonly receive enoxaparin (or other LMWHs) thrice weekly for blood purification procedures in an exceptionally high cumulative dose of about 8000 mg (800 000 IU!) a year. ^18,21 This population may be particularly affected by consequences of repeated enoxaparin-induced Scl release and its possible vascular depletion and therefore deserves particular attention and studies.

In conclusion, enoxaparin, often regarded as the useful but mere anticoagulant drug, may have a remarkable stimulating effect on the release of the potent calcification inhibitor Scl from its intravascular stores in humans. This novel action of popular enoxaparin may, hopefully, encourage vascular research and impact cardiovascular and renal medicine.