- Email: [email protected]
Damage of the endothelial glycocalyx in chronic kidney disease
Atherosclerosis 234 (2014) 335e343
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Damage of the endothelial glycocalyx in chronic kidney disease Jan-Sören Padberg a, Anne Wiesinger a, Giovana Seno di Marco a, Stefan Reuter a, Alexander Grabner a, Dominik Kentrup a, Alexander Lukasz a, Hans Oberleithner b, Hermann Pavenstädt a, Marcus Brand a, Philipp Kümpers a, * a Department of Medicine D, Division of General Internal Medicine, Nephrology, and Rheumatology, University Hospital Münster, Albert-Schweitzer-Campus 1 A1, 48149 Münster, Germany b Institute of Physiology II, University of Muenster, Robert-Koch-Straße 27 b, 48149 Münster, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 November 2013 Received in revised form 24 February 2014 Accepted 17 March 2014 Available online 29 March 2014
Background and objectives: The endothelial glycocalyx (eGC), a mesh of anionic biopolymers covering the luminal surface of endothelial cells, is considered as an intravascular compartment that protects the vessel wall against pathogenic insults in cardiovascular disease. We hypothesized that chronic kidney disease (CKD) is associated with reduced eGC integrity and subsequent endothelial dysfunction. Methods & results: Shedding of two major components of the eGC, namely syndecan-1 (Syn-1) and hyaluronan (HA), was measured by ELISA in 95 patients with CKD (stages 3e5) and 31 apparently healthy controls. Plasma levels of Syn-1 and HA increased steadily across CKD stages (5- and 5.5-fold, respectively P < 0.001) and were independently associated with impaired renal function after multivariate adjustment. Furthermore, Syn-1 and HA correlated tightly with plasma markers of endothelial dysfunction such as soluble fms-like tyrosine kinase-1 (sFlt-1), soluble vascular adhesion molecule-1 (sVCAM-1), von-Willebrand-Factor (vWF) and angiopoietin-2 (P < 0.001). Experimentally, excessive shedding of the eGC, evidenced by 11-fold increased Syn-1 plasma levels, was also observed in an established rat model of CKD, the 5/6-nephrectomized rats. Consistently, an atomic force microscopybased approach evidenced a significant decrease in eGC thickness (360 79 vs. 157 29 nm, P ¼ 0.001) and stiffness (0.33 0.02 vs. 0.22 0.01 pN/nm, P < 0.001) of aorta endothelial cell explants isolated from CKD rats. Conclusion: Our findings provide evidence for damage of the atheroprotective eGC as a consequence of CKD and potentially open a new avenue to pathophysiology and treatment of cardiovascular disease in renal patients. Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Endothelial glycocalyx Chronic kidney disease Shedding Syndecan-1 Hyaluronan Atomic force microscopy 5/6 Nephrectomy Cardiovascular disease Atherosclerosis
1. Introduction Patients with chronic kidney disease (CKD) exhibit endothelial dysfunction and accelerated vascular disease leading to increased morbidity and mortality due to cardiovascular events [1e3]. The underlying mechanisms, however, are still not fully understood. Endothelial dysfunction has long been ascribed to a malfunction of the endothelial cell itself. Recent studies, though, provided compelling evidence that the endothelium is protected against pathogenic insults by a highly hydrated negatively charged “firewall” called the glycocalyx. The endothelial glycocalyx (eGC), a carbohydrate-rich mesh of large anionic polymers, lines the luminal side of the endothelium along the entire vascular tree [4]. It
* Corresponding author. Tel.: þ49 25 18 34 49 51. E-mail address: [email protected] (P. Kümpers). http://dx.doi.org/10.1016/j.atherosclerosis.2014.03.016 0021-9150/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
is composed of core proteoglycans, especially those of the syndecan family, to which both highly sulfated glycosaminoglycans and hyaluronan are attached [5,6]. Measuring about 0.5e2 mm in thickness, the eGC forms an integral part of the vascular barrier alongside the endothelial cell itself [7]. Having been neglected over decades, important physiological properties were attributed to the eGC during the last years. Given its strategic location as the interface between blood and endothelium, the intact glycocalyx prevents transvascular protein leakage and reduces leukocyteeendothelial interactions [8e10]. Beyond that, the glycocalyx contributes to the regulation of redox state and is crucially involved in the mediation of shear-induced nitric oxide release as well as physiologic anticoagulation [4,6,11]. Its structure is fairly stable but also in healthy endothelium subject to a permanent dynamic equilibrium between biosynthesis of new components and shear dependent removal, the so-called shedding, of existing constituents [12].
336
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Damage to eGC, however, can occur under exposure to inflammatory or atherogenic noxae, such as endotoxin [13] or TNF-alpha [14,15], oxidized LDL [11,16], excess atrial natriuretic peptide [17,18], hyperglycaemia [19], and hypervolemia [20]. All of these eGCdeteriorating stimuli occur abundantly in CKD. Given the fact that the mechanisms of accelerated atherosclerosis and cardiovascular disease in CKD patients are still not fully understood, we aimed to investigate whether eGC damage might be a mediating factor in this process. Thus, we sought to assess the impact of chronic renal impairment on 1) circulating markers of eGC damage and endothelial dysfunction in CKD patients, as well as on 2) eGC structure (i.e. thickness and stiffness) itself of aorta endothelial cell explants isolated from CKD rats. 2. Materials & methods 2.1. Patients The medical ethics committee at University Hospital Münster (Münster, Germany) approved the study and written informed consent was obtained from all subjects. In total, 95 stable Caucasian patients with CKD stages 3e5 (eGFR 60 ml/min/ 1.73 m2) were included in this study. Patients with presumable eGC alterations due to acute or chronic infections, malignancy, history of organ transplantation, or pregnancy were excluded. CKD patients were divided according to the National Kidney Foundation’s classification of CKD. Subjects with eGFR 60 ml/
min/1.73 m2 and without evidence of proteinuria served as controls. The control group (n ¼ 31) consisted of healthy volunteers and patients scheduled for elective surgery at University Hospital Münster. All subjects continued their regular medication, including antihypertensive agents and statins. Demographic, clinical, and biochemical characteristics of patients and controls are summarized in Table 1. 2.2. Data collection and blood sampling Demographic data and details of medical history were collected at enrollment. Hypertension was defined as systolic blood pressure 140 mmHg and diastolic blood pressure 90 mmHg or current antihypertensive therapy. The clinical definition of diabetes included fasting glucose 126 mg/dl, intake of oral hypoglycaemic agents, or insulin use. Cardiovascular risk was estimated using the Framingham Risk Score. Estimated glomerular filtration rate (eGFR) was calculated using the chronic kidney disease epidemiology collaboration (CKD-EPI) equation, as previously described [21]. Blood samples from patients and control subjects were collected into ethylenediaminetetraacetic acid (EDTA) tubes, centrifuged, divided into aliquots and stored at 80 C. Serum creatinine, HbA1c, cholesterol and urinary protein were measured using standard laboratory techniques (Center for Laboratory Medicine, University Hospital Münster). In dialysis patients, blood samples were drawn pre-dialysis.
Table 1 Baseline characteristics of all participants.a Estimated GFR (ml/min/1.73 m2)e
Age (yr) Male gender [% (n)] Creatinine (mg/dl) eGFR (ml/min/1.73 m2)e Proteinuria (g/24 h) Cardiovascular risk factors Arterial hypertension [% (n)] SBP (mmHg) DBP (mmHg) BMI (kg/m2) Cholesterol (mg/dl) HDL-cholesterol (mg/dl) Calculated LDL-cholesterol (mg/dl) Diabetes [% (n)] HbA1c (%) Framingham risk score (%) Smokers [% (n)] Current medication ACE inhibitors [% (n)] AT1 blockers [% (n)] Erythropoietin [% (n)] Statins [% (n)] Markers of endothelial dysfunction Ang-2 (ng/ml) sFlt-1 (pg/ml) sVCAM-1 (pg/ml) vWF (U/ml) Circulating glycocalyx components Hyaluronan (ng/ml) Syndecan-1 (ng/ml)
P Value
60 (n ¼ 31)
30 to 59 (n ¼ 26)
15 to 29 (n ¼ 35)
<15 (n ¼ 34)
45 21 65 (20) 0.9 (0.7e1.0) 95 (79e110) 0.0 (0.0 - 0.0)
68 16 46 (12) 1.3 (1.1e1.5) 40 (36e48) 0.0 (0.0e0.7)
67 16 46 (16) 2.0 (1.8e2.3) 22 (18e25) 0.3 (0.2e1.5)
62 13 65 (22) 4.5 (3.3e5.5) 6 (5e10) 0.7 (0.2e1.3)
29 (9) 130 (125e135) 80 (80e80) 23.7 (22.1e28.0) 165 (134e211) 49 (45e56) 129 (109e166) 3 (1) 4.7 (4.5e5.0) 8 (1e20) 19 (6)
77 (20) 135 (129e150) 80 (75e80) 25.5 (23.1e29.7) 184 (134e210) 48 (39e56) 114 (97e156) 23 (6) 5.3 (5.0e5.7) 12 (4e17) 19 (5)
83 (29) 130 (120e140) 80 (70e80) 27.1 (23.4e30.9) 168 (142e198) 43 (36e50) 100 (68e127) 40 (14) 5.2 (4.8e6.5) 16 (5e22) 23 (8)
71 (24) 130 (125e140) 80 (78e80) 25.4 (22.5e27.5) 156 (130e199) 43 (39e48) 97 (67e116) 32 (11) 5.2 (5.0e5.7) 11 (6e16) 29 (10)
19 (6) 13 (4) 0 (0) 6 (2)
65 (17) 27 (7) 4 (1) 50 (13)
54 17 11 46
32 26 35 29
0.7 (0.6e1.2) 56 (44e79) 740 (544e874) 0.2 (0.1e0.5)
1.2 (0.9e1.8) 77 (61e107) 752 (660e1128) 0.3 (0.2e0.6)
1.6 (1.1e3.4) 113 (89e158) 817 (597e1458) 0.5 (0.2e0.9)
3.2 (1.7e5.7) 101 (83e149) 1679 (1028e2233) 0.7 (0.3e1.3)
<0.001c <0.001c <0.001c <0.001c
81 (48e171) 53 (41e92)
133 (62e224) 90 (61e344)
147 (86e230) 212 (100e355)
441 (163e1213) 270 (92e515)
<0.001c <0.001c
(19) (6) (4) (16)
(11) (9) (12) (10)
<0.001b 0.215d <0.001c <0.001c <0.001c 0.033d 0.457c 0.306c 0.416c 0.816c 0.031c 0.004c 0.044d 0.004c 0.113d 0.726d 0.034d 0.704d <0.001d 0.026d
ACE, angiotensin-converting enzyme; AT1, Angiotensin-II receptor; BMI, body mass index; DBP, diastolic blood pressure; HbA1c, glycohemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SBP, systolic blood pressure. a In case of a normal distribution, variables are presented as mean SD. In case of a skewed distribution, variables are presented as median (interquartile range). b Univariate ANOVA, post hoc test (Scheffé). c KruskaleWallis one-way ANOVA. d c2 test. e GFR values were calculated using CKD-EPI formula.
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
2.3. Quantification of circulating glycocalyx components As markers of eGC damage, we chose Syn-1 and HA as they probably reflect different aspects in glycocalyx composition (Syn-1 as core glycoprotein and hyaluronan as loosely attached substance). Human plasma syndecan-1 (Syn-1) levels were determined directly as reported previously (Gen-Probe Diaclone Research, Besancon, France [17,22,23]). Plasma hyaluronan (HA) measurements were performed using an enzyme-linked immunosorbent assay designed for detecting human hyaluronic acid (Echelon Biosciences Inc., Salt Lake City, UT, USA) as previously reported [19,23]. Preliminary experiments confirmed the stability of eGC components in plasma for 24 h at room temperature as well as their stability throughout five repeated freeze-thaw-cycles (data not shown). All measurements were performed in duplicates at the same day by the same investigators blinded to patients’ characteristics. Rat plasma Syn-1 levels were determined directly using a sandwich enzyme immunoassay with an antibody specific to rat Syn-1 (USCN Life Sciences, Wuhan, China) according to the manufacturer’s instructions. 2.4. Quantification of circulating markers of endothelial dysfunction Plasma levels of angiopoietin-2 (Ang-2) were measured by inhouse immuno-luminometric Assay (ILMA) methodology (R&D Systems, Minneapolis, MN, USA), as previously reported by our group in detail [24]. Plasma levels of human soluble fms-like tyrosine kinase-1 (sFlt-1, R&D Systems), soluble vascular cell adhesion molecule-1 (sVCAM-1, R&D Systems), and Von Willebrand factor-antigen (vWF, Technoclone, Vienna, Austria) were determined using commercially available ELISA kits according to the manufacturer’s instructions. Parts of the results of these measurements have been reported previously [25]. All measurements were performed in duplicates at the same day by the same investigators blinded to patients’ characteristics. 2.5. 5/6-Nephrectomy in rats All procedures were approved by a governmental committee on animal welfare (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen) and were performed according to international guidelines on animal experimentation. Male LewiseBrown Norway (LNB-F1) rats (200e270 g in body weight) were obtained from Janvier (Janvier, Le Genest Saint Isle, France). Rats had free access to standard chow and tap water, and were acclimated to the facility for at least a week before beginning an experiment. Male LBN-F1 rats underwent 5/6-nephrectomy as described before [26]. In brief, rats were anesthetized by isoflurane, ketamine 10% and xylazine 2% (both CEVA PLC, Duesseldorf, Germany). The right kidney was extirpated and 2 of 3 branches of the left renal artery were ligated. Fourteen days after the intervention, rats were anesthetized again and blood samples were drawn by left cardiac puncture. Aortas were removed for further fine preparation (see below). 2.6. Atomic force microscopy on rat aorta ex vivo Rat aortas were perfused with PBS (PAA Laboratories, Pasching, Austria) with 1% Penicillin and Streptomycin (Biochrom AG, Berlin, Germany) via cardiac puncture, isolated, immediately bathed in PBS with 1% Penicillin/Streptomycin (Pen/Strep) and freed from surrounding tissue (Fig. 1 A). Small patches (about 5 mm [2]) of the whole aorta were attached on Cell-TakÒ (BD Biosciences, Franklin Lakes, NJ, USA) coated glass, with the endothelial surface facing
337
upwards (Fig. 1 B). After preparation, the patches were bathed in minimal essential medium (MEM; Invitrogen Corp., La Jolla, CA, USA) supplemented with 20% fetal calf serum (FCS; PAA Clone, Coelbe, Germany), 1% MEM vitamins (Invitrogen), 1% MEM nonessential amino acids and 1% Pen/Strep (Biochrom AG, Berlin, Germany). Preservation of the endothelial cell layer on aorta preparations was approved by immunostaining of endothelial cell adhesion molecule PECAM-1/CD31 (Fig. 1 C) as reported elsewhere in detail [27]. To determine eGC’s thickness and stiffness we used the Atomic Force Microscope (AFM) nanoindentation technique. In Fig. 1 D, we illustrate the basic principles of this method. Using a Multimode AFM (Veeco, Mannheim, Germany) with a feedback-controlled heating device (Veeco), measurements were performed at 37 C as described previously [28]. In short, the central component of the AFM consists of a triangular cantilever with a mounted spherical tip (here: electrically uncharged polystyrene, diameter ¼ 10 mm, Novascan, Ames, IA, USA), which is a very sensitive mechanical nanosensor. This cantilever is used to periodically indent the cells and functions as a soft spring with a spring constant of 11 pN/nm. A laser beam is targeted at the gold-coated backside of the cantilever to be reflected to a position-sensitive quadrupled photodiode, which allows quantification of the cantilever deflection (V). To facilitate the calculation of the force (F) acting on the cantilever and, in turn, the force exerted by the cantilever to the sample, we previously determined the spring constant of the cantilever (Kcant) by the thermal tuning method and the deflection sensitivity (a) of the cantilever on bare glass coverslips.
F ¼ V$a$Kcant Knowing both the piezo displacement (xpiezo) and the deflection sensitivity (a), the indentation depth (deformation) of the sample (xsample) can consecutively be calculated.
xsample ¼ xpiezo ða$VÞ In the following, the indentation depth is hereafter called “thickness”. Of note, indentation depth represents the apparent thickness, not the anatomically exact thickness. Plotting the force (F) necessary to indent a single cell (indentation depth, xsample) leads to force indentation curves. The sample stiffness can be derived from Hook’s law.
Ksample ¼
F xsample
The stiffness (K) is the sample’s mechanical resistance against a defined deformation (e.g. indentation) and is strongly determined by the indentation depth and its location, due to the varying distribution of cellular organelles. The experimental parameters encompass an indentation velocity of 1 mm/s, a loading force of approximately 400 pN and an indentation frequency between 0.25 and 0.5 Hz. All measurements were performed in HEPES-buffered solution [standard composition 140 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 5 mM glucose, 10 mM HEPES (N-2-hydroxyethylpiperazineN0 -2-ethanesulfonic acid), pH 7.4] supplemented with 1% FCS in order to prevent eGC collapse [29].
2.7. Atomic force microscopy on rat aorta in vitro Additionally, indentation measurements on rat aorta preparations were performed after enzymatic eGC degradation in vitro by a mix of heparinase I (0.5 Sigma-U/ml), hyaluronidase (25.0 SeU/ml),
338
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Fig. 1. Atomic force microscopy. A) Preparation of rat aortas. B) Microscopic view of rat aortic endothelium. C) Immunostaining of endothelial cell adhesion molecule PECAM-1/ CD31. D) Original tracing of a forceedistance curve performed on the endothelial surface of a rat aorta ex vivo: 1) The AFM tip travels vertically towards the endothelial surface. 2) Upon indentation of the endothelial glycocalyx (eGC), the AFM cantilever, which serves as a soft spring, is deflected. The cantilever deflection is measured as a laser beam reflected from the back of the cantilever and is plotted as a function of sample position along the z-axis. The resulting curve is transformed into a force-versus-distance curve using the cantilever’s spring constant and the optical lever sensitivity. The slope of a forceedistance curve then directly reflects the stiffness (expressed in pN/nm), which is necessary to indent the eGC for a certain distance. The first slope indicates the stiffness (in this trace 0.31 pN/nm) of the very first layer, the (eGC). The second slope indicates the stiffness of the plasma membrane and the cortical actin web. The distance between the starting point of eGC indentation and the starting point of the second slope (projected to the x-axis) corresponds to the thickness of the eGC (in this trace 304.5 nm).
and chondroitinase (0.1 SeU/ml, all Sigma) for 60 min at 37 C. An equal amount of buffer (0.9% NaCl) served as control.
accepted at P < 0.05. Data analysis was performed using SPSS (IBMÒ, Armonk, NY, USA). Figures were prepared using GraphPad Prism (GraphPad Prism Software Inc., San Diego, CA, USA).
2.8. Statistical analysis 3. Results Data are presented as absolute numbers, percentages, means with corresponding standard deviations, or medians with corresponding 25th and 75th percentiles [inter-quartile range (IQR)]. Baseline characteristics of CKD patients and controls were compared using one-way ANOVA with post-hoc Scheffé’s test, KruskaleWallis test, and the c2-test as appropriate. The relationship between Syn-1, HA and clinical data as well as circulating markers of endothelial dysfunction was investigated using Pearson’s productemoment correlation. The association of circulating Syn-1 and HA with eGFR was further evaluated in adjusted linear regression models. Selection of variables to be included in the multivariable models was done a priori by determining probable confounders based both on differences in baseline characteristics in patients with different Syn-1 and HA levels and on theoretical considerations. All tests were two-sided and significance was
3.1. Circulating eGC plasma levels increase with progress of CKD The clinical characteristics of the study population are given in Table 1. Etiology of CKD was not different between CKD stages 3e5. As in previous studies, classical cardiovascular risk factors such as diabetes mellitus, levels of glycated hemoglobin or arterial hypertension increased significantly throughout the different stages of CKD (Table 1). Interestingly, omitting diabetic patients (n ¼ 32) from analysis did not substantially change the results displayed in Table 1 (Supplemental Table 1). Further, dividing the cohort into non-diabetic vs. diabetic patients revealed that diabetics had higher creatinine, proteinuria, prevalence of arterial hypertension, BMI, endothelial dysfunction as well as enhanced glycocalyx shedding (Supplemental Table 2).
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
339
Fig. 2. Levels of syndecan-1/hyaluronan increase throughout the different stages of CKD. Left: Box plots showing plasma levels of A) syndecan-1 (Syn-1) and C) hyaluronan (HA) measured by ELISA methodology in plasma samples from controls (n ¼ 31) and patients with different stages of CKD (n ¼ 95). The differences between patients and control subjects were assessed by one-way ANOVA with post-hoc Scheffé’s test (*P < 0.05 versus controls). Right: Scatter plots showing the relation (Spearman’s rank correlation coefficient) of estimated glomerular filtration rate (eGFR) to plasma levels of Syn-1 B) and HA D).
The median Syn-1 level was about 4-fold higher in CKD patients compared to controls (203 vs. 53 ng/mL; P < 0.001). The same was observed for HA (2.2-fold over controls, 179 vs. 81 ng/mL; P < 0.001). Next, we assessed circulating eGC components according to CKD stages. Median Syn-1 levels steadily increased across the following groups: controls: 53 (41e92) ng/mL, CKD 3: 90 (61e344) ng/mL, CKD 4: 212 (100e355) ng/mL, and CKD 5: 270 (92e515) ng/mL, respectively (non-parametric ANOVA P < 0.0001). Syn-1 levels were significantly elevated in patients with higher CKD stages (i.e. 4 and 5) compared to controls (P ¼ 0.047 and P ¼ 0.001, respectively, Fig. 2 A). Interestingly, HA showed a shallower increase but peaked distinctly in patients with CKD stage 5 (controls: 81 (48e171) ng/mL, CKD 3: 133 (62e224) ng/mL, CKD 4: 147 (86e 230) ng/mL, and CKD 5441 (163e1213) ng/mL, respectively; nonparametric ANOVA P < 0.0001). Median HA values in CKD 5 were significantly elevated compared to all other groups (P < 0.001, Fig. 2 C). Both parameters correlated significantly and inversely with the eGFR (Fig. 2 B/D). To identify associations between eGC damage and demographic and clinical variables, we stratified baseline characteristics of all participants according to tertiles of plasma Syn-1 (Table 2) and HA levels (Supplemental Table 3). Patients within the lowest Syn-1 tertile showed the highest eGFR (P < 0.001) and the lowest proteinuria (P < 0.001). Surprisingly, no significant differences were observed with regards to age, gender, medication, and major cardiovascular risk factors, except for diabetes (P ¼ 0.008). Basically, the same results were obtained for HA, with the exception that age
(P ¼ 0.005) and the use of erythropoietin (P ¼ 0.021) slightly increased across HA tertiles. To confirm the association between eGC shedding and renal function, we performed multivariate analysis using linear regression models and treating Syn-1 and HA as dependent variables. Table 3 shows that eGFR is independently associated with plasma Syn-1 and HA levels in all models, thus excluding significant bias from variables such as age, gender, arterial hypertension, smoking, diabetes and dyslipidaemia. 3.2. Circulating glycocalyx components correlate with endothelial dysfunction As initially hypothesized, we expected elevated soluble eGC components to be associated with endothelial dysfunction in CKD patients. We thus measured plasma levels of vWF, sVCAM-1, sFlt-1, and Ang-2 as established markers of endothelial activation and dysfunction [25,30e32]. As expected, markers of endothelial dysfunction steadily increased across different stages of CKD (Table 1). Similarly, circulating levels of Ang-2, sFlt-1, sVCAM-1 and vWF steadily increased across tertiles of plasma Syn-1 (all P 0.001, Table 2) and plasma HA (P ¼ 0.004, Supplemental Table 3). Bivariate correlations using Spearman’s test revealed moderate-to-strong correlations of both Syn-1 and HA with sFlt-1 (Syn-1: r ¼ 0.582; HA: r ¼ 0.446), sVCAM-1 (Syn-1: r ¼ 0.527; HA: r ¼ 0.709), vWF (Syn-1: r ¼ 0.373; HA: r ¼ 0.445), and Ang-2 (Syn-1: r ¼ 0.455; HA: r ¼ 0.523) with P < 0.001 for all calculations.
340
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Table 2 Baseline characteristics according to plasma Syn-1 levels.a Tertiles of syndecan-1 (ng/ml)
Age (yr) Male gender [% (n)] Creatinine (mg/dl) eGFR (ml/min/1.73 m2)d Proteinuria (g/24 h) Cardiovascular risk factors Arterial hypertension [% (n)] SBP (mmHg) DBP (mmHg) BMI (kg/m2) Cholesterol (mg/dl) HDL-cholesterol (mg/dl) Calculated LDL-cholesterol (mg/dl) Diabetes [% (n)] HbA1c (%) Framingham risk score (%) Smokers [% (n)] Current medication ACE inhibitors [% (n)] AT1 blockers [% (n)] Erythropoietin [% (n)] Statins [% (n)] Markers of endothelial dysfunction Ang-2 (ng/ml) sFlt-1 (pg/ml) sVCAM-1 (pg/ml) vWF (U/ml) Circulating glycocalyx components Hyaluronan (ng/ml) a b c d
P value
<78 (n ¼ 41)
79 to 241 (n ¼ 41)
>241 (n ¼ 43)
54 22 44 (18) 1.3 (0.9e3.1) 48 (19e84) 0.0 (0.0e0.8)
62 19 66 (27) 2.0 (1.3e2.9) 24 (15e41) 0.3 (0.1e0.7)
65 14 53 (23) 2.5 (1.8e5.1) 16 (5e27) 1.1 (0.4e1.3)
51 (21) 130 (125e148) 80 (80e80) 26.5 (23.3e29.0) 166 (134e186) 48 (43e54) 110 (97e138) 10 (4) 5.0 (4.7e5.3) 8 (1e17) 15 (6)
73 (30) 130 (121e149) 80 (76e80) 24.6 (21.2e29.0) 167 (143e201) 42 (38e56) 111 (89e132) 22 (9) 5.1 (4.8e5.7) 12 (7e20) 39 (16)
70 (30) 130 (128e140) 80 (70e80) 24.9 (23.4e30.0) 162 (130e204) 43 (38e48) 100 (70e119) 42 (18) 5.5 (5.0e6.6) 12 (4e20) 14 (6)
0.339c 0.827b 0.218b 0.309b 0.802b 0.128b 0.479b 0.008c 0.008b 0.084b 0.009c
27 (11) 29 (12) 7 (3) 22 (9)
54 20 12 29
42 14 21 42
0.144c 0.087c 0.285c 0.246c
1.1 (0.6e1.6) 77 (56e87) 736 (586e961) 0.2 (0.2e0.7)
1.4 (0.8e3.4) 88 (70e122) 789 (584e1039) 0.4 (0.3e0.7)
3.0 (1.4e6.3) 119 (99e411) 1577 (1192e2247) 0.5 (0.3e1.2)
<0.001b <0.001b <0.001b 0.004b
111 (50e184)
103 (75e247)
293 (158e766)
<0.001b
(22) (8) (5) (12)
(18) (6) (9) (18)
0.097b 0.168c <0.001b <0.001b <0.001b
Variables are presented as median (interquartile range) or as mean SD. KruskaleWallis one-way ANOVA. c2-test. GFR values were calculated using CKD-EPI formula.
3.3. 5/6-Nephrectomy leads to impaired nanomechanical properties of the eGC To confirm that elevated plasma levels of Syn-1 and HA reflect structural and/or functional alterations of the eGC, we used an established rat model of CKD, namely 5/6-nephrectomy. Physiologic data assessed 14 days after surgery is given in Supplemental Table 4. Nano-indentation measurements by AFM revealed a highly significant decrease of eGC thickness (360 vs. 157 nm, P ¼ 0.001) and stiffness (0.33 vs. 0.22 pN/nm, P < 0.001) in aortic explants from 5/6nephrectomized rats (n ¼ 4) compared to sham-operated animals (n ¼ 6) (Fig. 3 A, B). Additionally, we performed control measurements by incubating rat aortic endothelium in vitro with control buffer or an enzyme mix (consisting of heparinase, chondroitinase and hyaluronidase), respectively. The latter was used to strip all three
highly abundant components off the eGC, namely heparan sulfate, HA and chondroitin sulfate. As shown in Fig. 3 A and B, the decrease in glycocalyx thickness and stiffness after enzyme treatment was comparable to the alterations seen in 5/6-nephrectomized rats. As in CKD patients, Syn-1 plasma levels were elevated in 5/6nephrectomized rats (n ¼ 4) compared with sham-operated animals (n ¼ 6) (P < 0.001, Fig. 3 C). As in humans, plasma Syn-1 levels tightly correlated with plasma sFlt-1 levels (r ¼ 0.901, P ¼ 0.0008) and the creatinine-based approximation of GFR (r ¼ 0.783, P ¼ 0.01) in CKD rats. 4. Discussion Whilst special electron microscopic staining procedures uncovered that the luminal surface of the endothelium expressed a
Table 3 Adjusted linear regression model treating Syn-1/HA as dependent variables. Syndecan-1
Unadjusted eGFR Adjusted for model 1 eGFR Adjusted for model 2 eGFR Adjusted for model 3 eGFR
Hyaluronan
Stand. b
(95% CI)
P
Stand. b
(95% CI)
P
0.388
(6.07 to 1.855)
<0.001
0.382
(6.985 to 2.616)
<0.001
0.388
(6.07 to 1.855)
<0.001
0.382
(6.985 to 2.616)
<0.001
0.286
(5.156 to 1.125)
0.003
0.431
(7.326 to 3.077)
<0.001
0.294
(5.277 to 1.187)
0.002
0.416
(7.184 to 2.868)
<0.001
CI, confidence interval. Model 1: adjusted for age and gender. Model 2: adjusted for age, gender, arterial hypertension, diabetes, and smoking. Model 3: adjusted for age, gender, arterial hypertension, diabetes, smoking, HbA1c, SBP, DBP, body mass index, and Framingham risk score.
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Fig. 3. Endothelial glycocalyx in 5/6-nephrectomized rats. Bar charts showing thickness (A) and stiffness (B) of endothelial glycocalyx (eGC) as analyzed by atomic force microscopy (AFM) on living endothelium from rat aortic explants (ex vivo) harvested at day 14 after sham operation (n ¼ 6) or 5/6-nephrectomy (5/6Nx; n ¼ 4). From each aorta 2 aortic patches were prepared. On each patch 5 force indentation curves were recorded in each of 3 different tip positions, respectively. For statistical comparison, a single average value per tip position, aortic patch and aorta was calculated, respectively. For comparison, aortic explants from untreated rats were treated in vitro with either control buffer (n ¼ 3) or an enzyme mix (n ¼ 3) for 60 min to strip heparan sulfate, hyaluronan and chondroitin sulfate off the eGC. In addition C) Syn-1 plasma levels in sham-operated and 5/6-nephrectomized rats at day 14 after the procedure were measured by ELISA. *P < 0.001 versus sham-operated controls or buffer-treated controls, respectively.
carbohydrate layer by the mid-1940s, it was considered to be without major physiological relevance or functional significance for a long time. In the past two decades, however, intravital microscopy studies revealed that the eGC represents a substantial intravascular compartment contributing significantly to vascular wall homeostasis [9,33]. Since then, different groups have conclusively shown that the endothelial glycocalyx is seriously damaged in patients with diabetes [34,35], hypercholesterolemia [13,36], acute inflammation [13,36], or after ischemia reperfusion [22,37]. Only recently, Vlahu et al. found both, thinning and shedding of the glycocalyx in patients on maintenance dialysis [38]. However, the question whether impaired renal function per se is an independent risk
341
factor for glycocalyx impairment has not been formally addressed so far. This is particularly remarkable considering the fact, that CKD is an increasingly recognized contributor to cardiovascular morbidity and mortality. In the present study, we found a robust inverse association between reduced renal function and presumable shedding of the atheroprotective glycocalyx in a well-defined cohort of patients comprising different CKD stages. This association was hardly influenced by established triggers of glycocalyx damage such as dyslipidaemia and diabetesethe latter being the only other factor associated with increased glycocalyx shedding in univariate analysis (see also Supplemental Table 2). Interestingly, Syn-1 plasma levels were already increased in CKD 4, while the increase in HA became only evident in CKD 5. Given its unique and delicate composition, HA shedding may thus be indicative of more severe glycocalyx damage. It is thus conceivable that CKD is a new and potentially relevant risk factor for glycocalyx damage. Little is known about the half-life of circulating glycocalyx components. Early studies have shown that circulating HA (plasma half-life < 5 min) is rapidly eliminated by the liver, while renal excretion plays a negligible part in clearance [39,40]. Because the elimination of Syn-1 from the circulation has not been reported, we excluded renal accumulation of Syn-1 by comparing 24 h clearances of Syn-1 and creatinine in a subgroup of 21 patients (data not shown). Together, these data indicate the existence of a steady state between active refurbishment and on-going damage of the endothelial glycocalyx in CKD. Glycocalyx research has traditionally focused on thickness as a surrogate of structural integrity. Although highly intuitive in principle, this concept provides no conclusions concerning the mechanical properties of this highly elute and deformable layer. We have recently shown that AFM can reliably depict nanomechanics of the eGC on a variety of living endothelial cells by a linear approximation over the first hundreds of nanometres of the force indentation curves that represent the eGC [27]. Although heterogeneity and limitations of in vivo visualization methods probably impede any direct comparison of glycocalyx dimensions, we noted that the average nanomechanical thickness of the glycocalyx on ex vivo aortic preparations from sham-operated rats (360 nm) was compatible with many of the aforementioned reports [41e43]. Interestingly, here we could show by AFM that CKD induction (5/6 nephrectomy) in rats caused considerable softening and thinning (>50% from baseline) of the aortic glycocalyx, which was paralleled by Syn-1. In fact, nanomechanical changes after 5/6 nephrectomy were quite comparable to those seen after either enzymatic removal of its three main constituents e heparan sulfate, chondroitin sulfate, and hyaluronic acid e or during gram-negative endotoxemia [27]. Although findings from this animal study cannot be readily extrapolated to humans, it is conceivable to assume that elevated Syn-1/HA levels in CKD indeed reflect structural damage of the glycocalyx. Salmon and co-workers recently provided another indication for the importance of eGC in the context of chronic kidney disease and endothelial dysfunction [44]. In a rat model of spontaneous albuminuric CKD, they could conclusively show that loss of eGC is paralleled by defects in microvascular permeability. Specific mechanisms of eGC shedding in CKD remain poorly understood. It is now accepted that TNF-alpha [13,15] and hyperglycaemia [19] cause glycocalyx disruption probably via the action of proteases. In accordance with this notion, Vlahu et al. found elevated hyaluronidase activity in plasma from dialysis patients [38]. However, from a nephrologist’s perspective, two recently identified mediators of glycocalyx deterioration, namely hypervolemia [20] and sodium overload [28] might be of particular importance in CKD patients. Patients with impaired kidney function gradually lose their ability to excrete salt and water.
342
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Particularly in patients on maintenance dialysis (which had the highest Syn-1 and HA values in our cohort e Supplemental Table 5) the prevalence of volume overload has been estimated as high as 63% [45]. It is currently believed that hypervolemia is negatively affecting the glycocalyx via atrial natriuretic peptide (ANP), which is released after wall stress in the right atrium of the heart of humans [18]. However, the exact intracellular signal cascade leading to degradation of the glycocalyx by ANP is still unknown. Furthermore, Oberleithner and co-workers revealed that the eGC plays an important role as a significant buffer for sodium [7,28]. It is known, that the eGC preferentially binds sodium due to its overall net negative charge. They estimated that 700 mg of sodium, which is proportionate to the sodium amount of a single meal, can be transiently bound to the eGC within the human body to avoid sodium overload of the endothelium [7]. However, the increase of extracellular sodium concentration led to perturbation of sodium buffering capacity, as well as thinning and loss of negatively charged HS side chains in vitro. [28] Glycocalyx impairment in CKD might thus facilitate sodium entry into the endothelial cells. This could explain endothelial dysfunction and arterial hypertension observed in patients with impaired kidney function [46]. Consistent with this hypothesis, we found medium to strong correlations between Syn-1/HA plasma levels and several markers of endothelial dysfunction/activation. Intriguingly, eGC damage did not seem to correlate with the Framingham Risk Score. This goes along with the preliminary observation that only HA (299 vs. 387 ng/ml, P ¼ 0.021) but not Syn-1 (288 vs. 328 ng/ml, P ¼ 0.26) was significantly higher in patients with history of cardiovascular disease (i.e. “peripheral arterial disease” or “stroke” or “acute coronary syndrome” e data not shown)). However, interpretation of this finding is difficult given that (subclinical) cardiovascular disease was not thoroughly assessed and thus may be underdiagnosed in our patients. Some limitations regarding our study need to be addressed. First, the number of participants (n ¼ 126) might be too small. However, besides applying strict inclusion criteria to minimalize possible confounding effects and increase the validity of the obtained results, an animal model of CKD was used to avoid these confounders and confirm our results. Second, median Syn-1 plasma levels of control subjects were 56 ng/ml (5.6 mg/dl) and thus w2e4 fold higher than plasma Syn-1 levels from controls from the literature, which range between 12 ng/dl and 20.5 ng/dl in different studies using the same ELISA kit [47,48]. Median HA levels from controls subjects (81 ng/ml in our cohort) were comparable to controls from two previously published studies by Nieuwdorp and coworkers (HA 65 and 70 ng/ml) using the same ELISA kit [19,35]. A possible explanation for the slightly higher Syn-1 levels seen in our controls is that these were a mix of diseases controls and healthy volunteers (see M&M section) but not solely healthy volunteers as in the aforementioned studies. Third, inflammatory processes also play a part in eGC degradation. Though we did not systematically analyze a possible influence of (micro-) inflammation on eGC damage, a review of chart records and lab data revealed an only moderate correlation between Syn1 and CRP (r ¼ 0.401, P ¼ 0.013), whereas HA and CRP did not correlate significantly (r ¼ 0.201, P ¼ 0.22). Further studies focusing on this aspect would be highly desirable. Fourth, according to the cross-sectional study design, outcome data were not obtainable. Finally, due to the fact that the eGC is a well compressible structure, it is important to mention that the AFM approach of this study gives an indication of the apparent thickness, rather than the true anatomical thickness of the eGC. 5. Conclusion With our study, we support and extend findings by Vink’s group and show first evidence of shedding of the atheroprotective eGC in
(non-) dialysis CKD patients. Increased shedding correlates tightly with endothelial dysfunction and is independently associated with renal function impairment. This novel finding opens new perspectives on the pathophysiology of cardiovascular disease in CKD. Still, further studies are needed to fully understand the mechanisms behind eGC alterations in those patients. These, however, could potentially uncover how to protect the eGC and eventually reduce cardiovascular morbidity and mortality in CKD. Competing interests The authors declare that they have no competing interests. Acknowledgments This work was supported by the fund “Innovative Medical Research” of the University of Münster Medical School (KÜ111015). The study was partly supported by the Deutsche Forschungsgemeinschaft (CRC 656, project C7). Support by the Cluster of Excellence Cell in Motion (EXC 1003 e CiM) is gratefully acknowledged. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atherosclerosis.2014.03.016. References [1] Di Angelantonio E, Chowdhury R, Sarwar N, et al. Chronic kidney disease and risk of major cardiovascular disease and non-vascular mortality: prospective population based cohort study. BMJ 2010;341:c4986. [2] Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998;32:S112e9. [3] Schiffrin EL, Lipman ML, Mann JF. Chronic kidney disease: effects on the cardiovascular system. Circulation 2007;116:85e97. [4] Nieuwdorp M, Meuwese MC, Vink H, et al. The endothelial glycocalyx: a potential barrier between health and vascular disease. Curr Opin Lipidol 2005;16:507e11. [5] Pries AR, Secomb TW, Gaehtgens P. The endothelial surface layer. Pflugers Arch 2000;440:653e66. [6] Reitsma S, Slaaf DW, Vink H, et al. The endothelial glycocalyx: composition, functions, and visualization. Pflugers Arch 2007;454:345e59. [7] Oberleithner H. Two barriers for sodium in vascular endothelium? Ann Med 2012;44(Suppl. 1):S143e8. [8] Mulivor AW, Lipowsky HH. Role of glycocalyx in leukocyte-endothelial cell adhesion. Am J Physiol Heart Circ Physiol 2002;283:H1282e91. [9] Curry FE, Adamson RH. Endothelial glycocalyx: permeability barrier and mechanosensor. Ann Biomed Eng 2012;40:828e39. [10] Constantinescu A, Spaan JA, Arkenbout EK, et al. Degradation of the endothelial glycocalyx is associated with chylomicron leakage in mouse cremaster muscle microcirculation. Thromb Haemost 2011;105:790e801. [11] Vink H, Constantinescu AA, Spaan JA. Oxidized lipoproteins degrade the endothelial surface layer: implications for platelet-endothelial cell adhesion. Circulation 2000;101:1500e2. [12] Lipowsky HH. Microvascular rheology and hemodynamics. Microcirculation 2005;12:5e15. [13] Nieuwdorp M, Meuwese MC, Mooij HL, et al. Tumor necrosis factor-alpha inhibition protects against endotoxin-induced endothelial glycocalyx perturbation. Atherosclerosis 2009;202:296e303. [14] Henry CB, Duling BR. TNF-alpha increases entry of macromolecules into luminal endothelial cell glycocalyx. Am J Physiol Heart Circ Physiol 2000;279: H2815e23. [15] Chappell D, Hofmann-Kiefer K, Jacob M, et al. TNF-alpha induced shedding of the endothelial glycocalyx is prevented by hydrocortisone and antithrombin. Basic Res Cardiol 2009;104:78e89. [16] Constantinescu AA, Vink H, Spaan JA. Elevated capillary tube hematocrit reflects degradation of endothelial cell glycocalyx by oxidized LDL. Am J Physiol Heart Circ Physiol 2001;280:H1051e7. [17] Bruegger D, Jacob M, Rehm M, et al. Atrial natriuretic peptide induces shedding of endothelial glycocalyx in coronary vascular bed of guinea pig hearts. Am J Physiol Heart Circ Physiol 2005;289:H1993e9.
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343 [18] Jacob M, Saller T, Chappell D, et al. Physiological levels of A-, B- and C-type natriuretic peptide shed the endothelial glycocalyx and enhance vascular permeability. Basic Res Cardiol 2013;108:347. [19] Nieuwdorp M, van Haeften TW, Gouverneur MC, et al. Loss of endothelial glycocalyx during acute hyperglycemia coincides with endothelial dysfunction and coagulation activation in vivo. Diabetes 2006;55:480e6. [20] Rehm M, Haller M, Orth V, et al. Changes in blood volume and hematocrit during acute preoperative volume loading with 5% albumin or 6% hetastarch solutions in patients before radical hysterectomy. Anesthesiology 2001;95: 849e56. [21] Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604e12. [22] Rehm M, Bruegger D, Christ F, et al. Shedding of the endothelial glycocalyx in patients undergoing major vascular surgery with global and regional ischemia. Circulation 2007;116:1896e906. [23] Chappell D, Jacob M, Hofmann-Kiefer K, et al. Hydrocortisone preserves the vascular barrier by protecting the endothelial glycocalyx. Anesthesiology 2007;107:776e84. [24] Lukasz A, Hellpap J, Horn R, et al. Circulating angiopoietin-1 and angiopoietin2 in critically ill patients: development and clinical application of two new immunoassays. Crit Care 2008;12:R94. [25] Di Marco GS, Reuter S, Hillebrand U, et al. The soluble VEGF receptor sFlt1 contributes to endothelial dysfunction in CKD. J Am Soc Nephrol 2009;20: 2235e45. [26] Reuter S, Bangen P, Edemir B, et al. The HSP72 stress response of monocytes from patients on haemodialysis is impaired. Nephrol Dial Transpl 2009;24: 2838e46. [27] Wiesinger A, Peters W, Chappell D, et al. Nanomechanics of the endothelial glycocalyx in experimental sepsis. PloS One 2013;8:e80905. [28] Oberleithner H, Peters W, Kusche-Vihrog K, et al. Salt overload damages the glycocalyx sodium barrier of vascular endothelium. Pflugers Arch 2011;462: 519e28. [29] Thi MM, Tarbell JM, Weinbaum S, et al. The role of the glycocalyx in reorganization of the actin cytoskeleton under fluid shear stress: a “bumper-car” model. Proc Natl Acad Sci U S A 2004;101:16483e8. [30] David S, Kumpers P, Hellpap J, et al. Angiopoietin 2 and cardiovascular disease in dialysis and kidney transplantation. Am J Kidney Dis 2009;53: 770e8. [31] David S, John SG, Jefferies HJ, et al. Angiopoietin-2 levels predict mortality in CKD patients. Nephrol Dial Transplant 2012;27:1867e72. [32] Pequeriaux NC, Fijnheer R, Gemen EF, et al. Plasma concentration of von Willebrand factor predicts mortality in patients on chronic renal replacement therapy. Nephrol Dial Transpl 2012;27:2452e7.
343
[33] Chappell D, Westphal M, Jacob M. The impact of the glycocalyx on microcirculatory oxygen distribution in critical illness. Curr Opin Anaesthesiol 2009;22:155e62. [34] Broekhuizen LN, Lemkes BA, Mooij HL, et al. Effect of sulodexide on endothelial glycocalyx and vascular permeability in patients with type 2 diabetes mellitus. Diabetologia 2010;53:2646e55. [35] Nieuwdorp M, Mooij HL, Kroon J, et al. Endothelial glycocalyx damage coincides with microalbuminuria in type 1 diabetes. Diabetes 2006;55:1127e32. [36] Meuwese MC, Mooij HL, Nieuwdorp M, et al. Partial recovery of the endothelial glycocalyx upon rosuvastatin therapy in patients with heterozygous familial hypercholesterolemia. J Lipid Res 2009;50:148e53. [37] Bruegger D, Schwartz L, Chappell D, et al. Release of atrial natriuretic peptide precedes shedding of the endothelial glycocalyx equally in patients undergoing on- and off-pump coronary artery bypass surgery. Basic Res Cardiol 2011;106(6):1111e21. [38] Vlahu CA, Lemkes BA, Struijk DG, et al. Damage of the endothelial glycocalyx in dialysis patients. J Am Soc Nephrol 2012;23(11):1900e8. [39] Fraser JR, Laurent TC, Pertoft H, et al. Plasma clearance, tissue distribution and metabolism of hyaluronic acid injected intravenously in the rabbit. Biochem J 1981;200:415e24. [40] Fraser JR, Laurent TC, Engstrom-Laurent A, et al. Elimination of hyaluronic acid from the blood stream in the human. Clin Exp Pharmacol Physiol 1984;11: 17e25. [41] Gao L, Lipowsky HH. Composition of the endothelial glycocalyx and its relation to its thickness and diffusion of small solutes. Microvasc Res 2010;80: 394e401. [42] van den Berg BM, Vink H, Spaan JA. The endothelial glycocalyx protects against myocardial edema. Circ Res 2003;92:592e4. [43] Potter DR, Damiano ER. The hydrodynamically relevant endothelial cell glycocalyx observed in vivo is absent in vitro. Circ Res 2008;102:770e6. [44] Salmon AH, Ferguson JK, Burford JL, et al. Loss of the endothelial glycocalyx links albuminuria and vascular dysfunction. J Am Soc Nephrol JASN 2012;23: 1339e50. [45] Ebah LM, Wiig H, Dawidowska I, et al. Subcutaneous interstitial pressure and volume characteristics in renal impairment associated with edema. Kidney Int 2013;84:980e8. [46] Kusche-Vihrog K, Oberleithner H. An emerging concept of vascular salt sensitivity. F1000 Biol Rep 2012;4:20. [47] Steppan J, Hofer S, Funke B, et al. Sepsis and major abdominal surgery lead to flaking of the endothelial glycocalix. J Surg Res 2011;165:136e41. [48] Anttonen A, Leppa S, Ruotsalainen T, et al. Pretreatment serum syndecan-1 levels and outcome in small cell lung cancer patients treated with platinum-based chemotherapy. Lung Cancer 2003;41:171e7.
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Damage of the endothelial glycocalyx in chronic kidney disease Jan-Sören Padberg a, Anne Wiesinger a, Giovana Seno di Marco a, Stefan Reuter a, Alexander Grabner a, Dominik Kentrup a, Alexander Lukasz a, Hans Oberleithner b, Hermann Pavenstädt a, Marcus Brand a, Philipp Kümpers a, * a Department of Medicine D, Division of General Internal Medicine, Nephrology, and Rheumatology, University Hospital Münster, Albert-Schweitzer-Campus 1 A1, 48149 Münster, Germany b Institute of Physiology II, University of Muenster, Robert-Koch-Straße 27 b, 48149 Münster, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 November 2013 Received in revised form 24 February 2014 Accepted 17 March 2014 Available online 29 March 2014
Background and objectives: The endothelial glycocalyx (eGC), a mesh of anionic biopolymers covering the luminal surface of endothelial cells, is considered as an intravascular compartment that protects the vessel wall against pathogenic insults in cardiovascular disease. We hypothesized that chronic kidney disease (CKD) is associated with reduced eGC integrity and subsequent endothelial dysfunction. Methods & results: Shedding of two major components of the eGC, namely syndecan-1 (Syn-1) and hyaluronan (HA), was measured by ELISA in 95 patients with CKD (stages 3e5) and 31 apparently healthy controls. Plasma levels of Syn-1 and HA increased steadily across CKD stages (5- and 5.5-fold, respectively P < 0.001) and were independently associated with impaired renal function after multivariate adjustment. Furthermore, Syn-1 and HA correlated tightly with plasma markers of endothelial dysfunction such as soluble fms-like tyrosine kinase-1 (sFlt-1), soluble vascular adhesion molecule-1 (sVCAM-1), von-Willebrand-Factor (vWF) and angiopoietin-2 (P < 0.001). Experimentally, excessive shedding of the eGC, evidenced by 11-fold increased Syn-1 plasma levels, was also observed in an established rat model of CKD, the 5/6-nephrectomized rats. Consistently, an atomic force microscopybased approach evidenced a significant decrease in eGC thickness (360 79 vs. 157 29 nm, P ¼ 0.001) and stiffness (0.33 0.02 vs. 0.22 0.01 pN/nm, P < 0.001) of aorta endothelial cell explants isolated from CKD rats. Conclusion: Our findings provide evidence for damage of the atheroprotective eGC as a consequence of CKD and potentially open a new avenue to pathophysiology and treatment of cardiovascular disease in renal patients. Ó 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Endothelial glycocalyx Chronic kidney disease Shedding Syndecan-1 Hyaluronan Atomic force microscopy 5/6 Nephrectomy Cardiovascular disease Atherosclerosis
1. Introduction Patients with chronic kidney disease (CKD) exhibit endothelial dysfunction and accelerated vascular disease leading to increased morbidity and mortality due to cardiovascular events [1e3]. The underlying mechanisms, however, are still not fully understood. Endothelial dysfunction has long been ascribed to a malfunction of the endothelial cell itself. Recent studies, though, provided compelling evidence that the endothelium is protected against pathogenic insults by a highly hydrated negatively charged “firewall” called the glycocalyx. The endothelial glycocalyx (eGC), a carbohydrate-rich mesh of large anionic polymers, lines the luminal side of the endothelium along the entire vascular tree [4]. It
* Corresponding author. Tel.: þ49 25 18 34 49 51. E-mail address: [email protected] (P. Kümpers). http://dx.doi.org/10.1016/j.atherosclerosis.2014.03.016 0021-9150/Ó 2014 Elsevier Ireland Ltd. All rights reserved.
is composed of core proteoglycans, especially those of the syndecan family, to which both highly sulfated glycosaminoglycans and hyaluronan are attached [5,6]. Measuring about 0.5e2 mm in thickness, the eGC forms an integral part of the vascular barrier alongside the endothelial cell itself [7]. Having been neglected over decades, important physiological properties were attributed to the eGC during the last years. Given its strategic location as the interface between blood and endothelium, the intact glycocalyx prevents transvascular protein leakage and reduces leukocyteeendothelial interactions [8e10]. Beyond that, the glycocalyx contributes to the regulation of redox state and is crucially involved in the mediation of shear-induced nitric oxide release as well as physiologic anticoagulation [4,6,11]. Its structure is fairly stable but also in healthy endothelium subject to a permanent dynamic equilibrium between biosynthesis of new components and shear dependent removal, the so-called shedding, of existing constituents [12].
336
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Damage to eGC, however, can occur under exposure to inflammatory or atherogenic noxae, such as endotoxin [13] or TNF-alpha [14,15], oxidized LDL [11,16], excess atrial natriuretic peptide [17,18], hyperglycaemia [19], and hypervolemia [20]. All of these eGCdeteriorating stimuli occur abundantly in CKD. Given the fact that the mechanisms of accelerated atherosclerosis and cardiovascular disease in CKD patients are still not fully understood, we aimed to investigate whether eGC damage might be a mediating factor in this process. Thus, we sought to assess the impact of chronic renal impairment on 1) circulating markers of eGC damage and endothelial dysfunction in CKD patients, as well as on 2) eGC structure (i.e. thickness and stiffness) itself of aorta endothelial cell explants isolated from CKD rats. 2. Materials & methods 2.1. Patients The medical ethics committee at University Hospital Münster (Münster, Germany) approved the study and written informed consent was obtained from all subjects. In total, 95 stable Caucasian patients with CKD stages 3e5 (eGFR 60 ml/min/ 1.73 m2) were included in this study. Patients with presumable eGC alterations due to acute or chronic infections, malignancy, history of organ transplantation, or pregnancy were excluded. CKD patients were divided according to the National Kidney Foundation’s classification of CKD. Subjects with eGFR 60 ml/
min/1.73 m2 and without evidence of proteinuria served as controls. The control group (n ¼ 31) consisted of healthy volunteers and patients scheduled for elective surgery at University Hospital Münster. All subjects continued their regular medication, including antihypertensive agents and statins. Demographic, clinical, and biochemical characteristics of patients and controls are summarized in Table 1. 2.2. Data collection and blood sampling Demographic data and details of medical history were collected at enrollment. Hypertension was defined as systolic blood pressure 140 mmHg and diastolic blood pressure 90 mmHg or current antihypertensive therapy. The clinical definition of diabetes included fasting glucose 126 mg/dl, intake of oral hypoglycaemic agents, or insulin use. Cardiovascular risk was estimated using the Framingham Risk Score. Estimated glomerular filtration rate (eGFR) was calculated using the chronic kidney disease epidemiology collaboration (CKD-EPI) equation, as previously described [21]. Blood samples from patients and control subjects were collected into ethylenediaminetetraacetic acid (EDTA) tubes, centrifuged, divided into aliquots and stored at 80 C. Serum creatinine, HbA1c, cholesterol and urinary protein were measured using standard laboratory techniques (Center for Laboratory Medicine, University Hospital Münster). In dialysis patients, blood samples were drawn pre-dialysis.
Table 1 Baseline characteristics of all participants.a Estimated GFR (ml/min/1.73 m2)e
Age (yr) Male gender [% (n)] Creatinine (mg/dl) eGFR (ml/min/1.73 m2)e Proteinuria (g/24 h) Cardiovascular risk factors Arterial hypertension [% (n)] SBP (mmHg) DBP (mmHg) BMI (kg/m2) Cholesterol (mg/dl) HDL-cholesterol (mg/dl) Calculated LDL-cholesterol (mg/dl) Diabetes [% (n)] HbA1c (%) Framingham risk score (%) Smokers [% (n)] Current medication ACE inhibitors [% (n)] AT1 blockers [% (n)] Erythropoietin [% (n)] Statins [% (n)] Markers of endothelial dysfunction Ang-2 (ng/ml) sFlt-1 (pg/ml) sVCAM-1 (pg/ml) vWF (U/ml) Circulating glycocalyx components Hyaluronan (ng/ml) Syndecan-1 (ng/ml)
P Value
60 (n ¼ 31)
30 to 59 (n ¼ 26)
15 to 29 (n ¼ 35)
<15 (n ¼ 34)
45 21 65 (20) 0.9 (0.7e1.0) 95 (79e110) 0.0 (0.0 - 0.0)
68 16 46 (12) 1.3 (1.1e1.5) 40 (36e48) 0.0 (0.0e0.7)
67 16 46 (16) 2.0 (1.8e2.3) 22 (18e25) 0.3 (0.2e1.5)
62 13 65 (22) 4.5 (3.3e5.5) 6 (5e10) 0.7 (0.2e1.3)
29 (9) 130 (125e135) 80 (80e80) 23.7 (22.1e28.0) 165 (134e211) 49 (45e56) 129 (109e166) 3 (1) 4.7 (4.5e5.0) 8 (1e20) 19 (6)
77 (20) 135 (129e150) 80 (75e80) 25.5 (23.1e29.7) 184 (134e210) 48 (39e56) 114 (97e156) 23 (6) 5.3 (5.0e5.7) 12 (4e17) 19 (5)
83 (29) 130 (120e140) 80 (70e80) 27.1 (23.4e30.9) 168 (142e198) 43 (36e50) 100 (68e127) 40 (14) 5.2 (4.8e6.5) 16 (5e22) 23 (8)
71 (24) 130 (125e140) 80 (78e80) 25.4 (22.5e27.5) 156 (130e199) 43 (39e48) 97 (67e116) 32 (11) 5.2 (5.0e5.7) 11 (6e16) 29 (10)
19 (6) 13 (4) 0 (0) 6 (2)
65 (17) 27 (7) 4 (1) 50 (13)
54 17 11 46
32 26 35 29
0.7 (0.6e1.2) 56 (44e79) 740 (544e874) 0.2 (0.1e0.5)
1.2 (0.9e1.8) 77 (61e107) 752 (660e1128) 0.3 (0.2e0.6)
1.6 (1.1e3.4) 113 (89e158) 817 (597e1458) 0.5 (0.2e0.9)
3.2 (1.7e5.7) 101 (83e149) 1679 (1028e2233) 0.7 (0.3e1.3)
<0.001c <0.001c <0.001c <0.001c
81 (48e171) 53 (41e92)
133 (62e224) 90 (61e344)
147 (86e230) 212 (100e355)
441 (163e1213) 270 (92e515)
<0.001c <0.001c
(19) (6) (4) (16)
(11) (9) (12) (10)
<0.001b 0.215d <0.001c <0.001c <0.001c 0.033d 0.457c 0.306c 0.416c 0.816c 0.031c 0.004c 0.044d 0.004c 0.113d 0.726d 0.034d 0.704d <0.001d 0.026d
ACE, angiotensin-converting enzyme; AT1, Angiotensin-II receptor; BMI, body mass index; DBP, diastolic blood pressure; HbA1c, glycohemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SBP, systolic blood pressure. a In case of a normal distribution, variables are presented as mean SD. In case of a skewed distribution, variables are presented as median (interquartile range). b Univariate ANOVA, post hoc test (Scheffé). c KruskaleWallis one-way ANOVA. d c2 test. e GFR values were calculated using CKD-EPI formula.
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
2.3. Quantification of circulating glycocalyx components As markers of eGC damage, we chose Syn-1 and HA as they probably reflect different aspects in glycocalyx composition (Syn-1 as core glycoprotein and hyaluronan as loosely attached substance). Human plasma syndecan-1 (Syn-1) levels were determined directly as reported previously (Gen-Probe Diaclone Research, Besancon, France [17,22,23]). Plasma hyaluronan (HA) measurements were performed using an enzyme-linked immunosorbent assay designed for detecting human hyaluronic acid (Echelon Biosciences Inc., Salt Lake City, UT, USA) as previously reported [19,23]. Preliminary experiments confirmed the stability of eGC components in plasma for 24 h at room temperature as well as their stability throughout five repeated freeze-thaw-cycles (data not shown). All measurements were performed in duplicates at the same day by the same investigators blinded to patients’ characteristics. Rat plasma Syn-1 levels were determined directly using a sandwich enzyme immunoassay with an antibody specific to rat Syn-1 (USCN Life Sciences, Wuhan, China) according to the manufacturer’s instructions. 2.4. Quantification of circulating markers of endothelial dysfunction Plasma levels of angiopoietin-2 (Ang-2) were measured by inhouse immuno-luminometric Assay (ILMA) methodology (R&D Systems, Minneapolis, MN, USA), as previously reported by our group in detail [24]. Plasma levels of human soluble fms-like tyrosine kinase-1 (sFlt-1, R&D Systems), soluble vascular cell adhesion molecule-1 (sVCAM-1, R&D Systems), and Von Willebrand factor-antigen (vWF, Technoclone, Vienna, Austria) were determined using commercially available ELISA kits according to the manufacturer’s instructions. Parts of the results of these measurements have been reported previously [25]. All measurements were performed in duplicates at the same day by the same investigators blinded to patients’ characteristics. 2.5. 5/6-Nephrectomy in rats All procedures were approved by a governmental committee on animal welfare (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen) and were performed according to international guidelines on animal experimentation. Male LewiseBrown Norway (LNB-F1) rats (200e270 g in body weight) were obtained from Janvier (Janvier, Le Genest Saint Isle, France). Rats had free access to standard chow and tap water, and were acclimated to the facility for at least a week before beginning an experiment. Male LBN-F1 rats underwent 5/6-nephrectomy as described before [26]. In brief, rats were anesthetized by isoflurane, ketamine 10% and xylazine 2% (both CEVA PLC, Duesseldorf, Germany). The right kidney was extirpated and 2 of 3 branches of the left renal artery were ligated. Fourteen days after the intervention, rats were anesthetized again and blood samples were drawn by left cardiac puncture. Aortas were removed for further fine preparation (see below). 2.6. Atomic force microscopy on rat aorta ex vivo Rat aortas were perfused with PBS (PAA Laboratories, Pasching, Austria) with 1% Penicillin and Streptomycin (Biochrom AG, Berlin, Germany) via cardiac puncture, isolated, immediately bathed in PBS with 1% Penicillin/Streptomycin (Pen/Strep) and freed from surrounding tissue (Fig. 1 A). Small patches (about 5 mm [2]) of the whole aorta were attached on Cell-TakÒ (BD Biosciences, Franklin Lakes, NJ, USA) coated glass, with the endothelial surface facing
337
upwards (Fig. 1 B). After preparation, the patches were bathed in minimal essential medium (MEM; Invitrogen Corp., La Jolla, CA, USA) supplemented with 20% fetal calf serum (FCS; PAA Clone, Coelbe, Germany), 1% MEM vitamins (Invitrogen), 1% MEM nonessential amino acids and 1% Pen/Strep (Biochrom AG, Berlin, Germany). Preservation of the endothelial cell layer on aorta preparations was approved by immunostaining of endothelial cell adhesion molecule PECAM-1/CD31 (Fig. 1 C) as reported elsewhere in detail [27]. To determine eGC’s thickness and stiffness we used the Atomic Force Microscope (AFM) nanoindentation technique. In Fig. 1 D, we illustrate the basic principles of this method. Using a Multimode AFM (Veeco, Mannheim, Germany) with a feedback-controlled heating device (Veeco), measurements were performed at 37 C as described previously [28]. In short, the central component of the AFM consists of a triangular cantilever with a mounted spherical tip (here: electrically uncharged polystyrene, diameter ¼ 10 mm, Novascan, Ames, IA, USA), which is a very sensitive mechanical nanosensor. This cantilever is used to periodically indent the cells and functions as a soft spring with a spring constant of 11 pN/nm. A laser beam is targeted at the gold-coated backside of the cantilever to be reflected to a position-sensitive quadrupled photodiode, which allows quantification of the cantilever deflection (V). To facilitate the calculation of the force (F) acting on the cantilever and, in turn, the force exerted by the cantilever to the sample, we previously determined the spring constant of the cantilever (Kcant) by the thermal tuning method and the deflection sensitivity (a) of the cantilever on bare glass coverslips.
F ¼ V$a$Kcant Knowing both the piezo displacement (xpiezo) and the deflection sensitivity (a), the indentation depth (deformation) of the sample (xsample) can consecutively be calculated.
xsample ¼ xpiezo ða$VÞ In the following, the indentation depth is hereafter called “thickness”. Of note, indentation depth represents the apparent thickness, not the anatomically exact thickness. Plotting the force (F) necessary to indent a single cell (indentation depth, xsample) leads to force indentation curves. The sample stiffness can be derived from Hook’s law.
Ksample ¼
F xsample
The stiffness (K) is the sample’s mechanical resistance against a defined deformation (e.g. indentation) and is strongly determined by the indentation depth and its location, due to the varying distribution of cellular organelles. The experimental parameters encompass an indentation velocity of 1 mm/s, a loading force of approximately 400 pN and an indentation frequency between 0.25 and 0.5 Hz. All measurements were performed in HEPES-buffered solution [standard composition 140 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 5 mM glucose, 10 mM HEPES (N-2-hydroxyethylpiperazineN0 -2-ethanesulfonic acid), pH 7.4] supplemented with 1% FCS in order to prevent eGC collapse [29].
2.7. Atomic force microscopy on rat aorta in vitro Additionally, indentation measurements on rat aorta preparations were performed after enzymatic eGC degradation in vitro by a mix of heparinase I (0.5 Sigma-U/ml), hyaluronidase (25.0 SeU/ml),
338
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Fig. 1. Atomic force microscopy. A) Preparation of rat aortas. B) Microscopic view of rat aortic endothelium. C) Immunostaining of endothelial cell adhesion molecule PECAM-1/ CD31. D) Original tracing of a forceedistance curve performed on the endothelial surface of a rat aorta ex vivo: 1) The AFM tip travels vertically towards the endothelial surface. 2) Upon indentation of the endothelial glycocalyx (eGC), the AFM cantilever, which serves as a soft spring, is deflected. The cantilever deflection is measured as a laser beam reflected from the back of the cantilever and is plotted as a function of sample position along the z-axis. The resulting curve is transformed into a force-versus-distance curve using the cantilever’s spring constant and the optical lever sensitivity. The slope of a forceedistance curve then directly reflects the stiffness (expressed in pN/nm), which is necessary to indent the eGC for a certain distance. The first slope indicates the stiffness (in this trace 0.31 pN/nm) of the very first layer, the (eGC). The second slope indicates the stiffness of the plasma membrane and the cortical actin web. The distance between the starting point of eGC indentation and the starting point of the second slope (projected to the x-axis) corresponds to the thickness of the eGC (in this trace 304.5 nm).
and chondroitinase (0.1 SeU/ml, all Sigma) for 60 min at 37 C. An equal amount of buffer (0.9% NaCl) served as control.
accepted at P < 0.05. Data analysis was performed using SPSS (IBMÒ, Armonk, NY, USA). Figures were prepared using GraphPad Prism (GraphPad Prism Software Inc., San Diego, CA, USA).
2.8. Statistical analysis 3. Results Data are presented as absolute numbers, percentages, means with corresponding standard deviations, or medians with corresponding 25th and 75th percentiles [inter-quartile range (IQR)]. Baseline characteristics of CKD patients and controls were compared using one-way ANOVA with post-hoc Scheffé’s test, KruskaleWallis test, and the c2-test as appropriate. The relationship between Syn-1, HA and clinical data as well as circulating markers of endothelial dysfunction was investigated using Pearson’s productemoment correlation. The association of circulating Syn-1 and HA with eGFR was further evaluated in adjusted linear regression models. Selection of variables to be included in the multivariable models was done a priori by determining probable confounders based both on differences in baseline characteristics in patients with different Syn-1 and HA levels and on theoretical considerations. All tests were two-sided and significance was
3.1. Circulating eGC plasma levels increase with progress of CKD The clinical characteristics of the study population are given in Table 1. Etiology of CKD was not different between CKD stages 3e5. As in previous studies, classical cardiovascular risk factors such as diabetes mellitus, levels of glycated hemoglobin or arterial hypertension increased significantly throughout the different stages of CKD (Table 1). Interestingly, omitting diabetic patients (n ¼ 32) from analysis did not substantially change the results displayed in Table 1 (Supplemental Table 1). Further, dividing the cohort into non-diabetic vs. diabetic patients revealed that diabetics had higher creatinine, proteinuria, prevalence of arterial hypertension, BMI, endothelial dysfunction as well as enhanced glycocalyx shedding (Supplemental Table 2).
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
339
Fig. 2. Levels of syndecan-1/hyaluronan increase throughout the different stages of CKD. Left: Box plots showing plasma levels of A) syndecan-1 (Syn-1) and C) hyaluronan (HA) measured by ELISA methodology in plasma samples from controls (n ¼ 31) and patients with different stages of CKD (n ¼ 95). The differences between patients and control subjects were assessed by one-way ANOVA with post-hoc Scheffé’s test (*P < 0.05 versus controls). Right: Scatter plots showing the relation (Spearman’s rank correlation coefficient) of estimated glomerular filtration rate (eGFR) to plasma levels of Syn-1 B) and HA D).
The median Syn-1 level was about 4-fold higher in CKD patients compared to controls (203 vs. 53 ng/mL; P < 0.001). The same was observed for HA (2.2-fold over controls, 179 vs. 81 ng/mL; P < 0.001). Next, we assessed circulating eGC components according to CKD stages. Median Syn-1 levels steadily increased across the following groups: controls: 53 (41e92) ng/mL, CKD 3: 90 (61e344) ng/mL, CKD 4: 212 (100e355) ng/mL, and CKD 5: 270 (92e515) ng/mL, respectively (non-parametric ANOVA P < 0.0001). Syn-1 levels were significantly elevated in patients with higher CKD stages (i.e. 4 and 5) compared to controls (P ¼ 0.047 and P ¼ 0.001, respectively, Fig. 2 A). Interestingly, HA showed a shallower increase but peaked distinctly in patients with CKD stage 5 (controls: 81 (48e171) ng/mL, CKD 3: 133 (62e224) ng/mL, CKD 4: 147 (86e 230) ng/mL, and CKD 5441 (163e1213) ng/mL, respectively; nonparametric ANOVA P < 0.0001). Median HA values in CKD 5 were significantly elevated compared to all other groups (P < 0.001, Fig. 2 C). Both parameters correlated significantly and inversely with the eGFR (Fig. 2 B/D). To identify associations between eGC damage and demographic and clinical variables, we stratified baseline characteristics of all participants according to tertiles of plasma Syn-1 (Table 2) and HA levels (Supplemental Table 3). Patients within the lowest Syn-1 tertile showed the highest eGFR (P < 0.001) and the lowest proteinuria (P < 0.001). Surprisingly, no significant differences were observed with regards to age, gender, medication, and major cardiovascular risk factors, except for diabetes (P ¼ 0.008). Basically, the same results were obtained for HA, with the exception that age
(P ¼ 0.005) and the use of erythropoietin (P ¼ 0.021) slightly increased across HA tertiles. To confirm the association between eGC shedding and renal function, we performed multivariate analysis using linear regression models and treating Syn-1 and HA as dependent variables. Table 3 shows that eGFR is independently associated with plasma Syn-1 and HA levels in all models, thus excluding significant bias from variables such as age, gender, arterial hypertension, smoking, diabetes and dyslipidaemia. 3.2. Circulating glycocalyx components correlate with endothelial dysfunction As initially hypothesized, we expected elevated soluble eGC components to be associated with endothelial dysfunction in CKD patients. We thus measured plasma levels of vWF, sVCAM-1, sFlt-1, and Ang-2 as established markers of endothelial activation and dysfunction [25,30e32]. As expected, markers of endothelial dysfunction steadily increased across different stages of CKD (Table 1). Similarly, circulating levels of Ang-2, sFlt-1, sVCAM-1 and vWF steadily increased across tertiles of plasma Syn-1 (all P 0.001, Table 2) and plasma HA (P ¼ 0.004, Supplemental Table 3). Bivariate correlations using Spearman’s test revealed moderate-to-strong correlations of both Syn-1 and HA with sFlt-1 (Syn-1: r ¼ 0.582; HA: r ¼ 0.446), sVCAM-1 (Syn-1: r ¼ 0.527; HA: r ¼ 0.709), vWF (Syn-1: r ¼ 0.373; HA: r ¼ 0.445), and Ang-2 (Syn-1: r ¼ 0.455; HA: r ¼ 0.523) with P < 0.001 for all calculations.
340
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Table 2 Baseline characteristics according to plasma Syn-1 levels.a Tertiles of syndecan-1 (ng/ml)
Age (yr) Male gender [% (n)] Creatinine (mg/dl) eGFR (ml/min/1.73 m2)d Proteinuria (g/24 h) Cardiovascular risk factors Arterial hypertension [% (n)] SBP (mmHg) DBP (mmHg) BMI (kg/m2) Cholesterol (mg/dl) HDL-cholesterol (mg/dl) Calculated LDL-cholesterol (mg/dl) Diabetes [% (n)] HbA1c (%) Framingham risk score (%) Smokers [% (n)] Current medication ACE inhibitors [% (n)] AT1 blockers [% (n)] Erythropoietin [% (n)] Statins [% (n)] Markers of endothelial dysfunction Ang-2 (ng/ml) sFlt-1 (pg/ml) sVCAM-1 (pg/ml) vWF (U/ml) Circulating glycocalyx components Hyaluronan (ng/ml) a b c d
P value
<78 (n ¼ 41)
79 to 241 (n ¼ 41)
>241 (n ¼ 43)
54 22 44 (18) 1.3 (0.9e3.1) 48 (19e84) 0.0 (0.0e0.8)
62 19 66 (27) 2.0 (1.3e2.9) 24 (15e41) 0.3 (0.1e0.7)
65 14 53 (23) 2.5 (1.8e5.1) 16 (5e27) 1.1 (0.4e1.3)
51 (21) 130 (125e148) 80 (80e80) 26.5 (23.3e29.0) 166 (134e186) 48 (43e54) 110 (97e138) 10 (4) 5.0 (4.7e5.3) 8 (1e17) 15 (6)
73 (30) 130 (121e149) 80 (76e80) 24.6 (21.2e29.0) 167 (143e201) 42 (38e56) 111 (89e132) 22 (9) 5.1 (4.8e5.7) 12 (7e20) 39 (16)
70 (30) 130 (128e140) 80 (70e80) 24.9 (23.4e30.0) 162 (130e204) 43 (38e48) 100 (70e119) 42 (18) 5.5 (5.0e6.6) 12 (4e20) 14 (6)
0.339c 0.827b 0.218b 0.309b 0.802b 0.128b 0.479b 0.008c 0.008b 0.084b 0.009c
27 (11) 29 (12) 7 (3) 22 (9)
54 20 12 29
42 14 21 42
0.144c 0.087c 0.285c 0.246c
1.1 (0.6e1.6) 77 (56e87) 736 (586e961) 0.2 (0.2e0.7)
1.4 (0.8e3.4) 88 (70e122) 789 (584e1039) 0.4 (0.3e0.7)
3.0 (1.4e6.3) 119 (99e411) 1577 (1192e2247) 0.5 (0.3e1.2)
<0.001b <0.001b <0.001b 0.004b
111 (50e184)
103 (75e247)
293 (158e766)
<0.001b
(22) (8) (5) (12)
(18) (6) (9) (18)
0.097b 0.168c <0.001b <0.001b <0.001b
Variables are presented as median (interquartile range) or as mean SD. KruskaleWallis one-way ANOVA. c2-test. GFR values were calculated using CKD-EPI formula.
3.3. 5/6-Nephrectomy leads to impaired nanomechanical properties of the eGC To confirm that elevated plasma levels of Syn-1 and HA reflect structural and/or functional alterations of the eGC, we used an established rat model of CKD, namely 5/6-nephrectomy. Physiologic data assessed 14 days after surgery is given in Supplemental Table 4. Nano-indentation measurements by AFM revealed a highly significant decrease of eGC thickness (360 vs. 157 nm, P ¼ 0.001) and stiffness (0.33 vs. 0.22 pN/nm, P < 0.001) in aortic explants from 5/6nephrectomized rats (n ¼ 4) compared to sham-operated animals (n ¼ 6) (Fig. 3 A, B). Additionally, we performed control measurements by incubating rat aortic endothelium in vitro with control buffer or an enzyme mix (consisting of heparinase, chondroitinase and hyaluronidase), respectively. The latter was used to strip all three
highly abundant components off the eGC, namely heparan sulfate, HA and chondroitin sulfate. As shown in Fig. 3 A and B, the decrease in glycocalyx thickness and stiffness after enzyme treatment was comparable to the alterations seen in 5/6-nephrectomized rats. As in CKD patients, Syn-1 plasma levels were elevated in 5/6nephrectomized rats (n ¼ 4) compared with sham-operated animals (n ¼ 6) (P < 0.001, Fig. 3 C). As in humans, plasma Syn-1 levels tightly correlated with plasma sFlt-1 levels (r ¼ 0.901, P ¼ 0.0008) and the creatinine-based approximation of GFR (r ¼ 0.783, P ¼ 0.01) in CKD rats. 4. Discussion Whilst special electron microscopic staining procedures uncovered that the luminal surface of the endothelium expressed a
Table 3 Adjusted linear regression model treating Syn-1/HA as dependent variables. Syndecan-1
Unadjusted eGFR Adjusted for model 1 eGFR Adjusted for model 2 eGFR Adjusted for model 3 eGFR
Hyaluronan
Stand. b
(95% CI)
P
Stand. b
(95% CI)
P
0.388
(6.07 to 1.855)
<0.001
0.382
(6.985 to 2.616)
<0.001
0.388
(6.07 to 1.855)
<0.001
0.382
(6.985 to 2.616)
<0.001
0.286
(5.156 to 1.125)
0.003
0.431
(7.326 to 3.077)
<0.001
0.294
(5.277 to 1.187)
0.002
0.416
(7.184 to 2.868)
<0.001
CI, confidence interval. Model 1: adjusted for age and gender. Model 2: adjusted for age, gender, arterial hypertension, diabetes, and smoking. Model 3: adjusted for age, gender, arterial hypertension, diabetes, smoking, HbA1c, SBP, DBP, body mass index, and Framingham risk score.
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Fig. 3. Endothelial glycocalyx in 5/6-nephrectomized rats. Bar charts showing thickness (A) and stiffness (B) of endothelial glycocalyx (eGC) as analyzed by atomic force microscopy (AFM) on living endothelium from rat aortic explants (ex vivo) harvested at day 14 after sham operation (n ¼ 6) or 5/6-nephrectomy (5/6Nx; n ¼ 4). From each aorta 2 aortic patches were prepared. On each patch 5 force indentation curves were recorded in each of 3 different tip positions, respectively. For statistical comparison, a single average value per tip position, aortic patch and aorta was calculated, respectively. For comparison, aortic explants from untreated rats were treated in vitro with either control buffer (n ¼ 3) or an enzyme mix (n ¼ 3) for 60 min to strip heparan sulfate, hyaluronan and chondroitin sulfate off the eGC. In addition C) Syn-1 plasma levels in sham-operated and 5/6-nephrectomized rats at day 14 after the procedure were measured by ELISA. *P < 0.001 versus sham-operated controls or buffer-treated controls, respectively.
carbohydrate layer by the mid-1940s, it was considered to be without major physiological relevance or functional significance for a long time. In the past two decades, however, intravital microscopy studies revealed that the eGC represents a substantial intravascular compartment contributing significantly to vascular wall homeostasis [9,33]. Since then, different groups have conclusively shown that the endothelial glycocalyx is seriously damaged in patients with diabetes [34,35], hypercholesterolemia [13,36], acute inflammation [13,36], or after ischemia reperfusion [22,37]. Only recently, Vlahu et al. found both, thinning and shedding of the glycocalyx in patients on maintenance dialysis [38]. However, the question whether impaired renal function per se is an independent risk
341
factor for glycocalyx impairment has not been formally addressed so far. This is particularly remarkable considering the fact, that CKD is an increasingly recognized contributor to cardiovascular morbidity and mortality. In the present study, we found a robust inverse association between reduced renal function and presumable shedding of the atheroprotective glycocalyx in a well-defined cohort of patients comprising different CKD stages. This association was hardly influenced by established triggers of glycocalyx damage such as dyslipidaemia and diabetesethe latter being the only other factor associated with increased glycocalyx shedding in univariate analysis (see also Supplemental Table 2). Interestingly, Syn-1 plasma levels were already increased in CKD 4, while the increase in HA became only evident in CKD 5. Given its unique and delicate composition, HA shedding may thus be indicative of more severe glycocalyx damage. It is thus conceivable that CKD is a new and potentially relevant risk factor for glycocalyx damage. Little is known about the half-life of circulating glycocalyx components. Early studies have shown that circulating HA (plasma half-life < 5 min) is rapidly eliminated by the liver, while renal excretion plays a negligible part in clearance [39,40]. Because the elimination of Syn-1 from the circulation has not been reported, we excluded renal accumulation of Syn-1 by comparing 24 h clearances of Syn-1 and creatinine in a subgroup of 21 patients (data not shown). Together, these data indicate the existence of a steady state between active refurbishment and on-going damage of the endothelial glycocalyx in CKD. Glycocalyx research has traditionally focused on thickness as a surrogate of structural integrity. Although highly intuitive in principle, this concept provides no conclusions concerning the mechanical properties of this highly elute and deformable layer. We have recently shown that AFM can reliably depict nanomechanics of the eGC on a variety of living endothelial cells by a linear approximation over the first hundreds of nanometres of the force indentation curves that represent the eGC [27]. Although heterogeneity and limitations of in vivo visualization methods probably impede any direct comparison of glycocalyx dimensions, we noted that the average nanomechanical thickness of the glycocalyx on ex vivo aortic preparations from sham-operated rats (360 nm) was compatible with many of the aforementioned reports [41e43]. Interestingly, here we could show by AFM that CKD induction (5/6 nephrectomy) in rats caused considerable softening and thinning (>50% from baseline) of the aortic glycocalyx, which was paralleled by Syn-1. In fact, nanomechanical changes after 5/6 nephrectomy were quite comparable to those seen after either enzymatic removal of its three main constituents e heparan sulfate, chondroitin sulfate, and hyaluronic acid e or during gram-negative endotoxemia [27]. Although findings from this animal study cannot be readily extrapolated to humans, it is conceivable to assume that elevated Syn-1/HA levels in CKD indeed reflect structural damage of the glycocalyx. Salmon and co-workers recently provided another indication for the importance of eGC in the context of chronic kidney disease and endothelial dysfunction [44]. In a rat model of spontaneous albuminuric CKD, they could conclusively show that loss of eGC is paralleled by defects in microvascular permeability. Specific mechanisms of eGC shedding in CKD remain poorly understood. It is now accepted that TNF-alpha [13,15] and hyperglycaemia [19] cause glycocalyx disruption probably via the action of proteases. In accordance with this notion, Vlahu et al. found elevated hyaluronidase activity in plasma from dialysis patients [38]. However, from a nephrologist’s perspective, two recently identified mediators of glycocalyx deterioration, namely hypervolemia [20] and sodium overload [28] might be of particular importance in CKD patients. Patients with impaired kidney function gradually lose their ability to excrete salt and water.
342
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343
Particularly in patients on maintenance dialysis (which had the highest Syn-1 and HA values in our cohort e Supplemental Table 5) the prevalence of volume overload has been estimated as high as 63% [45]. It is currently believed that hypervolemia is negatively affecting the glycocalyx via atrial natriuretic peptide (ANP), which is released after wall stress in the right atrium of the heart of humans [18]. However, the exact intracellular signal cascade leading to degradation of the glycocalyx by ANP is still unknown. Furthermore, Oberleithner and co-workers revealed that the eGC plays an important role as a significant buffer for sodium [7,28]. It is known, that the eGC preferentially binds sodium due to its overall net negative charge. They estimated that 700 mg of sodium, which is proportionate to the sodium amount of a single meal, can be transiently bound to the eGC within the human body to avoid sodium overload of the endothelium [7]. However, the increase of extracellular sodium concentration led to perturbation of sodium buffering capacity, as well as thinning and loss of negatively charged HS side chains in vitro. [28] Glycocalyx impairment in CKD might thus facilitate sodium entry into the endothelial cells. This could explain endothelial dysfunction and arterial hypertension observed in patients with impaired kidney function [46]. Consistent with this hypothesis, we found medium to strong correlations between Syn-1/HA plasma levels and several markers of endothelial dysfunction/activation. Intriguingly, eGC damage did not seem to correlate with the Framingham Risk Score. This goes along with the preliminary observation that only HA (299 vs. 387 ng/ml, P ¼ 0.021) but not Syn-1 (288 vs. 328 ng/ml, P ¼ 0.26) was significantly higher in patients with history of cardiovascular disease (i.e. “peripheral arterial disease” or “stroke” or “acute coronary syndrome” e data not shown)). However, interpretation of this finding is difficult given that (subclinical) cardiovascular disease was not thoroughly assessed and thus may be underdiagnosed in our patients. Some limitations regarding our study need to be addressed. First, the number of participants (n ¼ 126) might be too small. However, besides applying strict inclusion criteria to minimalize possible confounding effects and increase the validity of the obtained results, an animal model of CKD was used to avoid these confounders and confirm our results. Second, median Syn-1 plasma levels of control subjects were 56 ng/ml (5.6 mg/dl) and thus w2e4 fold higher than plasma Syn-1 levels from controls from the literature, which range between 12 ng/dl and 20.5 ng/dl in different studies using the same ELISA kit [47,48]. Median HA levels from controls subjects (81 ng/ml in our cohort) were comparable to controls from two previously published studies by Nieuwdorp and coworkers (HA 65 and 70 ng/ml) using the same ELISA kit [19,35]. A possible explanation for the slightly higher Syn-1 levels seen in our controls is that these were a mix of diseases controls and healthy volunteers (see M&M section) but not solely healthy volunteers as in the aforementioned studies. Third, inflammatory processes also play a part in eGC degradation. Though we did not systematically analyze a possible influence of (micro-) inflammation on eGC damage, a review of chart records and lab data revealed an only moderate correlation between Syn1 and CRP (r ¼ 0.401, P ¼ 0.013), whereas HA and CRP did not correlate significantly (r ¼ 0.201, P ¼ 0.22). Further studies focusing on this aspect would be highly desirable. Fourth, according to the cross-sectional study design, outcome data were not obtainable. Finally, due to the fact that the eGC is a well compressible structure, it is important to mention that the AFM approach of this study gives an indication of the apparent thickness, rather than the true anatomical thickness of the eGC. 5. Conclusion With our study, we support and extend findings by Vink’s group and show first evidence of shedding of the atheroprotective eGC in
(non-) dialysis CKD patients. Increased shedding correlates tightly with endothelial dysfunction and is independently associated with renal function impairment. This novel finding opens new perspectives on the pathophysiology of cardiovascular disease in CKD. Still, further studies are needed to fully understand the mechanisms behind eGC alterations in those patients. These, however, could potentially uncover how to protect the eGC and eventually reduce cardiovascular morbidity and mortality in CKD. Competing interests The authors declare that they have no competing interests. Acknowledgments This work was supported by the fund “Innovative Medical Research” of the University of Münster Medical School (KÜ111015). The study was partly supported by the Deutsche Forschungsgemeinschaft (CRC 656, project C7). Support by the Cluster of Excellence Cell in Motion (EXC 1003 e CiM) is gratefully acknowledged. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atherosclerosis.2014.03.016. References [1] Di Angelantonio E, Chowdhury R, Sarwar N, et al. Chronic kidney disease and risk of major cardiovascular disease and non-vascular mortality: prospective population based cohort study. BMJ 2010;341:c4986. [2] Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998;32:S112e9. [3] Schiffrin EL, Lipman ML, Mann JF. Chronic kidney disease: effects on the cardiovascular system. Circulation 2007;116:85e97. [4] Nieuwdorp M, Meuwese MC, Vink H, et al. The endothelial glycocalyx: a potential barrier between health and vascular disease. Curr Opin Lipidol 2005;16:507e11. [5] Pries AR, Secomb TW, Gaehtgens P. The endothelial surface layer. Pflugers Arch 2000;440:653e66. [6] Reitsma S, Slaaf DW, Vink H, et al. The endothelial glycocalyx: composition, functions, and visualization. Pflugers Arch 2007;454:345e59. [7] Oberleithner H. Two barriers for sodium in vascular endothelium? Ann Med 2012;44(Suppl. 1):S143e8. [8] Mulivor AW, Lipowsky HH. Role of glycocalyx in leukocyte-endothelial cell adhesion. Am J Physiol Heart Circ Physiol 2002;283:H1282e91. [9] Curry FE, Adamson RH. Endothelial glycocalyx: permeability barrier and mechanosensor. Ann Biomed Eng 2012;40:828e39. [10] Constantinescu A, Spaan JA, Arkenbout EK, et al. Degradation of the endothelial glycocalyx is associated with chylomicron leakage in mouse cremaster muscle microcirculation. Thromb Haemost 2011;105:790e801. [11] Vink H, Constantinescu AA, Spaan JA. Oxidized lipoproteins degrade the endothelial surface layer: implications for platelet-endothelial cell adhesion. Circulation 2000;101:1500e2. [12] Lipowsky HH. Microvascular rheology and hemodynamics. Microcirculation 2005;12:5e15. [13] Nieuwdorp M, Meuwese MC, Mooij HL, et al. Tumor necrosis factor-alpha inhibition protects against endotoxin-induced endothelial glycocalyx perturbation. Atherosclerosis 2009;202:296e303. [14] Henry CB, Duling BR. TNF-alpha increases entry of macromolecules into luminal endothelial cell glycocalyx. Am J Physiol Heart Circ Physiol 2000;279: H2815e23. [15] Chappell D, Hofmann-Kiefer K, Jacob M, et al. TNF-alpha induced shedding of the endothelial glycocalyx is prevented by hydrocortisone and antithrombin. Basic Res Cardiol 2009;104:78e89. [16] Constantinescu AA, Vink H, Spaan JA. Elevated capillary tube hematocrit reflects degradation of endothelial cell glycocalyx by oxidized LDL. Am J Physiol Heart Circ Physiol 2001;280:H1051e7. [17] Bruegger D, Jacob M, Rehm M, et al. Atrial natriuretic peptide induces shedding of endothelial glycocalyx in coronary vascular bed of guinea pig hearts. Am J Physiol Heart Circ Physiol 2005;289:H1993e9.
J.-S. Padberg et al. / Atherosclerosis 234 (2014) 335e343 [18] Jacob M, Saller T, Chappell D, et al. Physiological levels of A-, B- and C-type natriuretic peptide shed the endothelial glycocalyx and enhance vascular permeability. Basic Res Cardiol 2013;108:347. [19] Nieuwdorp M, van Haeften TW, Gouverneur MC, et al. Loss of endothelial glycocalyx during acute hyperglycemia coincides with endothelial dysfunction and coagulation activation in vivo. Diabetes 2006;55:480e6. [20] Rehm M, Haller M, Orth V, et al. Changes in blood volume and hematocrit during acute preoperative volume loading with 5% albumin or 6% hetastarch solutions in patients before radical hysterectomy. Anesthesiology 2001;95: 849e56. [21] Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604e12. [22] Rehm M, Bruegger D, Christ F, et al. Shedding of the endothelial glycocalyx in patients undergoing major vascular surgery with global and regional ischemia. Circulation 2007;116:1896e906. [23] Chappell D, Jacob M, Hofmann-Kiefer K, et al. Hydrocortisone preserves the vascular barrier by protecting the endothelial glycocalyx. Anesthesiology 2007;107:776e84. [24] Lukasz A, Hellpap J, Horn R, et al. Circulating angiopoietin-1 and angiopoietin2 in critically ill patients: development and clinical application of two new immunoassays. Crit Care 2008;12:R94. [25] Di Marco GS, Reuter S, Hillebrand U, et al. The soluble VEGF receptor sFlt1 contributes to endothelial dysfunction in CKD. J Am Soc Nephrol 2009;20: 2235e45. [26] Reuter S, Bangen P, Edemir B, et al. The HSP72 stress response of monocytes from patients on haemodialysis is impaired. Nephrol Dial Transpl 2009;24: 2838e46. [27] Wiesinger A, Peters W, Chappell D, et al. Nanomechanics of the endothelial glycocalyx in experimental sepsis. PloS One 2013;8:e80905. [28] Oberleithner H, Peters W, Kusche-Vihrog K, et al. Salt overload damages the glycocalyx sodium barrier of vascular endothelium. Pflugers Arch 2011;462: 519e28. [29] Thi MM, Tarbell JM, Weinbaum S, et al. The role of the glycocalyx in reorganization of the actin cytoskeleton under fluid shear stress: a “bumper-car” model. Proc Natl Acad Sci U S A 2004;101:16483e8. [30] David S, Kumpers P, Hellpap J, et al. Angiopoietin 2 and cardiovascular disease in dialysis and kidney transplantation. Am J Kidney Dis 2009;53: 770e8. [31] David S, John SG, Jefferies HJ, et al. Angiopoietin-2 levels predict mortality in CKD patients. Nephrol Dial Transplant 2012;27:1867e72. [32] Pequeriaux NC, Fijnheer R, Gemen EF, et al. Plasma concentration of von Willebrand factor predicts mortality in patients on chronic renal replacement therapy. Nephrol Dial Transpl 2012;27:2452e7.
343
[33] Chappell D, Westphal M, Jacob M. The impact of the glycocalyx on microcirculatory oxygen distribution in critical illness. Curr Opin Anaesthesiol 2009;22:155e62. [34] Broekhuizen LN, Lemkes BA, Mooij HL, et al. Effect of sulodexide on endothelial glycocalyx and vascular permeability in patients with type 2 diabetes mellitus. Diabetologia 2010;53:2646e55. [35] Nieuwdorp M, Mooij HL, Kroon J, et al. Endothelial glycocalyx damage coincides with microalbuminuria in type 1 diabetes. Diabetes 2006;55:1127e32. [36] Meuwese MC, Mooij HL, Nieuwdorp M, et al. Partial recovery of the endothelial glycocalyx upon rosuvastatin therapy in patients with heterozygous familial hypercholesterolemia. J Lipid Res 2009;50:148e53. [37] Bruegger D, Schwartz L, Chappell D, et al. Release of atrial natriuretic peptide precedes shedding of the endothelial glycocalyx equally in patients undergoing on- and off-pump coronary artery bypass surgery. Basic Res Cardiol 2011;106(6):1111e21. [38] Vlahu CA, Lemkes BA, Struijk DG, et al. Damage of the endothelial glycocalyx in dialysis patients. J Am Soc Nephrol 2012;23(11):1900e8. [39] Fraser JR, Laurent TC, Pertoft H, et al. Plasma clearance, tissue distribution and metabolism of hyaluronic acid injected intravenously in the rabbit. Biochem J 1981;200:415e24. [40] Fraser JR, Laurent TC, Engstrom-Laurent A, et al. Elimination of hyaluronic acid from the blood stream in the human. Clin Exp Pharmacol Physiol 1984;11: 17e25. [41] Gao L, Lipowsky HH. Composition of the endothelial glycocalyx and its relation to its thickness and diffusion of small solutes. Microvasc Res 2010;80: 394e401. [42] van den Berg BM, Vink H, Spaan JA. The endothelial glycocalyx protects against myocardial edema. Circ Res 2003;92:592e4. [43] Potter DR, Damiano ER. The hydrodynamically relevant endothelial cell glycocalyx observed in vivo is absent in vitro. Circ Res 2008;102:770e6. [44] Salmon AH, Ferguson JK, Burford JL, et al. Loss of the endothelial glycocalyx links albuminuria and vascular dysfunction. J Am Soc Nephrol JASN 2012;23: 1339e50. [45] Ebah LM, Wiig H, Dawidowska I, et al. Subcutaneous interstitial pressure and volume characteristics in renal impairment associated with edema. Kidney Int 2013;84:980e8. [46] Kusche-Vihrog K, Oberleithner H. An emerging concept of vascular salt sensitivity. F1000 Biol Rep 2012;4:20. [47] Steppan J, Hofer S, Funke B, et al. Sepsis and major abdominal surgery lead to flaking of the endothelial glycocalix. J Surg Res 2011;165:136e41. [48] Anttonen A, Leppa S, Ruotsalainen T, et al. Pretreatment serum syndecan-1 levels and outcome in small cell lung cancer patients treated with platinum-based chemotherapy. Lung Cancer 2003;41:171e7.