Cardiovascular (CV) disease is the leading cause of morbidity and mortality in patients with chronic kidney disease (CKD) (Baigent et al., 2000; Matsushita et al., 2010; Mahmoodi et al., 2012; Cozzolino et al., 2017). Patients with CKD present with traditional CV risk factors, including hypertension, diabetes or dyslipidemia (Herzog et al., 2011; Gansevoort et al., 2013), but also with non-traditional CV risk factors (Go et al., 2004; El Nahas and Bello, 2005; Fox et al., 2012; Mahmoodi et al., 2012) of which endothelial dysfunction (ED) is established as a prominent one (Schächinger et al., 2000; Bonetti et al., 2003; Nagy et al., 2010).

A normally functioning endothelium modulates vascular permeability (Henry and Duling, 1999), controls vascular tone, and maintains blood flow [reviewed by Sandoo et al. (2015)]. Many of these endothelial functions are orchestrated by a properly functioning endothelial glycocalyx (eGC), which covers the luminal side of vascular endothelium. The eGC is a carbohydrate-rich mesh of glycoproteins, proteoglycans and glycosaminoglycans that interact with absorbed plasma proteins (Weinbaum et al., 2007). The highly sulfated and negatively charged properties of the eGC protect the endothelium and control cell migration (Constantinescu et al., 2003), and leakage of albumin (Ueda et al., 2004) and fluid (Van den Berg et al., 2003). Several studies indicated that damaged eGC in renal disease is associated with ED (Singh et al., 2007; Salmon et al., 2012; Vlahu et al., 2012; Padberg et al., 2014). For example, Vlahu et al. (2012) reported a diminished eGC thickness in hemodialysis patients by sidestream dark film imaging that was associated with elevated plasma markers of ED. For the development of ED, several interacting components seem to be involved. Next to oxidative stress and uremic toxins, the destructive effects of inflammation and volume overload on the vascular endothelium are increasingly receiving attention (Mitsides et al., 2018).

Patients with CKD often have elevated plasma markers of inflammation (Westphalen et al., 2020) which can have multiple underlying causes, including higher rates of various infections (De Jager et al., 2009). Additionally, accumulation of uremic toxins in patients with CKD promotes oxidative activity of leukocytes, which in turn leads to increased leukocyte-endothelial interactions, facilitating higher infiltration of macrophages and monocytes (Cozzolino et al., 2017; Liew et al., 2021). Importantly, the kidneys are a major source of antioxidants against reactive oxygen species causing diminished production of antioxidants to automatically translate in increased oxidative stress in CKD (Cozzolino et al., 2017). Next to directly inducing ED (Annuk et al., 2005), free radicals have been shown to contribute to inflammation (Kröller-Schön et al., 2014).

Overhydration, commonly seen in patients with end stage renal failure (Wizemann and Schilling, 1995; Wizemann et al., 1995), has been reported to associate with ED and microinflammation (Mitsides et al., 2017) contributing to arterial stiffness (Akdam et al., 2014), atherosclerosis, left ventricular hypertrophy (Bock and Gottlieb, 2010) and more CV events (Hung et al., 2014; Tsai et al., 2015). Interestingly, the interpretation of this association differs between authors with some concluding that ED leads to volume overload due to its role in water homeostasis (Levick and Michel, 2010; MacHnik et al., 2010; Paar et al., 2014; Dekker et al., 2017) whereas Colombo et al. posed that volume overload contributes to ED due to vascular stretch (Colombo et al., 2009, 2014). In fact, the damaging effect of overhydration on the endothelium may be due to increased shear stress as has been suggested by studies in patients with hypertension (Panza et al., 1990; Muiesan et al., 2001). Notably, endothelial cells can sense shear stress, hydrostatic pressure and stretch of the vessel wall through mechanosensors which can then be transferred into signaling pathways. Subsequently, gene and protein expression can be modified and endothelial function can deteriorate (Mammoto et al., 2012). The eGC functions as a mechanosensor of shear stress but at the same time pathologic shear stress might directly lead to eGC changes (Weinbaum et al., 2007) such as breakdown and shedding of eGC components (Zeng and Tarbell, 2014).

Clinical studies previously used indirect methods to demonstrate ED in CKD and its association with inflammation and volume overload (i.e., association analyses and imaging of the perfused boundary region) leaving open questions about possible targets for patient intervention. With the current work we aimed to investigate possible eGC loss and ED in skin biopsies of CKD patients by direct tissue staining and its association with inflammation and volume overload.

This study is the first to show direct evidence of dermal glycocalyx loss and ED in patients with CKD. In line with previous research, our results show that ED associates with volume overload rather than inflammation, emphasizing the importance of strict volume control in patients with CKD5. In fact, ED was associated with severe loss of the eGC marker UEA1 in endothelial cells of the dermal microcirculation of patients with CKD stage 5. Moreover, we found increased dermal endothelial expression levels of VEGFR2 in hemodialysis patients. A clear association between ED and volume overload was evidenced by elevated plasma levels of NT-proBNP and reduced plasma levels of sodium in CKD5 patients that significantly correlated with the tissue endothelial markers UEA1 and VEGFR2; and tissue vWF associating with plasma levels of NT-proBNP. We hardly found associations between plasma markers of inflammation and markers of ED in dermal biopsies. Together, these results suggest that dermal ED in CKD patients is associated with overhydration rather than with inflammation.

Chronic kidney disease stage 5 patients in our cohort clearly showed ED as evidenced by decreased UEA1 expression, and increased tissue VEGFR2 and vWF compared to healthy subjects. In this study, especially the breakdown of the eGC in CKD5 patients was striking as indicated by severe loss of the UEA1 glyco-epitope. Breakdown of eGC was significantly present in both CKD patient groups, but showed a tendency to be higher in patients with CKD5D than in those with CKD5ND. Breakdown of the eGC in CKD patients has been shown before by several indirect methods. Vlahu et al. (2012) used sidestream darkfield imaging to visualize the sublingual microcirculation. Other study groups showed similar results using the same technique in CKD patients (Dane et al., 2014; Mitsides et al., 2018). Several more publications reported glycocalyx components such as hyaluronan and syndecan-1 in plasma to be associated with eGC breakdown in CKD patients (Padberg et al., 2014; Liew et al., 2021). Moreover, Padberg et al. (2014) used atomic force microscopy to directly measure eGC in a CKD rat model. Here, they found decreased eGC thickness and stiffness which correlated with plasma biomarkers of ED such as vWF (Padberg et al., 2014).

In this study, we found strong associations between ED and volume overload in CKD patients. Increased plasma levels of NT-proBNP were present in all CKD patients as well as reduced sodium values, most outspoken in patients with CKD5D. NT-proBNP significantly associated with the tissue endothelial markers UEA1, VEGFR2, and vWF whereas plasma sodium associated with UEA1 and VEGFR2. The association between ED and volume overload has been shown before by several other groups that reported breakdown of the endothelial glycocalyx (Chappell et al., 2014; Mitsides et al., 2017) and elevated plasma levels of the endothelial markers vascular cell adhesion protein 1 (VCAM-1) and thrombomodulin (Mitsides et al., 2017). Indeed, Chappell et al. observed that inducing acute hypervolemia by infusing 20 mL/kg body weight of an iso-oncotic hydroxyethyl starch (HES) colloid solution (6% HES 130/0.4, Volulyte) led to shedding of the endothelial glycocalyx as evidenced by a significant increase in plasma levels of atrial natriuretic peptide and a rise in serum hyaluronan and serum and urine syndecan-1 (Chappell et al., 2014). In our study, we did not find differences between CKD patients and healthy controls with regard to dermal fluid content. Thus, we can only assume that volume overload was predominantly intravascular. We speculate that this could have led to vascular stretch and, subsequently, to ED. Previous research showed endothelial cell activation after vascular stretch from acute volume loading (central venous pressure ≥20 mmHg). After that, the vascular endothelium of healthy dogs showed increased mRNA levels of cyclooxygenase 2 and induced nitric oxide synthase which resulted from activation of the inflammatory/oxidative and hemostatic programs, similar to dogs suffering from heart failure (Colombo et al., 2009). More studies indicated ED and subsequent atherosclerosis after excessive vascular stretch (Hatami et al., 2013) as it is seen in hypertension, heart and kidney failure. Here, vascular remodeling took place after pathologic stretch with subsequent cell apoptosis, inflammation and the production of reactive oxygen species (Okada et al., 1998; Lemarié et al., 2010) and [reviewed by Jufri et al. (2015)]. In addition, cell differentiation from endothelial cell to smooth muscle cell has been observed after application of vascular stretch (Cevallos et al., 2006). In our cohort, endothelial VEGFR2 expression was significantly upregulated in dialysis patients. It has been demonstrated that VEGFR2 can get activated upon binding with its ligands VEGFA, VEGFC or VEGFD leading to endothelial cell proliferation, angiogenesis, cell survival but also to vascular permeability [reviewed by Lohela et al. (2009)]. Our study could neither show signs of angiogenesis nor of dermal edema. Nevertheless, higher values of VEGFR2 were only seen in CKD5D patients and a direct association between NT-proBNP and VEGFR2 was observed. It has been shown that stretching of the vasculature might directly induce VEGFR2 expression which can be interpreted as an endothelial stress response (Smith et al., 2001; Tian et al., 2016).

According to our data, dialysis patients showing eGC damage had significantly increased CRP values. In a previous publication on this cohort, we showed increased inflammation in dermal biopsies of CKD5D patients with increased levels of tissue macrophages and T-cells. Besides, dermal endothelial stainings of the chemokine C-C Motif Chemokine Ligand 2 (CCL2) and hyaluronan were performed. Moreover, we performed quantitative RT-PCR on CCL2, VCAM-1, and three hyaluronan enzymes. These data showed significant loss of hyaluronan staining, which is in full agreement with our data on UEA1 in the current report. Moreover, we showed a significant inverse association of quantitative mRNA expression of hyaluronan synthase 2 with the quantitative mRNA expression of CCL2. These data suggest that dermal inflammation in CKD5 patients is quantitatively associated with loss of hyaluronan (Hijmans et al., 2019). In the present study we found an association between plasma lymphocytes and tissue UEA1 suggesting an association between glycocalyx breakdown and lymphocytopenia, thus with inflammation which is in accordance with published research (Moseley et al., 1997; Vigetti et al., 2010; Oberleithner et al., 2011; Vlahu et al., 2012). In the (above-mentioned) study of Vlahu et al. (2012), the eGC breakdown was indeed related to inflammation in dialysis patients with elevated CRP values (>10 mg/L) having significantly increased perfused boundary regions. Besides, Kusche-Vihrog et al. (2011) have directly measured eGC stiffness of human endothelial cells (Eahy.924) by atomic force microscopy and demonstrated an increase in eGC stiffness after incubation with CRP. Subsequently, fluid permeability increased (Kusche-Vihrog et al., 2011). More in vitro studies reported reactive oxygen species and inflammatory cytokines such as tumor necrosis factor-α to lead to hyaluronan degradation in the glycocalyx of endothelial cells and polymorphonuclear leukocytes (Moseley et al., 1997; Vigetti et al., 2010). Importantly, inflammation might also be linked to (and probably induced by) overhydration as has been shown by this work and other researchers [reviewed by Colombo et al. (2008)]. Our study showed signs of generalized inflammation with increased CRP and lymphocytopenia, with the latter being associated with eGC breakdown. Thus, although the associations of increased vascular volume with ED are much stronger, we cannot exclude the potential contribution of a chronic inflammatory state to ED in patients with CKD.

This study has several limitations. Firstly, as we observed more co-morbidities such as hypertension, coronary heart disease or diabetes in CKD patients compared to healthy controls, we cannot exclude that our results were influenced by a difference in comorbidity burden. Nevertheless, high blood pressure seemed to be treated quite well considering the comparable blood pressure between CKD patients and healthy donors. Moreover, there was only one diabetic CKD patient in the CKD study group. Secondly, we observed variability within each group which can reflect biological variability in microvascular density and/or sampling error. Lastly, we only had access to a small number of patients which limits the validity of our results.

To conclude, this study is the first to show direct evidence of dermal glycocalyx loss and ED in patients with CKD. Moreover, plasma markers of volume overload were associated with tissue markers of endothelial breakdown (i.e., glycocalyx) and endothelial activation (i.e., VEGFR2 and vWF). Although our results argue for strict volume control in CKD patients, future studies should address whether such an approach is capable of attenuating the eGC loss and ED in these patients.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the study protocol is in accordance with the Dutch Medical Research Involving Human Subjects Act (WMO), and approved by the Medical Ethics Committee of the University Medical Center Groningen (METc 2014/077). All procedures were conducted in accordance with the declarations of Helsinki and Istanbul. Details on the study protocol have been described (Hijmans et al., 2019). The patients/participants provided their written informed consent to participate in this study.

JB, CF, RP, SB, and HP conceived and designed the study and made the skin biopsies and clinical data available. JK, RH, MO, and WD performed the laboratory measurements. JK, JB, and CF analyzed and interpreted the data and drafted the manuscript. All authors approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.