Endothelial barrier reinforcement relies on flow-regulated glycocalyx, a potential therapeutic target

Abstract

BACKGROUND:

The onset of many disease processes depends on the function of the endothelial cell (EC) glycocalyx (GCX) which acts as a flow-dependent barrier to cellular infiltration and molecular transport across the blood vessel wall.

OBJECTIVE:

This review aims to examine these processes with the potential end goal of implementing GCX repair to restore EC barrier function and slow the progression of disease.

METHODS:

Cell and mouse studies were employed to examine the state of EC GCX in healthy versus disruptive flow conditions. Correlates of observations of the GCX with a number of EC functions were sought with an emphasis on studies of trans-endothelial barrier integrity against vessel wall infiltration of cells and molecules from the circulation. To demonstrate the importance of GCX as a regulator of trans-endothelial infiltration, assays were performed using ECs with an intact GCX and compared to assays of ECs with an experimentally degraded GCX. Studies were also conducted of ECs in which a degraded GCX was repaired.

RESULTS:

In healthy flow conditions, the EC GCX was found to be thick and substantially covered the endothelial surface. GCX expression dropped significantly in complex flow conditions and coincided with a disease-like cellular and molecular accumulation in the endothelium or within the blood vessel wall. Therapeutic repair of the GCX abolished this accumulation.

CONCLUSIONS:

Regenerating the degraded GCX reverses EC barrier dysfunction and may attenuate the progression of vascular disease.

1. Background and motivation of the presented research

At Jagiellonian University, Kraków, Poland on July 2–6, 2018, the European Society for Clinical Hemorheology and Microcirculation, the International Society for Clinical Hemorheology, and the International Society of Biorheology held a joint meeting of the 3 societies. This joint meeting brought together scientists and clinicians from diverse areas of bio- and hemo-rheology research, to share their latest discoveries and engage in interactive discussions. This review paper is a summary of work that was shared by the Ebong research group in a session entitled “The Glycocalyx – Its Role in Disease.”

The Ebong laboratory studies mechanisms of mechanical regulation of endothelial cells (ECs), which normally line and protect blood vessels in order to maintain vascular health. These ECs are coated with molecules that together form a complex sugar structure, called glycocalyx (GCX). The GCX is a fibrous mesh that has been mainly attributed to the direct detection of blood flow-derived forces and transduction of these forces into EC responses to trigger events that are protective for the blood vessels in healthy conditions. Disruptions in blood flow contribute to vascular-associated disease conditions, including atherosclerosis and potentially cancer metastasis, which coincide with dysfunctional ECs [1]. The Ebong laboratory is focused on revealing that these disease conditions also coincide with and are in large part caused by degraded EC GCX.

Atherosclerosis is initiated by transition of ECs from an anti-inflammatory to a pro-inflammatory state [2,3]. Pro-inflammatory ECs exhibit impairment in their vasoregulatory functions [4,5], increased expression of cell adhesion receptors [6,7], and increased production of reactive oxygen species [8,9]. A major next step in the initiation and formation of atherosclerosis, after EC-dependent lipid trapping, is EC recruitment of circulating inflammatory cells that become retained in the blood vessel wall [10,11]. EC- dependent monocyte recruitment to affected vascular areas occurs in an attempt to reduce inflammation and, subsequently, reverse the disease process. However, in atherosclerosis, a chronic inflammatory state occurs whereby there is excessive blood vessel wall infiltration by inflammatory monocytes that become macrophages and promote lipid-filled, necrotic plaque formation. Previous studies have suggested that GCX degradation in vascular regions prone to atherosclerosis may lead to increased monocyte infiltration at the onset of plaque development [12–17].

Cancer disease mechanisms, particularly the formation of secondary tumors via metastasis, strongly resemble atherosclerosis disease mechanisms. Specifically, both diseases exhibit changes in expression of cell adhesion receptors, the influx of cells, and chronic inflammation, among other similarities [18]. This review draws attention to the metastatic stage of cancer, which is responsible for approximately 90% of cancer-related deaths [19–21]. During metastasis, primary tumor cells move from the parent tumor to neighboring or distant tissues to form secondary tumors [22]. This occurs when cells from the primary tumor migrate to and intravasate nearby blood vessels, travel through these vessels, and eventually extravasate the vessels at distant tissues where secondary tumors can form [23]. Preceding secondary tumor formation, the cancer cells must initially home to and cross the adhesive endothelium, in order to colonize and grow a secondary tumor. This trans-endothelial infiltration to form secondary tumors could occur due to dysfunction of the endothelium’s GCX coating [24,25].

The goal is to understand GCX-mediated EC mechanisms of these two diseases conditions, atherosclerosis and cancer metastasis, so that GCX-targeted treatments can be developed in a manner that heals ECs and effectively slows disease progression.

2. Introduction to the endothelial glycocalyx: Structure and function

The endothelial GCX is a mesh-like proteoglycan and glycoprotein layer on the luminal membrane of ECs [26,27] (Fig. 1). Structurally, the proteoglycans are the major component of the GCX. These proteoglycans consist of glycosaminoglycan (GAG) chains attached to the cell membrane via core proteins or receptors. GAGs of the EC GCX include heparan sulfate (HS), hyaluronic acid (HA), and chondroitin sulfate (CS) (Fig. 1). Of these, HS is the most abundant, constituting 50–90% of all GAGs [26]. HA is also a large constituent [28,29]. The core protein groups that many of these GAGs are attached to include syndecans, glypicans, perlecans, and versicans, among others [26,27]. However, syndecans and glypicans are the most prominent on ECs (Fig. 1). Furthermore, while syndecans and glypicans are attached to the endothelial membrane, perlecans and versicans are secreted and therefore likely do not provide structure to the surface GCX [30]. Additionally, various receptors act in lieu of core proteins when it comes to tethering HA to the membrane [26,27] (Fig. 1). Cluster of differentiation 44 (CD44), for example, is an important HA bound receptor [26,27] (Fig. 1). The other major structural components of the GCX include sialic acid (SA) moieties, glycoproteins, and soluble molecules [26,27] (Fig. 1). Glycoproteins of the GCX include the cell adhesion molecule groups selectins, integrins, and immunoglobulins [31,32]. Serum albumin is an example of a common soluble molecule. Serum albumin is adsorbed within the GCX mesh pores (which have been estimated at 7 nm or less in size [33]) and is believed to stabilize the GCX structure, sometimes even in the presence of GCX-degrading stimuli [34].

Fig. 1. Schematic of the EC GCX structure.

The GCX is mainly composed of glycosaminoglycans, such as heparan sulfate (HS), hyaluronic acid (HA), and chondroitin sulfate (CS), attached to core proteins or receptors, such as glypicans, syndecans, and CD44. The GCX also abundantly contains sialic acid (SA) residues and soluble proteins, such as serum albumin. Glycoproteins, including selectins, integrins, and immunoglobulins, lie at the base of the GCX. These components contribute to various functions of the GCX, which is mainly implicated in endothelial mechanotransduction and trans-endothelial barrier regulation. Through these functions, the GCX assists in regulating vascular activities and health.

The GCX is an inherently difficult structure to study because it is extremely susceptible to degradation by a variety of stimuli including dehydration [35], chemical reagents [35], inflammatory stimuli [36,37], and ischemia/reperfusion [16,38]. As such, its dimensions are largely debated. Initially, the GCX was estimated to be on the order of nanometers [39–41]. More recent reports have indicated that the GCX layer can extend to several microns in thickness [35]. In actuality, considering that morphological and functional differences exist between ECs of different vascular origin (with respect to species, vascular bed, etc.) [35,41], the same must be true when comparing the thicknesses of GCX from different origin.

GCX attachment to the cell membrane, its composition, organization, and its dimensions, as described above, play a role in determining its function. HS, HA, and SA have been the focus of many reports for their specific roles in dictating EC GCX function [28,29,42–44]. Glypican and syndecan core proteins have also been reported to play significant roles in several GCX functions, consistent with their direct links to and, sometimes, across the EC membrane, [26,27,45]. The role of CS is understudied. These components of the endothelial GCX work together, when intact, to promote endothelial and vascular health and, consequently, may prevent dysfunction and disease.

Under healthy conditions in which the GCX is abundantly expressed, the GCX functions by sensing and transducing blood flow-derived shear stresses into EC biochemical activities. This is known as EC mechanotransduction [46] and has been of great research interest. GCX-mediated mechanotransduction influences key endothelium functions. The GCX has been shown to regulate EC signaling pathways related to angiogenesis [47,48], vasoregulation [45,49], and regulation of inflammation [14,50]. One of the most well studied GCX-mediated mechanotransduction events is GCX-mediated flow-regulation of endothelial production of nitric oxide, which serves as a potent vasodilator while also promoting an anti-inflammatory endothelial state [28,45,49,51].

In addition to GCX’s function as an endothelial mechanotransducer, it also plays a vital function in the protection of ECs from interactions with molecules and cells that reside in the blood circulation. To adhere to and permeate the endothelium, circulating cells, such as leukocytes and neutrophils, require adhesion molecules including E-selectin, intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and a variety of other glycoprotein cell adhesion molecules that are contained within the GCX [15–17,52–54]. In cell culture and animal studies, decreased expression of the proteoglycan components of the GCX has been shown to facilitate increased leukocyte adhesion to the endothelium [15,16,52,55], a phenomenon that is thought to extend to other circulating cell types and molecules. An adhesive and permeable endothelial state is the underlying cause of a wide array of medical conditions. These conditions include atherosclerotic plaque formation [56], coagulation and thrombosis [57,58], and extravasation of metastatic cancer cells [54,59,60].

In this review, recent findings regarding GCX response to flow-derived shear stress, as indicated by flow-dependent differences in GCX structure on the EC surface, are discussed. Specifically, it will be established that disturbed flow (DF) patterns, compared to uniform flow (UF) patterns, can induce GCX degradation. Subsequently, it will be demonstrated that GCX, when it is disabled due to flow-induced degradation, adversely impacts EC function. EC mechanotransduction function is of great interest, but was the subject of a recent review paper [61] and numerous other review articles. Therefore, herein the focus will be on the implications of flow-induced GCX degradation on exposure of the endothelium and vascular wall to adhesion and accumulation of circulating cells and molecules. Data will be presented to show that GCX degradation coincides with blood vessel wall infiltration by inflammatory cells in regions of atherosclerosis development. It will also be shown that GCX degradation promotes circulating cancer cell adhesion to the endothelium, which is a first step in secondary tumor formation. Lastly, the possibility of developing new drugs that target the GCX to limit trans-endothelial permeability and potentially treat disease will be discussed.

3. Review of in vitro and in vivo observations of flow-induced GCX remodeling

Previous studies showed thinning of the GCX overlying atherosclerotic plaques when compared to the GCX overlying non-plaque filled blood vessel walls [14]. Therefore, one might assume that GCX thinning occurs as a consequence of the atherosclerotic process. However, in the present work it was considered that GCX thinning could occur prior to the onset of atherosclerosis. Interestingly, plaques consistently occur in predictable vascular regions, such as branches, constrictions, and curvatures [62–64]. Rheological studies of atheroprone vascular regions described the presence of unique blood flow patterns when compared to regions of infrequent atherosclerosis development [62,63]. Specifically, the 3-dimensional pattern of blood flow within atherosclerosis-prone areas is complex, characterized by low time-averaged shear stress magnitudes, high shear stress gradients, and occasional flow reversal, altogether referred to as DF [62,63]. In contrast, blood flow within atherosclerosis-resistant areas of the vasculature exhibits high time-averaged shear stress magnitudes, low shear stress gradients, and unidirectional flow, altogether referred to as UF [62,63]. A number of in vitro and in vivo studies performed by the research groups of Rubanyi [65], Gimbrone and Garcia-Cardena [66,67], Jo [68–72] and others have shown that DF patterns, in comparison to UF patterns, significantly induce atherosclerosis-associated endothelial inflammation. From these studies, clear indicators of DF-induced EC inflammation include, but are not limited to, impaired vasoregulatory functions, elevation of cytokines and chemokines and their receptors, induced expression of cell adhesion receptors, stimulation of morphogenesis and proliferation, increased expression and activation of inflammatory transcription factors, and production of reactive oxygen species via the reactive oxygen species synthesizing subunit p47^phox [4–9,65–72]. A GCX that is degraded by DF could be the open door to these pro-inflammatory events and, subsequently, to atherosclerotic plaque formation.

In the present work, it was hypothesized that GCX expression could be regulated by disturbed, atherosclerosis-prone flow patterns, resulting in a decrease in GCX expression. To test this hypothesis, both in vitro and in vivo methods were utilized to study GCX structure in relation to DF versus UF conditions [73,74]. This hypothesis was investigated using a custom, parallel-plate, in vitro flow chamber in which GCX-expressing ECs were exposed to flow patterns characteristic of the carotid artery bifurcation, one of the most common areas of atherosclerosis development [62,63,75]. In another experiment, this hypothesis was tested in vivo by performing surgical reconstruction of mouse carotid arteries to induce an acute disturbance in blood flow. This acute DF animal model, developed by the groups of Berk [76] and Jo [71], is well established. In a third experiment, the effects of chronic in vivo DF on EC GCX were assessed in mouse aortas along the inner curvature of the aortic arch and descending aorta, which are exposed to low shear stress magnitudes and DF that can predispose the vessel to atherosclerosis [77,78]. In the in vitro and in vivo experiments, EC GCX expression was also examined in UF conditions for reference. EC GCX expression was also examined in static (no-flow) control conditions for the in vitro experiments.

The cell culture experimental results confirmed expectations. Based on a measure of relative GCX fluorescence intensity, an increase in EC GCX coverage and thickness was observed after exposure to less than a day of UF when compared to static controls [73]. Specifically, GCX was labeled with wheat germ agglutinin (WGA) lectin, which broadly labels the GCX because it targets more than one GCX constituent, including SA and N-acetylglucosamine, a component of HS and HA. WGA-labeling of the GCX revealed that short term (less than a day) UF stimulated ECs express increased GCX coverage and thickness by 16% and 14%, respectively, when compared to static controls (Fig. 2A, 2C, 2G, 2H) [73]. The effects of UF on the GCX were further investigated by labeling the HS component exclusively, as it has previously been implicated in GCX mechanotransduction and vascular health. These experiments revealed a 48% and 10.8% UF-induced increase in HS coverage and thickness, respectively (Fig. 2D, 2F, 2I, 2J) [73]. Collectively, these WGA and HS results confirmed that UF patterns can increase GCX expression from baseline levels in static conditions [73]. Further, the results suggest individual GCX components may be differentially regulated by UF [73]. These results are in agreement with previous one-, three-, and seven- day UF studies conducted by Gouverneur et al. [79] and Koo et al. [80], which similarly found elevation of GCX expression in a GCX component-specific manner in UF conditions compared to static conditions [79,80]. More interesting results were obtained in studies of GCX coverage and thickness in response to DF, which is relevant to atherosclerosis. Compared to static controls, it was found that exposure of ECs to DF led to a 44% decrease in overall WGA-labeled GCX coverage and 14% decrease in WGA-labeled GCX thickness (Fig. 2B, 2G, 2H) [73]. Similarly, DF induced an 18% and 10.9% decrease in HS coverage and thickness, respectively (Fig. 2E, 2I, 2J) [73]. While these results are similar to observations reported by Koo et al. [80], who found three and seven days of DF to induce decreased expression of various GCX components [80], the present study is the first study to examine the effects of DF on GCX expression using a geometrically relevant flow model with juxtaposed DF and UF profiles.

Fig. 2. It has been found that, with exposure of cultured ECs to 6-hour flow conditions, GCX is differentially expressed in uniform versus disturbed flow (UF vs. DF) patterns.

[73]. A–C: WGA serves as a general GCX label because it targets more than one GCX constituent including SA and N-acetylglucosamine, a component of HS and HA. As determined by laser scanning confocal microscopy of fluorescent green WGA staining, GCX coverage and thickness on confluent monolayers of rat fat pad ECs is shown to be increased in regions of UF but decreased in DF (both compared to static conditions) [73]. D–F: Using an anti-HS antibody to look more specifically at the most abundant GCX component, HS, demonstrates similar trends in GCX coverage and thickness in response to flow [73]. G–J: Quantification of WGA- and anti-HS-labeled GCX coverage and thickness. Values were obtained using previously published methods [73]. Results were normalized to static controls. Statistically significant differences between static-, UF-, and DF-conditioned GCX are indicated as *P < 0.05 and ***P < 0.001, which were determined after performing one-way ANOVAs followed by post-hoc Tukey’s multiple comparison test (G–J). (This figure is adapted from [73], which is allowed by the publisher without formal written permission.)

The cell culture results were validated by two different types of mouse studies. The first study involved a study of two cohorts of C57BL/6-background apolipoprotein E knockout (ApoE KO) mice on a chow diet, in which the left carotid arteries of the first cohort were normal and the second were partially ligated to induce an acute disturbance in blood flow [74]. Typically, flow within the common carotid arteries is uniform. However, ligating 3 of the 4 caudal branches of the left carotid artery induces DF patterns characterized by low time-averaged shear stresses within the left common carotid artery [71,74]. Mitra et al. fluorescently labeled the GCX and identified a significant change in GCX expression on ligated left carotid arteries vs. non-ligated left carotid arteries (Fig. 3A–3C) [74]. Specifically, the chow-fed ApoE KO mice with ligated left carotid arteries, when compared to chow-fed ApoE KO mice with non-ligated left carotid arteries, exhibited approximately 50% less WGA-labelled GCX within one week after left carotid artery ligation (Fig. 3C) [74]. This work provides strong evidence that disturbed flow alone is sufficient to induce GCX degradation even in the absence of high fat diet conditions, which is a well-known disease risk factor and typically incorporated into mouse studies of vascular disease. In another experiment in which the GCX was fluorescently labeled with HA binding protein, performed by Tarbell and Cancel, the superimposition of high fat in the diet led to complete removal of GCX from the wall of the ligated left carotid artery [81].

Fig. 3. It has been found in vivo that GCX degradation in response to acute DF correlates to vascular wall infiltration by circulating cells.

[74]. Acute DF is induced in vivo, by using a partial carotid ligation mouse model. A–C: As shown by laser scanning confocal micrographs of fluorescent red WGA labeling, GCX is found to be expressed continuously and more significantly (ANOVA, ***P < 0.001) in non-ligated left carotid arteries, which experience UF, than in ligated left carotid arteries, which experience DF [74]. D–F: Here, inflammatory cells (fluorescent yellow) are examined because they represent an important circulating cell population. Uptake of macrophages, a key inflammatory cell type, was examined through immunohistochemistry of macrophage-specific CD68. Macrophages of interest were those residing in the vessel wall and not those that lie outside of the dotted line that marks the outer vessel wall. Macrophage uptake was found to correlate to GCX expression. Significantly less macrophage uptake (ANOVA, ***P < 0.001) was found in non-ligated, UF regions than in ligated, DF regions [74]. (This figure is adapted from [74], which is allowed by the publisher without formal written permission.)

The second mouse study involved en face imaging of various endothelium regions of the aorta in C57BL/6 mice [73]. Due to the geometry of the aorta, ECs along the aortic vessel wall are chronically exposed to UF patterns adjacent to DF patterns [77,78]. For example, the high curvature of the aortic arch leads to DF patterns within the inner curvature of the distal aortic arch and proximal descending aorta. Furthermore, this region is also susceptible to the formation of atherosclerotic plaques [82,83]. In contrast, the straight geometry of the abdominal aorta, which is resistant to atherosclerosis development, contains chronic UF patterns [84]. Fluorescent WGA labeling of the GCX within both the distal aortic arch and abdominal aorta corroborated previous in vitro findings: GCX expression was weaker within the distal aortic arch (Fig. 4A), which is exposed to chronic DF, than it was in the abdominal aorta (Fig. 4B), which is exposed to chronic UF [73]. While it is possible that variability in GCX expression along the aorta (and non-ligated vs. ligated carotid artery) is due to differences in pressure or hoop stresses, there has been no prior demonstration of a causal relationship between GCX expression and blood pressure or hoop stress. Considering the numerous studies that have presented evidence that shear stress regulates GCX expression [73,79,85], it is clear that the observed in vivo differences in GCX expression can be attributed in large part to changes in shear stresses (both acute and chronic).

Fig. 4. It has been shown in vivo that chronically different flow conditions differentially regulate GCX expression.

[73]. GCX expression is indicated by fluorescent green WGA labeling. A: GCX fluorescence intensity is weaker in the distal aortic arch of C57Bl/6 mice where flow patterns are chronically disturbed. B: GCX fluorescence is more intense in the abdominal aorta where flow patterns are chronically uniform [73]. (This figure is adapted from [73], which is allowed by the publisher without formal written permission.)

The collective in vitro and in vivo results summarized here [73,74] show that the endothelial GCX is degraded through exposure to atherosclerosis-prone blood flow. GCX degradation, therefore, precedes atherosclerosis and is not a consequence of atherosclerosis. The pronounced differences in GCX structure in UF conditions compared to DF conditions highlight the implications of pro- versus anti-atherosclerotic flow patterns on GCX remodeling, which should translate to implications for EC and vascular function. GCX degradation was previously implicated in inhibited flow-induced nitric oxide production [45,61,86], impaired inter-endothelial communication [61,86], uncontrolled reactive oxygen species expression [51], and up-regulated adhesion molecule expression [87], which are all associated with pro-atherosclerotic endothelial dysfunction. Further research will hopefully elucidate the relationship between GCX structure, its function, and endothelial phenotype.

4. Implications of GCX degradation on circulating cell adhesion to the endothelium

GCX degradation may be directly responsible for circulating cell adhesion to the endothelium. Degraded GCX expression could lead to a reduced physical adhesion barrier between ECs and circulating cells, as presented by Mitchell and King [88]. To examine this notion, the aforementioned experimental mouse model with ligated left carotid arteries was revisited. Left carotid arteries that were GCX-deficient due to DF conditions produced by artery ligation were compared to left carotid arteries that were GCX-rich due to the typical UF conditions present in the non-ligated artery (Fig. 3A–3C) [71,74]. Most related studies have been conducted using leukocytes, such as neutrophils and macrophages [15,16,52,55]. Consistent with this, immunohistochemical analysis of cluster of differentiation 68 (CD68), a macrophage marker, was performed. It was found that decreased GCX expression in ligated left carotid arteries coincides with increased uptake of macrophages (Fig. 3D–3F) derived from monocytes of the circulation [74]. Specifically, macrophage uptake increased 20-fold within GCX-deficient, DF-conditioned, ligated left carotid arteries compared to GCX-rich, UF-conditioned, non-ligated left carotid arteries [74]. Previous studies, although limited, are in agreement with these findings. For instance, Schmidt and colleagues found increased inflammatory neutrophil adhesion in areas of pulmonary sepsis-associated GCX degradation [17]. Similarly, in ischemia/reperfusion patients, Chappell and coworkers observed that inflammatory leukocyte adhesion correlated to shedding of HS and syndecan along with GCX thinning [16].

Additional experimentation was required to fully investigate the notion that GCX degradation is a direct cause for circulating cell adhesion to the endothelium and, further, to determine the specific GCX components responsible. Schmidt’s group laid the foundation for demonstrating the causal relationship by blocking GCX degradation and subsequent cell adhesion through inhibition of HS-specific heparinase III (Hep III) enzyme [17]. Contrarily, Ebong’s group explored the causal relationship by inducing EC GCX degradation. This was achieved through treatment of ECs with SA-specific Clostridium perfringens neuraminidase (Neur) enzyme [29,89], which degrades a broad variety of SA residues of the GCX. Subsequently, the GCX-deficient ECs were co-cultured with metastatic cancer cells as a model circulating cell (Mensah et al., manuscript in preparation). As previously mentioned, metastatic cancer cells are derived by cancer cell migration from primary tumor sites into the blood circulation. Preceding secondary tumor formation, the circulating cancer cells must adhere to the endothelium in order to begin their migration into the secondary tumor site. This process involves a range of endothelial and cancer receptors and ligands including selectins, cadherins, integrins, CD44, and immunoglobulin superfamily receptors [54,90]. Few studies have investigated the role of the EC GCX in circulating cancer cell adhesion to the endothelium [59,60].

Recent data revealed that after 4-hour exposure to GCX-degrading DF conditions, ECs interacting with circulating cancer cells for 1 hour resulted in an approximate 60% increase in cancer cell adhesion (Fig. 5B, 5J, 5M–5O) when compared to ECs pre-exposed to 4 hours of GCX-rich UF conditions (Fig. 5C, 5K, 5M–5O) obtained by Ebong’s group (Mensah et al., manuscript in preparation). To confirm GCX degradation as a direct cause of increased adhesion, the Neur enzyme was added to the EC environment in UF conditions. After Neur enzyme treatment, the Neur-containing UF perfusate was exchanged with media containing circulating cancer cells. It was found that Neur-induced degradation of the SA component of the EC GCX in UF conditions does indeed substantially increase circulating cancer cell attachment to the endothelium (Fig. 5D, 5L, 5O) (Mensah et al., manuscript in preparation). This confirms prior speculation that SA, in comparison to other GCX components, such as HS and HA, is an important mediator of cancer cell adhesion and potentially cancer cell biology and signaling [91,92]. Further, taking Ebong’s experiment together with Schmidt’s [17] shows that intact GCX plays an important role in blocking adhesion of any circulating cell to the endothelium and degradation of the HS and SA components of the GCX promotes adhesion to the endothelium.

Fig. 5. In a 4-hour flow study, DF-induced remodeling of the GCX, with respect to the SA component in particular, is found to enable adhesion of circulating cells to the endothelium.

(Mensah et al., manuscript in preparation). Fluorescent green WGA is again used as a general GCX label, to collectively visualize SA and N-acetylglucosamine, a component of HS and HA. A–C, M–N: Consistent with what is shown in Fig. 2, WGA labeled GCX is differentially regulated by DF versus UF in human umbilical vein ECs after 4 hours of flow. D, M–N: In UF conditions, treatment with Neur, which specifically degrades SA residues of the GCX, significantly reduces GCX expression as viewed by lesser WGA fluorescence. E–H: EC layers (in phase contrast images) maintain integrity in all conditions. E–L, O: Here, stage IV metastatic breast cancer cells (4T1; fluorescent red) are examined as a representative circulating cell population via static attachment over a 1-hour incubation period. Metastatic cancer cell adhesion differs under various GCX conditions. Specifically, a significant increase in adhesion was observed in GCX- deficient conditions of DF or GCX-deficient conditions of UF combined with Neur treatment. GCX expression and cancer cell attachment to ECs were quantified in accordance with previously published methods [96]. For statistical analysis of the data, all results were normalized to control conditions and reduced, ANOVA was performed between groups, and significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

It is plausible that the increase in GCX penetration by circulating cells can be attributed to an increased void ratio created by the eliminated HS and SA content. This increased void ratio will enable passive and non-receptor mediated interactions with ECs. It is also likely that there is GCX collapse [93], due to reduced electrostatic repulsive force that occurs with loss of negatively charged GCX. This GCX collapse will expose to circulating cells ECs adhesion ligands, such as e-selectin, for active and receptor mediated interactions. In addition to reducing the physical adhesion barrier between ECs and circulating cells [88], the degradation of GCX components likely impairs mechanotransduction, resulting in up-regulated expression of adhesion molecules [87], which further facilitates increased circulating cell adhesion to the endothelium and increased retention in the blood vessel wall.

5. Development of GCX-based therapeutics to prevent vascular-related diseases

This review has focused on EC GCX structure and function as a barrier to trans-endothelial infiltration, which breaks down in atherosclerosis and cancer. EC GCX degradation has also been implicated in the development of other vascular-related diseases and conditions such as diabetes [94], sepsis [17], and chronic kidney disease [95]. It is therefore of interest to investigate whether EC GCX restoration can improve vascular function.

Ebong’s research group is contributing to this endeavor by developing an approach to repair the EC GCX layer. Demonstration of the efficacy of using an EC GCX regeneration approach to block trans-endothelial circulating cell infiltration of the blood vessel wall is currently in progress. To date, proof of concept experiments have been performed in cell culture settings, demonstrating the feasibility of repairing degraded EC GCX and subsequently restoring nanoscale trans-endothelial barrier function [96] and inter-endothelial communication [86].

In an initial experiment, ECs with artificially regenerated GCX were exposed to nanoparticles [96], laying the foundation for experiments in which ECs would be exposed to circulating inflammatory and/or cancer cells in the future. Given that the GCX mesh pores are estimated at less than or equal to 7 nm in size [33], infiltration of the nanoparticles, which are much smaller than circulating inflammatory or cancer cells, can reveal GCX degradation as it begins to occur. Exclusion of the same nanoparticles, in contrast, can reveal complete restoration of the trans-endothelial barrier. A proof of concept experiment utilized 10-nm-sized, red fluorescent, polymer-coated gold nanoparticles [96]. To observe passive infiltration regulated exclusively by the GCX and not by non-GCX cell surface structures and adhesion molecules, the nanoparticles were not conjugated to any ligand, which would typically target endothelial surface adhesion molecules and encourage active nanoparticle internalization by the ECs [97]. First the effect of intact GCX on nanoparticle infiltration into the endothelium was investigated, and minimal infiltration was observed (Fig. 6) [96]. GCX degradation, achieved by treating ECs with Hep III, led to a more than 6-fold increase in nanoparticle infiltration into ECs (Fig. 6) [96]. This trend is similar to observations on the effect of GCX degradation on macrophage and cancer cell attachment [17,25,33,98–103]. To determine whether GCX restoration can reduce endothelium permeability to nanoparticles, endogenous HS was used to restore GCX expression on Hep III-treated ECs (Fig. 6) [96]. After restoration, nanoparticle infiltration of ECs was observed to rebound from infiltration levels of degraded samples and return to levels of nanoparticle infiltration in untreated controls (Fig. 6) [96].

Fig. 6. GCX regeneration may recover trans-endothelial barrier against molecular and cellular infiltration.

A: Here, HS expression and thickness on rat fat pad ECs, when compared to control conditions, is shown to be reduced by treatment of the cells with Hep III, which degrades HS from the GCX [96]. After Hep III treatment, HS is successfully regenerated through treatment of ECs with endogenous HS [96]. A, B: HS expression is correlated with infiltration of nanoparticles (in florescent red). Specifically, nanoparticle infiltration is significantly increased after Hep III-induced HS degradation and abolished with GCX regeneration due to endogenous HS treatment [96]. Infiltration values were obtained using previously published methods [96] and results were normalized to control conditions. To determine statistical significance of differences between infiltration on various GCX conditions, ANOVAs were performed. Significance is indicated as *P < 0.05. (This figure is adapted from [96], with formal written permission from the publisher.)

Prior studies have also successfully prevented GCX degradation or restored GCX expression to maintain endothelial function. In these studies, GCX-regenerating therapeutics included TNF-α inhibitors [58], angiopoetin-1 [104], sulodexide [94], sevofluorane [16], hydrocortisone [52], and heparin [105]. One prior study conducted by Fu’s group found that protection of the endothelial GCX through sphingosine- 1-phosphate (S1P) treatment or matrix metalloproteinase (MMP) inhibition resulted in regenerated GCX expression and reduced cancer cell adhesion [60]. This led Ebong and colleagues to incorporate S1P into the GCX regeneration strategy. The treatment of ECs with combined endogenous HS and S1P has been demonstrated, to date, to be effective in regenerating GCX and in restoring a complex EC function: cell-to-cell communication [86]. Preclinical animal studies are currently being performed to determine the potential of in vivo GCX regeneration to restore endothelial function. It is anticipated that the animal studies will show that GCX regeneration is able to restore the trans-endothelial infiltration barrier and a variety of other endothelial functions, ultimately promoting vascular health.

6. Conclusion

The endothelial GCX is a complex, protein-polysaccharide layer that assists in various endothelial functions including regulation of the barrier at the interface between the blood circulation and the tissue of the blood vessel wall. This review summarizes recent studies performed by Ebong’s group, which confirmed that EC GCX is thick and substantially covers the endothelial surface in healthy flow conditions. Ebong and colleagues have also shown that atherosclerosis relevant flow patterns can significantly impair GCX expression [73]. This decrease in GCX expression is correlated with an increase in macrophage accumulation within the blood vessel wall, a key step in the development of atherosclerosis plaques [74]. Additionally, GCX loss is linked to other vascular-related processes including recruitment of metastatic cancer cells to the endothelium (Mensah et al., manuscript in preparation). Studies of the effect of loss of function of certain GCX components via treatment with Neur and Hep III demonstrated that disease-like molecular and cellular accumulation in the endothelium or within the blood vessel wall can be attributed to loss of SA and HS [86,96]. Lastly, therapeutic strategies targeting the GCX have demonstrated an ability to restore vascular endothelial function in the presence of detrimental stimuli [86,96]. Future studies on GCX structure and function will further develop and substantiate the benefits of GCX-targeted therapeutics.

Abbreviations

4T1

Stage IV Metastatic Breast Cancer Cells

CD44

Cluster of Differentiation 44

CD68

Cluster of Differentiation 68 (a macrophage marker)

DF

Disturbed Flow

ECs

Endothelial Cells

GCX

Glycocalyx

GAGs

Glycosaminoglycan

Hep III

Heparinase III (heparan sulfate degradation enzyme)

HS

Heparan Sulfate

MMP

Matrix Metalloproteinase

Neur

Neuraminidase (sialic acid degradation enzyme)

S1P

Sphingosine-1-Phosphate

SA

Sialic Acid

UF

Uniform Flow

WGA

Wheat Germ Agglutinin (labels sialic acid and N-acetylglucosamine)