Dr M. Yvonne Alexander, Manchester Academic Health Science Centre (CMFT), The University of Manchester, School of Biomedicine, Cardiovascular Research Group, 3rd Floor Core Technology Facility, 46 Grafton St., Manchester M13 9NT, UK. Tel.: +44 161 275 1224; fax: +44 161 275 1183; e-mail: email@example.com
Aging poses one of the largest risk factors for the development of cardiovascular disease. The increased propensity toward vascular pathology with advancing age maybe explained, in part, by a reduction in the ability of circulating endothelial progenitor cells to contribute to vascular repair and regeneration. Although there is evidence to suggest that colony forming unit-Hill cells and circulating angiogenic cells are subject to age-associated changes that impair their function, the impact of aging on human outgrowth endothelial cell (OEC) function has been less studied. We demonstrate that OECs isolated from cord blood or peripheral blood samples from young and old individuals exhibit different characteristics in terms of their migratory capacity. In addition, age-related structural changes were discovered in OEC heparan sulfate (HS), a glycocalyx component that is essential in many signalling pathways. An age-associated decline in the migratory response of OECs toward a gradient of VEGF significantly correlated with a reduction in the relative percentage of the trisulfated disaccharide, 2-O-sulfated-uronic acid, N, 6-O-sulfated-glucosamine (UA[2S]-GlcNS[6S]), within OEC cell surface HS polysaccharide chains. Furthermore, disruption of cell surface HS reduced the migratory response of peripheral blood-derived OECs isolated from young subjects to levels similar to that observed for OECs from older individuals. Together these findings suggest that aging is associated with alterations in the fine structure of HS on the cell surface of OECs. Such changes may modulate the migration, homing, and engraftment capacity of these repair cells, thereby contributing to the progression of endothelial dysfunction and age-related vascular pathologies.
The average lifespan of humans is increasing and with it the percentage of people aged 65 years or over. By 2035, it is projected that approximately 20% of the population in the UK will be aged 65 years and over. Within this age group, cardiovascular disease (CVD) is the primary cause of death and disability; accounting for more than 40% of all deaths among those aged over 65 years (Ungvari et al., 2010). Aging is a major risk factor for the development and progression of vascular diseases, such as atherosclerosis. Indeed, the incidence rates of heart disease and stroke increase exponentially with age for both men and women. Furthermore, the cost associated with treatment is an ever increasing economic burden. Thus, it is vital to understand the changes that occur during the aging process that contribute to the high incidence of CVD in this population. The increased propensity toward vascular pathology with advancing age may be due, in part, to an imbalance between the magnitude of vascular injury and the capacity for repair. Accumulating evidence suggests that bone marrow-derived endothelial progenitor cells (EPCs) play an integral role in the cellular repair mechanisms for endothelial regeneration and maintenance (reviewed by George et al., 2011). Consistently, depletion of the circulating EPC pool has been shown to be a marker of cardiovascular damage and an independent predictor of cardiovascular events and death (Fadini et al., 2012). To participate in postnatal vasculogenesis or vascular repair, EPCs must be able to (i) respond to chemotactic signals that initiate mobilization from the bone marrow into the circulation; (ii) home to remote sites of vascular injury, ischemia or remodelling; (iii) extravasate from the circulation into such areas, and (iv) finally incorporate into the vasculature or exert paracrine support to the endothelium.
A number of studies have demonstrated that with age, EPCs are subject to changes that diminish their number in circulation and/or function (Vasa et al., 2001; Edelberg et al., 2002; Rauscher et al., 2003; Scheubel et al., 2003; Heiss et al., 2005). Such changes are believed to culminate in a decreased capacity for neovascularization or insufficient repair of the endothelium following injury, thereby facilitating the progression of endothelial dysfunction and subsequent vascular pathology. Of note, these studies utilized so called colony forming unit-Hill (CFU-Hill) cells or circulating angiogenic cells (CACs); two cell types which have previously been encompassed within the term ‘EPC’ and are obtained from the short-term culture of mononuclear cells (MNCs) on specific substrates in endothelial growth medium (reviewed by Hirschi et al., 2008). Despite evidence of a strong correlation between CFU-Hill and/or CAC numbers, endothelial function (measured by flow-mediated brachial artery dilatation) and cardiovascular events (Werner et al., 2005), it is now apparent that these cell populations are myeloid derived cells with pro-angiogenic properties, which can adopt an endothelial-like phenotype under the appropriate culture conditions (reviewed by Richardson & Yoder, 2011). Therefore, the term EPC is now being used to describe a cell type, which can be isolated from MNCs after a longer period in culture (between 14 and 21 days). This latter population of cells has been variously termed: outgrowth endothelial cells (Gulati et al., 2003), blood outgrowth endothelial cells (Lin et al., 2002); late EPCs (Hur et al., 2004), and endothelial colony forming cells (Ingram et al., 2004). For clarity, the term outgrowth endothelial cells (OECs) will describe the cell type used in this study. Outgrowth endothelial cells display functions such as in vitro clonal proliferative potential, in vitro tube formation, in vivo vessel formation with incorporation into the systemic circulation of immunodeficient mice, in vivo chimeric vessel formation into areas of ischemia and an unequivocal endothelial phenotype (Hur et al., 2004; Melero-Martin et al., 2007; Yoder et al., 2007; Mukai et al., 2008). Among all current putative EPC populations, this cell type displays the most features consistent with a human postnatal vasculogenic cell (Richardson & Yoder, 2011). Although there is evidence to suggest that the functional capacity of CFU-Hill cells and CACs declines with age, there is a lack of studies that have examined the impact of age on the function of OECs.
It is increasingly appreciated that the endothelial glycocalyx, a network of proteoglycans and glycoproteins on the cell surface, plays a complex role in vascular physiology and pathology (Broekhuizen et al., 2009). In the vasculature, heparan sulfate proteoglycans (HSPGs) constitute the predominant type of proteoglycan in the endothelial cell glycocalyx. Heparan sulfate proteoglycans are composed of a core protein to which one or more glycosaminoglycan (GAG) chains are covalently attached. Primarily it is this GAG component, heparan sulfate (HS), which is able to interact with a diverse array of protein ligands, including growth factors and their receptors, chemokines, enzymes, cell adhesion molecules, and various extracellular matrix proteins. These interactions can facilitate ligand-receptor binding, alter protein conformation, increase protein stability or modulate growth factor gradient formation. Consequently, HSPGs are known to regulate an array of cellular processes such as cell adhesion, proliferation, migration, and differentiation (Sarrazin et al., 2011). However, to date, the involvement of HSPGs as a possible mechanism underlying the age-associated reduction in OEC function has not been explored. Heparan sulfate is composed of repeating disaccharide units of uronic acid linked to glucosamine sugar residues, some of which are modified by the addition of sulfate groups on the 2-O position of uronic acid and the N-, 6-O, and 3-O positions of glucosamine. These modifications occur nonuniformly throughout the HS chain creating extensive sequence diversity in the final chain. In many cases, the density and pattern of these modified disaccharides is of special importance in determining HS-protein interactions. For example, a specific pentasaccharide structure, containing a central 3-O-sulfated glucosamine residue, within the HS chain is essential for high affinity interaction with antithrombin, and thus, for anticoagulant activity (Lindahl et al., 1980). Hence, structural changes of HS may perturb the binding of growth factors and other molecules that interact with HS, thereby altering the physiological effects of the ligand and having important functional consequences for the cell.
By interacting with cytokines and chemokines, HSPGs have been shown to be important in the homing of hematopoietic progenitor cells to the bone marrow endothelium and in regulating their retention and proliferation within this niche (Netelenbos et al., 2003). Similarly, HS could play a role in recruiting and activating EPCs to sites of vascular damage and repair. The aim of this study was to determine if the aging process induces structural changes of HS on the surface of OECs, and if these changes correlate with functional alterations of these progenitor cells. Gaining an insight into the mechanisms underlying the age-associated reduction in vascular health is important for defining new treatments to reduce cardiovascular mortality in an aging population.
Isolation and phenotypic characterization of OECs derived from cord and peripheral blood
Outgrowth endothelial cells were successfully isolated from cord (n = 5) and adult peripheral blood (n = 13) samples and expanded in vitro. Outgrowth endothelial cell colonies were detected between 14 and 21 days and were ready to expand after a further 3–4 weeks of culture (Fig. S1). In addition to exhibiting a cobblestone morphology typical of endothelial cells [Fig. 1A (i)], cell phenotype was verified by several means. First, using immunofluorescence, OECs were uniformly positive for several endothelial markers including VE-CAD [Fig. 1A (ii)], von Willebrand factor (vWF) [Fig. 1A (iii)], CD31 [Fig. 1A (iv)], able to incorporate acetylated-LDL [Fig. 1A (v)] and bind the lectin UEA-1 (data not shown). Importantly, OECs lacked expression of the hematopoietic marker CD45 [Fig. 1A (vi)]. Secondly, flow cytometric analysis revealed that OECs expressed the endothelial cell surface antigens CD31, CD146 and, to a lesser extent CD34 and VEGFR-2, as well as being positive for Dil-Ac-LDL uptake and UEA-1 lectin binding. Importantly, OECs were found to lack expression of the hematopoietic antigens CD45 and CD133 (Fig. 1B). These results are consistent with those reported by other investigators (Medina et al., 2010). Further validation of endothelial gene expression was obtained using RT–PCR: endothelial primers for CD34, VECAD, VEGFR-2, vWF, CD31 showed positive amplification. By contrast, no amplicons were generated in OECs when using primers for hematopoietic markers CD14 and CD45, while monocytes, used as a positive control, yielded positive bands for the latter (Fig. 1C). Primer details are listed in Table S5. Finally, OECs were able to form a network of tubelike structures in Matrigel assays (Fig. 1D).
The proliferative, survival and tube-forming capacity of OECs is not significantly impaired with age
To determine whether aging is accompanied by a decline in OEC function, we measured the proliferative, survival, migratory, and tube forming capacity of these cells isolated from cord blood and peripheral blood samples of young (20–30 years) and old (50–70 years) healthy subjects; non smokers who were not receiving medication for any clinical diagnosis.
The proliferative activity of OECs was determined using an MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay over a 48 h period. Although a significant difference in proliferation was seen between cord (n = 5) and peripheral blood OECs (n = 9; P = 0.029), no statistically significant difference was detected between OECs isolated from young (n = 5) and old (n = 4) subjects (Fig. S2A).
To investigate whether or not the survival capacity of OECs in response to apoptotic stimuli changes with age, OEC apoptosis was measured by examining cellular morphology and the levels of endogenous cleaved caspase-3 protein, under normal growth conditions, growth factor/serum deprivation, and staurosporine (STS; 25 nm) treatment. Following STS treatment, morphological changes including cytoplasmic shrinkage, nuclear condensation, and the formation of apoptotic bodies were observed in all three groups of cells (data not shown). There was no significant difference in baseline concentrations of intracellular active caspase-3 between OECs from cord blood (n = 3) and those isolated from peripheral blood samples of young (n = 3) and old (n = 3) volunteers. Similarly, growth factor/serum deprivation and STS-induced intracellular active caspase-3 concentrations were not significantly different between the three groups (Fig. S2B).
(iii) In vitro tube-forming assay
The tube-forming potential of OECs from cord (n = 4) and peripheral blood samples of young (n = 4) and old (n = 4) individuals was evaluated in growth factor reduced Matrigel. In all cases, OECs aligned to form branching, anastomosing tubes with multicentric junctions that created a network of capillary-like structures (Fig. S3A). Several morphological features were measured to characterize these capillary-like networks. Quantification of the number of closed loops, junctions, and branches, as well as, the total length of the network revealed no significant differences in tube-forming capacity among the three groups (Fig. S3B).
The migratory response of OECs declines with age
The migratory capacity of OECs was evaluated using an in vitro wound healing assay in which cells migrate into an area that has been mechanically denuded of cells. Cells at the wound edge migrated into the cell-free area, closing it over time. The rate of migration was determined by calculating the percentage of area closure, as measured by the pre migration and post migration time points. A significant reduction in OEC migration was evident with age (Fig. 2); at 12 hrs 84.87% closure of the wound by cord blood OECs [n = 5; Fig. 2A (i) and B] was observed compared to 64.95% closure by OECs of young volunteers [n = 5;P = 0.027; Fig. 2A (ii) and B] and only 39.61% closure from OECs of older subjects [n = 3; P = 0.034; Fig. 2A (iii) and B]. After 24 hrs, OECs from cord blood and peripheral blood samples of young subjects migrated and refilled the cell-free space (99.3% and 100% respectively). By contrast, OECs from older subjects demonstrated a significant reduction in cell migration with only 87% closure (P = 0.01; Fig. 2C).
The ability of OECs to migrate toward SDF-1α (100 ng/mL) or VEGF (50 ng/mL), two factors which have been implicated in the mobilization and homing of EPCs, was evaluated using a transwell migration assay. The concentration of VEGF and SDF-1α was selected from the linear region of a dose response curve (data not shown). Cells were allowed to migrate for 6 hrs in response to the chemoattractant, after which time, OECs that had migrated onto the lower side of the membrane were counted. Cell migration was expressed as the fold increase relative to the baseline (no chemoattractant). A significant reduction in both SDF-1α and VEGF-induced migration was detected in OECs isolated from old subjects (n = 4) as compared to those from cord blood samples (n = 4) and from peripheral blood samples of younger subjects (n = 4; Fig. 3). Migration toward SDF-1α and VEGF was similar between cord blood OECs and those isolated from peripheral blood samples of young subjects; SDF-1α induced a 1.93 and 1.92 fold increase in cell migration, respectively, but only a 1.5 fold increase in migration of OECs from old subjects (P = 0.044). In response to VEGF, migration was increased 2.14 and 2.04 fold in OECs from cord blood and peripheral blood of young subjects, although migration was significantly lower, at only 1.62 fold, in OECs of old subjects (P = 0.029).
Next, we explored the possibility that the reduction in OEC migration with age is due to changes in the expression of the cell surface receptors for SDF-1α and VEGF. Age-related changes in OEC expression of mRNA encoding CXCR4 and VEGFR-2 was quantified using reverse transcription quantitative real-time-PCR (RT-qPCR). A significant decline in the expression of CXCR4 was detected in OECs isolated from old subjects (n = 4; P = 0.037) in comparison to OECs from cord blood (n = 4; P = 0.037) and peripheral blood of young subjects (n = 4; P = 0.034), although not between cord blood and young peripheral blood OECs. This reduction in CXCR4 expression by OECs from old subjects may contribute to the decline in migration toward SDF-1α observed in the transwell assays. By contrast, mRNA levels of VEGFR-2 did not appear to significantly differ between the three groups of OECs (Fig. S4). As the reduction in OEC migration toward VEGF could not be attributed to a reduction in VEGFR-2 expression by OECs with increasing age, we investigated alternative underlying mechanisms.
Age-associated changes of heparan sulfate structure on OECs
To further explore the reduced migratory capacity of OECs, we investigated the possible link to structural changes of HS with age. Heparan sulfate was extracted from OECs isolated from cord blood (n = 5) and peripheral blood samples of young (n = 5) and old (n = 4) subjects. Heparan sulfate was degraded into disaccharides and the composition analyzed using reverse phase (RP)-high performance liquid chromatography (HPLC) with fluorescence detection using 2-aminoacridone (AMAC) according to a modified version of the method described by Deakin & Lyon, 2008 (Fig. S5). The contribution of each type of disaccharide to the HS sample extracted from OECs was calculated and compared (Fig. 4A and Table S1). The results of the analysis demonstrate that a major component of OEC HS in all three groups was the disaccharide unit ΔUA-GlcNAc, accounting for 44–54% of the total HS (Fig. 4A). However, a significant reduction in the proportion of the trisulfated disaccharide ΔUA[2S]-GlcNS[6S] was detected within HS samples isolated from peripheral blood of young and old subjects as compared to HS extracted from cord blood OECs (P = 0.020). A significant reduction in the abundance of this disaccharide unit was also evident within HS from old subjects in comparison to their younger counterparts (P = 0.014). The disaccharide units ΔUA-GlcNS[6S] and ΔUA-GlcNAc[6S] were significantly reduced in OEC-derived HS from peripheral blood of young (P = 0.047 and P = 0.028, respectively) and old (P = 0.014) subjects as compared to the HS samples from cord blood OECs. By contrast, a significant increase in the abundance of the disaccharide units ΔUA[2S]-GlcNS and ΔUA-GlcNS was detected within HS isolated from OECs of old subjects as compared to the HS samples isolated from OECs of cord blood (P = 0.05 and P = 0.014, respectively) and peripheral blood of young subjects (P = 0.027 and P = 0.014, respectively). Calculation of the overall extent of 2-O-sulfation and 6-O-sulfation indicated that the age-associated decline in the proportion of the ΔUA[2S]-GlcNS[6S] unit was due to a significant decrease in the levels of 6-O-sulfate substitution of GlcNS residues with increasing age (Pearson's correlation coefficient; −0.830, significant at the P < 0.01 level), whereas 2-O-sulfation of GlcA or IdoA residues did not significantly change (n = 18) (Pearson's correlation coefficient; 0.158) (Fig. S6).
The size of the HS chains on the cell surface was also analyzed using Sepharose CL-6B gel filtration chromatography. The approximate size of OEC-derived HS was within the range of 31.05–39.5 kDa and not significantly different among the three groups of OECs analyzed (Fig. S7).
HSPG disruption reduces OEC migration
Given the decline detected in the migratory response of OECs with age, and the significant changes in O-sulfate distribution within OEC HS, we hypothesized that the structural changes of HS may influence the migratory capacity of these cells. First, the relationship between HS structure and migration was investigated. Given the strong, negative correlation observed between subject age and the overall extent of 6-O-sulfation of OEC HS (Fig. S6; r = −0.830, P < 0.01), initially the association between the total level of 6-O-sulfation within OEC HS and the migratory response of OECs toward VEGF or SDF-1α was examined. For both SDF-1α and VEGF-induced migration and 6-O-sulfation levels within HS, a strong, positive correlation between these two variables was evident; with higher levels of 6-O-sulfation of HS associated with higher levels of cell migration toward these chemotactic agents (Table S2 and Fig. S8) (SDF-1α; r = 0.537, P = 0.048 and VEGF; r = 0.767, P < 0.01). Next, the question as to whether the changes in OEC migration were associated with changes in the expression of specific 6-O-sulfated disaccharides within the HS chains, rather than simply an overall change in the extent of 6-O-sulfation within HS, was addressed (Table S2). A strong, positive correlation was found between the percentage expression of the disaccharide UA-[2S]GlcNS[6S] within HS chains on the cell surface and OEC migration toward both SDF-1α and VEGF (SDF-1α; r = 0.682, P = 0.007 and VEGF; r = 0.840, P < 0.01). As the levels of expression of this trisulfated disaccharide increase, so does the relative fold in migration of OECs toward VEGF and SDF-1α (Fig. 5). By contrast, the correlation between the expression levels of the disaccharides UA-GlcNS[6S] (Fig. S9) and UA-GlcNAc[6S] (Fig. S10) within OEC HS and both VEGF and SDF-1α-induced cell migration did not reach statistical significance.
To gain further insight into the role of HS in SDF-1α and VEGF-mediated chemotaxis of OECs, a combination of heparinase I and heparinase III was used to degrade HS on the cell surface and the effect on migration assessed. To verify that a substantial fraction of cell surface HS was removed by heparinase treatment, both treated and untreated cells were labelled with an anti-HS antibody and visualized with a fluorescently labelled secondary antibody (Fig. 6A). Figure 6B shows that heparinase treatment of peripheral blood derived OECs from young subjects significantly reduced their migratory response toward VEGF, such that levels of migration became similar to those observed by OECs isolated from older individuals; VEGF induced a 1.62 fold increase in migration of Hep treated OECs versus 2.07 fold increase in migration in untreated cells (P = 0.046). Consistent with the notion that changes in CXCR4 expression with age, rather than structural changes of HS, may account for the reduction in SDF-1α -induced migration of OECs from old subjects, heparinase pretreatment did not significantly affect OEC migration toward SDF.
It is well accepted that a characteristic of the aging process is the development of endothelial cell dysfunction, rendering the vasculature susceptible to the development of atherosclerosis and subsequent cardiovascular events. A deterioration of endogenous EPC function with age is thought to contribute to this increased risk of vascular disease. This study demonstrates that the migratory capacity of human OECs toward VEGF and SDF-1α significantly declines with increasing age. Furthermore, the reduction in the chemotactic response of OECs significantly correlates with structural alterations of HS. This is the first study to show a decline in the functional capacity of OECs during human aging and a correlation to cell surface glycosaminoglycan structure.
To determine whether or not OEC function changes with age, we investigated the proliferative, survival, migratory and tube-forming capacity of OECs isolated from cord blood and peripheral blood samples of young and old healthy subjects. In support of previous studies (Ingram et al., 2004), we show a significant difference in the proliferative capacity of cord blood cells and peripheral blood OECs. However, of note, no significant difference in proliferation was detected between OECs from young and old subjects. Also, under the conditions tested in this study, the survival and tube forming capacities of OECs did not significantly decline with age.
Of the parameters assessed, a diminished migratory capacity toward SDF-1α and VEGF was detected by peripheral blood derived OECs isolated from old subjects using in vitro transwell migration assays. Both VEGF and SDF-1α have been widely implicated in the mobilization of EPCs from the bone marrow and for guiding the homing of these cells to sites of injury or ischemia (Kalka et al., 2000; Abbott et al., 2004; Ceradini et al., 2004; Walter et al., 2005). Thus, the impaired migratory response detected in peripheral blood derived OECs from older individuals could limit the availability of these reparative cells at the site of injury, hindering repair and neovascularization. Indeed, support for the concept that age has a negative impact on the migratory capacity of OECs in vivo comes from a recently published study by Xia et al. The investigators demonstrated that OECs from elderly subjects had a significantly reduced capacity in vivo to promote re-endothelialization of injured arteries after transplantation into nude mice with carotid artery denudation injury, as compared to OECs from young subjects. Upon examination of the injured tissue significantly fewer OECs were detected within the site of injury, suggesting that OECs from elderly subjects have a reduced capacity to home to the target site and consequently reendothelialization capacity (Xia et al., 2012).
Heparan sulfate proteoglycans are known to participate in an array of cellular processes, such as cell migration, due to their ability to modulate the physiological activities of a wide range of protein ligands including SDF-1α and VEGF. Numerous studies have investigated the structural features of HS that are required for the binding of VEGF (Ono et al., 1999; Robinson et al., 2006; Zhao et al., 2012). These studies suggest that that both N- and 6-O-sulfate groups within HS are important for the binding of VEGF, whereas 2-O-sulfate groups contribute to a lesser extent for binding, but maybe required for signalling. Much less is known regarding the structural features of HS required for efficient binding of SDF-1α, but it has been suggested that both N- and 2-O-sulfate groups are important (Sadir et al., 2001). Given that both VEGF and SDF-1α appear to require specific structural features of HS for binding, it is possible that any change in the fine structure of HS during the aging process could alter the binding of these proteins and, subsequently, their ability to promote cell migration.
Age-associated structural changes of HS on the surface of OECs were evident by disaccharide analysis, which revealed a significant reduction in the percentage expression of the trisulfated UA[2S]-GlcNS[6S] disaccharide unit within OEC HS polysaccharide chains. Furthermore, the levels of SDF-1α and VEGF induced migration of OECs significantly correlated with levels of 6-O-sulfation within HS; with higher levels of 6-O-sulfation of HS, particularly the abundance of ΔUA[2S]-GlcNS[6S] residues, associated with higher levels of cell migration toward these chemotactic agents. This data line indicates that 6-O-sulfate groups within OEC HS may be important for SDF-1α and VEGF-binding and this interaction could subsequently mediate the physiological activities of these ligands to promote cell migration.
To verify the role of HS in the directional migration of OECs toward SDF-1α and VEGF a substantial amount of cell surface HS was enzymatically removed and the effect on migration assessed. A role for HS in the migratory response of OECs toward VEGF was supported by the significant reduction in cell migration toward this chemotactic agent following HSPG disruption. By contrast, the migratory response of OECstoward SDF-1α was not significantly altered following HSPG disruption. It is possible that HS-independent mechanisms of cell migration are altered during aging that contribute to a greater extent to the decline in the migratory response of OECs toward SDF-1α. For example, we observed a decline in the expression of CXCR4 on the surface of OECs with age which could be the major contributor factor to the decline in SDF-1α-induced migration of OECs isolated from old subjects.
The binding of VEGF by HS may modulate the biological activity of this protein to promote cell migration in several ways. Firstly, the interaction of HS with VEGF may enhance its interaction with VEGFR-2 to initiate signalling pathways such as MAPK/ERK(1/2) and PI3K/Akt cascades that are known to be involved in endothelial cell migration. Alternatively, the interaction of HS with VEGF may immobilize the protein at the cell surface to enhance the probability of receptor binding (Ruhrberg et al., 2002). In other cases, the HS-VEGF interaction may prolong VEGF activity by sequestration away from degradatory pathways increasing its availability for VEGFR-2 binding (Soker et al., 1993; Gengrinovitch et al., 1999). Given the importance of 6-O sulfate groups within HS for VEGF-binding and activity, it is possible that the decline in the abundance of these residues within OEC HS during the aging process perturbs the ability of HS to bind VEGF and, subsequently, the activity of this protein in mediating cell migration. The structural changes of HS detected during aging may reflect an age-dependent alteration of the enzymes regulating the biosynthesis of HS. Structures containing 6-O-sulfate groups in HS are regulated during the process of biosynthesis by sulfotransferases, and also at the post biosynthesis stage through the action of sulfatase (Sulf)-1 and Sulf-2. We found a trend for an increase in expression of both Sulf-1 and Sulf-2 with age (Fig. S11). However, further work is required to determine how the age-associated changes in 6-O-sulfation detected within OEC HS translate into an alteration of the enzymatic machinery regulating HS biosynthesis.
A great deal of research is aimed at harnessing the regenerative capacity of EPCs to treat CVD, yet results to date have yielded inconsistent outcomes in terms of therapeutic benefit. Among the challenges associated with the use of these cells as a therapeutic tool to promote neovascularization and vascular repair include problems with the retention and viability of these cells at the site of injection. Understanding the role of HS modification on OECs with age could help in the design of more successful strategies to optimize the therapeutic properties of these cells. Heparan sulfate is able to bind and immobilize and/or activate a diverse array of regulatory factors including growth factors and cytokines, which may modulate EPC homing. Recently, a Star PEG-heparin hydrogel system has been successfully used to generate a localized SDF-1α gradient that was capable of attracting early EPCs and subsequently improving neovascularization (Prokoph et al., 2012). Thus, there is great interest in this type of approach, as well as in the use of HS mimetics (Frescaline et al., 2012), to create a permissive microenvironment to recruit stem and progenitor cells to the site of injury/ischemia to promote repair and regeneration.
Outgrowth endothelial cell isolation and in vitro culture
Peripheral venous blood samples (50 mL) were collected in ethylenediaminetetraacetic acid (EDTA)-containing tubes from (n = 13) healthy subjects; non smokers, who were not receiving medication for any clinical diagnosis. All protocols had been approved by The Institutional Review Board at the University of Manchester, and informed consent was obtained from all volunteers [Ref No. (10/H1011.21)].
Outgrowth endothelial cells were isolated as previously described (Ingram et al., 2004), with minor modifications. Full details in the online Suppl. Outgrowth endothelial cells at early passages (4–6) were used for all experiments.
Cell migration assay
The migratory capacity of OECs was determined using an in vitro scratch assay in which the cells migrate into an area that has been mechanically denuded of cells. Outgrowth endothelial cells were seeded at a density of 20 000 cells per well of a 24 well tissue culture plate precoated with collagen I and allowed to grow to confluence. Following an overnight incubation in 1% FBS, a scratch wound was inflicted using a p200 pipette. Cell monolayers were then washed twice and fresh EGM-2 medium was added. The wounds were observed using time-lapse microscopy on an inverted microscope (Leica AS MDW) and images were taken at regular intervals over the course of 0–24 h. Migration was determined by calculating the percentage of area closure, as measured by the pre migration and migration time points. Images were processed using ImageJ software.
Transwell migration assay
Migration toward a chemotatic gradient was carried out in a 24-well, 8 μm pore collagen-coated transwell system (Corning Costar, Cambridge, MA, USA). Outgrowth endothelial cells at a density of 2 × 104 cells/mL in 100 μL of medium (endothelial basal medium containing 1% FBS) were placed in the upper chamber of the transwell assembly. The lower chamber contained 650 μL of medium with 100 ng/mL SDF-1α (Miltenyi Biotech) or 50 ng/mL VEGF. After incubation at 37°C and 5% CO2 for 6 h, the upper surface of the membrane was gently scraped to remove nonmigrating cells and washed with PBS. The membrane was then fixed in ice cold methanol for 15 min and stained in hematoxylin. The number of migrating cells was determined by counting five random fields of view per insert under the microscope at × 100 magnification. Migration experiments were always performed in duplicate.
Analysis of OEC-derived HS by reverse phase HPLC
Heparan sulfate was extracted from OECs as described previously (Chen et al., 2005). Full details in the online Suppl. AMAC-labelled HS disaccharides were loaded onto a Kinetex C18 HPLC column (Phenomenex, UK), equilibrated in 0.1 m ammonium acetate. Disaccharides were eluted from the column with a linear gradient of 5–15% acetonitrile at a flow rate of 1.0 mL/min. The eluate was detected by in-line fluorescence at excitation 488 and 520 nm emission. Disaccharides were identified by comparison to HS standards of known concentration (10 pmol) that were separated under the same conditions in sequential runs. Results were analyzed using ChemStation software.
Heparinase treatment of OECs
A combination of heparinase I and III was used to cleave HS on the OEC cell surface. Briefly, OECs were treated in suspension with 10 mIU/mL of heparinase I and III for 30 min at 37°C. Cells were then washed extensively with PBS, counted, and seeded into transwell assays as described. Cell migration was expressed as the fold increase relative to untreated cells (in buffer alone).
All data are expressed as mean values ± standard deviation (SD). Pair-wise comparisons were analyzed by Mann–Whitney U tests; P = <0.05 (2-tailed) was considered significant. Pearson's correlation coefficients were used for correlation analysis. All statistical analysis was performed using SPSS 16.0 software (IBM). Data are presented as mean values:errors bars are the standard error of the mean.
We thank Dr Mark Wareing Director of the University of Manchester Biobank for use of the cord blood cells (Approval No 08/H1010/55). Support from The Manchester Academic Health Science Centre is acknowledged. This study was supported by The Universities UK and The BBSRC Research Studentship awarded to KW. The authors disclose no potential conflicts of interest.