Developmental Biology 369 (2012) 330–339

Contents lists available at SciVerse ScienceDirect

Developmental Biology journal homepage: www.elsevier.com/locate/developmentalbiology

The glycocalyx is present as soon as blood flow is initiated and is required for normal vascular development Caitlin E. Henderson-Toth a, Espen D. Jahnsen a,b, Roya Jamarani a, Sarah Al-Roubaie a, Elizabeth A.V. Jones a,b,n a b

McGill University, Department of Chemical Engineering, 3610 University St, Montreal, QC, Canada H3A 2B2 Lady Davis Institute, McGill University, 3755 Cote Ste Catherine, Montreal, QC, Canada H3T 1E2

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 December 2011 Received in revised form 2 July 2012 Accepted 10 July 2012 Available online 20 July 2012

The glycocalyx, and the thicker endothelial surface layer (ESL), are necessary both for endothelial barrier function and for sensing mechanical forces in the adult. The goal of this study is to use a combination of imaging techniques to establish when the glycocalyx and endothelial surface layer form during embryonic development and to determine the biological significance of the glycocalyx layer during vascular development in quail embryos. Using transmission electron microscopy, we show that the glycocalyx layer is present as soon as blood flow starts (14 somites). The early endothelial glycocalyx (14 somites) lacks the distinct hair-like morphology that is present later in development (17 and 25 somites). The average thickness does not change significantly (14 somites, 182 nm 733 nm; 17 somites, 218 7 30 nm; 25 somites, 212 7 32 nm). The trapping of circulating fluorescent albumin was used to evaluate the development of the ESL. Trapped fluorescent albumin was first observed at 25 somites. In order to assess a functional role for the glycocalyx during development, we selectively degraded luminal glycosaminoglycans. Degradation of hyaluronan compromised endothelial barrier function and prevented vascular remodeling. Degradation of heparan sulfate down regulated the expression of shear-sensitive genes but does not inhibit vascular remodeling. Our findings show that the glycocalyx layer is present as soon as blood flow starts (14 somites). Selective degradations of major glycocalyx components were shown to inhibit normal vascular development, examined through morphology, vascular barrier function, and gene expression. & 2012 Elsevier Inc. All rights reserved.

Keywords: Glycocalyx Endothelial surface layer Mechanotransduction Vascular remodeling Vascular development

Introduction Endothelial cells form a barrier within blood vessels, between the circulating blood and surrounding tissues. In the adult, the luminal side of endothelial cells expresses a thin, gel-like layer called the glycocalyx. The glycocalyx is composed of membrane-bound proteoglycans and glycosaminoglycans (GAGs) such as hyaluronan (HA), chondroitin sulfate (CS), and heparan sulfate (HS), with terminal sialic acids (SA) (Pries et al., 2000). Blood proteins, such as albumin and superoxide dismutase, become entrapped in the glycocalyx and create a thicker layer known as the endothelial surface layer (ESL) (Becker et al., 1994; Pries et al., 2000; Shimada et al., 1991). The glycocalyx forms hair-like extensions into the vascular lumen that are anchored directly to actin stress fibers of the cytoskeleton and are believed to transmit shear stress, a force created by circulating blood, to the cytoskeleton (Florian et al.,

n Corresponding author at: Lady Davis Institute, McGill University, 3755 Cote Ste Catherine, Montreal, QC, Canada H3T 1E2. Fax: þ1 514 398 6678. E-mail address: [email protected] (E.A.V. Jones).

0012-1606/$ - see front matter & 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ydbio.2012.07.009

2003; Pahakis et al., 2007; Thi et al., 2004; Yao et al., 2007). The negatively charged glycocalyx also creates both a charge and a size barrier for protein diffusion to the endothelial cell surface and is vital for endothelial barrier function (Florian et al., 2003; Vink and Duling, 2000). Selective degradation of the adult glycocalyx has been shown to increase vascular permeability (Henry and DeFouw, 1995), cause attenuated nitrous oxide availability (Mochizuki et al., 2003), and increase adhesion of leukocytes and platelets to the endothelium (Constantinescu et al., 2003). Thinning of the glycocalyx has been linked to diseases such as diabetes mellitus, where dysfunction of the glycocalyx is believed to contribute to an increase in vascular permeability (Nieuwdorp et al., 2006a). Though the glycocalyx plays a significant role in adult vascular homeostasis, the role of the glycocalyx during vascular development has not yet been investigated. During embryonic development the vasculature forms as an immature capillary plexus that requires blood flow to mature (Chapman, 1918; le Noble et al., 2004; Lucitti et al., 2007). Mechanical forces, such as shear stress, must be present for vascular development to occur normally (le Noble et al., 2004; Lucitti et al., 2007). As the heart begins to beat, vessel permeability plays a large role in the type of

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

flow that is present. Since liquids flow in the direction of least resistance, if no barrier to flow into the interstitium is present, the blood plasma will flow into surrounding tissues and reduce the flow in the blood vessels themselves. Therefore, since the glycocalyx is important in the adult both for permeability and for mechanotransduction, it seems likely that the glycocalyx plays a significant role in vascular development. The stage at which both the glycocalyx and the ESL form during vascular development is not known. Many of the proteins that produce the GAGs of the glycocalyx are expressed before the onset of blood flow (Klewer et al., 2006). Lectins are proteins that bind to specific disaccharides in glycosaminoglycan chains (Barker et al., 2004; Florian et al., 2003; Mulivor and Lipowsky, 2004). Lectin binding affinities change during early vascular development (Henry and DeFouw, 1995) and therefore, there is indirect evidence that the composition of the glycocalyx is variable during development. Beyond this, not much is known about the presence of the glycocalyx in the embryo. Even if the glycocalyx and the ESL are present before the onset of circulation, blood flow itself is likely to modify both. Proteins that produce the GAGs of the glycocalyx, such as hyaluronan synthases, are upregulated when endothelial cells are exposed to flow in vitro (Gouverneur et al., 2006). The ESL, on the other hand, is created by the incorporation of flowing plasma protein and therefore the onset of circulation could alter the composition and structure of this layer. In this work, we have investigated the formation and role of the glycocalyx and the ESL during vascular development using the quail embryo as our model system. Transmission electron microscopy (TEM) was used to visualize the glycocalyx proper. A fluorescent albumin method was used to visualize the ESL. We found that the glycocalyx is present when blood flow begins. Large protrusions of glycosaminoglycans are present early in development, though these often follow irregularities in the endothelial cell membrane. By 25 somites, the glycocasminoglycan structure shows a regular thickness and distinct hair-like structure similar to the adult. We could detect the presence of an ESL starting at 25 somites. By immunohistochemistry, we investigated the presence of HA, CS, HS, and SA in the glycocalyx during vascular development. We then used intra-vascular injection of bead-immobilized enzymes to selectively degrade the luminal GAGs that make up the glycocalyx. We looked at the effect of glycocalyx degradation on vascular network morphology, permeability, gene expression and smooth muscle cell recruitment.

Materials & methods Transmission electron microscopy Fertilized quail eggs (Coturnix japonica) were incubated at 37 1C and approximately 60% humidity for 40–72 h as noted. Embryos were fixed for TEM using a modified version of a previously published protocol (van den Berg et al., 2003). For more information, please see supplemental materials. Samples were imaged on a FEI Tecnai 12 120 kV equipped with an AMT XR80C CCD Camera. Because the endothelial lining was irregular, the thickness of the glycocalyx was calculated by measuring the area of the glycocalyx divided by the length of the endothelium in regions that lacked large protrusions. It was necessary to ignore these large protrusions since they overstated the thickness covering most of the vessel. A protrusion was defined as a structure greater than twice the average thickness. Ex ovo fluorescent albumin method Embryos were placed in culture using a modified version of a previously published ex ovo method (Krull and Kulesa, 1998).

331

Acetylated low-density lipoprotein (AcLDL, Invitrogen L35353) was injected directly into the embryonic heart and allowed to circulate for 15 min to label endothelial cells. AcLDL is taken up by scavenger receptors expressed on endothelial cells (Adachi et al., 1997). After AcLDL labeling, fluorescently-tagged albumin from bovine serum (AlexaFluor 488; Invitrogen A13100) was injected directly into the embryonic heart. Embryos were returned to the incubator for one hour to allow the albumin time to incorporate into the ESL. Embryos were imaged on a heated confocal microscope. Due to the lack of growth media, embryos were imaged for only 1–3 h. Immunohistochemistry For whole-mount, embryos were dissected and fixed in 4% paraformaldehyde solution. Embryos were blocked and incubated with primary antibodies. Endothelial cell were labeled with QH1 (Developmental Studies Hybridoma Bank, 1:100). The heparan sulfate proteoglycan perlecan was labeled with a-perlecan, (Developmental Studies Hybridoma Bank, 5C9, 1:100). Smooth muscle cells were labeled with a-SMA-Cy3 (clone 1A4, Sigma Aldrich; 1:100). Replicating cells were labeled with a-PH3 (New England Biolabs, 1:100). Fluorescent secondary antibodies were purchased from Invitrogen (AlexaFluor 488; 1:400). For sections, embedded embryos were cut into 7–10 mm sections and antigen retrieval was performed using citrate buffer unmasking solution (Vector Labs). Biotinylated primary antibodies against hyaluronan (HABP, EMD Chemicals, 1:100) and sialic acid (Flavus Limax, EY Laboratories, 1:100) were used and detected with Cy3-Streptavidin (Sigma, 1:400). For more information, please see supplemental materials. Analysis of blood flow Fluospheres (1 mm, Invitrogen) were injected into circulation using a picospritzer and a fine pulled needle. Embryos were transferred to a fluorescent microscope equipped with an Ultima APX-RS camera and the arterial region of the yolk sac was imaged for 1000 frames at 500 fps. From the series of 1000 images, the period of systole was assessed visually and four frames were extracted from the observed peak velocities. Several individual vessel segments of approximately equal diameter were cropped out of the larger images. Using uraPIV, an open source MatLab software developed by Roi Gurka, the speckle pattern in the vessel segments was analyzed for all image pairs. The interrogation window size and spacing was optimized for the vessel segment and based on a laminar flow profile the data from the best image pair was selected. Only peak velocities in the center of the vessel are reported. Enzymatic degradation Embryos were incubated until they reached 14 somites (approximately 40 h). A small window was cut into the top of the eggshell to expose the embryo. Embryos were injected intravascularly using a pulled quartz needle filled with either enzyme in solution or enzymes immobilized on beads. Enzyme immobilization was achieved using sulfated Fluospheres (Invitrogen). Immobilization was necessary for longer exposures to the enzymes since proteins in solution could leach out of the blood vessels. The sulfated Fluorospheres were mixed with concentrated enzyme solution for 8 h. The suspension was then centrifuged at 3000 rpm for 10 min, washed, re-centrifuged and resuspended in 40,000 MW Texas Red dextran. After injection, eggs were sealed and returned to the incubator for 14–18 h. All enzymes were purchased from Sigma (Hyaluronidase, H1136;

332

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

Chondroitinase ABC, C2905; Heparinase III, H8891; Neuraminidase Type V, N2876). Albumin was immobilized on beads as a control. AnnexinV analysis Apoptosis was analyzed after embryos had been exposed to bead-immobilized enzymes for 18 h using AnnexinV staining. AnnexinV binds phosphatidylserine which is expressed on the cell surface of apoptotic cells (Koopman et al., 1994). Embryos were injected intra-vascularly with AnnexinV-AlexaFluor555 (Invitrogen) and returned to the incubator for 30 min. Embryos were dissected and fixed in 4% paraformaldehyde solution. Immunohistochemistry was performed against QH1 that detects quail endothelial cells (Pardanaud et al., 1987). Analysis of vascular permeability Embryos were injected in ovo with Ringer’s (control) and enzyme solution and returned to the incubator. After one hour of incubation, embryos were injected with FITC-dextran (Sigma, FD40S), and observed every hour until vessels were indistinguishable from avascular regions. For more information, please see supplemental materials. RNA extraction and gene expression analysis Quail embryos were injected with immobilized enzyme at 14 somites and dissected at 21–26 somites. Three embryos were combined per sample (n ¼6–9 samples, for all treatments), and

Fig. 1. The Glycocalyx is present as soon as blood flow initiates. Transmission electron microscopy of the endothelial glycocalyx in the dorsal aorta of 14 somite quail embryos (A). At this stage, the thickness is variable and protrusions of glycosaminoglycans are visible. Fewer protrusions are present at 17 somites and the structure of the glycosaminoglycans displays a regular hair-like structure (B). By 25 somites, these protrusions are absent and a very regular hair-like structure is present (C). Scale bar 500 nm.

treated with collagenase to acquire a single cell suspension. QH1 conjugated Dynabeads (Invitrogen) were used to isolate endothelial cells. Primers for qPCR were designed based on regions of high sequence conservation between mouse, rat, human and chicken, using sequence information from the chicken. For more information, please see supplemental materials.

Results The glycocalyx and endothelial surface layer are present from the onset of blood flow We investigated the formation of the glycocalyx of the dorsal aorta by TEM at 14 somites (onset of blood flow), 17 somites (early vascular remodeling) and 25 somites (late vascular remodeling). At 14 somites, we observed that the glycocalyx is present, although the thickness was not uniform (average value¼ 182 nm722 nm, n¼3 embryos) and luminal protrusions were observed (Fig. 1A). Although the glycocalyx is present at this stage, the structure of the glycosaminoglycans did not show a regular hair-like pattern as has been previously observed for adult blood vessels in frogs and rats (Squire et al., 2001; van den Berg et al., 2003). At 17 somites, the glycocalyx was similar in thickness that of the 14-somite embryo (average value¼218730 nm, n¼3), but fewer protrusions into the vascular lumen were present and a hair-like morphology was present (Fig. 1B). At 25 somites, no differences in glycocalyx thickness were observed when compared to younger embryos and luminal protrusions were no longer present (Fig. 1C, average value¼ 212732 nm, n¼3). The ESL is believed to consist largely of entrapped of proteins, particularly albumin (Pries et al., 1998). We therefore injected fluorescently-tagged albumin into the extra-embryonic vasculature of quail embryos to image luminal immobilized albumin (Fig. 2). Extra-embryonic vessels were imaged since the dorsal aorta is too

Fig. 2. The Endothelial Surface Layer (ESL) is visible at 25 and 35 Somites. The endothelial surface layer, in vessels of the extra-embryonic yolk sac of a living embryo, was imaged using fluorescently conjugated albumin (green). The ESL is visible by this method at 25 (A) and 35 somites (B). Endothelial cells lining vessel wall were labeled with AF555-AcLDL (red, C & D). Large patch of immobilized albumin (green) are present extending into the vessel lumen from the endothelial cell surface (red). Scale bar 50 mm.

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

deep in the tissue for confocal microscopy. Endothelial cells also took up fluorescent albumin so we co-labeled using AcLDL (which labels the endothelial wall) to differentiate between albumin inside endothelial cells and luminally-trapped albumin. At 14 and 17 somites, the fluorescent albumin did not indicate the presence of the ESL layer rather it had become incorporated into the endothelial cells (data not shown, n¼6). At 25 and 35 somites, it was possible to visually identify patches of fluorescence in the lumen of vessels associated with the endothelial lining but entrapped albumin was always observed in patches and not as a continuous layer (arrows, n¼6, Fig. 2A, B). Using immunohistochemistry, we investigated the presence of the most abundant components of the glycocalyx (HA, CS, HS and SA) during vascular development (Fig. 3). We stained whole mount and tissue sections for these components at 14, 17, 25, and 35 somites of development using immunohistochemistry and lectin-labeling of extra-embryonic vessels (n¼3 for each stage and each component). HA detection showed positive staining at all stages investigated. For 14 and 17 somite embryos, the intensity of the staining around blood vessels was the same as surrounding tissue (Fig. 3A–B). By 25 somites, the blood vessel wall was stained more strongly for HA than the surrounding tissues (Fig. 3C, arrows). The strong vascular staining

333

was only present on a subset of blood vessels, which is clearly visible in the image of the embryo after 4 days of incubation (435 somites Fig. 3D). The staining for CS revealed this component’s presence between 14 somites and 35 somites (Fig. 3E–H). However, the CS is difficult to visualize by immunohistochemistry, and does not appear to present any developmental pattern over the stages we examined. Perlecan is a heparin sulfate proteoglycan that is localized to the basement membrane of endothelial cells (Dziadek et al., 1985) but that is also secreted into the ESL (Haraldsson et al., 2008). The perlecan staining was performed in whole-mount and showed a vascular specific staining during development in extra-embryonic vessels (Fig. 3I–L). Thickened regions of perlecan were observed at 17 and 25 somites, during vascular remodeling (Fig. 3J–K). We stained for the presence of SA using a lectin from Limax flavus, which is specific for this GAG (Miller, 1982). Though staining was present at all stages observed, the intensity was not any stronger near blood vessels than in other parts of the tissue (Fig. 3M–P). Degradation of the glycocalyx affects vascular morphology In order to help identify the functional role of the glycocalyx during vascular development, we degraded individual components of

Fig. 3. Immunohistochemical staining of components of the glycocalyx during development. The extra-embryonic yolk sac was stained for HA (A–D, sections), CS (E–H, sections), a HS called perlecan (I–L, whole-mount) and SA (M–P, sections) at different developmental stages. Hyaluronan was present at all stages but the intensity was the same for vascular and non-vascular tissue at 14 and 17 somites (A–B). Stronger perivascular signal was present at 25 and 35 somites for some vessels (arrows, C–D). L¼lumen. Scale bars 100 mm.

334

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

this structure via enzyme-mediated degradation in vivo, then evaluated changes in vascular development. First, Hyaluronidase (Hyal) from S. hyalurolyticus was used to degrade HA. This enzyme is specific to HA, and unlike other hyaluronidases, does not degrade other GAGs (Ohya and Kaneko, 1970). Chondroitinase ABC (ChABC) was used to degrade both CS and dermatan sulfate. Heparinase III (HepIII) was used to specifically degrade HS, and Neuraminidase Type V (NeurV) was used to degrade SA. Ringer’s saline solution was injected as a control. All enzymes were tested for specificity in vitro prior to use in the embryo. We tested whether any of the four other major components were degraded at concentrations that degrade the target GAG. The enzymes were applied to a monolayer of endothelial cells, and then immunohistochemical analyses of all four major components of the glycocalyx were performed postdegradation to investigate the enzymatic effects (n ¼3 for each enzyme, please see supplemental materials Figs. 1–4). Though we found some enzymes, such as Hyal, only caused a partial degradation of their target component (please see supplemental materials Fig. 1E), we did not observe any non-specific enzymatic degradation of the other major components. We then explored whether gross cardiovascular function was affected by enzymatic degradation of the glycocalyx. We injected enzyme solutions into circulation and then injecting fluorescent microspheres one hour later. We imaged the movement of these spheres with a high-speed microscope camera (500 fps) and used micro-Particle Image Velocimetry to calculate blood flow velocities in the center of blood vessels at peak systole (n¼ 3 per enzyme, see Methods). Blood flow in Hyal-treated embryos was significantly reduced as compared to control embryos (Fig. 4, Supplemental materials Movie 1–2). Injection of ChABC solution (1 U/mL) also caused a decrease in the blood flow velocity (Fig. 4, Movie 3). Furthermore, the heart of ChABC injected embryos stop beating abruptly 1–2 h after injection of the enzyme solution. Blood flow in both the HepIII injected embryos (Movie 4) and the NeurV injected embryos (Movie 5) was similar to control injected embryos.

Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.ydbio.2012.07.009. In order to observe changes in the vascular network morphology, embryos were exposed to vascular-specific enzyme degradation for 16 h and stained with QH1, an antibody that labels quail endothelial cells (Pardanaud et al., 1987) (Fig. 5A–E). Many components of the glycocalyx are also present in the extracellular matrix (ECM) and we were concerned that over a period of 16 h, the enzymes might leach from the blood vessels and affect GAGs in the ECM. To circumvent this problem, we immobilized the proteins on 1 mm fluorescent microspheres before injection. This also allowed us to confirm by fluorescence that the beads remained in circulation for the entire treatment period. We verified that GAGs in the ECM were not degraded by immunohistochemistry on sections (data not shown). Bead injection did not hinder normal vascular remodeling in control embryos (n ¼3, Fig. 5A). Degradation of HA resulted in hyperfused vasculature in which hierarchical branching was absent (Fig. 5B, n¼7). In some embryos (n ¼4 of 7), the vitelline artery formed but its growth was stunted and the artery did not extend far beyond the embryo proper. Unlike injection of ChABC in solution, immobilized-ChABC did not cause the heart to stop beating. ChABCtreated embryos remodeled (Fig. 5C, n ¼6), however a lower vascular density was observed near the large vitelline artery in some embryos (arrow, n ¼3 of 6). HepIII treated embryos underwent vascular remodeling (Fig. 4D, n¼3) as did the embryos treated with immobilized NeurV (Fig. 5E, n ¼4). We looked whether endothelial proliferation and endothelial apoptosis were affected by enzymatic degradation. The number of replicating cells per mm2 was assessed by co-staining for QH1 and for phospho-histone 3, which marks cells during mitosis. No statistically significant difference in endothelial replication was found in any of the enzymatic degradations as compared to control (please see supplemental materials). The number of apoptotic cells was analyzed by injection of AnnexinV, a protein that binds to phosphatidylserine, which is expressed on the surface of apoptotic cells. There was no significant difference between the amount of cell death in control injected and experimental embryos (please see supplemental materials Fig. 5B). Vascular remodeling is accompanied with the recruitment of smooth muscle cells to larger arteries and does not occur if blood flow is stopped. The ability of endothelial cells to recruit smooth muscle cells was examined after enzymatic degradation using the bead-immobilized enzymes. With all treatments, we found no effect on smooth muscle cell recruitment (please see Supplemental Fig. 6). Degradation of hyaluronan causes an increase in vascular permeability

Fig. 4. Blood flow velocity is reduced after injection of hyaluronidase or chondroitinase. Embryos were injected with concentrated enzyme solution (or albumin for control) and then injected with fluorescent microspheres one hour later. Flow was imaged using a high-speed fluorescent microscopy (please see supplemental movies) and the velocity at the center of vessel segments during systole was analyzed using micro-particle image velocimetry. Both Hyal injection and ChABC injection caused a decrease in blood flow velocity as compared to control. Within 1–2 h of ChABC injection, the embryo’s heart stopped beating. HepIII and NeurV injection had no effect on blood flow velocity. *p o 0.05.

One of the most important roles of the glycocalyx in the adult vasculature is to control vascular permeability (Vink and Duling, 2000). Embryonic vessels are often described as ‘‘leaky’’. We therefore sought to investigate the role of the glycocalyx in controlling permeability in embryonic vessels. To establish the stage at which endothelial barrier function could first be observed, quail embryos between 14 and 26 somites were injected in ovo with FITC-40K neutral dextran. Embryos were observed each hour in order to assess the time required for the dextran to leach out from the vessels (Fig. 6A). The endpoint was chosen as the time at which vessels could no longer be differentiated from avascular tissues by fluorescence. At 14 somites, dextran leached out of the blood vessels within 1–2 h. The time required for the dye to leach from the vasculature increased until embryos reached 16 somites, after which dextran took around 6 h to leach out of vessels regardless of the stages observed.

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

335

Fig. 5. Vessel branching is altered by selective degradation of the glycocalyx. Whole-mount embryos were stained with QH1 antibody following 16-hour exposure to bead-immobilized enzymes (A–E). Hyal treatment induced regions of hyperfused vessels and stunted the growth of the vitelline artery (arrow, B). Injection of beadimmobilized ChABC resulted in a decrease in vascular density near the vitelline artery (arrow, C). Injection of HepIII and NeurV did not inhibit vascular remodeling (D, E). Images were obtained by capturing 40–60 images of the extra-embryonic yolk sac on a confocal microscope and reassembled using Adobe Photoshop’s photomerge function. The beads used for protein immobilization are fluorescent and create a speckle pattern in the blood vessels. Scale bar 1 mm.

Fig. 6. Hyaluronidase is the main component which controls permeability. Diagram detailing hours required for FITC-dextran to leach from embryonic vessels at different somite stages during vascular development (A). Graph detailing the hours required for leaching of FITC-dextran in embryos after enzymatic degradation (B). Dextran-injected embryos initially (C,E) and after 3 h (D, F) for control and Hyal embryos. Treatment with Hyal (50 U/mL) caused a decrease in the time required for dye to leach out. ChABC treatment (0.001 U/mL) caused a small, but statistically significant increase, in the time required for dye to leach out. ***p o 0.005. Scale bar 1 mm.

336

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

To investigate the role of the various components in establishing vascular barrier function, we examined the time required for dextran to leach from vessels after each of the enzymatic solution degradations. Quail embryos were injected at 19–20 somites since a stable level of vascular barrier was present by this stage (Fig. 6A). As expected from our analysis of permeability changes with somite stage, FITC-dextran needed an average of 6 h to leach out of vessels in control-injected embryos (n¼13, Fig. 6B). In embryos treated with Hyal at 1 U/mL, FITC-dextran leached out much more quickly and was no longer visible in the blood vessels within 3 h post-injection (n¼ 3, Fig. 6E–F). At this time, control embryos still retained the dye within the vasculature (Fig. 6D). For ChABC, we performed a serial dilution to find a concentration that did not stop the heart from beating. Injection of ChABC at 0.001 U/mL resulted in a small, but statistically significant, increase in the time required for FITC-dextran to leach out (n ¼7, 8 h versus 6 h for control). Injections of HepIII and of NeurV did not cause statistically significant differences in leaching times (n ¼6, for both). Therefore, only degradation of hyaluronan caused an increase in vascular permeability.

Endothelial gene expression following selective degradation of the glycocalyx Because of the reported role of the glycocalyx in mechanotransduction, we investigated whether the expression of genes known to be affected by shear stress were altered in embryos subjected to enzymatic degradation (Fig. 7A). We also investigated whether genes known to be involved in angiogenesis were affected since we observed defects in vascular branching patterns (Fig. 7B). We injected the immobilized enzymes at the onset of flow (14 somites) and isolated endothelial cells 16 h later using magnetic beads (Dynabeads) coated with QH1, an antibody specific to quail endothelial cells and macrophages. Dynabeads are able to isolate cell types to 95% purity (Gomm et al., 1995). Though the population will not be purely endothelial, the isolated cells are highly enriched for endothelial cells. Genes used to assess changes in mechanotransduction included: endothelial ¨ nitrous oxide synthase (nos3), kruppel-like factor 2 (klf2), nuclear factor kB (nfkb1), platelet derived growth factor b (pdgfb), and endothelin-1 (edn1). Genes used to analyze effects on angiogenesis included: uncoordinated 5b (unc5b), vascular endothelial growth factor (vegfa), delta-like 4 (dll4), and hif1a (hif1a). Vegfa and hif1a were chosen since they are central to vascular health. Dll4 is expressed by sprouting endothelial cells and is a marker of angiogenesis (Hellstrom et al., 1999; Lobov et al., 2007; Suchting et al., 2007). Unc5b, conversely, inhibits angiogenesis and sprouting (Larrivee et al., 2007). A subset of mechanotransduced genes was affected by HepIII and Hyal degradation (Fig. 7A), whereas the other treatments had no effect on these mechanotransduced genes. For instance, HepIII treatment caused a reduction in nos3 and nfkb1 expression but did not affect the expression of klf2, pdgfb or edn1. Hyal treatment affected only the expression of edn1. While this down-regulation was statistically significant, the magnitude was never larger than 2-fold. The expression of genes involved in angiogenesis was altered by three of the enzymatic treatments with immobilized enzymes (Fig. 7B). Degradation of CS induced a reduction in the expression of both unc5B and vegfa. Injection of immobilized HepIII resulted in a 2-fold decrease in expression of vegfa by endothelial cells. NeurV treatment also decreased unc5b expression. While a decrease in unc5b expression was also observed with Hyal and HepIII, this decrease was not statistically significant.

Fig. 7. Expression of mechanotransduced and angiogenesis genes after degradation of the glycocalyx. Endothelial cell gene expression for shear-induced genes (A) nos3, klf2, nfkb1, pdgfb, and edn1, and for genes linked to angiogenesis (B) unc5b, vegfa, dll4 and hif1a after enzymatic degradation. Both nos3 and nfkb1 expression were significantly downregulated with injection of bead-immobilized HepIII. Injection of immobilized Hyal caused a reduction in expression of edn1. Unc5B was downregulated by ChABC and NeurV treatment. Vegfa expression was downregulated by ChABC and HepIII treatment. Gene expression normalized to two housekeeping genes (hprt, actb). *p o 0.05, ***p o 0.005.

Discussion The importance of the glycocalyx in adult vasculature has received a significant amount of attention, yet little is known about its role during vascular development. It has been shown that in the adult the glycocalyx is essential for maintaining normal endothelial barrier function and for transduction of the mechanical signals produced by blood flow (Florian et al., 2003; Hecker et al., 1993; van den Berg et al., 2003; Vink and Duling, 2000). Interestingly, both of these functions are essential for proper vascular development (Lucitti et al., 2007; Madri et al., 2003). Our results establish that the glycocalyx is present in the quail embryo as soon as blood flow begins at 14 somites. We find that the early glycocalyx does not show a regular hair-like structure as is seen later in development. Protrusions of glycosaminoglycans into the vascular lumen are evident at 14 and 17 somites. At these stages, however, the surface of endothelial cells is also irregular and protrusions are often associated with irregularities in the endothelial cell surface. The structure of the glycocalyx becomes more uniform as vascular remodeling proceeds, and fewer protrusions are visible (25 somites). The average thickness does not change significantly during development.

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

We were unable to detect the ESL by imaging immobilized luminal albumin in the developing vasculature when blood flow first starts. The entrapped albumin was present in clumps at 25 and 35 somites. We observed entrapped masses of albumin as thick as 10–12 mm, which is significantly thicker than ever observed in the adult vasculature (Pries et al., 2000). At this point, we do not know whether thicker regions of the ESL have a functional purpose. When we time-lapsed embryos co-labeled with fluorescent albumin and AcLDL, the thickened regions did not correspond to any specific morphological events, such as angiogenic sprouting or increasing vessel diameter (data not shown). The glycocalyx is composed of a large array of GAGs. For the purpose of this study we chose to test four of the most prevalent components in the adult glycocalyx (Reitsma et al., 2007). While it is possible that the glycocalyx composition differs between development and the adult, this possibility is outside the scope of this study. The glycocalyx is known to shed GAGs under normal circumstances, and blood naturally contains components that damage and degrade parts of the glycocalyx. GAGs of the glycocalyx are experiencing constant degradation and synthesis (Mulivor and Lipowsky, 2004). By keeping fluorosphere beads with enzyme in circulation until just prior to dissection, we ensure that partial degradation of the glycocalyx component occurs continuously throughout the incubation period. We believe, however, that our treatments lead to only a partial degradation of the glycocalyx. ChABC in solution caused a decrease in the velocity of blood flow and caused the heart to stop beating. When ChABC is delivered immobilized on beads to ensure the enzyme is confined to the luminal side of the vasculature, we observed a healthy heartbeat and vascular remodeling occurred. This might indicate that the CS component in the ECM of the heart is required for cardiac contraction. Because the glycocalyx and ESL are known to be important for mechanotransduction, we investigated whether degradation of the glycocalyx during vascular development affects the expression of mechanotransduced genes. Our results showed a two-fold reduction in nos3 expression after HepIII degradation. A 44% percent reduction in NO production after heparinase treatment has previously been reported for endothelial cells exposed to flow in culture (Pahakis et al., 2007). The authors also found that degradation of HA, SA and CS influenced NO production (Pahakis et al., 2007). Our work, conversely, did not show an effect on nos3 expression with any of the other enzymatic treatments. We did not measure eNOS phosphorylation or NO production and it has been reported that it is the level of phosphorylated eNOS and not the total levels of eNOS levels that are affected by glycocalyx degradation (Tarbell and Ebong, 2008). Though the expression of two important mechanotransduced genes (nos3 and nfkb1) was reduced with HepIII treatment, vascular remodeling was normal. We cannot rule out the possibility that HS was only partially degraded and that more significant defects would be present if all luminal HS was removed. The role of mechanotransduced genes during development is also controversial. Mouse embryos in which all three nitric oxide synthases are ablated are viable (Tsutsui et al., 2006) and undergo vascular remodeling. PECAM-1 is believed to be required for mechanotransduction (Tzima et al., 2005), however genetic ablation of this protein does not affect vascular development (Duncan et al., 1999). Therefore, our results are in agreement with the literature that embryonic development can still occur when certain aspects of mechanotransduction are ablated. Since mechanical signals from blood flow are required for vascular development (Lucitti et al., 2007), this may indicate that redundant pathways for mechanotransduction are present or that the effects by HepIII degradation in our system were too subtle to induce significant defects in remodeling. While it has been

337

previously established that HA in the ECM is involved in the migration and proliferation of smooth muscle cells, we did not observe any down regulation in shear-induced genes after treatment with hyaluronidase. Therefore it is not surprising that shear dependent recruitment of smooth muscle cells occurred normally. Because we observed changes in the branching pattern of blood vessels (Fig. 5B, C), we examined whether genes involved in endothelial cell sprouting and guidance were affected by enzymatic degradation. HepIII is known to be anti-angiogenic (Van Sluis et al., 2010) and we found that degrading luminal HS caused a decrease in vegfa expression. This was not associated with any specific vascular defects in the HepIII treated embryos. Surprisingly, the gene most affected by degradation of the glycocalyx was unc5b, whose signaling is known to be inhibitory to angiogenesis (Larrivee et al., 2007). Only treatment with CS showed any reduction in vascular density. In the nervous system, CS is inhibitory to neuritic outgrowth and the ligand for Unc5B, Netrin (Leonardo et al., 1997), has been shown to bind cell surface HS and CS. Interestingly, these are the two treatments that led to a statistically significant decrease in unc5b expression. This raises the interesting possibility that CS plays a similar role in the cardiovascular system by guiding angiogenic sprouts, similar to its established role in the nervous system. It is not clear how luminal GAGs could affect endothelial sprouting into tissues. Netrin-1 is secreted into blood plasma (Ramesh et al., 2011) and luminal GAGs could sequester Netrin and prevent its entry into surrounding tissues (Shipp and Hsieh-Wilson, 2007). When we looked at the importance of the glycocalyx components in establishing vascular permeability we found that only Hyal treatment increased vascular permeability in the embryo. Interestingly, the appearance of vascular-specific staining for HA occurs between 17 and 25 somites (Fig. 3B–C), which coincides with the stage at which we see vascular barrier function established in the embryonic vasculature (Fig. 6A). Fluid shear stress has previously been shown to induce incorporation of HA into the glycocalyx (Gouverneur et al., 2006) and the vascular-specific HA staining in the embryo appears just shortly after the onset of blood flow. Degradation of HA in the adult vasculature increases vascular permeability, rendering the glycocalyx permeable to 70 and 145 kDa high-molecular weight dextrans, which are normally excluded from the glycocalyx (Henry and Duling, 1999). Degradation of the glycocalyx through perfusion of the enzyme pronase results in a 2.5  increase in the hydraulic permeability of blood vessels (Adamson, 1990). Our results confirm a vital role for HA in controlling the permeability of the embryonic vasculature. We find that dextrans leach out of blood vessels twice as fast after hyaluronidase treatment as compared to control. The degradation of hyaluronan also prevents vascular remodeling and results in a hyperfused vascular plexus. The branching patterns observed in the embryos treated with Hyal solution are similar to what occurs when flow is stopped (Chapman, 1918) (data not shown). Since Hyal treatment also affects blood flow, we cannot separate effects due to permeability from those due to decreased shear stress, though we believe the reduced flow is a secondary defect induced by increased permeability. We saw that treatment with ChABC caused an increase in time required for dye to leach out of blood vessels. We do not believe this increase is due to altered barrier function. Though the heartbeat was normal at the concentrations of ChABC used to test for permeability (0.001 U/mL), we cannot state that more subtle defects in blood flow were absent. Since transmural flow is a function of both hydrostatic pressure in the vessels and vessel permeability, we believe that blood pressure may have been slightly lowered, leading to a decrease in hydrostatic pressure and therefore a decrease in the transmural flowrate out of the vessel. Interestingly, this presents itself as a sensitive and simple

338

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

method to investigate the presence of mild defects in heart function during vascular development. We did not observe an increase in permeability due to the partial degradation of HS, which makes up the great majority of GAGs in the endothelial glycocalyx (Oohira et al., 1983). Though it is surprising that degradation of a major component of the glycocalyx does not affect permeability, these results are supported by findings in the adult whereby degradation of HS has no effect on permeability (Rehm et al., 2004). Several diseases have been linked to the thinning of the glycocalyx. Diabetes mellitus is a disease wherein high levels of oxygen radicals cause the glycocalyx to become thinned (Nieuwdorp et al., 2006b). In this disease, the thinned glycocalyx is also made hyperpermeable. Incubation of embryos in high glucose inhibits vascular remodeling in a manner similar to what we observed due to HA degradation (Fig. 5B) (Madri et al., 2003). Since HA determines barrier function in the embryo, it is likely that it is a component affected by maternal diabetes.

Conclusions Our results show the presence of the glycocalyx before the onset of blood flow. Using TEM microscopy, we also observe that the early glycocalyx (14 and 17 somites) shows significant protrusion into the vascular lumen though these protrusions are often accompanied by heterogeneity in the endothelial cell lining. At 14-somites, the distinct hair-like structure of the glycocalyx is absent. Additionally, we demonstrated the functionality of the individual glycocalyx components in the embryo through the use of selective enzymatic degradation. We have found that HA is the component mainly responsible for controlling barrier function and that degradation of this component causes a morphological effect similar to the effects of maternal diabetes. We also showed that while some mechanotransduction genes were down regulated by HepIII treatment, embryonic vascular development remained normal. These results indicate that the glycocalyx has a functional role in vascular development.

Sources of funding This work was supported by grants from the Sick Kids Foundation of Canada (NI12-029), NSERC Discovery Program (342134), and Canadian Foundation for Innovation and the Canada Research Chair Program. CHT was supported by the CIHR Banting and Best Master’s Award.

Acknowledgments A special thanks to the staff at the Facility for Electron Microscopy Research for their kind assistance with electron microscopy preparation and imaging. Additional thanks to Bahar Kasaai for her kind assistance.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.ydbio.2012.07.009.

References Adachi, H., Tsujimoto, M., Arai, H., Inoue, K., 1997. Expression cloning of a novel scavenger receptor from human endothelial cells. J. Biol. Chem. 272, 31217–31220.

Adamson, R.H., 1990. Permeability of frog mesenteric capillaries after partial pronase digestion of the endothelial glycocalyx. J. Physiol. 428, 1–13. Barker, A.L., Konopatskaya, O., Neal, C.R., Macpherson, J.V., Whatmore, J.L., Winlove, C.P., Unwin, P.R., Shore, A.C., 2004. Observation and characterisation of the glycocalyx of viable human endothelial cells using confocal laser scanning microscopy. Phys. Chem. Chem. Phys. 6, 1006–1011. Becker, M., Menger, M.D., Lehr, H.A., 1994. Heparin-released superoxide dismutase inhibits postischemic leukocyte adhesion to venular endothelium. Am. J. Physiol. 267, H925–H930. Chapman, W.B., 1918. The effect of the heart-beat upon the development of the vascular system in the chick. Am. J. Anat. 23, 175–203. Constantinescu, A.A., Vink, H., Spaan, J.A., 2003. Endothelial cell glycocalyx modulates immobilization of leukocytes at the endothelial surface. Arterioscler Thromb. Vasc. Biol. 23, 1541–1547. Duncan, G.S., Andrew, D.P., Takimoto, H., Kaufman, S.A., Yoshida, H., Spellberg, J., Luis de la Pompa, J., Elia, A., Wakeham, A., Karan-Tamir, B., Muller, W.A., Senaldi, G., Zukowski, M.M., Mak, T.W., 1999. Genetic evidence for functional redundancy of platelet/endothelial cell adhesion molecule-1 (PECAM-1): CD31-deficient mice reveal PECAM-1-dependent and PECAM-1-independent functions. J. Immunol. 162, 3022–3030. Dziadek, M., Fujiwara, S., Paulsson, M., Timpl, R., 1985. Immunological characterization of basement membrane types of heparan sulfate proteoglycan. Embo. J. 4, 905–912. Florian, J.A., Kosky, J.R., Ainslie, K., Pang, Z.Y., Dull, R.O., Tarbell, J.M., 2003. Heparan sulfate proteoglycan is a mechanosensor on endothelial cells. Circ. Res. 93, E136–E142. Gomm, J.J., Browne, P.J., Coope, R.C., Liu, Q.Y., Buluwela, L., Coombes, R.C., 1995. Isolation of pure populations of epithelial and myoepithelial cells from the normal human mammary gland using immunomagnetic separation with Dynabeads. Anal. Biochem. 226, 91–99. Gouverneur, M., Spaan, J.A.E., Pannekoek, H., Fontijn, R.D., Vink, H., 2006. Fluid shear stress stimulates incorporation of hyaluronan into endothelial cell glycocalyx. Am. J. Physiol. Heart Circ. Physiol. 290, H458–H462. Haraldsson, B., Nystrom, J., Deen, W.M., 2008. Properties of the glomerular barrier and mechanisms of proteinuria. Physiol. Rev. 88, 451–487. Hecker, M., Mulsch, A., Bassenge, E., Busse, R., 1993. Vasoconstriction and increased flow: two principal mechanisms of shear stress-dependent endothelial autacoid release. Am. J. Physiol. 265, H828–H833. Hellstrom, M., Kal n, M., Lindahl, P., Abramsson, A., Betsholtz, C., 1999. Role of PDGF-B and PDGFR-beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development 126, 3047–3055. Henry, C.B., DeFouw, D.O., 1995. Differential lectin binding to microvascular endothelial glycoconjugates during normal angiogenesis in the chick chorioallantoic membrane. Microvasc. Res. 49, 201–211. Henry, C.B.S., Duling, B.R., 1999. Permeation of the luminal capillary glycocalyx is determined by hyaluronan. Am. J. Physiol. Heart Circ. Physiol. 277, H508–H514. Klewer, S.E., Yatskievych, T., Pogreba, K., Stevens, M.V., Antin, P.B., Camenisch, T.D., 2006. Has2 expression in heart forming regions is independent of BMP signaling. Gene Expr. Patterns 6, 462–470. Koopman, G., Reutelingsperger, C.P., Kuijten, G.A., Keehnen, R.M., Pals, S.T., van Oers, M.H., 1994. Annexin V for flow cytometric detection of phosphatidylserine expression on B cells undergoing apoptosis. Blood 84, 1415–1420. Krull, C.E., Kulesa, P.M., 1998. Embryonic explant and slice preparations for studies of cell migration and axon guidance. Curr. Top. Dev. Biol. 36, 145–159. Larrivee, B., Freitas, C., Trombe, M., Lv, X., Delafarge, B., Yuan, L., Bouvree, K., Breant, C., Del Toro, R., Brechot, N., Germain, S., Bono, F., Dol, F., Claes, F., Fischer, C., Autiero, M., Thomas, J.L., Carmeliet, P., Tessier-Lavigne, M., Eichmann, A., 2007. Activation of the UNC5B receptor by Netrin-1 inhibits sprouting angiogenesis. Genes Dev. 21, 2433–2447. le Noble, F., Moyon, D., Pardanaud, L., Yuan, L., Djonov, V., Matthijsen, R., Breant, C., Fleury, V., Eichmann, A., 2004. Flow regulates arterial-venous differentiation in the chick embryo yolk sac. Development 131, 361–375. Leonardo, E.D., Hinck, L., Masu, M., Keino-Masu, K., Ackerman, S.L., Tessier-Lavigne, M., 1997. Vertebrate homologues of C. elegans UNC-5 are candidate netrin receptors. Nature 386, 833–838. Lobov, I.B., Renard, R.A., Papadopoulos, N., Gale, N.W., Thurston, G., Yancopoulos, G.D., Wiegand, S.J., 2007. Delta-like ligand 4 (Dll4) is induced by VEGF as a negative regulator of angiogenic sprouting. Proc. Natl. Acad. Sci. U S A 104, 3219–3224. Lucitti, J.L., Jones, E.A., Huang, C., Chen, J., Fraser, S.E., Dickinson, M.E., 2007. Vascular remodeling of the mouse yolk sac requires hemodynamic force. Development 134, 3317–3326. Madri, J.A., Enciso, J., Pinter, E., 2003. Maternal diabetes: effects on embryonic vascular development—a vascular endothelial growth factor-A-mediated process. Pediatr. Dev. Pathol. 6, 334–341. Miller, R.L., 1982. A sialic acid-specific lectin from the slug Limax flavus. J. Invertebr. Pathol. 39, 210–214. Mochizuki, S., Vink, H., Hiramatsu, O., Kajita, T., Shigeto, F., Spaan, J.A., Kajiya, F., 2003. Role of hyaluronic acid glycosaminoglycans in shear-induced endothelium-derived nitric oxide release. Am. J. Physiol. Heart Circ. Physiol. 285, H722–H726. Mulivor, A.W., Lipowsky, H.H., 2004. Inflammation- and ischemia-induced shedding of venular glycocalyx. Am. J. Physiol. Heart Circ. Physiol. 286, H1672–H1680.

C.E. Henderson-Toth et al. / Developmental Biology 369 (2012) 330–339

Nieuwdorp, M., van Haeften, T.W., Gouverneur, M.C., Mooij, H.L., van Lieshout, M.H., Levi, M., Meijers, J.C., Holleman, F., Hoekstra, J.B., Vink, H., Kastelein, J.J., Stroes, E.S., 2006b. Loss of endothelial glycocalyx during acute hyperglycemia coincides with endothelial dysfunction and coagulation activation in vivo. Diabetes 55, 480–486. Nieuwdorp, M., Mooij, H.L., Kroon, J., Atasever, B., Spaan, J.A., Ince, C., Holleman, F., Diamant, M., Heine, R.J., Hoekstra, J.B., Kastelein, J.J., Stroes, E.S., Vink, H., 2006a. Endothelial glycocalyx damage coincides with microalbuminuria in type 1 diabetes. Diabetes 55, 1127–1132. Ohya, T., Kaneko, Y., 1970. Novel hyaluronidase from streptomyces. Biochim. Biophys. Acta 198, 607–609. Oohira, A., Wight, T.N., Bornstein, P., 1983. Sulfated proteoglycans synthesized by vascular endothelial cells in culture. J. Biol. Chem. 258, 2014–2021. Pahakis, M.Y., Kosky, J.R., Dull, R.O., Tarbell, J.M., 2007. The role of endothelial glycocalyx components in mechanotransduction of fluid shear stress. Biochem. Biophys. Res. Commun. 355, 228–233. Pardanaud, L., Altmann, C., Kitos, P., Dieterlen-Lievre, F., Buck, C.A., 1987. Vasculogenesis in the early quail blastodisc as studied with a monoclonal antibody recognizing endothelial cells. Development 100, 339–349. Pries, A.R., Secomb, T.W., Gaehtgens, P., 2000. The endothelial surface layer. Pflugers Arch. 440, 653–666. Pries, A.R., Secomb, T.W., Sperandio, M., Gaehtgens, P., 1998. Blood flow resistance during hemodilution: effect of plasma composition. Cardiovasc. Res. 37, 225–235. Ramesh, G., Berg, A., Jayakumar, C., 2011. Plasma netrin-1 is a diagnostic biomarker of human cancers. Biomarkers 16, 172–180. Rehm, M., Zahler, S., Lotsch, M., Welsch, U., Conzen, P., Jacob, M., Becker, B.F., 2004. Endothelial glycocalyx as an additional barrier determining extravasation of 6% hydroxyethyl starch or 5% albumin solutions in the coronary vascular bed. Anesthesiology 100, 1211–1223. Reitsma, S., Slaaf, D.W., Vink, H., van Zandvoort, M.A., oude Egbrink, M.G., 2007. The endothelial glycocalyx: composition, functions, and visualization. Pflugers Arch. 454, 345–359.

339

Shimada, K., Kobayashi, M., Kimura, S., Nishinaga, M., Takeuchi, K., Ozawa, T., 1991. Anticoagulant heparin-like glycosaminoglycans on endothelial cell surface. Jpn. Circ. J. 55, 1016–1021. Shipp, E.L., Hsieh-Wilson, L.C., 2007. Profiling the sulfation specificities of glycosaminoglycan interactions with growth factors and chemotactic proteins using microarrays. Chem. Biol. 14, 195–208. Squire, J.M., Chew, M., Nneji, G., Neal, C., Barry, J., Michel, C., 2001. Quasi-periodic substructure in the microvessel endothelial glycocalyx: a possible explanation for molecular filtering? J. Struct. Biol. 136, 239–255. Suchting, S., Freitas, C., le Noble, F., Benedito, R., Breant, C., Duarte, A., Eichmann, A., 2007. The Notch ligand Delta-like 4 negatively regulates endothelial tip cell formation and vessel branching. Proc. Natl. Acad. Sci. U S A 104, 3225–3230. Tarbell, J.M., Ebong, E.E., 2008. The endothelial glycocalyx: a mechano-sensor and transducer. Sci. Signal. 1, pt 8. Thi, M.M., Tarbell, J.M., Weinbaum, S., Spray, D.C., 2004. 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 101, 16483–16488. Tsutsui, M., Shimokawa, H., Morishita, T., Nakashima, Y., Yanagihara, N., 2006. Development of genetically engineered mice lacking all three nitric oxide synthases. J. Pharmacol. Sci. 102, 147–154. Tzima, E., Irani-Tehrani, M., Kiosses, W.B., Dejana, E., Schultz, D.A., Engelhardt, B., Cao, G., DeLisser, H., Schwartz, M.A., 2005. A mechanosensory complex that mediates the endothelial cell response to fluid shear stress. Nature 437, 426–431. van den Berg, B.M., Vink, H., Spaan, J.A.E., 2003. The endothelial glycocalyx protects against myocardial edema. Circ. Res. 92, 592–594. Van Sluis, G.L., Nieuwdorp, M., Kamphuisen, P.W., van der Vlag, J., Van Noorden, C.J., Spek, C.A., 2010. A low molecular weight heparin inhibits experimental metastasis in mice independently of the endothelial glycocalyx. PLoS One 5, e11200. Vink, H., Duling, B.R., 2000. Capillary endothelial surface layer selectively reduces plasma solute distribution volume. Am. J. Physiol. Heart Circ. Physiol. 278, H285–H289. Yao, Y., Rabodzey, A., Dewey Jr., C.F., 2007. Glycocalyx modulates the motility and proliferative response of vascular endothelium to fluid shear stress. Am. J. Physiol. Heart Circ. Physiol. 293, H1023–H1030.