Glomerular Endothelial Surface Layer Acts as a Barrier against Albumin Filtration

Abstract

Glomerular endothelium is highly fenestrated, and its contribution to glomerular barrier function is the subject of debate. In recent years, a polysaccharide-rich endothelial surface layer (ESL) has been postulated to act as a filtration barrier for large molecules, such as albumin. To test this hypothesis, we disturbed the ESL in C57Bl/6 mice using long-term hyaluronidase infusion for 4 weeks and monitored albumin passage using immunolabeling and correlative light-electron microscopy that allows for complete and integral assessment of glomerular albumin passage. ESL ultrastructure was visualized by transmission electron microscopy using cupromeronic blue and by localization of ESL binding lectins using confocal microscopy. We demonstrate that glomerular fenestrae are filled with dense negatively charged polysaccharide structures that are largely removed in the presence of circulating hyaluronidase, leaving the polysaccharide surfaces of other glomerular cells intact. Both retention of cationic ferritin in the glomerular basement membrane and systemic blood pressure were unaltered. Enzyme treatment, however, induced albumin passage across the endothelium in 90% of glomeruli, whereas this could not be observed in controls. Yet, there was no net albuminuria due to binding and uptake of filtered albumin by the podocytes and parietal epithelium. ESL structure and function completely recovered within 4 weeks on cessation of hyaluronidase infusion. Thus, the polyanionic ESL component, hyaluronan, is a key component of the glomerular endothelial protein permeability barrier. Glomerular endothelium is highly fenestrated, and its contribution to glomerular barrier function is the subject of debate. In recent years, a polysaccharide-rich endothelial surface layer (ESL) has been postulated to act as a filtration barrier for large molecules, such as albumin. To test this hypothesis, we disturbed the ESL in C57Bl/6 mice using long-term hyaluronidase infusion for 4 weeks and monitored albumin passage using immunolabeling and correlative light-electron microscopy that allows for complete and integral assessment of glomerular albumin passage. ESL ultrastructure was visualized by transmission electron microscopy using cupromeronic blue and by localization of ESL binding lectins using confocal microscopy. We demonstrate that glomerular fenestrae are filled with dense negatively charged polysaccharide structures that are largely removed in the presence of circulating hyaluronidase, leaving the polysaccharide surfaces of other glomerular cells intact. Both retention of cationic ferritin in the glomerular basement membrane and systemic blood pressure were unaltered. Enzyme treatment, however, induced albumin passage across the endothelium in 90% of glomeruli, whereas this could not be observed in controls. Yet, there was no net albuminuria due to binding and uptake of filtered albumin by the podocytes and parietal epithelium. ESL structure and function completely recovered within 4 weeks on cessation of hyaluronidase infusion. Thus, the polyanionic ESL component, hyaluronan, is a key component of the glomerular endothelial protein permeability barrier. The glomerular endothelium is highly fenestrated, and its pores are estimated to be approximately 60 to 80 nm in diameter. Although fenestrae may facilitate the formation of high volumes of ultrafiltrate, this would also imply loss of macromolecules, such as albumin, as suggested by intravital microscopy studies.1Russo L.M. Sandoval R.M. Campos S.B. Molitoris B.A. Comper W.D. Brown D. Impaired tubular uptake explains albuminuria in early diabetic nephropathy.J Am Soc Nephrol. 2009; 20: 489-494Crossref PubMed Scopus (216) Google Scholar In vivo, endothelium is covered with a polysaccharide-protein gel-like structure, the glycocalyx, or the endothelial surface layer (ESL). The composition and biological activity of this surface layer can widely differ in a tissue-specific manner. The main constituents of the ESL are protein cores, typically syndecans, with large heparan sulfate side chains (±60%) or chondroitin sulfate (±15%).2Reitsma S. Slaaf D.W. Vink H. van Zandvoort M.A. oude Egbrink M.G. The endothelial glycocalyx: composition, functions, and visualization.Pflugers Arch. 2007; 454: 345-359Crossref PubMed Scopus (1353) Google Scholar In addition, plasma proteins, such as albumin and orosomucoid, bind to these negatively charged proteoglycans and glycosaminoglycans.3Friden V. Oveland E. Tenstad O. Ebefors K. Nystrom J. Nilsson U.A. Haraldsson B. The glomerular endothelial cell coat is essential for glomerular filtration.Kidney Int. 2011; 79: 1322-1330Crossref PubMed Scopus (104) Google Scholar Hyaluronan integrated into this gel-like mesh anchors to CD44 on the cell membrane.4Weinbaum S. Tarbell J.M. Damiano E.R. The structure and function of the endothelial glycocalyx layer.Annu Rev Biomed Eng. 2007; 9: 121-167Crossref PubMed Scopus (877) Google Scholar Hyaluronan consists of repeating disaccharides composed of N-acetylglucosamine and glucuronic acid and is the largest polysaccharide found in vertebrates. Together, these polysaccharides form a biologically active surface layer, which may stretch out up to 500 nm into the vascular lumen.2Reitsma S. Slaaf D.W. Vink H. van Zandvoort M.A. oude Egbrink M.G. The endothelial glycocalyx: composition, functions, and visualization.Pflugers Arch. 2007; 454: 345-359Crossref PubMed Scopus (1353) Google Scholar Many growth factors, such as fibroblast growth factor, vascular endothelial growth factor, and circulating complement inhibitors (eg, factor H), require prior engagement with this surface layer before they can interact with their functional ligands.5Allen B.L. Filla M.S. Rapraeger A.C. Role of heparan sulfate as a tissue-specific regulator of FGF-4 and FGF receptor recognition.J Cell Biol. 2001; 155: 845-858Crossref PubMed Scopus (129) Google Scholar, 6Fromm J.R. Hileman R.E. Weiler J.M. Linhardt R.J. Interaction of fibroblast growth factor-1 and related peptides with heparan sulfate and its oligosaccharides.Arch Biochem Biophys. 1997; 346: 252-262Crossref PubMed Scopus (45) Google Scholar, 7Lehtinen M.J. Rops A.L. Isenman D.E. van der Vlag J. Jokiranta T.S. Mutations of factor H impair regulation of surface-bound C3b by three mechanisms in atypical hemolytic uremic syndrome.J Biol Chem. 2009; 284: 15650-15658Crossref PubMed Scopus (84) Google Scholar Heparan sulfate repeats can also activate antithrombin-III, and bind extracellular superoxide dismutases, whereas after activation of the endothelial cell, they function as receptors for platelet and leukocyte adhesion.8Chappell D. Dorfler N. Jacob M. Rehm M. Welsch U. Conzen P. Becker B.F. Glycocalyx protection reduces leukocyte adhesion after ischemia/reperfusion.Shock. 2010; 34: 133-139Crossref PubMed Scopus (139) Google Scholar, 9Mulivor A.W. Lipowsky H.H. Role of glycocalyx in leukocyte-endothelial cell adhesion.Am J Physiol Heart Circ Physiol. 2002; 283: H1282-H1291Crossref PubMed Scopus (288) Google Scholar, 10Vink H. Constantinescu A.A. Spaan J.A. Oxidized lipoproteins degrade the endothelial surface layer: implications for platelet-endothelial cell adhesion.Circulation. 2000; 101: 1500-1502Crossref PubMed Scopus (314) Google Scholar, 11Rops A.L. Jacobs C.W. Linssen P.C. Boezeman J.B. Lensen J.F. Wijnhoven T.J. van den Heuvel L.P. van Kuppevelt T.H. van der Vlag J. Berden J.H. Heparan sulfate on activated glomerular endothelial cells and exogenous heparinoids influence the rolling and adhesion of leucocytes.Nephrol Dial Transplant. 2007; 22: 1070-1077Crossref PubMed Scopus (32) Google Scholar, 12Rops A.L. van den Hoven M.J. Baselmans M.M. Lensen J.F. Wijnhoven T.J. van den Heuvel L.P. van Kuppevelt T.H. Berden J.H. van der Vlag J. Heparan sulfate domains on cultured activated glomerular endothelial cells mediate leukocyte trafficking.Kidney Int. 2008; 73: 52-62Crossref PubMed Scopus (56) Google Scholar The pivotal role of the ESL is illustrated by the fact that, to our knowledge, no successful genetic mice models could be generated in which structural polysaccharide components of the ESL are lacking. These properties have put the ESL in the center of interest as a modulator of endothelial cell function. The ESL has also been postulated to act as a barrier against albumin filtration. The combination of negatively charged heparan sulfates and the mesh-like structure of the less charged hyaluronan may theoretically prevent albumin filtration through the fenestrae.13Henry C.B. Duling B.R. Permeation of the luminal capillary glycocalyx is determined by hyaluronan.Am J Physiol. 1999; 277: H508-H514PubMed Google Scholar In fact, the Frank Starling filtration equilibrium in nonrenal microvasculature is generated by the inability of albumin to permeate the ESL and its resulting oncotic gradients over this layer.14Levick J.R. Michel C.C. Microvascular fluid exchange and the revised Starling principle.Cardiovasc Res. 2010; 87: 198-210Crossref PubMed Scopus (680) Google Scholar Moreover, acute infusion of high dosages of enzymes that degrade the ESL results in increased fractional albumin clearance, extrapolating this concept also to the kidney.15Jeansson M. Haraldsson B. Glomerular size and charge selectivity in the mouse after exposure to glucosaminoglycan-degrading enzymes.J Am Soc Nephrol. 2003; 14: 1756-1765Crossref PubMed Scopus (117) Google Scholar, 16Jeansson M. Haraldsson B. Morphological and functional evidence for an important role of the endothelial cell glycocalyx in the glomerular barrier.Am J Physiol Renal Physiol. 2006; 290: F111-F116Crossref PubMed Scopus (171) Google Scholar In the current study, we explore the effects of long-term chronic disruption of the ESL on albumin filtration over the glomerular endothelium and the possible effects on glomerular morphological characteristics. To this end, we use a new electron microscopy technique that allows visualization of the ESL and albumin transport within the entire glomerular section at nanometer resolution.17Faas F.G. Avramut M.C. van den Berg B.M. Mommaas A.M. Koster A.J. Ravelli R.B. Virtual nanoscopy: generation of ultra-large high resolution electron microscopy maps.J Cell Biol. 2012; 198: 457-469Crossref PubMed Scopus (93) Google Scholar To study the functional role of the ESL with respect to prevention of albumin filtration, it was perturbed by long-term low-dose infusion of hyaluronidase, a hyaluronan degrading enzyme, which has been shown to increase the permeation of the ESL (eg, in the cremaster muscle).13Henry C.B. Duling B.R. Permeation of the luminal capillary glycocalyx is determined by hyaluronan.Am J Physiol. 1999; 277: H508-H514PubMed Google Scholar All materials are from Sigma-Aldrich (St. Louis, MO), unless stated otherwise. Experiments were performed using C57Bl/6 mice (B6, Charles River, France) at 14 weeks of age. Mice were randomly divided into two groups and implanted with an osmotic minipump (Alzet, Cupertino, CA) containing hyaluronidase (group 1, total n = 26) or inactivated hyaluronidase (group 2, total n = 22). Anesthesia during surgical procedures was performed using an i.p. injection of 0.2 mL of a 1 mg/mL cocktail of midazolam (Roche Nederland BV, Woerden, The Netherlands), 50 μg/mL and μg/mL in 4 mice from were the of mice were and these were to for an 4 In a of mice = and mice with an osmotic minipump containing hyaluronidase for 4 weeks = the glomerular ESL was using cupromeronic In addition, mice an osmotic minipump containing of hyaluronidase for weeks to the = or an osmotic minipump containing hyaluronidase = or inactivated hyaluronidase = for weeks for blood pressure All to and The was by the of The hyaluronidase and blood pressure and The All was performed in with the hyaluronidase was into an osmotic which its at a of into the for 4 The of of the hyaluronidase is after infusion into the not In a of of and in were and into the for In hyaluronidase for at by was as a M.C. M. S. E. M.J. M. Vink H. van den Berg B.M. surface layer by chronic hyaluronidase infusion in 2010; PubMed Scopus (45) Google Scholar Mice were in for before and before the of the and 4 or Mice from the groups were in addition, in before and weeks before was and at was to albumin in from M.C. of ferritin by cells and its regulation by and 1997; PubMed Google were determined by a using a and a (Roche were determined by the using a (Roche of mice of ferritin the before the of In the was and to the The was at the and by in before of the The of the was and the was in was at in in for of albumin and and the other was in and in at for of ferritin three were with active and three with hyaluronidase for 4 Both were and in for electron of ESL in the the was with albumin and in at to by cupromeronic blue in containing and for at of the the was removed and the was into V. Oveland E. Tenstad O. Ebefors K. Nystrom J. Nilsson U.A. Haraldsson B. The glomerular endothelial cell coat is essential for glomerular filtration.Kidney Int. 2011; 79: 1322-1330Crossref PubMed Scopus (104) Google and at in the containing cupromeronic as These cupromeronic were in and for 1 in acid in were and in electron microscopy were on a The from nm as using the was into with a BV, The of the are with and in for on by at with albumin to in in approximately 1 mL of mL to was for at and for 1 with and in in and in were and into nm were on a covered with and a for and on a on a for microscopy E. the and electron Biol. 2006; PubMed Scopus Google Scholar The was with on an were with a were with an of and before and were at an of on a transmission electron with an The To the in their are as by F.G. Avramut M.C. van den Berg B.M. Mommaas A.M. Koster A.J. Ravelli R.B. Virtual nanoscopy: generation of ultra-large high resolution electron microscopy maps.J Cell Biol. 2012; 198: 457-469Crossref PubMed Scopus (93) Google Scholar The for the were using and with at the to a size at the results in a with an of the in which can into high using this correlative microscopy the can be visualized with and Although allows for the of from from localization of the at high resolution.17Faas F.G. Avramut M.C. van den Berg B.M. Mommaas A.M. Koster A.J. Ravelli R.B. Virtual nanoscopy: generation of ultra-large high resolution electron microscopy maps.J Cell Biol. 2012; 198: 457-469Crossref PubMed Scopus (93) Google Scholar for of were with BV, The containing and were with μg/mL of lectins and and or μg/mL mouse CA) in after a with μg/mL or in were to the of a and were using a and a of the were using The were with the the and of were by the from the of the to the of the or a of at in and in were into with a were with and for 1 with and in in and in and were and into were at an of on a were in in for in and for at in The Netherlands), and in were for with 1 to μg/mL against mouse or The were and for with μg/mL or BV, The were with with in and in The The of were in 50 on a to no 1, of groups were determined by using was weeks of of and were performed to systemic vascular blood cell and total vascular as Vink H. endothelial glycocalyx permeability and of with blood Physiol. PubMed Scopus (139) Google Scholar ESL permeability and the of with Spaan Vink glycocalyx barrier properties and in Scholar In were with and with a of injection of mL of this blood were at 2, and were was and plasma was at was determined using vascular plasma was using the M. M.C. Vink H. J.B. The endothelial glycocalyx: a barrier and vascular PubMed Scopus Google Scholar, M. van M.C. van M. J.B. Vink H. of endothelial glycocalyx during acute with endothelial and activation in 2006; PubMed Scopus Google Scholar In plasma was at nm using a The of were with a function to pressure was in using the of by the two of were a was the successful The from was for and in are as and are as groups were determined by using was complete of the ESL in can in and we for a to study the early effects of ESL loss on glomerular permeability osmotic minipump filled with hyaluronidase was implanted with a into the to degrade hyaluronan from the ESL within the blood Although a of in vascular and within a not a of in approximately a in systemic ESL after weeks as by the in of which can the the of which the or in or were observed in to infusion of active or hyaluronidase The in ESL were visualized at 4 weeks of long-term infusion of active hyaluronidase, with hyaluronidase as a Renal were with lectins that bind to and or and These lectins were shown to in the ESL that with at the A.W. Lipowsky H.H. and of J Physiol Heart Circ Physiol. PubMed Scopus Google Scholar To ESL the from the of the of the endothelial cell membrane to the was determined with long-term infusion of inactivated hyaluronidase, and were after 4 weeks of long-term hyaluronidase infusion. of the osmotic minipump and of the of the glomerular ESL was and lectins and to In to structures the endothelial with which 1, was by hyaluronidase infusion = for 4 weeks and after 4 weeks of To in glomerular endothelial surface structures at the ultrastructure were for a with the cationic cupromeronic and for In this polysaccharides are in a the cupromeronic blue was observed at within the permeability barrier may be from the on the endothelial surface and the filled fenestrae could be observed 2, and The ESL dense in the endothelial fenestrae and a dense layer the endothelial which was by a layer 2, and In the capillary the endothelial a layer of which of ESL in other vascular den Berg B.M. Spaan J.A. Vink H. and glycocalyx and increase at J Physiol Heart Circ Physiol. 2006; 290: PubMed Scopus Google Scholar, den Berg B.M. Vink H. Spaan J.A. The endothelial glycocalyx against Res. 2003; PubMed Scopus Google Scholar The and their was the of the dense from the endothelial fenestrae blue the structures the endothelial whereas the structures the podocytes and their were not by long-term hyaluronidase infusion. in the hyaluronidase infusion to have removed the structures from the glomerular To an of hyaluronidase on the glomerular basement membrane barrier retention of cationic ferritin by binding in the was ferritin ferritin in the and ferritin ferritin of that the hyaluronidase not the composition and permeability of the To the of the ESL by hyaluronidase on the filtration barrier for albumin, were for albumin. The were with to and with to be to albumin with and the were with albumin could be All luminal proteins, albumin, were during and for for albumin were visualized using of was and were was at the glomerular filtration barrier. 90% of the were for albumin after hyaluronidase as observed by the capillary and and In no albumin was in the three and after the three using a technique to the found by in high a in were F.G. Avramut M.C. van den Berg B.M. Mommaas A.M. Koster A.J. Ravelli R.B. Virtual nanoscopy: generation of ultra-large high resolution electron microscopy maps.J Cell Biol. 2012; 198: 457-469Crossref PubMed Scopus (93) Google Scholar The allows at nanometer resolution of a glomerular section in an this for The and the were to the of the albumin. In the albumin was the in the and In to the no albumin could be found the endothelial layer and the glomerular basement membrane in and after 4 weeks of the albumin that was found after the hyaluronidase was no To be to the perturbed filtration barrier shown with also in and a was were and before the osmotic minipump after 4 weeks of hyaluronidase and during the 4 weeks of and Although albumin was shown to be to the filtration barrier in the after hyaluronidase, no could be found in or albumin also to podocytes and parietal cells in the this may that uptake and of albumin have as To albumin binding would also in to and protein was Although there was a a and such as can be observed during and high during hyaluronidase treatment, the was to demonstrate this that glomerular fenestrae are filled with dense negatively charged polysaccharide infusion of hyaluronidase, in a that the polysaccharide surfaces of other glomerular cells the polysaccharides in the endothelial albumin passage across the endothelium can be observed in of the albumin passage was not observed for the with low-dose hyaluronidase is in the ESL in the whereas the glomerular surface layer unaltered. and J. K. in fenestrae of glomerular of the filtration 2002; PubMed Scopus Google Scholar have suggested the presence of in the glomerular current that the glomerular fenestrae are with The fact that hyaluronan is less charged the heparan sulfates in ESL may the to a dense mesh in the hyaluronan have been shown to the of through the surface J.R. J. PubMed Scopus Google Scholar, J.R. The structure and function of an Cell Biol. PubMed Scopus Google Scholar Although hyaluronidase theoretically could also the polysaccharide on other glomerular the and the no structural the of the ESL also not the glomerular ultrastructure and basement membrane whereas the in ESL and albumin filtration were completely on cessation of the hyaluronidase infusion. These to the that the glomerular ESL is and that its in endothelial function. the passage of albumin was not by albuminuria or is in to our in mice on a which were with M.C. M. S. E. M.J. M. Vink H. van den Berg B.M. surface layer by chronic hyaluronidase infusion in 2010; PubMed Scopus (45) Google Scholar possible for this is that the mice a loss of ESL was on vascular and is by the fact that in that could be by of the recent study using in also could ESL loss to the of H. D. J. of the endothelial glycocalyx albuminuria and vascular Am Soc Nephrol. 2012; PubMed Scopus Google Scholar In the current study, we to the role of the ESL in glomerular filtration and at a ESL in in fact, the hyaluronan mesh in the ESL was Although this allows a on the of the glomerular ESL to prevent albumin filtration, the may be less to study the endothelial and its with Although we not increased protein in the we albumin to podocytes and parietal epithelium. are to the and is in and glomerular Dial Transplant. 2012; PubMed Scopus Google Scholar, U. M.A. of in podocytes in 2011; PubMed Scopus Google Scholar also such as and the receptor for the that albumin was by as has also been for tubular epithelium. uptake could and to and as was postulated by M. S. S. M. S. P. In to protein podocytes and for of chronic J PubMed Scopus Google Scholar In our the albumin passage the was with a loss of and increased the was to morphological in the In addition, that hyaluronidase has of polysaccharide surface structures on the cells in such a that albumin could to these our cupromeronic blue and of the surface layer not morphological after hyaluronidase the tubular this albumin was not uptake however, not be in to the glomeruli, we could not the tubular of our study was the fact that we not the glomerular ESL the surface layer in other vascular endothelial respect to albumin filtration, the not in in of the be a to test the functional role of the glomerular our in would that, in vivo, of the ESL are of ESL results in as can also be observed with our using high of In addition, of the surface layer has been shown to in with of ferritin in the glomerular permeability as a of of the Int. PubMed Scopus Google Scholar In this study, these morphological are not we a of the glomerular filtration barrier. that the of the endothelial surface layer in this study an early in the of glomerular and In the glomerular endothelial surface layer as a protein permeability a of the key component, hyaluronan, within this layer albumin passage across the endothelial layer and the the We and and for on the blood pressure and of the for the use of the electron and with of ESL and and of ESL in mice after weeks of with or of hyaluronidase in ESL was by and and and at the and at the of the hyaluronidase for in and was by with in the of ferritin in the in and mice after a = were in the total surface of the as of ferritin or at the as of ferritin of are = to in of ferritin was by with of glomerular filtration barrier for albumin using albumin and and with and after 4 weeks of or active hyaluronidase = These can be found at with in localization of and in of or mice with hyaluronidase for 4 of are = 50 three in was by a of the as a against of the of The of the up with for as in the on be in the on of cationic ferritin by binding in the was be of ferritin within the was The to the of ferritin within the glomerular basement membrane was with