PERSPECTIVES www.jasn.org

What Is the Glomerular Ultrafiltration Barrier?

William H. Fissell1 and Jeffrey H. Miner2

1Division of Nephrology and Hypertension, Vanderbilt University Medical Center, Nashville, Tennessee; and 2Division of Nephrology, Washington University School of Medicine, St. Louis, Missouri

J Am Soc Nephrol 29: ccc–ccc, 2018. doi: https://doi.org/10.1681/ASN.2018050490

Since the unique features of the three- (2) Exclusion of and antibodies superstructure with two distinct - layered glomerular wall between from the subepithelial GBM if foot 521 and two distinct IV networks4 the and the were first visual- processes are intact but their ap- that are organized separately by ized by electron microscopy, the identity pearance in Bowman’sspacewhen and endothelial cells. Although laminin and of the definitive components that restrict either foot processes become ef- collagen IV networks do not interact with the filtration of albumin, Igs, and other faced or blood flow is halted. each other directly, there are multiple bridg- large plasma constituents has been contro- ing molecules that stabilize them, presum- versial. Because is the single (3) Accumulation of certain antibodies ably forming a dense, gel-like structure that strongest predictor of progression from (e.g., anti-phospholipase A2 receptor) impedes the passage of macromolecules. CKD to renal failure,1 an understanding at the interface between podocytes Additional data combining laminin and of what constitutes the ultrafiltration bar- and the GBM. collagenIVmutationssupportanindirect rier and how it is maintained is critical. physical interaction that cross-stabilizes the fi It is evident that all three layers of the (4) Near-perfect delity of glomerular two networks to maintain GBM architec- glomerular capillary wall—fenestrated en- permselectivity over decades without ture and glomerular permselectivity.5 dothelium and its associated , deterioration, clogging, or fouling. In 1998, the identification of nephrin, a glomerular (GBM), glomerular slit diaphragm protein that, (5) Absence of nephrotic syndrome even and foot processes with their slit when mutated, causes catastrophic with defects in GBM composition diaphragms—participate in the mainte- nephrotic syndrome, brought the slit dia- (with the exception of laminin). nance of the glomerular filtration barrier phragm back to the forefront. In the sub- 2 sequent two decades, the list of podocyte (GFB). However, the physical chemistry (6) Heavy proteinuria either accompanied proteins essential to healthy podocyte ar- of the GFB is not completely understood. or caused by derangement of chitecture and maintenance of the GFB Moreover, the mechanisms of action of the foot processes. most commonly used and most effective has grown substantially due mainly to treatments to reduce proteinuria, such as (7) Increase in GFR associated with de- breakthroughs in the genetics of nephrotic - system blockade and cal- creased glomerular capillary wall syndrome. It is indisputable that proper cineurin inhibitors, remain incompletely macromolecular permeability in podocyte architecture is necessary to main- defined. New mechanism-based treatments healthy rats. tain the GFB over the long term. This ledto a for renal disease will rely on new mechanis- simple heuristic model wherein slit dia- tic insights into an old problem. (8) Increase in existing proteinuria with phragms, stabilized by connections to the Comprehensive models of the GFB increase in systolic BP. should explain all of the following clinical and laboratory observations of glomerular Many models, ranging from simple to Published online ahead of print. Publication date anatomy and physiology in both health and complex, have been proposed to explain available at www.jasn.org. disease. the extraordinary GFB. Despite intriguing and highly suggestive ultrastructural data Correspondence: Dr. William H. Fissell, Division of Nephrology and Hypertension, Vanderbilt Univer- 3 (1) Nearly complete retention of albumin regarding the slit diaphragm, before the sity Medical Center, 1161 21st Avenue, S-3223 and other macromolecules in both mid-1990s, hindered transport through MCN, Nashville, TN 37232, or Dr. Jeffrey H. Miner, health and nephrotic syndrome, the unique double-thickness GBM was Division of Nephrology, Washington University School of Medicine, 4523 Clayton Avenue, CB with severe nephrotic syndrome the concept that attracted attention as the 8126, St. Louis, MO 63110. Email: William.Fissell@ resulting from a quantitatively mi- likely dominant factor in plasma protein vanderbilt.edu or [email protected]

nor change in glomerular permse- retention. Recent super-resolution imag- Copyright © 2018 by the American Society of lectivity. ing of the GBM reveals a highly ordered Nephrology

J Am Soc Nephrol 29: ccc–ccc, 2018 ISSN : 1046-6673/2909-ccc 1 PERSPECTIVES www.jasn.org

to the podocyte. This “gel compression” is well documented in other biologic sys- tems.10 The resulting anisotropic density of the GBM explains the observed spatial distribution of macromolecules therein and the very limited—but not zero— transmission of plasma proteins to the primary urine. With disease or injury and the cytoskeletal rearrangements that typically accompany them, the decreased contractile strength of injured podocytes allows for a subtle shift from radial com- pression to circumferential tension within the GBM, opening the meshwork of the GBM gel, increasing average pore size, and permitting leakage of macromolecules into Bowman’sspace(Figure1). In this model, the compressed GBM hinders transport of macromolecules but allows water and small solutes to pass freely. The slit diaphragms and whatever aberrant cell-cell junctions form after ef- facement are nonrestrictive, as supported by cryoelectrontomography data.11 This model explains why mutations that affect the slit diaphragm, its linkage to the cyto- skeleton, or the cytoskeleton itself cause nephrotic syndrome and importantly, takes into account mechanisms involving cellular regulation of GBM protein orga- nization through integrin-a3b1. The model, like a recently presented one,12 also has explanatory power regarding the varied phenotypes arising from mutations in GBM components.13 Mutations affect- Figure 1. A new model of the glomerular filtration barrier links glomerular basement mem- brane (GBM) compression to permselectivity. Schematics of hydrodynamic forces in the ing laminin-521 disproportionately glomerular capillary wall. In a healthy (upper panel), water filtration (blue arrows) derange the GFB compared with colla- 13 through the GBM (dark blue) compresses the GBM constituents (gold lines) against the gen IV and agrin. Laminin-521 is the podocyte foot processes (orange). Podocyte cell-cell attachment and contraction forces (large primary tether for podocytes to the green double arrows) act to buttress (white arrows) the GBM against the distending force of matrix and may uniquely mediate trans- circulation pressure. The compressed GBM cannot admit albumin (purple ovals) except possibly mission of podocyte traction forces to the in the subendothelial space, where gel compression is least pronounced. In proteinuric glo- GBM while at the same time, ensuring the merular disease with podocyte foot process effacement (lower panel), cell contraction is de- proper organization of other GBM compo- creased (small green arrows), and podocytes no longer form as strong a buttress against nents.5 By analogy, laminin is to the GBM which the GBM is compressed. With reduced gel compression, the mesh size of the what nephrin is to the slit diaphragm. GBM gel becomes large enough to allow albumin to transit through to Bowman’sspace. actin cytoskeleton, form a molecular-scale Several other explanations have emerged. ACKNOWLEDGMENTS colander separating plasma from urine. Here, we expand on an older proposition9 This model, however, predicts a spatial and then link podocyte biology to the The authors regret being unable to discuss and distribution ofplasma proteins in the GBM physical chemistry of the GBM. cite additional relevant literature due to space at odds with observations by multiple in- In our model (Figure 1), the fluid flow of constraints and acknowledge funding from Na- vestigators, and it is not easily reconciled glomerular filtration compresses the GBM tional Institutes of Health grants U01EB021214 with continuous, clogfree operation over against the restraining foot processes, likely (to W.H.F.), R01DK058366 (to J.H.M.), and the 2.5 billion heartbeats of a lifetime.6–8 in an increasing gradient from the lumen R01DK078314 (to J.H.M.).

2 Journal of the American Society of Nephrology J Am Soc Nephrol 29: ccc–ccc,2018 www.jasn.org PERSPECTIVES

DISCLOSURES 5. Funk SD, Bayer RH, Malone AF, McKee KK, 9. Robinson GB, Walton HA: Glomerular base- Yurchenco PD, Miner JH: Pathogenicity of a ment membrane as a compressible ultrafilter. None. human laminin b2 mutation revealed in models Microvasc Res 38: 36–48, 1989 of Alport syndrome. JAmSocNephrol29: 10. Holmes MH, Mow VC: The nonlinear char- – REFERENCES 949 960, 2018 acteristics of soft gels and hydrated connec- 6. Lund U, Rippe A, Venturoli D, Tenstad O, tive tissues in ultrafiltration. JBiomech23: Grubb A, Rippe B: Glomerular filtration rate 1145–1156, 1990 1. Palmer BF: Proteinuria as a therapeutic target dependence of sieving of albumin and 11. Grahammer F, Wigge C, Schell C, Kretz O, in patients with chronic disease. Am J some neutral proteins in rat kidneys. Am J Patrakka J, Schneider S, et al.: A flexible, Nephrol 27: 287–293, 2007 Physiol Renal Physiol 284: F1226–F1234, multilayered protein scaffold maintains the 2. Haraldsson B, Nyström J, Deen WM: Prop- 2003 slit in between glomerular podocytes. JCI erties of the glomerular barrier and mecha- 7. Ryan GB, Karnovsky MJ: Distribution of Insight 1: e86177, 2016 nisms of proteinuria. Physiol Rev 88: endogenous albumin in the rat glomer- 12. Lawrence MG, Altenburg MK, Sanford R, 451–487, 2008 ulus: Role of hemodynamic factors in Willett JD, Bleasdale B, Ballou B, et al.: Per- 3. Rodewald R, Karnovsky MJ: Porous substructure glomerular barrier function. Kidney Int 9: meation of macromolecules into the renal of the glomerular slit diaphragm in the rat 36–45, 1976 glomerular basement membrane and capture and mouse. JCellBiol60: 423–433, 1974 8. Fujigaki Y, Nagase M, Kobayasi S, Hidaka S, by the tubules. Proc Natl Acad Sci U S A 114: 4. Suleiman H, Zhang L, Roth R, Heuser JE, Miner Shimomura M, Hishida A: Intra-GBM site of 2958–2963, 2017 JH, Shaw AS, et al.: Nanoscale protein archi- the functional filtration barrier for endog- 13. Suh JH, Miner JH: The glomerular basement tecture of the kidney glomerular basement enous proteins in rats. Kidney Int 43: membrane as a barrier to albumin. Nat Rev membrane. eLife 2: e01149, 2013 567–574, 1993 Nephrol 9: 470–477, 2013

J Am Soc Nephrol 29: ccc–ccc, 2018 The Glomerular Filter 3