Properties of The Glomerular Barrier and Mechanisms

Physiol Rev 88: 451– 487, 2008;
doi:10.1152/physrev.00055.2006.
Properties of the Glomerular Barrier and Mechanisms

of Proteinuria
BÖRJE HARALDSSON, JENNY NYSTRÖM, AND WILLIAM M. DEEN
Department of Molecular and Clinical Medicine/Nephrology, Institute of Medicine, Göteborg University,

Sweden; and Department of Chemical Engineering, Massachusetts Institute of Technology,
Cambridge, Massachusetts
I. Introduction 452
II. Components of the Glomerular Barrier 453
A. Podocytes 453
B. Basement membrane 453
C. Endothelium 453
D. The endothelial cell surface layer 454
E. Mesangial cells 455
F. Plasma components 456
III. Methods to Study the Glomerular Barrier 457
A. Urinalysis in vivo 457
B. Micropuncture of single nephrons 457
C. The isolated perfused kidney 457
D. Tissue-uptake techniques 458
E. The isolated glomerulus 458
F. Isolated glomerular cells 458
G. Glomerular endothelial cells 459
H. Proteoglycan and GAG production 459
I. Isolated GBMs 460
J. Artificial membranes 460
IV. Solutes Used to Study Glomerular Permeability 461
A. Theoretically ideal solutes 461
B. Proteins 461
C. Dextran polymers 462
D. Ficoll polymers 462
E. Proteins versus synthetic polymers 462
V. Solute Properties That Affect Transglomerular Passage 464
A. Molecular size 464
B. Molecular charge 464
C. Configuration 466
D. Relative importance of size and charge selectivity 467
VI. Physical Principles and Theoretical Models 467
A. Hindered transport and membrane size selectivity 468
B. Sieving in homogeneous membranes 469
C. Effects of molecular charge, shape, and concentration on membrane selectivity 472
D. Sieving in multilayer membranes 473
VII. Mechanisms Behind Proteinuria and Nephrotic Syndromes 476
A. What is proteinuria? 476
B. Where do proteins in the urine come from? 476
C. The albumin retrieval hypothesis 476
D. Human diseases and animal models 477
VIII. Summary 478
A. What is the relevance of the individual components of the glomerular barrier? 478
B. What do we need to know more about? 479
C. Concluding remarks 479
www.prv.org 0031-9333/08 $18.00 Copyright © 2008 the American Physiological Society 451
Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

452 HARALDSSON, NYSTRÖM, AND DEEN
Haraldsson B, Nyström J, Deen WM. Properties of the Glomerular Barrier and Mechanisms of Proteinuria.
Physiol Rev 88: 451– 487, 2008; doi:10.1152/physrev.00055.2006.—This review focuses on the intricate properties of
the glomerular barrier. Other reviews have focused on podocyte biology, mesangial cells, and the glomerular
basement membrane (GBM). However, since all components of the glomerular membrane are important for its
function, proteinuria will occur regardless of which layer is affected by disease. We review the properties of
endothelial cells and their surface layer, the GBM, and podocytes, discuss various methods of studying glomerular
permeability, and analyze data concerning the restriction of solutes by size, charge, and shape. We also review the
physical principles of transport across biological or artificial membranes and various theoretical models used to
predict the fluxes of solutes and water. The glomerular barrier is highly size and charge selective, in qualitative
agreement with the classical studies performed 30 years ago. The small amounts of albumin filtered will be
reabsorbed by the megalin-cubulin complex and degraded by the proximal tubular cells. At present, there is no
unequivocal evidence for reuptake of intact albumin from urine. The cellular components are the key players in
restricting solute transport, while the GBM is responsible for most of the resistance to water flow across the
glomerular barrier.
I. INTRODUCTION
The intricate properties of the glomerular barrier

have fascinated researchers for decades (87). Today,
many of the molecular components of the barrier have
been revealed (234), but their functional interactions re-
main to be elucidated. The glomerular barrier is by far the
most complex biological membrane, with properties that
allow for high filtration rates of water, nonrestricted pas-
sage of small and middle-sized molecules, and almost
total restriction of serum albumin and larger proteins. In
humans, close to 180 liters of primary urine are produced
each day at capillary pressures far exceeding those in
other organs. Despite this tremendous work load, the
FIG. 1. Schematic drawing of the glomerular barrier. Podo, podo-
glomerulus remains intact year after year. Indeed, several
cytes; GBM, glomerular basement membrane; Endo, fenestrated endo-
anticlogging mechanisms have been proposed, including thelial cells; ESL, endothelial cell surface layer (often referred to as
protein consumption by the mesangial cells (87), charge the glycocalyx). Arrows indicate the filtration of plasma fluid across the
repulsion in the glomerular basement membrane (GBM) glomerular barrier, forming primary urine at a glomerular filtration rate
(GFR) of 125 ml/min in humans. The plasma flow rate (Qp) is close to
(146), a tangential-flow filtration mechanism (260), and a 700 ml/min, giving a filtration fraction of 20%. Also shown are the
gel permeation hypothesis (292, 346, 347). concentrations of albumin in serum (40 g/l) and estimated concentration
This review focuses on the functional properties of in primary urine 4 mg/l (i.e., 0.1% of that in plasma). The sieving coef-
ficient of albumin across the glomerular barrier in humans is estimated
the glomerular barrier and the mechanisms of proteinuria to be 10% of that in rodents.
and nephrotic syndromes. Although recent discussions of
proteinuria have tended to focus on the role of the podo-
cyte (319), a more complete understanding of the under-
itations. Misconceptions about the properties of the mac-
lying mechanisms requires careful consideration of the
romolecular tracers used to analyze permeability have led
other glomerular components (323). To this end, we pro-
to some confusion in the field. Therefore, we discuss the
vide an overview of the role of glomerular endothelial
cells and the glycocalyx surrounding them and discuss the “ideal solute” and compare its properties with those of the
GBM. For details on podocyte biology, readers are re- solutes currently available for research. Subsequently, we
ferred to Pavenstädt et al. (234). Also considered are address the renal clearance of various solutes in humans
interactions between the barrier and plasma proteins, and intact animals and under different experimental con-
such as orosomucoid and albumin, which further illus- ditions, discussing the results in terms of solute size,
trate the multifaceted nature of the glomerular mem- shape, and charge. At the end of that section, the recent
brane. The principal components of the glomerular bar- controversy on glomerular charge selectivity is discussed
rier are illustrated in Figure 1, together with the approx- and analyzed.
imate glomerular filtration rate, plasma flow rate, and Finally, we review physical principles and various
albumin concentrations in humans. theoretical models that reflect our understanding of how
Next we review the various methods used to study the glomerular barrier carries out its remarkable func-
glomerular permeability and discuss their virtues and lim- tions. “Simple” and multilayer barriers are discussed,
Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

GLOMERULAR BARRIER AND PROTEINURIA 453
along with the implications of their behaviors for inter- ␣5␤␥1) (205), and nidogen/entactin, together with proteo-
preting in vivo and in vitro data. glycans such as agrin and perlecan, as well as glycopro-
teins. The basement membrane of the glomerular cap-
illaries is much thicker (240 –370 nm) (163) than the
II. COMPONENTS OF THE basement membranes in other vascular beds (40 – 80
GLOMERULAR BARRIER nm) (288).
The collagen IV network may be described as the
A. Podocytes backbone of the basement membrane. Mutations in the
collagen chains give rise to severe pathological condi-
The epithelial cells of the glomerulus have attracted tions, the most well-known being Alport’s syndrome (he-
considerable interest in recent years. These specialized, reditary glomerulonephritis) (14, 170, 195). Depending on
highly differentiated cells line the outside of the glomer- the location of the mutation, less severe conditions such
ular capillaries and thus face the Bowman’s capsule and as thin glomerular membrane disease may occur (320).
the primary urine. The cells have a large cell body and The basement membrane contains large amounts of
long, extending cytoplasmic foot processes, which are proteoglycans, with mainly heparan sulfate chains at-
separated by a filtration slit that is ⬃25– 60 nm wide (150, tached. These chains may contribute to the selective
269, 342) and covered by a diaphragm. The molecular properties of the barrier (118, 145). However, the impor-
components of the diaphragm have been extensively stud- tance of the GBM in renal permselectivity has been de-
ied, and some proteins are vital for the maintenance of bated. For a long time, it was considered to be the prin-
normal glomerular permselectivity: the restriction of the cipal barrier (15, 43, 94, 303). Then in vitro studies of
permeation of macromolecules across the glomerular bar- isolated basement membranes suggested that additional
rier on the basis of molecular size, charge, and physical components were required to maintain glomerular perm-
configuration. One such molecule is nephrin, which when selectivity (20, 65). In another study, the inert probe Ficoll
mutated causes massive leakage of protein and severe and its negatively charged counterpart Ficoll sulfate were
consequences for patients with congenital nephrotic syn- used to investigate the charge effect of isolated GBMs
drome of the Finnish type (153, 154). (28). There was no difference between the sieving curves
The podocyte is a major component of the glomeru- of negative and neutral Ficolls. This suggests that the bulk
lar barrier, but its contribution to the restriction of fluid of effective charge density within the glomerular barrier
and protein transport is unclear. The slit membrane has lies in the endothelial or epithelial cell layers. However,
porelike structures (342) whose dimensions are postu- the GBM is still an important part of the glomerular
lated to be 40 ⫻ 140 Å (254, 319), corresponding to a slit barrier, since it accounts for most of the restriction of the
half-width of 20 Å. With a Stokes-Einstein radius (the fluid flux (65, 73). Moreover, recent data on laminin ␤2 in
most common measure of molecular size; see sect. VIA) of patients (350) and knockout mice (132) seem to indicate
36 Å, albumin cannot reach the urinary space across such that the GBM may restrict solute flux as well. In the mice,
slits, except through “large pores” or “shunt pathways.” proteinuria preceded alterations in podocyte morphology,
Moreover, the slit membrane dimensions described above suggesting the GBM to be an important part of the glo-
also are too small to explain the sieving of other solutes. merular barrier (132). However, it could not be ruled out
The sieving coefficient (⌰) of a solute is a measure of its that endothelial or epithelial functions were affected by
transport rate in relation to that of water. Myoglobin with the loss of laminin and normal GBM structure (116).
a Stokes-Einstein radius of 18 Å has a glomerular sieving
coefficient close to 0.8 (348), but a slit half-width of 20 Å
C. Endothelium
implies total restriction (116), i.e., ⌰ close to zero. It is
therefore most likely that the slit dimensions were under-
estimated in Reference 254. In addition, several studies in The endothelial cells of the glomerulus have not been
both mice and humans show that proteinuria can occur well investigated, likely because they are located within
without effacement of podocyte foot processes (296, 330). the glomerulus and because they are difficult to maintain
Indeed, proteinuria may occur regardless of which layer in culture without altering their phenotype. Glomerular
of the glomerular wall is damaged (116, 144). endothelial cells are unusually flattened, with a height
around the capillary loops of ⬃50 –150 nm (315). Capil-
laries with continuous endothelium are the most common
B. Basement Membrane type and are typically found in skeletal muscle, cardiac
muscle, and skin (288). However, the glomerular capillar-
The basement membrane is composed of a fibrous ies have endothelial cells with a large fenestrated area
network consisting of type IV collagen (collagen ␣3, ␣4, constituting 20 –50% of the entire endothelial surface (40).
and ␣5 chains) (193, 271), laminin (mainly laminin 11; The unusually high density of fenestrae is thought to

allow high permeability to water and small solutes in the

glomerulus (73). A scanning electron micrograph of a
mouse glomerulus and a fenestrated capillary is shown in
Figure 2.
The fenestrae are ⬃60 nm in diameter (316), and
albumin, which is restricted by the glomerular wall, has a
diameter of only 3.6 nm. These measurements have been
used to suggest that endothelial cells do not contribute to
the permselectivity of the glomerular barrier. However,
the endothelial cell coat has charge-selective properties
and the barrier probably begins at the endothelial level.
For instance, the presence of a diaphragm across the
fenestrations in glomerular capillaries has been a matter
of debate. Rostgaard and Qvortrup (260) described a high
density of fibers in the fenestrae, probably consisting of
negatively charged proteoglycans. Other studies support a
permselective layer at the level of the endothelium (11, 52,
70, 134, 135, 212, 302). These studies include both mor- FIG. 3. A human glomerulus with podocytes stained green with
phological (267, 268) and functional data and provide antibodies against synaptopodin and endothelial cells stained red with
support for the notion that glomerular endothelial cells Ulex europaeaus agglutinin I lectin.
and their cell coat should be considered in evaluating the
permselective properties of the glomerular barrier. A hu- D. The Endothelial Cell Surface Layer
man glomerulus with the endothelial and podocyte cells
stained with different markers is shown in Figure 3. On the luminal side, blood vessel walls are covered
Thus the endothelium seems to be an important com- with an endothelial cell surface layer (ESL) that is in-
ponent of the glomerular barrier, restricting macromol- volved in blood coagulation (16, 174), modulation of an-
ecules despite the presence of numerous large giogenesis (39, 96), rheology (62), and capillary barrier
fenestrae. Recently, there has been a renewed interest function (129, 134, 135, 301). This layer has two compo-
in the cell surface coat that is produced by, and sur- nents: the glycocalyx, which most commonly refers to the
rounds, the endothelial cells. plasma-membrane-bound part of the layer, and the endo-
FIG. 2. Scanning electron micrograph showing a mouse glomerulus (A) with several capillary loops, capillary lumen, and podocytes with their
foot processes. To the right (B) is a fenestrated glomerular capillary with its fenestrated endothelium surrounded by podocyte foot processes. Scale
bars: 10 ␮m (A) and 1 ␮m (B).

thelial cell coat, a larger, more loosely associated part of contributes to the high permselectivity of the glomerular
the layer. Originally described in the 1940s as a thin, wall.
noncellular coat (44, 63), the glycocalyx was first visual-
ized by staining with ruthenium red, a cationic dye that 1. The podocyte glycocalyx
binds to negatively charged molecules (178). The ESL is
composed of negatively charged glycoproteins, glycos- Podocytes are covered by a glycocalyx containing
aminoglycans (GAGs), and membrane-associated and se- sulfated molecules such as GAG and sialyzated glycocon-
creted proteoglycans, and can be visualized by using dif- jugates (232). Most studies of the podocyte glycocalyx
ferent dyes (Fig. 4). were done using cationic dyes, which have, for example,
It has been challenging to visualize the glycocalyx shown that heparan sulfate GAGs are present at the slit
and the cell coat because of technical issues (e.g., dehy- membrane (261). The major known sialoprotein on podo-
dration and disruption during fixation and staining) and cytes is podocalyxin, a heavily glycosylated protein (151).
because shear stress and plasma components are impor- In rats treated with puromycin to mimic a nephrotic syn-
tant aspects of the in vivo situation. Moreover, associated drome, the sialic acid content of podocalyxin was low-
ered, indicating that foot process effacement is linked to
plasma proteins such as orosomucoid and albumin affect
a reduced negative charge on the podocyte surface (152).
the composition (278, 279) and physical properties (114,
Glomerular epithelial cells also produce proteoglycans
115, 117) of the endothelial cell coat. Earlier work using
such as glypican-1 and syndecan-4 (238). The surface
different dyes to visualize the glycocalyx indicated a
anionic charge on podocytes is essential for maintaining
thickness of 50 –100 nm (112, 289). However, intravital
the foot process structure (25). It is also involved in
microscopy studies with fluorescent tracers indicated a
keeping a distance between parietal and visceral epithe-
thickness of 0.5 ␮m (339), while other measurements
lial cells, thereby helping to maintain glomerular structure
suggested an even thicker layer, up to 1 ␮m (77). These
and function (150).
differences probably reflect the binding of the dyes to the
membrane-bound glycocalyx, which is present after fixa-
tion, whereas intravital microscopy reveals the thicker E. Mesangial Cells
ESL, which is present when the surrounding environment
is physiological. In two studies (121, 135), intralipid drop- The main function of the glomerular mesangium is to
lets were infused, and the distance from the droplets to maintain the structure and function of the glomerular
the vessel wall was calculated. The results indicated a barrier. The mesangium is composed of two entities: mes-
large ESL of ⬃200 – 400 nm. Digestion of proteoglycans angial cells and the extracellular matrix. Like smooth
with hyaluronidase, heparinase, or chondroitinase re- muscle cells, mesangial cells have contractile properties
duced the thickness of the ESL (121, 135). Treatment with (162, 277). Although this contractility might suggest that
chondroitinase also increased the fractional clearance of mesangial cells influence filtration, their overall contribu-
albumin (135). Note that the enzymes used mainly af- tion to permselectivity is probably small. Rather, the con-
fected the ESL and not the GBM or other structures of the tractility may regulate glomerular distensibility in response
barrier, since the high molecular weights of the enzymes to pressure.
made them remain mostly in the vascular compartment Mesangial cells produce components of the mesan-
during the experiment (135). Thus the ESL seems to be a gial matrix, which is thought to provide support for glo-
rather thick, negatively charged structure that most likely merular capillaries. The mesangial matrix consists of col-
FIG. 4. Electron microcraphs show-

ing the glomerular barrier, with the cap-
illary lumen above and the urinary space
below. The endothelial cell surface coat
(ESL) can be visualized with different
techniques (121). A: Cupromeronic Blue
was used to visualize the charged struc-
tures of the barrier. With this technique,
the GBM appears to have a homoge-
neous structure. B: staining with lantha-
num results in higher contrast and
shows “bushlike” structures in the fenes-
trate that extend into the capillary lu-
men. Scale bars: 100 nm.

lagens (type I, III, IV, and V), laminin, fibronectin, and other capillary beds, low concentrations of albumin (⬍1%
proteoglycans with both heparan sulfate and chondroitin of normal) increase permeability to water (63, 191). In the
sulfate chains (189, 215). Mesangial cells secrete several kidney, however, the absence of albumin seems to in-
growth factors, including interleukin (IL)-1, platelet-decrease glomerular permeability to macromolecules as
rived growth factor (PDGF), and insulin-like growth fac- well (92). Orosomucoid (a plasma protein and acute
tor (IGF) (1, 8, 352). PDGF is a potent mitogen for mes- phase reactant normally present in concentrations be-
angial cells. Transforming growth factor (TGF)-␤ has the tween 0.1 and 1.0 g/l in human plasma) has a similar
opposite effect and inhibits cell proliferation; it also stim- permselective effect in several vascular beds (58, 114,
ulates proteoglycan synthesis in both mesangial cells and 115), including the kidney (117, 139). The mechanism is
podocytes in culture. not clear, but it may involve fiber matrix interaction
Mesangial cell proliferation and matrix expansion are (168), endothelial cell production of orosomucoid
seen in several pathological conditions, such as IgA ne- (299), binding of orosomucoid to endothelial cells
phritis (143, 165) and diabetic nephropathy (102, 133, (279), or an anti-inflammatory cAMP-mediated receptor
276). As the mesangial matrix expands, it affects the activation by orosomucoid (300).
glomerular capillaries by reducing the area available for In other vascular beds, plasma proteins increase the
filtration and may eventually even occlude the capillary thickness of the endothelial cell surface coat (often de-
lumen. Matrix expansion due to cell proliferation also noted as glycocalyx) (2). Any sufficiently concentrated
leads to glomerulosclerosis. protein (such as albumin in normal plasma) would be
expected to increase the equilibrium partition coefficient
F. Plasma Components for itself and other macromolecules in glycocalyx and
possibly GBM, which in turn would influence glomerular
The idea that components of plasma influence the sieving (168) (see sect. VI). Thus there seem to be ample
glomerular barrier is not entirely new (164), but it has possibilities for plasma proteins, such as orosomucoid, to
mainly concerned “permeability factors” that induce pro- affect the glomerular barrier.
teinuria (98). Several candidates have been suggested, but Figure 5 shows a schematic drawing of the glomeru-
causal associations have not been demonstrated (34). In lar barrier with focus on the endothelium. Podocytes
FIG. 5. Schematic drawing of the glomerular barrier with components of the glomerular endothelium [e.g., the integrins, Tie2, VEGF receptor
1 (VEGFR1), VEGFR2] and the endothelial cell surface coat (ESL). Several, or all, of the components of the ESL are produced by the endothelium
at certain rates and removed to blood or urine at similar rates during steady-state conditions. Membrane-bound proteoglycans (PG), such as
syndecan and glypican, which carry chondroitin sulfate (CS) side chains (syndecan) and/or heparan sulfate (HS) side chains (glypican and syndecan)
form the glycocalyx. The ESL is formed by secreted proteoglycans such as perlecan (mainly HS) and versican (mainly CS) together with secreted
glycosaminoglycans (GAG) (e.g., hyaluronan) and adsorbed plasma proteins (e.g., albumin and orosomucoid). Several other macromolecules not
shown in the figure are important components of the ESL and the glycocalyx. The podocytes produce substances such as ang1 and VEGF that affect
the endothelial cells. For a more detailed picture of the podocyte biology, please consult Reference 234.

influence endothelial properties by secreting hormones transport. For analysis of macromolecules such as al-
such as vascular endothelial growth factor (VEGF) and bumin, however, certain technical problems emerge.
ANG I. The glycocalyx is composed of the membrane- For example, if samples are obtained from the proximal
bound proteoglycans such as syndecan and glypican. The tubules and not from Bowman’s capsule, the urinary
ESL is a much thicker structure, which also consists concentration of albumin will be underestimated. Ad-
of secreted proteoglycans (e.g., versican and perlecan), sorption of protein to the surface of the glass pipette
hyaluronan, and adsorbed plasma proteins (e.g., oroso- may cause further underestimation. In addition, the
mucoid and albumin). sampling procedure itself may damage the delicate bar-
rier. In an attempt to circumvent some of these limita-
III. METHODS TO STUDY THE tions, Tojo and Endou (317) plotted the albumin con-
GLOMERULAR BARRIER centration against the tubular fluid to plasma concen-
tration ratio (TF/P) of inulin (317). Retropolating to a
Several experimental approaches have been used to TF/P of 1.0 for inulin, they estimated the sieving coef-
unravel the complexities of glomerular permeability, in- ficient for albumin to be 0.0006. Furthermore, their data
cluding urinalysis in vivo, micropuncture of single show that the albumin concentration is underestimated
nephrons, isolated perfused kidneys, tissue-uptake tech- by one order of magnitude when the TF/P is 2. In
niques, isolated glomeruli, isolated glomerular cells, iso- contrast, the normal urinary excretion of albumin in
lated basement membranes, and artificial membranes. humans is ⬍20 mg/day, giving apparent sieving coeffi-
Each approach has merits and drawbacks. cients of ⬎10⫺5, reflecting the extensive tubular modi-
fication of the urinary content.
A. Urinalysis In Vivo
C. The Isolated Perfused Kidney
In vivo urinalysis is readily available for studies in
humans. The use of metabolically inert solutes such as This technique has been used for many years (343)
inulins (oligosaccharides derived from plants) is the gold but with several modifications. For example, the kidney
standard for estimating the glomerular filtration rate can be perfused at normal temperatures (37°C) (30,
(GFR) and for analyzing solute sieving. 180, 257) or at reduced temperatures [4 – 8°C, cooled
It might seem more physiological to analyze endoge- isolated perfused kidney (cIPK)] (173, 240). As perfus-
nous proteins (308, 309), but such data are less useful ate, one can use blood (33, 60, 248) or artificial solu-
(203) due to the extensive reabsorption, possible degra- tions containing albumin (140), dextran (297), or other
dation, or secretion by tubular cells. Attempts have been colloids (38). Tubular modification of the primary urine
made to correct for these limitations, but the accuracy of has been minimized by using chemical agents (222,
the corrections is still uncertain (310, 311). Data concern- 228), low temperatures (140, 213, 240, 302), and chem-
ing glomerular size selectivity in humans are mainly based ical fixation (51–53). Finally, the kidneys can be per-
on studies using dextran (3, 106, 243, 244) or Ficoll (26). fused with recirculating (33) or single-passage sys-
Human glomerular charge selectivity has been studied tems (140).
using sulfated dextran (105). Perfusion with blood is theoretically more physiolog-
Patients with defects in the reabsorption machinery ical but is technically difficult for several reasons, such as
of the proximal tubule (i.e., Fanconi syndrome) excrete the risk for hemolysis, complement activation, and throm-
the proteins filtered across the glomerular wall. In a series bocyte degranulation. Therefore, most researchers inter-
of elegant studies in such patients, modern proteomics ested in glomerular permeability have used artificial so-
approaches such as matrix-assisted laser desorption/ lutions based on albumin or other colloids. However, one
ionization time of flight mass spectrometry (MALDI-TOF problem with a blood-free solution is that the maximum
MS) were used to analyze urine composition (59, 207, amount of oxygen that can be dissolved in water does not
208). The results showed that the megalin-cubulin com- meet the metabolic demands of the kidney, as reflected by
plex in the proximal tubules was impaired (206), and morphological evidence of renal ischemia (173). Fluoro-
protein excretion was increased almost 100-fold (208). carbons can be used (91), but such agents might affect the
This value is compatible with a highly selective glomeru- permeability per se. In the warm IPK, renal hypoxia leads
lar barrier (45– 48, 68, 73, 116, 317). to albuminuria that gradually increases over time. The
kidneys may be better preserved if the perfusate contains
B. Micropuncture of Single Nephrons mannitol and amino acids (180, 333). Kidneys perfused at
reduced temperatures (8 –27°C) do not show ischemic
Micropuncture of single nephrons is an excellent damage (173, 177); nevertheless, albumin clearance in-
technique for studying pressures, flow, and small-solute creased gradually over time (117).

To study glomerular filtration, one must either inhibit theoretical osmotic pressure being equal to the osmotic
tubular cell activity or use inert tracers that are neither reflection coefficient (␴). Therefore, the authors interpret
secreted nor reabsorbed in the tubules (30). Tubular cell changes in glomerular volume in terms of alterations in
toxins, such as lysine (150 mM), ammonium chloride (10 ␴alb. This is not an adequate term, however, since numer-
mM), chlorochine (0.1 mM), and cytochalasin B (0.02 mM) ous factors (e.g., cellular contraction) can affect the glo-
(228), increase glomerular permeability (211) and cannot merular volume apart from the permeability of albumin.
be recommended. The cIPK model is most useful for Moreover, reported ␴alb values of 0.5 or less in response
studies of macromolecular transport, since the equivalent to damage (61, 188, 281, 285, 286) are unrealistically low.
pore dimensions are similar in the cIPK model and in vivo Such values would imply urinary albumin losses for ne-
(123), a finding reported for warm IPKs as well (30) using phrotic patients in the range of several kilograms of albu-
dextran as the inert tracer. min per day (180 l/d ⫻ 40 g/l ⫻ 0.5), far more than is
compatible with life. However, the model may still be
useful as a diagnostic tool in search for proteinuric fac-
D. Tissue-Uptake Techniques
tors in serum (42, 282–284, 318). The diffusional perme-
ability of macromolecules has also been studied in iso-
Glomerular function can also be studied by analyzing
lated glomeruli (64, 65, 81, 82). In certain animals, it has
the tissue uptake of radiolabeled tracers administered
even been possible to microdissect and perfuse single
in vivo. However, this approach assumes that the filtered
glomeruli for studies of glomerular transport (88).
tracer is present in the urine, the tubular system, or the
tubular cells (19), at least shortly after administration
(312, 313). Radioactivity is counted in plasma, urine, and F. Isolated Glomerular Cells
the removed kidneys, allowing for estimates of clearance.
Comparisons between experimental groups of animals
Cells cannot be used to analyze the properties of the
(12) are hampered by the fact that there may be differ-
complex glomerular barrier. However, studies in cells are
ences in hemodynamic conditions and in the number of
crucial for understanding how it is created and main-
nephrons, and hence differences at the single-nephron
tained. Glomerular endothelial (13, 23, 99, 233, 272, 273,
level affecting the filtration rate (179, 251). Such differ-
298, 324), epithelial (125, 199, 200), and mesangial cells
ences may affect the clearance of tracers even in the
(156, 275) have been cultured and studied in isolation.
absence of alterations at the membrane level, a phe-
Human podocytes in culture are shown in Figure 6. Fewer
nomenon sometimes neglected, as will be discussed in
studies have focused on interactions between the differ-
section VI.
ent cell types (e.g., mesangial and endothelial cells) (156).
Glomerular cells produce molecules that may have endo-
E. The Isolated Glomerulus crine and autocrine functions, such as VEGF (66, 103),
and their signaling effects are most likely important for
Isolated glomeruli have been used to estimate albu- glomerular function (79, 85, 125, 142, 182, 304).
min permeability (274), based on osmotically induced To study the importance and contribution of the
changes in glomerular volume. The underlying theory is different cell types in the glomerulus, methods have been
that solutes exert between 0 and 100% of the theoretical developed to separate the various kinds of cells. Glomer-
osmotic pressure across membranes, the fraction of the uli are easy to harvest from kidney samples of humans
FIG. 6. Human podocytes in culture.

A: image obtained by light microscopy
shows small buds of foot processes begin-
ning to form (white arrows). B: actin fila-
ments stained by phalloidin.

and the most commonly used species such as rats and only a few passages, so reliable information can only be
mice. They are readily visualized by light microscopy and gained from low-passage primary cultures. Most of the
may be microdissected from renal tissue. For a more earlier work on endothelial cells and their properties was
large-scale approach, glomeruli can be obtained by pass- done on cells derived from large vessels (e.g., aortic and
ing renal tissue through stainless steel sieves with specific umbilical vein endothelial cells). Although such studies
mesh sizes (13) or by using magnetic beads (305). To have yielded important information, there are differences
single out the different cell types, the glomeruli must be both between species and between different vascular
broken and the cells released (e.g., with collagenase or beds. A recent study comparing bovine aortic endothelial
other enzymes). For human kidney tissue, the age of the cells to glomerular endothelial cells revealed significant
donor is also of importance for the outcome (99), with the transcriptional differences between cells of macro- and
cells from a young donor maintaining their phenotype microvascular origin (280). To circumvent problems with
better. The cells may be cultured and then subcloned, or dedifferentiation, immortalized lines of glomerular endo-
specific cell sorting may be performed through FACS or thelial cells were created.
Percoll gradients. Another important factor is shear stress. Endothelial
Culturing the cells poses different challenges de- cells showed marked differences in expression and mor-
pending on the cell type. Mesangial cells were the first phology in the presence and absence of shear stress (196).
cells of glomerular origin to be cultured. They rapidly Nevertheless, cultures of glomerular endothelial cells are
overgrow other cells in a coculture and are thus easier to an important source of information for cell function. For
maintain in vitro than the other two cell types (89). It is all studies on glomerular endothelium, it is important to
vital that the cultured cells are properly evaluated. Cells characterize the cells carefully using specific markers
in primary cultures may lose their phenotype and stop (Fig. 7).
expressing cell-specific markers. This has lately been ad-
dressed by creating immortalized lines of both human and
H. Proteoglycan and GAG Production
mouse glomerular endothelial cells (272, 331) and podo-
cytes (200, 201, 270). Still, it is important to characterize
the cells for specific markers. Is an endothelial cell with- Proteoglycans are versatile molecules found in di-
out specific markers such as platelet/endothelial cell ad- verse areas of the body. The hallmark of a proteoglycan is
a protein core carrying at least one GAG chain. GAGs are
hesion molecule 1 still an endothelial cell? Is it relevant to
sulfated polysaccharides consisting of repeating disac-
study a podocyte that does not express nephrin and
charide units of uronic acid (glucuronic acid or iduronic
CD2AP?
acid) and amino sugar (galactosamine or glucosamine).
The GAG moieties of the proteoglycans are divided into
G. Glomerular Endothelial Cells groups. Chondroitin sulfate and dermatan sulfate contain
galactosamine and are called galactosaminoglycans. Hep-
Primary cultures of glomerular endothelial cells are arin and heparan sulfate contain glucosamine and are
difficult to work with. The cellular phenotype is lost after therefore called glucosaminoglycans.
FIG. 7. Human glomerular endothelial

cells in primary culture. A: light micro-
scopic image of normal cells. B: cells la-
beled with Dil-Ac-LDL, an endothelial cell-
specific agent. C: cells labeled with an en-
dothelial cell-specific lectin from Ulex
europaeus agglutinin I.

Several enzymes regulate GAG chain synthesis and proteoglycan expression in glomerular endothelial cells
sulfatation. The amount of attached sulfate groups deter- and induced a nephrotic syndrome in rats (24).
mines how negatively charged the GAG will be. The galac-
tosaminoglycans (chondroitin sulfate and dermatan sul-
fate) are modified by sulfotransferases (4-O or 6-O), I. Isolated GBMs
which add sulfate groups to the carbohydrate chain, and
2-O-sulfotransferase, which adds sulfate groups to the Isolated GBM has been used extensively to study the
uronic acid residue. The N-deacetylase N-sulfotransferase transport of fluid and water-soluble solutes (65, 82, 93,
enzyme family is involved in glucosaminoglycan (heparin 130). These studies have been of great importance for our
and heparin sulfate) sulfatation. There are also several understanding of sieving across gels or fiber matrices. In
O-transferases that add sulfate groups to the glucuronic isolated GBM, charge selectivity has been studied using
acid moiety. In addition to the different GAG chains, there dextrans (32) and Ficolls (28), and the effect of plasma
are also several proteoglycan core proteins (Fig. 8). proteins on the selectivity of the GBM has been analyzed
In the glomerulus, the most described proteoglycans (168). When extrapolating data from the isolated base-
are perlecan (131) and agrin (110, 239), both expressed in ment membranes to the GBM in vivo, one must consider
the GBM together with another proteoglycan, collagen the possibility that certain components may have been
VIII (113, 202). Perlecan carries heparan sulfate and some- lost during the isolation procedure (28). Moreover, the
times chondroitin sulfate, but agrin carries only heparan hydraulic conductivity and the sieving properties of iso-
sulfate. These molecules contribute to the negative lated GBM can be modulated by elevating the hydrostatic
charge of the GBM and may thus contribute to the perm- pressure on both sides of the membrane (253). There are,
selectivity of the glomerular barrier. However, mice lack- however, strong arguments to suggest that the GBM has
ing perlecan heparan sulfate chains do not develop pro- similar properties in vivo and in vitro. The hydraulic
teinuria or any other renal defects (258). This may be due conductance of isolated GBM accounts for most of the
to compensatory mechanisms, but more likely it further fluid restriction of the intact glomerular barrier (65, 73).
strengthens the findings that the GBM alone cannot be Therefore, it seems safe to conclude that the GBM nor-
responsible for the permselective properties of the glo- mally restricts water transport but only marginally re-
merular barrier. Proteoglycans are present on the surface stricts the passage of solutes (28).
of endothelial cells and podocytes and in the mesangial
cell matrix as well. Cultured immortalized human podo-
cytes express both mRNA and protein for syndecan-1, J. Artificial Membranes
versican, and perlecan (24). The negative charges on
podocytes are thought to contribute to charge selectivity, Synthetic membranes of various types have been
podocyte stability, and signaling. Bovine glomerular en- used to examine the solute and membrane properties that
dothelial cells synthesize perlecan, a synthesis that may govern the transport of proteins and other macromolec-
be altered by treatment with TGF-␤ (148) or glucose ular solutes. Results for pores or microchannels of pre-
(109). Biglycan is expressed by glomerular endothelial cisely controlled dimensions have been reviewed (67, 69).
cells in tissue sections, and human glomerular endothelial The desire to better understand and mimic materials such
cells in culture express several proteoglycans (syndecan, as GBM has motivated several studies of transport
versican, glypican, and perlecan) along with their regula- through hydrogels, including agarose and agarose-dextran
tory enzymes (sulfotransferases) (23). In addition, treat- composites (138, 141, 158 –160, 167). Such studies are
ment with puromycin aminonucleoside downregulated gradually refining our understanding of the physics of
FIG. 8. Glomerular cells produce

various proteoglycans. These images
are from human podocytes in culture.
A: production of perlecan, labeled
green with specific antibodies. B: perle-
can (green) and actin are labeled red
using phalloidin.

hindered transport through networks of cross-linked urine are shown in Table 1. Micropuncture of rat proximal
polymers with fluid-filled interstices. tubules (317) gives values for ⌰alb (assuming a serum
albumin concentration of 30 g/l) that are 10-fold higher
IV. SOLUTES USED TO STUDY than estimates from patients with Fanconi (Dents) syn-
GLOMERULAR PERMEABILITY drome (i.e., 8 ⫻ 10⫺5) (208). The latter value represents a
lower limit for the sieving coefficient, since there still may
be a residual reabsorption of albumin. Lower values (3 ⫻
A. Theoretically Ideal Solutes 10⫺4) were obtained in vivo in rats in which tubular
reabsorption was (partially? Ref. 314) inhibited with
The ideal solute for studies of transport is a solid lysine (310). The ⌰alb values from cIPK (8°C ) are two- to
inert sphere that does not affect the organism in any way threefold higher (122, 123, 176, 301, 302), but still close to
and is not metabolized during the observation period. For
0.001. In contrast, estimates from IPK at 37°C are 10 times
studies of glomerular charge selectivity, solid spheres
higher (225, 228) and increase further, to 0.08, after treat-
should carry a known net charge without distorting mo-
ment with a tubular cell inhibitor (228). The latter value is
lecular size or shape and should remain metabolically
not compatible with the notion of a highly selective glo-
inert. To assess the effect of shape, the solid molecules
should be made ellipsoid in a controlled and stable man- merular barrier and gave rise to the so-called albumin
ner. Unfortunately, the ideal solute does not exist. There- retrieval hypothesis (see sect. VII).
fore, in interpreting the results of clearance studies with Glomerular sieving data for other proteins during
“real world” solutes, their nonideal properties must be controlled tubular cell activity are scarce. In vivo, lysine
taken into account. At present, Ficoll seems to be closest has been administered to produce proteinuria by inhibit-
to the “ideal” solute (see sect. VI). ing tubular cells (310). However, the albumin sieving co-
efficient calculated from such data is only half of that
obtained with micropuncture (317) or tissue-uptake tech-
B. Proteins niques (19, 179) or from patients with Fanconi syndrome
(209) (Table 1). Indeed, lysine has variable and incom-
Endogenous proteins are highly relevant to study, but
plete effects, and its mechanism of action is still unclear
they deviate from the ideal solute in many respects. By
(314). Therefore, the sieving data for ␤2-microglobin,
definition, proteins are not inert, are never solid, seldom
orosomucoid, IgG, and ␣2-macroglobin are likely to be
have spherical shapes, are often compressible, and are
underestimated as well (310). To circumvent some of the
made of amino acids with a neutral, negative, or positive
charge, giving them a certain pH-sensitive net charge. To problems with administration of lysine, Lund et al. (179)
complicate things further, charges may be expressed on used a tissue-uptake technique to estimate glomerular
the surface or not, and the protein core may carry GAG sieving of certain proteins. Although attractive, this tech-
chains. The most common issue with respect to glomer- nique is an indirect approach based on certain assump-
ular filtration is, however, the tubular handling of pro- tions, as described in detail elsewhere (312). The clear-
teins. The most reliable estimates of the glomerular siev- ance data from tissue-uptake studies (18, 19, 179) are
ing of proteins come from studies in which reabsorption similar to those derived from measurements using neutral
is controlled. Ficoll in vivo (26, 219). These values in turn are similar
Estimated sieving coefficients for albumin (⌰alb) to those from cIPK studies (122, 123, 176, 301, 302)
from studies controlled for tubular modification of the (Table 2).
TABLE 1. Reported sieving coefficients for albumin in studies with “controlled” tubular modification
Species Albumin Sieving Coefficient Technique Reference Nos.
Rat 0.00033 In vivo ⫹ lysine 310

Rat 0.00066 Tissue uptake 179
Rat 0.00080 Micropuncture 317
Humans, Fanconi 0.00008 Urine proteomics 208
Rat 0.0015 cIPK 123, 213, 214, 301, 302
Mouse 0.0023 cIPK 134
Rat 0.0080 IPK no tubular cell inhibitor 228
Rat 0.0087 Fixed kidney 53
Rat 0.0450 IPK after tubular cell inhibitor 225
Rat 0.0800 IPK after tubular cell inhibitor 228

TABLE 2. Sieving coefficients for various proteins in selected studies
Type of Protein SE Radius Sieving Coefficient Charge References
Neutral or slightly cationic proteins

nAlbumin 35.5 0.0330 0 228
nAlbumin 35.5 0.0260 0 19
Kappa-dimer 28.4 0.1490 0 179
nHRP 32.0 0.0700 0 348
nHRP 32.0 0.1100 0 302
nMyoglobin 19.6 0.7700 0 179
cMyoglobin 17.5 0.7400 2 348
IgG 54.0 0.0023 0 18
LDH-5 46.0 0.0056 2 176
Anionic proteins
Albumin 35.5 0.0021 ⫺23 176
Albumin 35.5 0.0015 ⫺23 214
Albumin 35.5 0.0006 ⫺23 19
Albumin 35.5 0.0007 ⫺23 179
Ovalbumin 27.4 0.0770 ⫺13 301
Orosomucoid 40.5 0.0036 ⫺24 301
aHRP 32.0 0.0690 ⫺6 301
aHRP 32.0 0.0450 ⫺11 302
aHRP 32.0 0.0170 ⫺11 224
aHRP 32.0 0.0100 ⫺14 348
aMyoglobin 17.5 0.5100 ⫺6 348
LDH-1 46.0 0.0011 ⫺19 176
C. Dextran Polymers D. Ficoll Polymers
Dextran has been extensively used for transport Ficolls are neutral, heavily cross-linked, sucrose-epi-
studies in the kidney and other organs (250). It is an chlorohydrin copolymers. The molecules behave as rigid
␣-D-1,6-glucose-linked polymer with a few short side hydrated spheres and can be labeled with isotopes (26) or
branches that behave as a flexible hydrated random coil fluorescent dyes (213). The normal molecular Stokes-Ein-
rather than as a solid sphere. In the kidneys, dextran was stein radius for Ficoll 70 is 10 –70 Å (213). Ficolls are inert
used 30 years ago in the classical studies of glomerular and do not seem to be reabsorbed, secreted, or degraded
size and charge selectivity (17, 36, 45– 48). However, in kidneys in vivo. Indeed, neutral Ficoll is close to the
pore dimensions based on neutral dextran sieving data ideal solute for studies of transport across biological and
were overestimated (219), owing in part to dextran’s artificial membranes. Ficolls, including sulfated Ficoll, do
flexibility (218). Thus, compared with a rigid solute not bind to plasma proteins (28). Whereas sulfated Ficoll
such as Ficoll or a globular protein, the sieving coeffi- seems to behave according to hydrodynamic laws (28),
cient of dextran may be an order of magnitude too there have been reports of anomalous behavior of car-
high (219). boxymethylated Ficoll (10, 107). The mechanism respon-
Sulfated dextran raises other concerns. Notably, it sible for the high clearance for carboxymethylated Ficoll
does not appear to be fully inert. Some batches of dextran remains to be elucidated (10, 107). The sieving coeffi-
sulfate seem to bind to plasma proteins, and certain tu- cients for Ficoll from intact rats in vivo (219) and cIPK in
bular cells may bind and incorporate the solute. As a mice (134) are plotted against their molecular Stokes-
result of either effect, the concentration in urine will be Einstein radius in Figure 9.
underestimated, and charge selectivity will be overesti-
mated. Indeed, the glomerular charge density estimated
from studies of dextran sulfate is ⬃120 –170 meq/l (76) E. Proteins Versus Synthetic Polymers
but only 30 –50 meq/l in studies of proteins (176, 302, 346,
347) or Ficoll (123, 134, 301). Other elongated and/or Some investigators suggest that proteins behave dif-
flexible molecules that behave like dextran have sieving ferently from Ficoll in their transport (264) across biolog-
coefficients larger than expected from their Stokes- ical and artificial membranes (335). This view is based on
Einstein radii, such as hyaluronan (214) and bikunin three arguments: 1) that Ficoll and protein glomerular
(175, 214). Thus there are good reasons to avoid dex- sieving data differ; 2) that Ficoll may not be uncharged;
tran for studies of transport across biological mem- and 3) that the Ficoll solute radius may change in re-
branes. sponse to changes in ionic strength.

FIG. 9. Glomerular size selectivity is illus-

trated by a plot of sieving coefficients for neu-
tral Ficoll versus molecular radius. Biological
data from two studies are presented together
with simulated curves. The experiments were
performed on intact rats (E) with a lognormal
pore distribution model (219) and data from
cIPK in mice (F) together with a fiber model
(134). Both models describe the biological
data with reasonable accuracy, as does a two-
pore model (212, 213, 301) not shown in the
graph.
Concerning the first argument, the glomerular sieving during routine serum protein electrophoresis and be-
data depend on the Peclet number (see sect. VI), and it haves as a neutral solute even though FITC has a charge
may be difficult to compare observations obtained under of ⫺1 (cited in text in Ref. 301). Concerning the third
different experimental conditions. It would be more rele- argument, HPLC gel filtration of sharp fractions of
vant to determine if neutral spherical proteins differ from Ficoll at normal, low, and high ionic strengths (as used
Ficoll under controlled conditions, such as using artificial in Refs. 301, 302) showed that the Ficoll Stokes-Ein-
membranes or isolated GBM. Indeed, such studies suggest stein radius was rather stable. It was not significantly
that proteins and Ficoll are rather similar (28, 71, 136, altered by reducing the ionic strength to 20 mM (⫺0.02 ⫾
141). Furthermore, neutral proteins seem to behave sim- 0.18 Å, n ⫽ 10) or by increasing it to 520 mM (0.61 ⫾ 0.26
ilarly to Ficoll in IPKs if studied under similar and stan- Å, n ⫽ 10, not significant) (B. Haraldsson and J. Nyström,
dardized conditions (Fig. 10). unpublished observation).
The second argument, that Ficoll may be charged We conclude that Ficoll in many respects behaves as
(335), is easily addressed. FITC-Ficoll moves slowly an ideal solute. Its transport across artificial and biologi-
FIG. 10. Glomerular sieving coeffi-

cients for neutral Ficoll, neutral proteins,
and anionic proteins plotted against mo-
lecular radius. Two Ficoll sieving curves
are presented based on data from cIPK
studies in rats (301) and mice (134). The
neutral and anionic proteins can be found
in Table 2 together with references. The
anionic proteins differ considerably in net
charge, even for the same protein, depend-
ing on the experimental conditions (see,
e.g., anionic HRP in Table 2).

cal membranes is quite similar to that of uncharged spher- both size and charge selectivity (52) and it fits well with
ical proteins. biological data (Fig. 9; see also Fig. 1 in Ref. 134). The
equations are well developed for the partition coefficient,
but the model is less mature in terms of estimations of the
V. SOLUTE PROPERTIES THAT AFFECT
reflection coefficients (see Ref. 351 and below for more
TRANSGLOMERULAR PASSAGE
information). The size selectivity seen in Figure 9 for
Ficoll is from three different studies from rats in vivo
A. Molecular Size (219), rat cIPK (301), and mouse cIPK (134). Furthermore,
Figure 10 shows that neutral proteins behave as neutral
Solutes are retarded in their passage across the glo- Ficoll (the two curves are from Refs. 134, 301).
merular barrier depending on their molecular size. This is In Table 2, the sieving coefficients for various neutral
evident from a large number of human and experimental proteins plotted in Figure 10 are given together with the
studies using dextran (47), Ficoll (26, 213, 219), and pro- reference. There are two values for “neutral” albumin, one
teins (179). Glomerular size selectivity can be expressed obtained with a tissue-uptake technique (19) and the
in terms of several different theoretical models. Accord- other with IPKs (228). The latter value is likely to be an
ing to the two-pore theory, the equivalent small pore underestimate, since tubular cell activities were not com-
radius is close to 45–50 Å (26, 212, 213), the large pore pletely blocked. The charge-modified albumin may form
radius is close to 80 –100 Å (176, 212, 213, 310), and the dimers, spontaneously regain its negative charge, and
large pore pathway represents 0.003 or less of the total bind to tissue structures, as described in detail elsewhere
hydraulic conductance (213). For the lognormal pore dis- (108). Indeed, values that are only 20% of those of Refer-
tribution model, the mean pore size seems to be close to ence 228 have been reported for neutral albumin (179).
35– 45 Å, with a distribution parameter of 1.3–1.1; the From the latter study, the value for ␬-dimer is obtained
lower pore radius value is then combined with the together with data for neutral myoglobin. Similar data for
broader distribution parameter (26, 213, 219) (Fig. 9). neutral horseradish peroxidase (HRP) have been reported
Finally, glomerular size selectivity has been described by in some studies (246, 302, 348), and 30 –50% lower values
the charged-fiber model, according to which the fiber in others (179, 222). Finally, data have been reported for
volume is 6 –7% (52, 134) (Fig. 9). All these theoretical IgG (18), lactate dehydrogenase-5 (176), and slightly cat-
models describe the sieving of solutes more or less to the ionic myoglobin (348). The protein sieving data of Lund
same degree. However, they evolved over time, and only et al. (314) are systematically lower than expected. As
the latter model incorporates both size and charge effects. described above, the inhibition of tubular uptake by lysine
As mentioned above, dextrans give falsely high val- infusion seems to be incomplete and variable (314), lead-
ues for the small pore radius, which typically is estimated ing to underestimations of the transglomerular passage of
to be ⬃50 – 60 Å (37, 242, 263), almost 10 Å greater than proteins. We conclude that neutral Ficoll and neutral
the “true” equivalent pore radius. However, pore size can proteins behave similarly in their passage across the
be markedly underestimated as well. Reports based on highly size-selective glomerular barrier.
protein analysis in vivo suggest a small pore radius of 30
Å or less (310). In the latter study, the work of Remuzzi
B. Molecular Charge
et al. (242) was cited erroneously as supporting such a
small pore size. The underestimation of the pore size
The first experimental evidence of capillary charge
reflects the limitations of in vivo protein analysis, even in
selectivity was reported in 1969 by Areekul, who used
the presence of tubular inhibition with lysine (310),which
neutral and sulfated dextran (7). Several now-classical
is only partial (314). Indeed, using a tissue-uptake tech-
studies suggested strong charge selectivity in the kidney,
nique, authors from the same group estimated the pore
with anionic macromolecules being more retarded than
radius to be 38 Å (179). Naturally, the estimates of pore
their neutral counterparts of similar sizes (35, 36, 45– 48,
radii depend on the biological conditions during the ex-
76). Similar effects were found using anionic, cationic,
periment, and occasionally low pore radii are found even
and neutral HRP (246). This view prevailed for decades. In
with Ficoll (217).
1991, however, the use of dextran sulfate was called into
The studies above indicate that, for neutral solutes,
question (306). This led to a series of critical papers
the sieving coefficient decreases as molecular size in-
questioning the very existence of glomerular charge se-
creases. This relationship can be described adequately
lectivity.
with at least four different transport theories: the two-
pore model, the lognormal pore distribution model (with
1. Is there any glomerular charge selectivity?
or without shunt), (neutral) fiber matrix theory, and the
negatively charged fiber model (136, 137). As mentioned Over the past decade, a large number of studies,
above, the negatively charged fiber model incorporates mainly from one laboratory, have suggested that glomer-

ular charge selectivity does not exist at all (264). In argu- 10-fold for anionic albumin (228). The fact that the tubular
ing against a glomerular charge barrier, Comper and col- cell inhibitors raised the clearance for the neutral solutes
leagues make several assertions: 1) that all studies using and for the anionic proteins suggests toxic effects on
dextran sulfate are invalid, due to problems with desulfa- glomerular size selectivity. Indeed, toxic effects on glo-
tion in the urine (57), binding to plasma proteins, glomer- merular size and charge selectivity were found by Ohlson
ular cell uptake, and binding to GBM (306, 340, 341); 2) et al. (213), which seriously questions the validity of the
that their own data on the glomerular clearance of neutral original interpretations (222, 224, 228).
and anionic HRP (aHRP) (222) and of neutral and charge- The third argument, that albumin is filtered in large
modified albumin (228) provide evidence against a charge quantities and reabsorbed in intact form, is dealt with
barrier (264); 3) that albumin is filtered in large quantities below (see sect. VII). It seems safe to conclude that there
across the glomerular wall and reabsorbed in intact form are ample experimental data to refute the hypothesis.
(228); 4) that proteins are degraded in the urine and The fourth argument is that proteins such as albumin
excreted as small peptides not detected by most routine are degraded (223, 227), causing protein excretion to be
techniques (223, 227); 5) that anionic dextrans and Ficolls underestimated (100). The validity of this argument has
behave anomalously when filtered across glomerular been tested (209) and is discussed in detail in section VII.
walls (10, 41, 107); and 6) that evidence of charge selec- The fifth argument is that anionic dextrans and Fi-
tivity is artifactual, due to limitations of the investigative colls behave anomalously when filtered across glomerular
techniques, mainly cIPK (264). walls. The properties of Ficoll and dextran have been
Are any of these arguments valid? Let us start with discussed earlier in this review. There also have been
the questions regarding dextran sulfate. Indeed, dextran reports of anomalous behavior of dextran sulfate (41),
sulfate has been shown to bind to fibronectin (210) and to mainly suggested to be due to increased desulfation in
plasma proteins, depending on the molecular weight of diabetic animals. For charge-modified anionic Ficoll, the
the dextran molecules (105). In the latter study, after clearance is reported to be even more anomalous (107).
measurement and correction of such effects, Guasch et al. Thus carboxymethylated Ficoll and neutral Ficoll have
(105) still found evidence of a negative charge barrier. similar clearances for solutes with a Stokes-Einstein ra-
The data on cellular uptake and binding of dextran sulfate dius ⬍50 Å; however, for larger molecules, the sieving
by glomerular cells and its subsequent desulfation were coefficients differed by more than 100-fold (at 75 Å) (107).
reviewed recently in detail (73). In brief, the cellular Indeed, similar findings of facilitated transport of car-
processing adduced by Comper and co-workers (266) is boxymethylated Ficoll were reported over a broad molec-
rapidly saturable and therefore would not influence ular size range (15– 80 Å) (10). Both papers conclude that
steady-state clearance data, such as those reported by anionic Ficoll is unsuitable for studies of solute transport
Guasch et al. (105). We conclude that the use of dextran (10, 107). At least two important questions remain unre-
sulfate overestimates the charge density of the glomerular solved. First, is there active tubular secretion of car-
barrier, but the data are qualitatively correct. boxymethylated Ficoll? Second, is the anomalous behav-
Next, let us examine the second argument, that the ior specific for carboxymethylated Ficoll or does it hold
data on neutral HRP (nHRP), aHRP, and albumin provide also for the Ficoll sulfate used in studies of the GBM (28)?
evidence against a negative charge barrier. The fractional Both these question are crucial, since Ficoll sulfate actu-
clearance of nHRP was 2.4 times higher than that of aHRP ally seems to behave quite like neutral Ficoll apart from
under control conditions (224), a ratio identical to that the net charge (28).
reported by Sörensson et al. (302) but less than the ratio In the sixth argument, all data supporting the concept
of 7 (348) or 9 (246) reported in earlier studies. Adding of a negative charge barrier are dismissed as simply being
150 mM lysine (n ⫽ 4) or 10 mM NH4Cl (n ⫽ 4) to the flawed (107, 264). Still, the fractional clearance for albu-
perfusate elevated the fractional clearance of nHRP by min at 37°C reported by Ohlson et al. (213) is close to the
70% and of aHRP by 170%, giving a residual charge selec- value of 0.0075 reported by Comper et al. (228) during
tivity ratio of 1.6 (224). Also for albumin, the original data control situations in the absence of tubular inhibitors.
of Osicka et al. (228) do suggest charge selectivity; in the However, the evidence for glomerular charge selectivity
IPK perfused at 37°C, the fractional clearance was 0.0075 does not rest solely on experiments with the cIPK, com-
for native albumin and 0.0330 for neutral albumin (228). paring neutral Ficoll and anionic proteins, as suggested by
The clearance of native albumin is 5–10 times higher than some authors (10, 107, 264, 335). In Table 2 and Figure 10,
values reported for cIPK or in vivo with controlled tubular data from several experiments are presented for neutral
reabsorption (Table 1). Similar levels of clearance were and anionic solutes together with sieving data for neutral
reported for neutral albumin by Bertolatus and Hunsicker Ficoll. The charge-selective effect is obvious, even though
using a tissue-uptake technique (19), but their clearance the observations were obtained under different condi-
ratio was 43. After treatment with 150 mM lysine, fractions. We conclude that the evidence against a negative
tional clearance increased 3-fold for neutral albumin and glomerular charge barrier is weak at best. Let us there-

fore review recent data regarding glomerular charge ular cells have a key role in charge selectivity (65). In-
selectivity. deed, our current understanding of the GBM indicates
that it has little charge selectivity (28).
2. Current understanding of glomerular Attempts have been made to alter the glomerular
charge selectivity charge barrier by using enzymes to digest GAGs (4, 101,
145, 256). Heparan sulfate proteoglycans are major
Figure 10 illustrates molecular sieving as a function components of the GBM (255); in particular, perlecan
of Stokes-Einstein radius for neutral Ficoll and for several (197, 258) and agrin (161, 255) have been reported to
neutral and anionic proteins. The data were from several affect glomerular permeability. The endothelial cell sur-
studies using a variety of techniques, but with more or face coat (339) also contains GAGs. Heparanase, chon-
less controlled tubular activity. It is evident that the neu- droitinase, and hyaluronidase may increase the sieving
tral proteins are close to the Ficoll sieving curve, reflect- coefficient for albumin by reducing the charge density
ing similar properties for neutral proteins and Ficoll. This (134) and reducing the thickness of the endothelial cell
is in line with data obtained during highly controlled surface coat (135). In a recent report, mice with adria-
conditions in artificial membranes or isolated GBM (28, mycin-induced nephrosis expressed heparinase in their
73). The sieving coefficients are considerably lower for all glomeruli, which the control mice did not (161). We
anionic proteins than for neutral proteins, providing conclude that the data obtained with various GAG-
strong support for glomerular charge selectivity. Thus, degrading enzymes support the notion of glomerular
irrespective of theoretical modeling, the net charge of a charge selectivity.
molecule dramatically affects its transport. In Table 2, Let us use one of the theoretical models to predict
sieving data for several neutral and anionic proteins are the glomerular sieving coefficient for solutes of different
shown. molecular size and charge. As mentioned below, all cur-
Please note that, before correction, the raw data of rent models have limitations. In a recent paper, a hetero-
Comper and co-workers (222, 228) actually support the geneous charged fiber model was used to analyze the
notion of charge selectivity (Table 2). The authors beneficial effects of orosomucoid on rats with puromycin-
reached other conclusions based on the effect of the induced nephrotic syndrome (122). In Figure 11, the results
tubular inhibitors, which reduced the clearance ratios of such an analysis are presented as a plot of sieving
(neutral over anionic) for albumin and HRP (222, 228). coefficient versus Stokes-Einstein radius. Curves are
There are, however, reasons to believe that the agents shown for neutral solutes, slightly positively or negatively
affect glomerular permeability per se (213). Indeed, the charged solutes, and solutes with strong negative charge
reported dextran sieving values cover a rather narrow densities similar to that of serum albumin. Naturally,
range of molecular radii, and the sieving coefficients are these curves represent an oversimplification of the glo-
somewhat larger than reported for intact animals (47). merular barrier, due to the limitations of current transport
This is unexpected because perfused kidneys normally equations for a charged fibrous network. There is, how-
have a lower area available for diffusion, which reduces ever, a reasonable fit between the theoretical curves and
the sieving coefficient for small and intermediate-sized the biological data presented in Figure 9.
solutes (123). On the other hand, inhibiting tubular cell
activity by reducing the temperature to 4 – 8°C lowers the
urine-to-plasma concentration ratio for Cr-EDTA from C. Configuration
20 –100 to 1.1–1.2 without significantly altering glomerular
charge and size selectivity (213). The effect of molecular shape on sieving was recog-
The concept of glomerular charge selectivity has nized early by Rennke and Venkatachalam (247), who
been experimentally tested by altering the ionic strength noted a sevenfold greater sieving coefficient for dextran
and pH of the perfusate. Sörensson et al. (302) found that than for nHRP. It was later found that dextran is less
reducing the ionic strength from 152 to 34 mM lowered suitable for studies of permeability, since it behaves as a
the sieving coefficient (⌰) for aHRP by 40%; ⌰ for nHRP flexible random coil (218) (see sect. IV). The effect of
was reduced by 15% (302). Similar effects were found for solute shape is not, however, limited to dextran (72). The
other anionic proteins, including orosomucoid and albu- plasma protein bikunin is an extremely elongated anionic
min (301). Moreover, reducing the pH of the perfusate protein with a diffusion constant (and hence a Stokes-
increased the sieving coefficient significantly, in relation Einstein radius) close to that of albumin (290). However,
to the diminished net negative charge of the protein, and bikunin crosses the glomerular barrier 80 times faster
pretreatment with protamine had a similar effect (53), than albumin, despite similarities in molecular size and
consistent with previous observations (149, 334). Prota- charge (175). Furthermore, the sieving coefficient for the
mine has prominent effects on the intact glomerular bar- GAG hyaluronan was three times that of bikunin and 240
rier but not on isolated GBM (64), suggesting that glomer- times that of albumin despite similarities in size and

FIG. 11. The fractional clearance or sieving coefficient of a solute plotted against its Stokes-Einstein radius. From the top, the curves represent
slightly positive solutes (with charge densities of 0.002 Coul/m2), neutral solutes, negatively charged solutes (with charge densities of ⫺0.006
Coul/m2), and highly negative solute charge similar to that of albumin (⫺0.022 Coul/m2). The clearance ratio of neutral over strongly negative solutes
was 20:1 or higher in the molecular radius range of 30 – 40 Å. The curves are based on a heterogeneous charged fiber model (122) based on the work
of Johnson and Deen (136), with a fiber radius of 4.5 Å, a relative fiber volume of 0.055, and a fiber charge density of ⫺0.2 Coul/m2. The charged
fiber gel consists of two parallel regions that differ in fiber density. One of the regions has only 10% of the fiber density of the rest of the gel. However,
this “nonselective” region constitutes only a minute fraction of the total gel and contributes to ⬃0.1% of the total hydraulic conductance of the
charged fiber gel. In other models, such high-permeable regions have been named “shunt pathways” or “large pores.” For more detail, please consult
http://kidney.med.gu.se/9.
charge (214). In the latter study, the frictional ratios for The molecular radius interval between 30 and 40 Å is also
anionic solutes were related to their “apparent neutral where charge selectivity is greatest. Thus a solute charge
molecular radii” (214). The frictional ratio is 1.0 for a density similar to that of albumin reduces the sieving coef-
sphere (like Ficoll) and increases with the degree of ficient for those solutes to 5–10% of that for a neutral solute.
elongation (1.3 for albumin, 1.8 for bikunin, and 2.3 for The charge effect will depend on the charge density
hyaluronan). As calculated by Ohlson et al. (214), the ap- of the solute and will occur over the entire molecular size
parent neutral molecular radii for the four solutes are, range. The molecular configuration may, however, over-
respectively, 35.5, 55.1, 33.0, and 23.5 Å. We conclude rule the effects of size and charge, and elongated solutes
that molecular configuration substantially affects sol- of the size and charge of albumin may have sieving coef-
ute transport across the glomerular barrier if the sol- ficients between 0.1 and 1.0 (214). Therefore, to predict
utes are elongated or random coils (71, 73, 214, 247). the transglomerular passage of a solute, one must know
However, smaller deviations from the ideal sphere its molecular size (diffusion constant, i.e., Stokes-Einstein
shape do not seem to affect transport to any significant radius), its net charge (or charge density), and its config-
degree (28). urations, as revealed, for example, by its frictional ratio
(214). For solutes that are not random coils or markedly
elongated, shape does not significantly affect their trans-
D. Relative Importance of Size
port (28). The sieving coefficients presented in Figure 11
and Charge Selectivity
are therefore reasonably accurate for most solutes and
predict the effects of molecular size and charge.
Size selectivity is most significant for neutral sol-
utes in the steep portion of the sieving curve (i.e.,
Stokes-Einstein radius of 30 –50 Å; see solid curve in VI. PHYSICAL PRINCIPLES
Fig. 11). In that part of the curve, the sieving coefficient AND THEORETICAL MODELS
falls by one order of magnitude for every 10-Å increase
in molecular radius. For neutral solutes up to the SE- In reviewing the physical principles that govern the
radius of 30 Å, the sieving coefficient only falls from 1.0 to interpretation of data on the glomerular permeability to
0.1. Also, increasing the molecular radius from 50 to 70 Å proteins and other macromolecules, our emphasis is on
reduces the sieving coefficient by a factor of five (Fig. 11). factors that influence the sieving coefficient of a given

molecule. For ultrafiltration processes in general, the A. Hindered Transport and Membrane
sieving coefficient is the concentration in the filtrate di- Size Selectivity
vided by that in the retentate; for the glomerulus, it is
usually defined as the concentration in Bowman’s space The defining feature of an ultrafiltration membrane is
divided by that in arterial plasma. The sieving coefficient that it can sieve macromolecules on the basis of molecu-
of a solute is determined by multiple factors: the intrinsic lar size. Solutes rejected in ultrafiltration have molecular
selectivity of the membrane based on solute size, charge, masses of ⬃1–1,000 kDa. However, the linear dimensions
and shape; the filtration rate of water; solute concentra- of a molecule are more directly relevant to its ability to
tions (if the retentate is not dilute); and, for a multilayer pass through a small pore. The most common such mea-
barrier such as the glomerular capillary wall, the arrange- sure of molecular size is the Stokes-Einstein radius (rs),
ment of the layers. which is related to the diffusivity in bulk solution (D⬁) as
In general, one must distinguish between the concen-
tration adjacent to the upstream surface of an ultrafiltra- kBT
rs ⫽ (1)
tion membrane and the average concentration in the re- 6␲␮D⬁
tentate. In any sieving process, the concentration of re-
jected molecules tends to increase next to the upstream where kB is Boltzmann’s constant, T is temperature, and ␮
membrane surface, a phenomenon called concentration is the viscosity of the solvent. This is the radius of a rigid
polarization. This increase drives the diffusion of rejected sphere that would have the given diffusivity. For water at
solutes back into the bulk retentate. In industrial ultrafil- 37°C, a molecule with rs ⫽ 1.0 nm has a diffusivity of
tration, concentration polarization often leads to a D⬁ ⫽ 3.3 ⫻ 10⫺10 m2/s. For the glomerulus, the range of
marked reduction in water filtration rates, due to the sizes of most interest for sieving is 2 ⱕ rs ⱕ 6 nm (as
osmotic effects of the retained solute and/or membrane discussed in sect. IV), with rs ⫽ 3.6 nm for serum albumin.
fouling (349). However, an analysis based on the condi- The local solute flux (N) in a porous or fibrous mem-
tions in glomerular capillaries indicates that the effects brane can be written as
there are negligible, because of the small capillary dimen-
sions, moderate filtrate velocities, and mixing effects of dC
N⫽⫺KdD⬁ ⫹ KcvC (2)
red blood cells (75). Accordingly, concentration polariza- dx
tion in the capillary lumen will not be considered here,
although an analogous phenomenon within a multilayer where C is concentration, x is position, and v is fluid
membrane will be discussed. velocity. This is simply Fick’s first law with a term added
Another key distinction involving retentate concen- for convection (bulk flow) and allowances made for hin-
trations concerns their variations along the length of a drances to solute movement caused by the membrane.
permeable tube, such as a glomerular capillary. As plasma The coefficients Kd and Kc are hindrance factors that
moves from the afferent to the efferent end, the concen- describe the effects of the membrane on solute diffusion
tration of a selectively retained solute must increase be- and convection, respectively. Thus the apparent diffusiv-
cause of the proportionately greater removal of water; the ity within the membrane is KdD⬁, and the solute velocity
greater the filtration fraction of water and the smaller the due to bulk flow is Kcv. When Equation 2 is applied to
membranes with discrete pores, N, v, and C are inter-
sieving coefficient, the larger the increase. Such axial
preted as averages over the pore cross-section; for mate-
variations in concentration have been incorporated into
rials of more complex structure (e.g., arrays of fibers with
glomerular filtration models since the early 1970s (48, 72,
fluid-filled interstices), N and v are based on total surface
74, 78), and they have been important for understanding
area and C is usually based on total volume (fluid plus
the effects of plasma flow rate and other hemodynamic solid). If C is in molar units (mol/m3), then N has the units
variables on single-nephron GFR and sieving coefficients. mol䡠m⫺2 䡠s⫺1. Alternatively, mass units can be used for C
Thus a complete model of the glomerular filtration pro- and N.
cess must include mass balance equations that describe The basis for membrane size-selectivity is described
concentration variations along a capillary. Nonetheless, by theories that predict the values of Kd and Kc from rs
we have chosen not to discuss axial concentration and the quantities that define the membrane nanostruc-
variations in detail and to focus instead on the deter- ture. A thorough discussion of such theories is beyond the
minants of the local sieving coefficient (i.e., the filtrate- scope of this review, but an introduction to the main ideas
to-plasma concentration ratio at a given position along will make the subsequent material on sieving more under-
a capillary). Those local factors are most crucial for standable. In general, theories for hindered transport are
understanding barrier selectivity and are also most con- based on a combination of hydrodynamic and steric con-
troversial. siderations. A macromolecule in solution behaves in part

as a hydrodynamic particle, and the presence of fixed other words, diffusion is severely hindered well before
objects (e.g., pore walls or membrane fibers) tends to the solute becomes a tight fit in the pore.
increase the hydrodynamic drag on a particle and thereby The pattern for convection is more complicated, in
reduce its mobility (lowering Kd). Interactions between that Kc first increases and then decreases as solute size
the particle and the wall (or fiber) are mediated by vis- increases. The initial increase is due to the fact that the
cous stresses and pressure variations in the fluid. Such finite size of the particle excludes its center from the layer
hydrodynamic interactions also cause the velocity of a of fluid next to the pore wall, and therefore prevents it
freely suspended particle to deviate from the mean fluid from experiencing the most slowly moving fluid. Eventu-
velocity (affecting Kc). The finite size of the particle limits ally, though, the hydrodynamic retardation due to the
the positions that its center can occupy, bringing steric pore wall dominates, and Kc declines. Note that convec-
considerations into the calculation of the hindrance fac- tion tends to be less hindered than diffusion, such that
tors. Kc/Kd ⱖ 1, with that ratio increasing with relative solute
The theory of hindered transport is most completely size. A large value of Kc/Kd tends to aid sieving, as will be
developed for the special case of spherical molecules in seen. The functions H(␭), W(␭), and ⌽(␭), which are also
long cylindrical or slit-shaped pores; diffusion and con- shown in Figure 12, will be discussed shortly.
vection in fibrous media are less completely understood. Although the results in Figure 12 are for cylindrical
The theory for uniform pores, originating with the work of pores, the qualitative behavior of Kd and Kc with increas-
Pappenheimer, Renkin, and co-workers more than 50 ing molecular size is similar in slit pores (flat-wall chan-
years ago (230, 245), has been reviewed elsewhere (67, nels) and even in fibrous media, where the transport paths
69). Included in those reviews are examples of results are more tortuous. Only one aspect of those results is
from experiments using membrane pores or microfabri- unique to cylindrical pores, the convective hindrance of a
cated channels of carefully controlled size, which gener- solute that nearly fills the pore (␭ 3 1). Whereas Kc 3 1
ally confirm the theoretical predictions. The predictions in that limit for cylindrical pores, for slits and other pore
shapes Kc 3 0 for large solutes (67). The peculiar aspect
for cylindrical pores are shown in Figure 12, where the
of a tightly fitting sphere in a cylindrical pore is that it
hindrance factors are plotted as a function of relative
completely occludes the channel. With fluid unable to
solute size (␭ ⫽ rs/rp, where rp is pore radius). Reflecting
bypass it, the sphere acts as a piston and moves at the
the progressively greater hindrances to diffusion as rela-
mean fluid velocity, rather than becoming immobile.
tive solute size increases, Kd declines monotonically from
Hence, Kc 3 1 instead of 0.
unity to zero. The intrapore diffusivity reaches half of the
There have been many applications of pore theory to
free-solution value (i.e., Kd ⫽ 0.50) when ␭ ⫽ 0.24. In
the glomerulus, as reviewed previously (181). Fiber-ma-
trix theory is less completely developed. For hindered
transport through fibrous media, noteworthy recent stud-
ies include a discussion of diffusion in randomly oriented
fiber matrices (235) and a calculation of the osmotic
reflection coefficient for flow along the axes of parallel
fibers (351).
B. Sieving in Homogeneous Membranes
The experimental measure of membrane selectivity

in ultrafiltration is the sieving coefficient (⌰). This section
begins by relating ⌰ to the hindrance factors just dis-
cussed, and then illustrates how sieving is influenced also
by the operating conditions of the ultrafiltration process.
This initial discussion is for a homogeneous or single-
layer membrane; multilayer barriers are considered later.
Suppose that a membrane consists of a single layer of
FIG. 12. Hindrance factors and partition coefficients for spherical porous or fibrous material, extending from x ⫽ 0 to x ⫽ L.
molecules in cylindrical pores plotted against relative solute size
(␭ ⫽rs/rp, i.e., solute over pore radius), illustrating membrane size The concentrations in the retentate and filtrate are de-
selectivity. The local hindrance factors of the membrane for diffusion noted as C0 and CL, respectively. If the membrane thick-
(Kd) and for convection (Kc) are plotted together with the partition ness greatly exceeds the pore radius or interfiber spacing,
coefficient (⌽). Also shown are the overall hindrance factors for diffu-
sion (H⫽ ⌽ 䡠 Kd) and for convection (W⫽ ⌽ 䡠 Kc). The expressions used then near-equilibrium conditions will exist at the two
for Kd, Kc, H, W, and ⌽ are all given in Ref. 67. boundaries, such that

C共0兲 C共L) Of course, the diffusional hindrance factor (H),filtrate

⫽⌽⫽ (3) velocity (v), membrane thickness (L), and solute diffusiv-
C0 CL
ity in bulk solution (D⬁) all influence Pe, but none of those
Thus ⌽, termed the partition coefficient, equals the equi- quantities affects sieving independently. The dependence
librium ratio of internal (membrane) to external concen- of ⌰ on Pe for several values of W is illustrated in
tration. If both external solutions are dilute and the solute Figure 13. In each case, ⌰ decreases from 1 to W as Pe
molecule is uncharged, the partition coefficient at both increases from 0 to ⬁; Pe ⫽ 5 is large enough to be
sides is the same, as assumed in Equation 3. For spheres practically infinite. Recalling that Pe ⬀ vL, the depen-
in cylindrical pores, ⌽ ⫽ (1 ⫺ ␭)2, which is plotted in dence of ⌰ on Pe indicates that the selectivity of an
Figure 12. The theoretical effects of solute charge, shape, ultrafiltration membrane will be fully evident only if the
and concentration on partitioning are discussed in sec- product of filtrate velocity and membrane thickness is
tion VIC. large enough to make convection dominant. Helping to
For ultrafiltration into a dead-end chamber (such as amplify the effects of convection is the fact that, typically,
Bowman’s space), the filtrate concentration is determined W/H ⱖ 1 for porous or fibrous membranes, as discussed in
by the ratio of the solute flux to the volume flux. That is, connection with Figure 12. When Pe is not large, the time
CL ⫽ N/v. With the boundary conditions now established, scale for diffusion is comparable to or less than the transit
Equation 2 is integrated over the membrane thickness, time through the membrane, thereby allowing at least
using the requirement from conservation of mass that N partial equilibration of the filtrate with the retentate. If Pe
and v be independent of x. The result for the sieving 3 0, then ⌰ 3 1 and the intrinsic selectivity of the
coefficient (⌰ ⫽ CL/C0) is membrane (as embodied in W) is completely masked.
The foregoing suggests that ultrafiltration will be an
ineffective separation process unless Pe for each macro-
⌽K c W
⌰⫽ ⫺Pe ⫺Pe ⫽ (4) molecule of interest exceeds some minimum value. As
1 ⫺ e ⫹ ⌽Kce 1 ⫺ e ⫹ We⫺Pe
⫺Pe
indicated in Figure 13, the minimum Pe needed to achieve
a desired sieving coefficient depends on W. For example,
⌽KcvL WvL the value of Pe required to make ⌰ as small as 0.3 clearly
Pe ⫽ ⫽ (5)
⌽KdD⬁ HD⬁ increases with increasing W. Expanding the exponentials
in Equation 4 for small Pe shows that
where Pe, the Peclet number, is a dimensionless param-
eter that measures the importance of convection relative W
to diffusion. From the first forms of Equations 4 and 5, it ⌰⫽ (Pe ⬍⬍ 1) (6)
W⫹Pe(1⫺W)
is evident that the effects of membrane selectivity on
sieving are contained fully in the products ⌽Kc and ⌽Kd. This result, which is accurate to within a few percent for
In other words, sieving coefficients depend only on those Pe ⬍ 0.1, indicates that ultrafiltration is effective even for
two products, and not on the three individual quantities
involved. Accordingly, when analyzing sieving, it is advan-
tageous to switch from the local hindrance factors Kc and
Kd to overall hindrance factors defined as W ⫽ ⌽Kc and
H ⫽ ⌽Kd. That substitution yields the second forms of
Equations 4 and 5. An expression equivalent to Equation
4 was presented originally by Spiegler and Kedem (293)
some 40 years ago.
The behavior of W and H for spherical solutes in
cylindrical pores is shown in Figure 12. Because ⌽ ⬍ 1,
each overall hindrance factor is smaller than the corre-
sponding local one; the differences are amplified as the
relative molecular size increases, because ⌽ becomes
smaller. When measured in terms of H, the rate of diffu-
sion in a cylindrical pore is half that in free solution when
␭ ⫽ 0.14. In terms of W, convective transport is halved
when ␭ ⫽ 0.39. Thus solute transport in pores can be
hindered severely even when rs does not closely appro-
ach rp. FIG. 13. Sieving coefficient (⌰) as a function of membrane Peclet
number (Pe) for four values of the convective hindrance factor (W).
Equation 4 shows that the sieving coefficient de- These predictions for a homogeneous membrane are based on Equa-
pends on just two dimensionless parameters, W and Pe. tion 4.

small Pe, as long as Pe greatly exceeds W. More precisely, monitored to avoid mistaking a flow-induced shift in a
the requirement is that Pe ⬎⬎ W/(1 ⫺ W). In other words, sieving curve for one caused by a change in the mem-
thin membranes or membranes subjected to low filtration brane properties.
rates must be especially selective (i.e., have very small The foregoing discussion of sieving in homogeneous
values of W) to exhibit sieving. membranes provides the background needed for a critical
The combined effects of pore radius and filtration examination of a model for albumin sieving proposed
rate on sieving curves (plots of ⌰ vs. rs) are illustrated in recently by Smithies (292). The fundamental assumptions
Figure 14. The baseline condition is a hypothetical situa- in that model are as follows: 1) barrier selectivity resides
tion in which rp ⫽ 5.0 nm and the product vL is such that entirely in the GBM, 2) albumin transport across the GBM
Pe ⫽ 1 for rs ⫽ 3.0 nm. The other predictions correspond is mainly by diffusion, and 3) diffusion there proceeds
to a selective increase in pore radius (to 6.0 nm) or a independently of convection (water filtration). An impli-
selective decrease in filtrate velocity (to one-half of the
cation of assumptions 2 and 3 is that variations in GFR
baseline v). In each case, ⌰ declines from 1 to 0 as
have almost no effect on the filtered load of albumin; if
molecular size increases. Because Pe is not uniformly
GFR is increased or decreased, the albumin concentration
large, none of the curves in Figure 14 has quite the same
in Bowman’s space simply would be diluted to a greater
shape as the W curve in Figure 12. A selective increase in
or lesser degree, with the concentration-GFR product
pore radius shifts the entire sieving curve to the right, as
one might expect. The effects of a selective decrease in remaining nearly the same. If assumptions 1 and 2 are
filtration rate are more complicated, in that the sieving correct, then the glomerular barrier behaves as a homo-
curve of small molecules resembles that for an increased geneous membrane and Pe for albumin is small, making
pore radius, but the curve for large molecules is un- Equation 6 applicable. For ⌰ to be small, as is true for
changed. albumin (Table 1), Equation 6 requires that W ⬍⬍ Pe, in
To understand these effects of filtration rate, one which case it reduces to ⌰⬃W/Pe. Because Pe is propor-
must realize that each curve corresponds to a wide tional to v and therefore (if the number of nephrons
range of Pe values, since Pe increases with molecular remains constant) proportional to GFR, this is consistent
size due both to decreases in D⬁ and increases in W/H. with the idea that filtered load will not vary with GFR if
If the initial value of Pe is moderate, a reduction in it the barrier properties (i.e., W) are not affected. However,
will increase the sieving coefficient, as shown in Figure the data of Rippe et al. (251) for Ficoll sieving in rats in
13. In Figure 14, the absence of any noticeable change which GFR was manipulated show a dependence of ⌰ on
in ⌰ for the larger molecules, when v was reduced, GFR that is far too weak to keep filtered loads constant.
indicates that their Pe values remained large enough to Similar findings were reported for the sieving of albumin
be in the plateau region of Figure 13. A practical con- and other proteins in rat kidneys (179). Thus the available
clusion from Figure 14 is that filtrate velocities must be data do not support the idea that the filtered load of
albumin or similar-sized macromolecules is nearly inde-
pendent of GFR.
Given the ample evidence that both endothelium and
epithelium participate in glomerular barrier selectivity
(see sects. VI and VII), assumption 1 in the Smithies model
is its most obvious oversimplification. However, assump-
tion 3 is also incorrect in general, as may be seen by
returning to Equation 2. As noted earlier, steady-state
conservation of mass requires that N and v be indepen-
dent of x. Because C decreases with increasing x, the
absolute value of the concentration gradient (⫺dC/dx)
must increase with x, leading to a C(x) profile that is
concave downward. The amount of curvature in C(x) is
determined by Pe. Thus flow influences diffusion by alter-
ing the shape of the concentration profile. Moreover, the
contributions of convection and diffusion to N each vary
FIG. 14. Sieving in a homogeneous membrane for three hypothet-
with x so that, in general, it is meaningless to assert that
ical conditions. The baseline case corresponds to a pore radius (rp) of a specific fraction of the transmembrane flux is due to one
5.0 nm. The other curves are those predicted for a selective increase mechanism or the other. The exception is Pe ⫽ 0, where
in pore radius to 6.0 nm or a selective decrease in the filtrate ve-
locity to one-half of the baseline level. See text for other parameter there is no convection at all, but then there is also no
values. sieving!

C. Effects of Molecular Charge, Shape, and uted in some manner over its surface or volume, rather
Concentration on Membrane Selectivity than concentrated at a point. By computing position-
dependent electrostatic potentials using the differential
The effects of molecular charge and shape are incom- equations that describe diffuse double layers, ⌽ has
pletely understood, even for cylindrical or slit pores. A been evaluated for solid or porous spheres in cylindri-
major impediment to theoretical progress has been the cal pores (291) and for solid spheres in dilute fiber
difficulty in solving the sets of differential equations that matrices (136, 137).
describe the hydrodynamic forces acting on charged An advantage of these more detailed calculations is
and/or aspherical particles in small channels. Even more that they include the interplay between the effects of
formidable computational problems arise when the con- molecular size and molecular charge. As the results show,
centration of one or more macromolecules is large the effects of size and charge on ⌽ are not entirely sepa-
enough to make solute-solute interactions important. rable. A limitation, in common with attempts to predict
Given this incomplete state of knowledge, it has been size selectivity, is that the actual structure of a membrane
common to assume that Kc and Kd are affected mainly by may be much more complex than the available idealiza-
the Stokes-Einstein radius of the molecule and to focus on tions (e.g., uniform cylindrical pores or randomly oriented
how the other factors might affect the equilibrium parti- fibers). The fiber model has been suggested as an alter-
tion coefficient ⌽. As already seen (Eq. 4), partitioning native to the Donnan model in describing glomerular
influences sieving via the overall convective hindrance charge selectivity (52, 122, 134).
factor, W ⫽ ⌽Kc; it does not affect Pe.
2. Shape effects
1. Charge effects
Equilibrium partition coefficients in uncharged sys-
The simplest way to describe the effects of charge on tems have been computed for many combinations of sol-
⌽ is to model the solute of interest as a point charge with
ute shape and membrane structure. These include various
a specified valence and to assume Donnan equilibria be-
shapes of rigid particles in pores of uniform cross-section
tween the membrane and adjacent solutions. In such
(95, 155), spherical (211) or ellipsoidal (166) solutes in
calculations, the membrane is treated as a solution that
random fiber matrices, and freely jointed chains in pores
contains a certain concentration of fixed charges, in ad-
(80) or random fiber matrices (344). The fiber matrices
dition to mobile ions. The fixed charges may arise from
considered include randomly positioned and oriented fi-
covalently attached acidic or basic groups or adsorbed
bers of a single radius (211, 344) or mixtures with multiple
ions. Structural information such as pore radius, fiber
fiber radii (166). It is reasonable to approximate globular
concentration, and fiber radius is neglected in Donnan
proteins and Ficoll as rigid particles, whereas the freely
models. If an external solution contains nearly imper-
meant ions, such as large proteins, then they are viewed jointed chain may be a better model for dextran or other
as fixed charges in that solution. An excess of fixed neg- linear polymers. Such molecules do not have a fixed
ative charges in the membrane creates a negative in- shape in solution and behave more as random coils.
tramembrane potential, which reduces ⌽ for anionic mac- Owing to the gel-like nature of the endothelial glyco-
romolecules and increases it for cationic ones. There have calyx and GBM and the physiological importance of glob-
been several applications of Donnan models to the glo- ular proteins, the partitioning of rigid particles in fiber
merulus (53, 76, 301, 302, 313, 327, 348). matrices is especially pertinent. In such systems, the ef-
A more realistic approach is to recognize that mem- fects of molecular shape seem to be of little importance,
brane fixed charges will be localized on pore walls or on unless the solutes are markedly elongated or they are
the surface of membrane fibers, rather than distributed random coils (see sect. IV). For example, the hydrody-
uniformly over the membrane volume. A surface charge namic properties of albumin resemble those of a prolate
creates a region of opposing charge in the adjacent solu- (elongated) ellipsoid with a major semiaxis of 7.0 nm and
tion called the diffuse double layer, and it is only within a minor semiaxis of 2.1 nm (167). Using the excluded
that region that the electrical potential is perturbed by the volume theory (166), we can compare the partition
surface charge. The double-layer thickness is on the order coefficient of such a molecule with that of a sphere of
of the Debye length, which for physiological saline solu- the same Stokes-Einstein radius, 3.6 nm. In a fiber
tions is ⬃0.8 nm. Thus, within a porous or fibrous mem- matrix with a fiber radius of 1.0 nm, ⌽ for the ellipsoid
brane that freely passes molecules up to 2 nm in radius, is predicted to be within 1% of that for the sphere of
and allows some passage of macromolecules as large as 6 equal diffusivity, for all fiber volume fractions from 0
nm, the electrical potential is not spatially uniform, as and 0.2. This is consistent with partitioning (167) and
assumed in the Donnan approach. Moreover, the charge sieving studies (141) in agarose gels, in which the re-
on a multivalent macromolecular solute will be distrib- sults for albumin and certain other globular proteins

were nearly identical to those for Ficolls of equivalent been calculated for cylindrical pores (194), but not for
Stokes-Einstein radius. fiber matrices.
3. Concentration effects D. Sieving in Multilayer Membranes

In most applications of hindered transport theory to
ultrafiltration, whether in the glomerulus or other kinds of Because the glomerular filtration barrier consists of
barriers, each solute is assumed to cross the membrane three layers in series (fenestrated endothelium with its
independently. However, repulsive forces among like glycocalyx, GBM, and epithelial filtration slits with their
molecules, due to steric and electrostatic interactions, slit diaphragms), it is important to understand how siev-
cause ⌽ to increase with increasing concentration (5, 97). ing in a multilayer membrane differs from that in a homo-
For uncharged spheres, this effect is noticeable even geneous one. We begin quite generally. Suppose that a
when the solute occupies only a small percentage of the membrane consists of n layers arranged in series, num-
solution volume. Solute-solute interactions exist also bered 1, 2, . . . n from the upstream to the downstream
among unlike solutes. For example, increasing the con- side. The convective hindrance factor and Peclet number
centration of an abundant solute is predicted to increase for layer i are denoted as Wi and Pei, respectively. The
both ⌽ of that solute and ⌽ of a second molecule present derivation of Equation 4 can be generalized to give an
at tracer concentrations (166). This prediction has been expression for the sieving coefficient in layer i.Denoted as
⌰i, that sieving coefficient equals the downstream-to-up-
confirmed quantitatively by partitioning measurements
stream concentration ratio within that particular layer.
for BSA (abundant solute) and Ficoll (tracer) in agarose
The result is (68)
gels (167), and also helps to explain the effects of BSA on
Ficoll sieving in filters constructed from isolated GBM
(168). When ⌽ is concentration dependent, the partition Wi
⌰i ⫽ (7)
coefficient at the upstream side of an ultrafiltration mem- ⌰i⫹1⌰i⫹2 . . . ⌰n共1 ⫺ e⫺Pei兲 ⫹ Wie⫺Pei
brane must be distinguished from that at the downstream
side (168); in Equation 4, the upstream value of ⌽ (or W)
where ⌰n⫹1 ⫽ 1 by definition. For the most downstream
must be used in the numerator and the downstream value
layer (i ⫽ n), Equation 7 reduces to Equation 4, indicat-
in the denominator.
ing that ⌰n is a function only of Wn and Pen. In other
Although rarely included in ultrafiltration models,
words, the sieving coefficient in the last layer is indepen-
concentration effects may be physiologically significant.
dent of the selectivity or flow conditions in the preceding
For example, using a two-fiber model for GBM and as-
layers. However, as indicated by the product of sieving
suming a solution protein concentration of 6 g/dl, the
coefficients in the denominator of Equation 7, all of the
partition coefficients of spherical tracers with 2.0 ⱕ rs ⱕ other sieving coefficients are coupled to some extent,
5.0 nm were predicted to be 1.4 –3.9 times greater than in each being affected by events downstream. This coupling
the absence of protein, the percentage increases being is most complete for the first layer (i ⫽ 1), in that ⌰1 is
greater for the larger molecules (168). The effects of pure influenced by all of the other sieving coefficients. The fact
BSA, or a 1:1 mixture of BSA and IgG, were predicted to that the sieving coefficient in a given layer influences all of
be nearly identical. Under these conditions, the partition the upstream ones, but none of the downstream ones,
coefficient for BSA itself was predicted to increase 2.1- indicates that the ordering of the layers is important. The
fold, relative to a very dilute BSA solution. For the glo- effects of ordering in a two-layer membrane will be ex-
merulus in vivo, the effects of protein concentration on amined shortly.
plasma protein or exogenous tracer sieving will depend The interdependence of the layer sieving coefficients
on the extent to which large proteins such as albumin stems from two physical constraints. One is that the water
penetrate the capillary wall. If high protein levels exist and solute fluxes across each layer must be the same.
only in the capillary lumen, the effect will be limited to the Once a steady state has been reached, there can be no
partition coefficient between the lumen and endothelial time-dependent accumulation or depletion of water or
glycocalyx (168). This issue is discussed further after solute at any boundary between layers. The other con-
Equation 8. straint is that the concentration at the downstream side of
The aforementioned estimates of the concentration layer i is linked to that at the upstream side of layer i ⫹ 1.
dependence of ⌽ are for uncharged solutes and mem- The solute concentration profile in each layer must adjust
branes. Electrostatic interactions are absent in agarose accordingly, and Equation 7 is a consequence of those
gels (167) and minor in isolated GBM (28) but may be simultaneous adjustments.
quite important in the GAG-rich glycocalyx. The com- For a multilayer barrier and dilute retentate, the over-
bined effects of solute charge and concentration have all sieving coefficient is calculated as

⌰ ⫽ ⌰ 1⌰ 2 . . . ⌰ n (8) slit diaphragm is a significant contributor to albumin siev-

ing, the GBM is not (73).
Of course, ⌰ measures how well the barrier functions as The conclusion that the GBM does not influence
a whole. If the retentate is not dilute, one or more addi- albumin sieving, just described, is at odds with the finding
tional, multiplicative factors will appear in Equation 8 that selective alterations in mouse GBM can lead to pro-
(168), depending on the extent to which abundant solutes teinuria (132). If in reality the GBM contributes signifi-
penetrate the barrier. For example, if the major plasma cantly to albumin sieving in normal animals, W1 would
proteins are excluded from the endothelial glycocalyx, need to be orders of magnitude smaller than that mea-
such that high protein concentrations occur only in the sured in vivo. (For example, to yield ⌰1 ⫽ 0.1 at Pe1 ⫽
lumen, the additional factor equals the relative increase in 0.14, then W1 ⫽ 0.0014 for ⌰2 ⫽ 0.1 and W1 ⫽ 0.00014 for
⌽ for the glycocalyx that is due to the plasma proteins ⌰2 ⫽ 0.01.) It is possible that loss of macromolecular
(168). In other words, if the luminal protein concentration constituents during GBM isolation led to a “looser” mate-
is sufficient to double ⌽ in the glycocalyx, Equation 8 will rial with an elevated W. However, theories of flow through
contain an additional factor of 2. For simplicity, we will fibrous media indicate that loss of, say, GAGs, also would
proceed as if all solutions are dilute. greatly elevate the hydraulic permeability (27). What is
Examining the behavior of a two-layer membrane most difficult to reconcile is that the hydraulic permeabil-
provides insight into how pairs of layers of the glomerular ity in isolated GBM (measured simultaneously with Ficoll
barrier may interact. For n ⫽ 2, Equation 7 applied to the sieving) was low enough to account for nearly all of the
upstream layer becomes known resistance to water filtration in vivo (73). That is,
a much “tighter” GBM in vivo would likely give unrealis-
tically low values of single-nephron GFR.
W1 In a multilayer barrier, physical interactions among
⌰1 ⫽ (9)
⌰2共1 ⫺ e 兲 ⫹ W1e⫺Pe1
⫺Pe1
the layers also influence the dependence of the overall
sieving coefficient on the water filtration rate. As dis-
The influence of the downstream layer on sieving in the cussed already, ⌰ for a homogeneous membrane invari-
upstream one is most striking for Pe1 ⬎⬎ 1, in which case ably declines as the filtration rate increases, until at high
Equation 9 reduces to ⌰1 ⫽ W1/⌰2. If the upstream layer Pe it reaches a minimum value equal to W. That is not
is less selective than the downstream one, in the sense necessarily the case for a multilayer membrane, as will be
that W1 ⬎ ⌰2, then ⌰1 ⬎ 1. This indicates that an internal illustrated next.
concentration polarization is created, such that the solute For simplicity, we again consider a barrier with just
concentration in layer 1 increases in the direction of two layers. In general, the value of the Peclet number in a
flow. Concentration polarization in the upstream layer is given layer may be small, moderate, or large, suggesting
avoided if W1 ⱕ ⌰2. However, for large Pe1, the overall that there are nine combinations to consider when explor-
sieving coefficient for the two layers is always ⌰ ⫽ ⌰1⌰2 ⫽ ing how two-layer membranes behave. However, small
W1. It is remarkable that, under these circumstances, ⌰ is values of Pe1 or Pe2 are less interesting, because a layer
independent of the properties or flow conditions in the with a small Peclet number will tend to be “invisible”
downstream layer. For Pe1 ⬍⬍ 1, Equation 9 reduces to (unless Wi ⬍ Pei). Of the four Peclet number combina-
tions that remain, only three give different results for the
overall sieving coefficient, as discussed recently (68) and
W1
⌰1 ⫽ (10) explained in the three cases that follow. We will use
W1 ⫹ Pe1(⌰2 ⫺ W1) “Pe ⬃1” to indicate a moderate Pe and “Pe ⬎⬎ 1” to
indicate a large Pe. More precisely, Pe ⬃1 means that
which is the two-layer analog of Equation 6. In this case, neither exp(⫺Pe) nor 1 ⫺ exp(⫺Pe) is negligible com-
both ⌰1 and ⌰ depend on the properties of each mem- pared with unity. If errors of ⬍20% are acceptable, that
brane layer and are also affected (through Pe1) by the will be true for Pe values ranging from ⬃0.2 to 2.
filtration rate of water.
Suppose that layer 1 is the GBM and layer 2 is the 1. Case 1
epithelium. For a molecule the size of albumin, it has been
estimated that Pe1 ⫽ 0.14 (73), based on the sieving and If Pe1 ⬎⬎ 1, then it follows from Equations 7 and 8
diffusion properties of isolated rat GBM and the dimen- that ⌰1 ⫽ W1/⌰2 and
sions and flow rates for rats in vivo. With the use of Pe1 ⫽
0.14, together with the finding in isolated rat GBM that ⌰ ⫽ W1 (11)
W1 ⫽ 0.08 for a Ficoll molecule the size of albumin (73),
either Equation 7 or Equation 10 indicates that ⌰1 ⬃1 for As already discussed, the overall sieving coefficient is
any small value of ⌰2. Thus it was concluded that if the independent of the filtration velocity in this case and is

determined entirely by the selectivity of the upstream

layer. The downstream layer has no effect here, even if
Pe2 is not small. To obtain this result, Pe1 must be large
enough to make W1exp(⫺Pe1) negligible compared
with ⌰2.
2. Case 2
If Pe1 ⬃1 and Pe2 ⬎⬎ 1, then ⌰2 ⫽ W2 and
W1W2
⌰⫽ (12)
W2 ⫹ 共W1 ⫺ W2兲e⫺Pe1
Here the sieving coefficient is flow dependent and is

influenced by the selectivity of both layers. Since Pe1 is
proportional to the filtration rate of water, ⌰ will increase FIG. 15. Sieving in single-layer and two-layer membranes. The
with the filtration rate if W1 ⬎ W2 and decrease with overall sieving coefficient (⌰) is plotted as a function of the Peclet
number in layer 1 (Pe1). Curves are shown for a single-layer membrane
filtration rate if W1 ⬍ W2. (W1 ⫽ 0.01); a membrane in which a second, identical layer has been
added upstream or downstream from the first (W1 ⫽ W2 ⫽ 0.01); and a
3. Case 3 membrane with a second, less selective layer added upstream from the
first (W1 ⫽ 0.1, W2 ⫽ 0.01). In the two-layer membranes, it was assumed
that Pe1 ⫽ Pe2.
If Pe1 ⬃1 and Pe2 ⬃1, there is no algebraic simplifi-
cation, and the resulting expression for ⌰ is too unwieldy
to offer much insight. However, if Pe1 ⫽ Pe2 ⫽ Pe, the
effect (a 45% reduction in ⌰) occurs at a Pe value of ⬃0.1.
result is
Because the two layers have the same intrinsic selectivity,
the double membrane is the same as a homogeneous one
W1W2 with twice the original thickness. Recalling that the Peclet
⌰⫽ (13)
W2 ⫹ 共W1 ⫺ W2兲e⫺Pe ⫺ W1共1 ⫺ W2兲e⫺2Pe number is proportional to membrane thickness, it is ap-
parent that for the double membrane the Peclet number
which is valid for any Pe. As in case 2, the sieving coef- to be used in Equation 4 must increase from Pe1 to 2Pe1.
ficient is flow dependent and influenced by the properties The reduction in ⌰ occurs only at small-to-moderate
of both layers. If Pe 3 0 in Equation 13, then ⌰ 3 1, as Pe1, because when Pe1 is large, there is no benefit in
seen before. If Pe ⬎⬎ 1, Equation 13 reduces to Equation doubling it. That is, for large Pe1 the sieving coefficient
11. A comparison of Equations 12 and 13 reveals an is already at its minimum value, ⌰ ⫽ W. The assertion
additional term in the denominator of the latter, which is that adding an identical second layer merely doubles
negative and which therefore tends to increase ⌰ at any the apparent value of Pe can be verified by setting W1 ⫽
intermediate value of Pe. Thus the intrinsic selectivity of W2 ⫽ W in Equation 13.
the two layers is used less efficiently in the conditions If the additional layer is an order of magnitude less
assumed for Equation 13 than in those for Equation 12. selective, the results are very sensitive to its position. If it
It is informative to synthesize a two-layer membrane is placed downstream, such that W1 ⫽ 0.01, W2 ⫽ 0.1, and
mathematically, starting with one layer and then adding a Pe2 ⫽ Pe1, barrier function is improved. However, the
second with the same or different properties. Keeping in values of ⌰ obtained from Equation 13 (not shown) are at
mind that the sieving coefficients for large molecules most 7% lower than those for the single-layer membrane,
(e.g., albumin) should be as small as possible, we can changes that are too small to be evident on the log scale
inquire whether the overall barrier function is improved in Figure 15. If the additional layer is placed upstream,
or worsened by adding a second layer. The results of such such that W1 ⫽ 0.1, W2 ⫽ 0.01, and Pe2 ⫽ Pe1, barrier
calculations are given in Figure 15, which shows the function is worsened. In this case, Equation 13 predicts
overall sieving coefficient as a function of the Peclet marked elevations in ⌰ at both moderate and large values
number in layer 1. The curve for the single membrane was of Pe1, relative to the single-layer membrane. For Pe1 ⬎ 5
generated using W ⫽ 0.01 in Equation 4 and is identical to (somewhat beyond the range of Fig. 15), the overall siev-
the bottom curve in Figure 13. If we now add a second ing coefficient closely approaches that of the upstream
layer that has the same properties as the first, creating a layer, consistent with Equation 11.
“double membrane,” there tends to be a downward shift The functional degradation seen when the less selec-
in ⌰ for any given value of Pe1. For the arbitrary param- tive layer is placed upstream reflects internal concentra-
eter values that we chose for Figure 15, the maximum tion polarization within the first membrane layer. In gen-

eral, the hallmark of internal polarization is a sieving alterations in glomerular structures. It has therefore been
coefficient in one or more membrane layers that exceeds natural to assume that the proteinuria is caused by de-
unity. For the example under consideration (W1 ⫽ 0.1, fects in the glomerular barrier. Recently, this view has
W2 ⫽ 0.01, Pe2 ⫽ Pe1), ⌰1 ⫽ 2.1 at Pe1 ⫽ 1.0. Thus, in been questioned and an alternative hypothesis has been
moving across this two-layer membrane, the solute con- proposed, the so-called albumin retrieval hypothesis (84).
centration first increases and then decreases. At low val- The hypothesis has attracted certain interest, and it is
ues of Pe1, the initial increase is negligible, and the two- therefore discussed below, together with the overwhelm-
layer membrane functions nearly as well as the single- ing evidence against it.
layer membrane (Fig. 15).
Overall, the results in Figure 15 suggest that ⌰ varies
inversely with filtration rate only if sieving in the most C. The Albumin Retrieval Hypothesis
downstream layer is not dominant. Accordingly, the ten-
dency for ⌰ to vary inversely with single-nephron GFR (or According to the albumin-retrieval hypothesis, the
with GFR at constant nephron number), which has been proteinuria that occurs with diabetic nephropathy (226,
observed for dextran (47), Ficoll (251), and proteins 229) or other renal disorders (265) reflects defects in the
(179), supports the view that the endothelium and/or tubular system with intact glomerular permeability (265,
GBM contribute significantly to sieving. This does not 266). The morphological changes in the glomerulus are
exclude a role for the epithelium in sieving; it suggests therefore considered coincidental and of no pathophysi-
only that it is not the single dominant structure. ological significance. The hypothesis rests on four as-
sumptions: 1) that albumin is retrieved from the tubule
VII. MECHANISMS BEHIND PROTEINURIA AND fluid in intact form (84); 2) that the glomerular sieving
NEPHROTIC SYNDROMES coefficient for albumin is close to 0.08 (228), which is two
to three orders of magnitude higher than other estimates
(see Fig. 10); 3) that the uptake of albumin is mediated by
A. What is Proteinuria? the megalin-cubulin complex (264); and 4) that albumin is
markedly degraded and fragmented in the urine (84).
Proteinuria is said to be present when the urine None of these assumptions is valid.
contains more than 300 mg protein per day (or 200 mg/l). First, the question of whether there is any tubular
In the nephrotic syndrome, ⬎3.5 g/day of proteins are uptake of intact albumin is not a new one. In 1966,
excreted in the urine. Urine with daily protein excretion Maunsbach studied the tubular uptake of autologous rat
of 30 –300 mg (20 –200 mg/l) reflects microalbuminuria. albumin (185) and concluded that such uptake must be
Individual laboratories may give slightly different defini- extremely small or nonexistent (184). Others have con-
tions based on the technique used and patient character- firmed and extended these findings. Measurements of the
istics (e.g., age and sex). Many clinicians prefer to use kinetics of albumin uptake in proximal tubules (231) led
albumin excretion in relation to that for creatinine to to the conclusion that albumin is taken up by the tubular
correct for differences in urine dilution. Normally, the cells at the luminal side, and its amino acids are released
albumin-creatinine concentration ratio is ⬍30 ␮g/mg (128, on the other side to blood (231). The molecular mecha-
183). Today, protein analysis of the urine is used more as nism is the megalin-cubulin complex (21, 22).
a prognostic index for cardiovascular disease than as a Second, a glomerular sieving coefficient for albumin
means of detecting patients with renal disorders (31, 111, close to 0.08 is not compatible with life and must be due
119, 336). to flaws in the experimental technique. A simple example
illustrates the magnitude of the problem. Thus, for a
B. Where do Proteins in the Urine Come From? person with a glomerular filtration rate of 180 liters/day,
a serum concentration of albumin close to 40 g/l, and a
Proteins in the final urine have different origins. First, sieving coefficient for albumin of 0.08, the daily losses of
proteins cross the glomerular barrier, and tubular cell albumin would be more than half a kilogram (180 ⫻ 40 ⫻
reabsorption modifies their final concentration. Second, 0.08 ⫽ 576 g). With the assumption that a reuptake of
there is tubular secretion of proteins from the blood. intact albumin exists (despite compelling evidence
Third, proteins may be synthesized by the cells them- against it, see above), such amounts are higher than the-
selves and released into the urine. Fourth, the proteins oretically achievable using detailed models of tubular
may be added to the urine at a later stage (e.g., by excre- uptake (169).
tion from the prostate gland in men). Let us examine Third, it has been suggested that the retrieval of
these alternatives in more detail. intact albumin is mediated by the megalin-cubulin com-
Patients with glomerulonephritis have proteinuria, plex (264). Indeed, megalin and cubulin are strongly
hematuria, elevated blood pressure, and morphological linked to the uptake of albumin and several other impor-

tant solutes in many species, including humans (49, 337). from focal segmental glomerulosclerosis, and others have
It is also known that the turnover of the megalin-cubulin primary or secondary forms of membranoproliferative
complex is controlled by a chloride channel, ClC-5 (50, glomerulonephritis. In all these cases, the diagnosis is
124), which opens new therapeutic avenues (157). How- obtained from morphological analysis of renal biopsies. In
ever, the megalin-cubulin complex cannot be responsible the future, quantitative PCR will be used to assess the
for the reuptake of intact albumin, since these complexes expression profiles of different proteins (325).
are internalized into lyzosomes, and the albumin is de- Today we know the molecular defect underlying
graded (49). many genetic disorders (321). Patients with nephrotic
Fourth, one research group concluded that albumin syndrome of the Finnish type have a defect in the nephrin
fragments are present in urine and require special tech- gene, NPHS1(153). Children with steroid-resistant ne-
niques to be detected (54, 56, 84, 100, 120, 221, 265, 295). phrotic syndrome are likely to have a problem with the
That conclusion is based on the assumption that their podocin gene, NPHS2(29). LMX1B, a gene that is mutated
method, which uses both the Biuret reagent and an HPLC in patients with nail-patella syndrome, is required for
technique, is superior to the methods used by others. The podocyte differentiation (192). Indeed, several proteins
Biuret reagent is assumed to be sensitive and specific for are known to be important for an intact glomerular bar-
protein detection. Norden et al. (209) tested this assump- rier (see Refs. 83, 90, 103, 104, 142, 190, 237, 249, 320, 321,
tion using state-of-the-art proteomics matrix-assisted la- 326, 345). However, many renal disorders do not seem to
ser desorption/ionization time of flight mass spectrometry be explained by genetic factors, and the underlying mech-
and Nano liquid chromatography electrospray tandem anisms remain unknown.
mass spectrometry, to select 14 nonpeptide components Because of our rudimentary understanding of renal
from urine, mix them, and test the samples using the pathophysiology, most treatments are nonspecific, cyto-
Biuret reagent. toxic, and anti-inflammatory, with potentially harmful
Although there were no proteins present in these side effects. As a result, many patients with glomerulone-
samples, the Biuret reagent gave a protein value of 0.4 g/l, phritis are given conservative treatments such as diuretics
and similar errors were found with the HPLC technique and antihypertensive medicine alone. Preeclampsia is of
(209). Moreover, a proteomics analysis of the fate of particular interest because the disease involves endothe-
albumin in patients with Fanconi syndrome and in normal lial dysfunction (147, 186). Women with preeclampsia
controls showed no evidence of significant fragmentation have glomerular endotheliosis with swelling and vacuol-
of albumin in the urine (209). As mentioned in section IV, ization of the cells (187). Podocytes are not affected and
patients with Fanconi syndrome excrete large quantities the GBM is intact, suggesting that the proteinuria is due
of amino acids and proteins due to a defect in their solely to endothelial damage. Moreover, VEGF is in-
tubular reabsorption process. These findings provide ad- volved in the process (86, 186). Proteinuria also occurs
ditional support for our current understanding of the in Fanconi syndrome due to defective tubular reabsorp-
megalin-cubulin complex, in which albumin is internal- tion of filtered protein (207–209) (see above).
ized by lyzosomes, degraded, and released to the blood in Proteinuria also occurs after ischemia-reperfusion,
the form of amino acids (49, 207). Considering the paucity which is evident from studies on transplanted kidneys (9,
of experimental support, it is surprising that the albumin 294). A longer period of ischemia (60 min) causes nonse-
retrieval hypothesis has survived so many years. lective damage to the glomerular barrier (252). If, how-
ever, the ischemia is shorter (15–20 min), the damage
following reperfusion mainly affects glomerular charge
D. Human Diseases and Animal Models selectivity (6, 252). These changes occur with no apparent
damage to the podocytes, the GBM, or the endothelial
Proteinuria is a hallmark of renal disease, and pa- cells, suggesting more subtle effects of the endothelial
tients may suffer from glomerulonephritis, diabetic ne- cell surface layer (6). In other organs, similar short ische-
phropathy, and numerous other conditions. Among the mic periods have been shown to selectively damage the
conditions that cause nephrotic syndrome (with heavy endothelial glycocalyx (198, 236, 262). Also, perfused kid-
proteinuria, edema, low serum albumin concentrations, neys have been reported to lose sulfated proteoglycans
and hyperlipidemia), minimal change nephrosis is the and GAGs after ischemia (333), supporting the idea that
most common disorder in children, while membranous ischemia reperfusion causes proteinuria through damage
glomerulonephritis is predominant in elderly patients. Di- of the endothelial surface layer.
abetes mellitus affects 170 million people worldwide, and Several animal models have been used to mimic kid-
the number is expected to double over the next 25 years ney disease in humans. Most studies of nephrotic syn-
(102). This makes diabetic nephropathy the most com- drome have been done using nephrosis-inducing agents
mon cause of proteinuria, and occasionally the amounts such as puromycin aminonucleoside (122, 123, 216, 220,
excreted reach nephrotic levels. Some patients suffer 225) and adriamycin (18, 19, 161). More recently, studies

in mouse knockout models have provided more informa- uous capillaries, such as those in skeletal muscle (250).
tion on the specific actions of individual proteins such as Indeed, the sieving coefficient for albumin in peripheral
laminin (350) and lamb2 (132), suggesting a major role for capillaries is ⬍0.1, due to the properties of the endothelial
the GBM in proteinuric conditions. Furthermore, induc- surface coat or to the fiber matrix nature of underlying
tion of B7-1 in podocytes gives a nephrotic syndrome in basement membrane-like structures (55, 259). In Figure 5,
mice (241). Mice lacking the podocyte slit membrane we have illustrated the endothelial cell surface layer with
CD2-associated protein develop congenital nephrotic syn- some key components. The endothelial surface coat has
drome (287). There are mice that develop a disease re- been studied quite extensively in other organs (204, 328,
sembling focal segmental glomerulosclerosis (125, 126) 329, 332, 338), but the visualization is mainly indirect, and
and many other models (127). Most investigators try to disease models are lacking (323). Nevertheless, it seems
develop tissue-specific inducible genetic models to sup- safe to conclude that glomerular capillary endothelium is
press or overexpress a certain protein. More sophisti- at least as selective as that in intestine and in salivary and
cated techniques can be used to study glomerular size and pancreatic glands. If so, the concentration of albumin in
charge selectivity in these mice (133–135), rather than the fluid entering the GBM is likely to be 10% or less of
simply determining if proteinuria is present. The variety that in plasma.
of animal models and the continuing discovery of genes The highly differentiated and specialized podocyte is
mutated in patients with kidney disease give us a firm of utmost importance for an intact barrier. Several genetic
platform for elucidating the mechanisms of proteinuria. defects in the podocyte affect its structure (321, 322) and
cause diseases with proteinuria. In addition, through the
production of growth factors such as VEGF, the podocyte
VIII. SUMMARY
affects the GBM and the endothelium indirectly (66).
Regardless of where the defect is, the result is proteinuria.
In life, we search for simple answers to complex
Thus the glomerular barrier with its specialized compo-
questions. With respect to the glomerular barrier, scien-
nents truly acts as an integrative and delicate structure.
tists have presented data suggesting that the GBM, the
Over the next decade, a deeper understanding of the
podocyte, or the endothelium is the one and only impor-
barrier will lead to insight into the pathogenic mecha-
tant component. In this review, we argue for an integra-
nisms of kidney diseases and provide specific therapies.
tive view of the glomerular barrier and of the mechanisms
The glomerular barrier is highly size and charge selective
behind proteinuria. From the theoretical analysis pre-
in a manner that can be predicted using modern transport
sented above, it is evident that defects in any of the three
theories (see Fig. 11). In particular, it markedly restricts
glomerular layers may result in proteinuria. Indeed, a
the passage of large anionic proteins such as albumin.
quest is under way for a more global analysis of glomer-
If these capillaries were as permeable as peripheral cap-
ular proteins affected by disease (305).
illaries in skeletal muscle, a normal man would lose close
to a kilogram of albumin per day in the urine (116).
A. What is the Relevance of the Individual Fortunately, we lose ⬍1% of that amount into the primary
Components of the Glomerular Barrier? urine and ⬍30 mg/day of albumin in the final urine. The
components of the glomerulus that create such a magnif-
The slit membrane of the podocytes is the most icent barrier are the podocytes, the GBM, and the fenes-
complex structure in the glomerular barrier and definitely trated endothelial cells with their endothelial surface
of great importance for glomerular selectivity. As dis- layer of GAGs.
cussed already, several diseases in humans and animals Most known pathologies have been found in podo-
are caused by gene polymorphisms that affect proteins cytes, in particular in proteins of the podocyte slit mem-
involved in the slit membrane. Whether the podocyte brane (319, 323). The GBM constitutes most of the glo-
layer contributes to both charge and size selectivity is a merular hydraulic resistance (65), while being highly per-
matter of debate. There simply are no data to support meable to solutes. Thus the GBM does not seem to be
opinions on this matter. The GBM, on the other hand, particularly charge or size selective (28). Still, some GBM
shows some size selectivity but little charge selectivity, at lesions, such as mutations of the laminin ␤2-chain (350),
least when studied in isolated form using current proto- produce proteinuria even in the absence of detectable
cols (28). Finally, the capillary endothelium is gaining changes in the cellular components (132). This finding is
more and more attention (323). The high frequency of not readily compatible with functional studies of the GBM
large fenestrae has led many investigators to assume that (68), unless laminin is present in the endothelial cell
the endothelium does not contribute to the barrier, apart surface layer as well. The endothelium has hitherto been
from keeping blood cells from the GBM. However, data neglected (68, 70, 116). However, recent data on pre-
from other capillary beds with fenestrated endothelia eclampsia (147, 171, 172, 187) show that this endothelial
(307) suggest these structures are as selective as contin- disease may affect glomerular capillaries and cause pro-

teinuria through the actions of VEGF (147). Indeed, glo- ACKNOWLEDGMENTS

merular lesions similar to those in human preeclampsia Address for reprint requests and other correspondence: B.
were found in a mouse model with altered expression of Haraldsson, Dept. of Molecular and Clinical Medicine/Nephrol-
VEGF-A in the podocytes (85). Studies on kidneys sub- ogy, The Sahlgrenska Academy at Göteborg University, SU/
jected to mild ischemia support the notion that protein- Sahlgrenska, SE-413 45 Gothenburg, Sweden (e-mail: borje.
uria may be caused by defects in the endothelial surface haraldsson@gu.se).
layer (6).
GRANTS
We are grateful for financial support of this work from the
B. What Do We Need to Know More About? Swedish Research Council 9898, the Ingabritt and Arne Lund-
berg Research Foundation, ALF/LUA funds of the VG region, the
Swedish Diabetes Association, and the Swedish National Kidney
Despite recent advances, the mechanisms underlying Foundation.
acquired glomerular disorders are poorly understood. As
a result, the therapeutic arsenal is limited and blunt.
Regarding the properties of the glomerular barrier, more REFERENCES
biological data are needed on the transglomerular pas-
1. Abboud HE, Poptic E, DiCorleto P. Production of platelet-de-
sage of proteins that differ in size, charge, and shape. rived growth factorlike protein by rat mesangial cells in culture.
Different technical approaches should be adopted, since J Clin Invest 80: 675– 683, 1987.
each method has advantages and drawbacks. In particu- 2. Adamson RH, Clough G. Plasma proteins modify the endothelial
cell glycocalyx of frog mesenteric microvessels. J Physiol 445:
lar, the impact of molecular configuration must be studied 473– 486, 1992.
in detail. Only limited amounts of biological data have 3. Ala-Houhala I, Pasternack A. Fractional dextran and protein
clearances in glomerulonephritis and in diabetic nephropathy. Clin
been obtained under conditions of controlled tubular cell Sci 72: 289 –296, 1987.
activity. We need to learn more about the individual com- 4. Amon H, Gayer J. Electron microscope studies on the influence of
ponents of the barrier. The podocyte is the most complex hyaluronidase on the basal membrane of the glomerular capillaries
with special reference to the problem of permeability. Klinische
and fascinating part of the glomerular membrane, and Wochenschrift 41: 163–172, 1963.
despite a decade of intense research, much remains to be 5. Anderson JL, Brannon JH. Concentration dependence of the
learned about this structure. For example, it will be im- distribution coefficient for macromolecules in porous media. J
Polymer Sci 19: 405– 421, 1981.
portant to understand in greater detail the communica- 6. Andersson M, Nilsson UA, Hjalmarsson C, Haraldsson B, Nys-
tion between podocytes and the endothelium. It will also trom Sorensson J. Mild renal ischemia-reperfusion reduces
be important to elucidate the endothelial and epithelial charge and size selectivity of the glomerular barrier. Am J Physiol
Renal Physiol 292: F1802–F1809, 2007.
production of proteoglycans and GAGs. What substances 7. Areekul S. Reflection coefficients of neutral and sulphate-substi-
are produced by these cells and at what rates? What is the tuted dextran molecules in the isolated perfused rabbit ear. Acta
Soc Med Ups 74: 129 –138, 1969.
composition of the endothelial cell surface layer as shown 8. Aron DC, Rosenzweig JL, Abboud HE. Synthesis and binding of
in Figure 5, and what is the turnover rate of this layer? insulin-like growth factor I by human glomerular mesangial cells.
What is the role of the mesangial cell? We need to know J Clin Endocrinol Metab 68: 585–591, 1989.
9. Artz MA, Dooper PM, Meuleman EJ, van der Vliet JA, Wetzels
how various agents affect glomerular permeability during JF. Time course of proteinuria after living-donor kidney transplan-
health and disease. Finally, we must use the information tation. Transplantation 76: 421– 423, 2003.
at hand to develop better therapies for the many patients 10. Asgeirsson D, Rippe B, Venturoli D, Rippe C. Increased glomer-
ular permeability to negatively charged Ficoll relative to neutral
with kidney disease. Ficoll in rats. Am J Physiol Renal Physiol 291: F1083–F1089, 2006.
11. Avasthi PS, Koshy V. The anionic matrix at the rat glomerular
endothelial surface. Anat Rec 220: 258 –266, 1988.
C. Concluding Remarks 12. Bakoush O, Tencer J, Torffvit O, Tenstad O, Skogvall I, Rippe
B. Increased glomerular albumin permeability in old spontaneously
hypertensive rats. Nephrol Dial Transplant 19: 1724 –1731, 2004.
We have presented an integrative view of the glomer- 13. Ballermann BJ. Regulation of bovine glomerular endothelial cell
growth in vitro. Am J Physiol Cell Physiol 256: C182–C189, 1989.
ular barrier, according to which each component of the 14. Barker DF, Hostikka SL, Zhou J, Chow LT, Oliphant AR,
barrier must maintain its integrity to preserve normal Gerken SC, Gregory MC, Skolnick MH, Atkin CL, Tryggvason
K. Identification of mutations in the COL4A5 collagen gene in
function. The GBM appears to be less selective than the Alport syndrome. Science 248: 1224 –1227, 1990.
cellular components. The statement that podocyte dam- 15. Batsford SR, Rohrbach R, Vogt A. Size restriction in the glomer-
age leads to proteinuria is still accurate; it is, however, ular capillary wall: importance of lamina densa. Kidney Int 31:
710 –717, 1987.
equally true that damage to the endothelium causes pro- 16. Benedict CR, Pakala R, Willerson JT. Endothelial-dependent
teinuria. In the final section of this review, we have high- procoagulant and anticoagulant mechanisms. Recent advances in
lighted important questions that need to be addressed for understanding. Texas Heart Inst J 21: 86 –90, 1994.
17. Bennett CM, Glassock RJ, Chang RL, Deen WM, Robertson
a better understanding of one of nature’s great wonders, CR, Brenner BM, Troy JL, ueki IR, Rasmussen B. Permselec-
the glomerular barrier. tivity of the glomerular capillary wall. Studies of experimental

glomerulonephritis in the rat using dextran sulfate. J Clin Invest skeletal muscles in relation to angiogenesis. Exp Physiol 81: 1043–
57: 1287–1294, 1976. 1046, 1996.
18. Bertolatus JA, Abuyousef M, Hunsicker LG. Glomerular siev- 40. Bulger RE, Eknoyan G, Purcell DJ 2nd, Dobyan DC. Endothe-
ing of high molecular weight proteins in proteinuric rats. Kidney lial characteristics of glomerular capillaries in normal, mercuric
Int 31: 1257–1266, 1987. chloride-induced, gentamicin-induced acute renal failure in the rat.
19. Bertolatus JA, Hunsicker LG. Glomerular sieving of anionic and J Clin Invest 72: 128 –141, 1983.
neutral bovine albumins in proteinuric rats. Kidney Int 28: 467– 41. Burne MJ, Adal Y, Cohen N, Panagiotopoulos S, Jerums G,
476, 1985. Comper WD. Anomalous decrease in dextran sulfate clearance in
20. Bertolatus JA, Klinzman D. Macromolecular sieving by glomer- the diabetic rat kidney. Am J Physiol Renal Physiol 274: F700 –
ular basement membrane in vitro: effect of polycation or biochem- F708, 1998.
ical modifications. Microvasc Res 41: 311–327, 1991. 42. Cattran D, Neogi T, Sharma R, McCarthy ET, Savin VJ. Serial
21. Birn H, Christensen EI. Renal albumin absorption in physiology estimates of serum permeability activity and clinical correlates in
and pathology. Kidney Int 69: 440 – 449, 2006. patients with native kidney focal segmental glomerulosclerosis.
22. Birn H, Willnow TE, Nielsen R, Norden AG, Bonsch C, J Am Soc Nephrol 14: 448 – 453, 2003.
Moestrup SK, Nexo E, Christensen EI. Megalin is essential for 43. Caulfield JP, Farquhar MG. The permeability of glomerular cap-
renal proximal tubule reabsorption and accumulation of transco- illaries to graded dextrans. Identification of the basement mem-
balamin-B(12). Am J Physiol Renal Physiol 282: F408 –F416, 2002.
brane as the primary filtration barrier. J Cell Biol 63: 883–903, 1974.
23. Björnson A, Moses J, Ingemansson A, Haraldsson B, Sörens-
44. Chambers R, Zweifach BW. Intercellular cement and capillary
son J. Primary human glomerular endothelial cells produce pro-
permeability. Physiol Rev 27: 436 – 463, 1947.
teoglycans, puromycin affects their posttranslational modification.
45. Chang RLS, Deen WM, Robertson CR, Bennett CM, Glassock
Am J Physiol Renal Physiol 288: F748 –F756, 2005.
24. Björnson Granqvist A, Ebefors K, Saleem MA, Mathieson PW, RJ, Brenner BM, Troy JL, Ueki IF, Rasmussen B. Permselec-
Haraldsson B, Sörensson Nyström J. Podocyte proteoglycan tivity of the glomerular capillary wall. Studies of experimental
synthesis is involved in the development of nephrotic syndrome. glomerulonephritis in the rat using neutral dextran. J Clin Invest
Am J Physiol Renal Physiol 292: F722–F730, 2006. 57: 1272–1286, 1976.
25. Blau EB, Haas JE. Glomerular sialic acid and proteinuria in 46. Chang RLS, Deen WM, Robertson CR, Brenner BM. Permselec-
human renal disease. Lab Invest 28: 477– 481, 1973. tivity of the glomerular capillary wall. III. Restricted transport of
26. Blouch K, Deen WM, Fauvel JP, Bialek J, Derby G, Myers BD. polyanions. Kidney Int 8: 212–218, 1975.
Molecular configuration and glomerular size selectivity in healthy 47. Chang RLS, Ueki IF, Troy JL, Deen WM, Robertson CR, Bren-
and nephrotic humans. Am J Physiol Renal Physiol 273: F430 – ner BM. Permselectivity of the glomerular capillary wall to mac-
F437, 1997. romolecules. II. Experimental studies in rats using neutral dextran.
27. Bolton GR, Deen WM. Limitations in the application of fiber- Biophys J 15: 887–906, 1975.
matrix models to glomerular basement membrane. In: Membrane 48. Chang RLS, Robertson CR, Deen WM, Brenner BM. Permselec-
Transport and Renal Physiology, edited by Layton HE, Weinstein tivity of the glomerular capillary wall to macromolecules. I. Theo-
AM. New York: Springer-Verlag, 2002, p. 141–156. retical considerations. Biophys J 15: 861– 886, 1975.
28. Bolton GR, Deen WM, Daniels BS. Assessment of the charge 49. Christensen EI, Birn H. Megalin and cubilin: synergistic endo-
selectivity of glomerular basement membrane using Ficoll sulfate. cytic receptors in renal proximal tubule. Am J Physiol Renal
Am J Physiol Renal Physiol 274: F889 –F896, 1998. Physiol 280: F562–F573, 2001.
29. Boute N, Gribouval O, Roselli S, Benessy F, Lee H, Fuchshu- 50. Christensen EI, Devuyst O, Dom G, Nielsen R, Van der Smis-
ber A, Dahan K, Gubler MC, Niaudet P, Antignac C. NPHS2, sen P, Verroust P, Leruth M, Guggino WB, Courtoy PJ. Loss of
encoding the glomerular protein podocin, is mutated in autosomal chloride channel ClC-5 impairs endocytosis by defective trafficking
recessive steroid-resistant nephrotic syndrome. Nature Genet 24: of megalin and cubilin in kidney proximal tubules. Proc Natl Acad
349 –354, 2000. Sci USA 100: 8472– 8477, 2003.
30. Boyce NW, Holdsworth SR. Glomerular permselectivity in the 51. Ciarimboli G, Bokenkamp A, Schurek HJ, Fels LM, Kilian I,
isolated perfused rat kidney. Am J Physiol Renal Fluid Electrolyte Maess B, Stolte H. The “fixed” charge of glomerular capillary wall
Physiol 249: F780 –F784, 1985. as determinant of permselectivity. Renal Failure 23: 365–376, 2001.
31. Brantsma AH, Bakker SJ, Hillege HL, de Zeeuw D, de Jong 52. Ciarimboli G, Hjalmarsson C, Bokenkamp A, Schurek HJ,
PE, Gansevoort RT. Urinary albumin excretion and its relation Haraldsson B. Dynamic alterations of glomerular charge density
with C-reactive protein and the metabolic syndrome in the predic- in fixed rat kidneys suggest involvement of endothelial cell coat.
tion of type 2 diabetes. Diabetes Care 28: 2525–2530, 2005. Am J Physiol Renal Physiol 285: F722–F730, 2003.
32. Bray J, Robinson GB. Influence of charge on filtration across 53. Ciarimboli G, Schurek HJ, Zeh M, Flohr H, Bokenkamp A,
renal basement membrane films in vitro. Kidney Int 25: 527–533, Fels LM, Kilian I, Stolte H. Role of albumin and glomerular
1984.
capillary wall charge distribution on glomerular permselectivity:
33. Breimer ME, Svalander CT, Haraldsson B, Björck S. Physio-
studies on the perfused-fixed rat kidney model. Pflügers Arch 438:
logical and histological characterisation of a pig kidney in vitro
883– 891, 1999.
perfusion model for xenotransplantation studies. Scand J Urol
54. Clavant SP, Sastra SA, Osicka TM, Comper WD. The analysis
Nephrol 30: 213–221, 1996.
34. Brenchley PEC. Vascular permeability factors in steroid-sensitive and characterisation of immuno-unreactive urinary albumin in
nephrotic syndrome and focal segmental glomerulosclerosis. Neph- healthy volunteers. Clin Biochem 39: 143–151, 2006.
rol Dial Transplant 18 Suppl 6: 21–25, 2003. 55. Clough G. Relationship between microvascular permeability and
35. Brenner BM, Baylis C, Deen WM. Transport of molecules across ultrastructure. Prog Biophys Mol Biol 55: 47– 69, 1991.
renal glomerular capillaries. Physiol Rev 56: 502–534, 1976. 56. Comper WD, Osicka TM. Detection of urinary albumin. Adv
36. Brenner BM, Bohrer MP, Baylis C, Deen WM. Determinants of Chronic Kidney Dis 12: 170 –176, 2005.
glomerular permselectivity: insights derived from observations in 57. Comper WD, Tay M, Wells X, Dawes J. Desulphation of dextran
vivo. Kidney Int 12: 229 –237, 1977. sulphate during kidney ultrafiltration. Biochem J 297: 31–34, 1994.
37. Bridges CR Jr, Rennke HG, Deen WM, Troy JL, Brenner BM. 58. Curry FE, Rutledge JC, Lenz JF. Modulation of microvessel wall
Reversible hexadimethrine-induced alterations in glomerular struc- charge by plasma glycoprotein orosomucoid. Am J Physiol Heart
ture and permeability. J Am Soc Nephrol 1: 1095–1108, 1991. Circ Physiol 257: H1354 –H1359, 1989.
38. Brink HM, Moons WM, Slegers JF. Glomerular filtration in the 59. Cutillas PR, Chalkley RJ, Hansen KC, Cramer R, Norden AG,
isolated perfused kidney. I. Sieving of macromolecules. Pflügers Waterfield MD, Burlingame AL, Unwin RJ. The urinary pro-
Arch 397: 42– 47, 1983. teome in Fanconi syndrome implies specificity in the reabsorption
39. Brown MD, Egginton S, Hudlicka O, Zhou AL. Appearance of of proteins by renal proximal tubule cells. Am J Physiol Renal
the capillary endothelial glycocalyx in chronically stimulated rat Physiol 287: F353–F364, 2004.

60. Cuypers Y, Vandenreyt I, Bipat R, Toelsie J, Van Damme B, albumin to the rat renal vein—the albumin retrieval pathway. Renal
Steels P. The functional state of the isolated rabbit kidney per- Failure 23: 347–363, 2001.
fused with autologous blood. Pflügers Arch 440: 634 – 642, 2000. 85. Eremina V, Quaggin SE. The role of VEGF-A in glomerular de-
61. Dahly-Vernon AJ, Sharma M, McCarthy ET, Savin VJ, Ledbet- velopment and function. Curr Opin Nephrol Hypertens 13: 9 –15,
ter SR, Roman RJ. Transforming growth factor-beta, 20-HETE 2004.
interaction, glomerular injury in Dahl salt-sensitive rats. Hyperten- 86. Eremina V, Sood M, Haigh J, Nagy A, Lajoie G, Ferrara N,
sion 45: 643– 648, 2005. Gerber HP, Kikkawa Y, Miner JH, Quaggin SE. Glomerular-
62. Damiano ER. The effect of the endothelial-cell glycocalyx on the specific alterations of VEGF-A expression lead to distinct congen-
motion of red blood cells through capillaries. Microvasc Res 55: ital and acquired renal diseases. J Clin Invest 111: 707–716, 2003.
77–91, 1998. 87. Farquhar MG. Editorial: the primary glomerular filtration bar-
63. Danielli JF. Capillary permeability and oedema in the perfused rier— basement membrane or epithelial slits? Kidney Int 8: 197–
frog. J Physiol 98: 109 –129, 1940. 211, 1975.
64. Daniels BS. Increased albumin permeability in vitro following 88. Fels LM, Sanz-Altamira PM, Decker B, Elger B, Stolte H.
alterations of glomerular charge is mediated by the cells of the Filtration characteristics of the single isolated perfused glomerulus
filtration barrier. J Lab Clin Med 124: 224 –230, 1994. of Myxine glutinosa. Renal Physiol Biochem 16: 276 –284, 1993.
65. Daniels BS, Hauser EB, Deen WM, Hostetter TH. Glomerular 89. Foidart JB, Dechenne CA, Mahieu P, Creutz CE, de Mey J.
basement membrane: in vitro studies of water and protein perme- Tissue culture of normal rat glomeruli. Isolation and morphological
ability. Am J Physiol Renal Fluid Electrolyte Physiol 262: F919 – characterization of two homogeneous cell lines. Invest Cell Pathol
F926, 1992. 2: 15–26, 1979.
66. Datta K, Li J, Karumanchi SA, Wang E, Rondeau E, Mukho- 90. Franceschini N, North KE, Kopp JB, McKenzie L, Winkler C.
padhyay D. Regulation of vascular permeability factor/vascular NPHS2 gene, nephrotic syndrome and focal segmental glomerulo-
endothelial growth factor (VPF/VEGF-A) expression in podocytes. sclerosis: a HuGE review. Genet Med 8: 63–75, 2006.
Kidney Int 66: 1471–1478, 2004. 91. Franke H, Malyusz M, Runge D. Improved sodium and PAH
67. Dechadilok P, Deen WM. Hindrance factors for diffusion and transport in the isolated fluorocarbon-perfused rat kidney.
convection in pores. Ind Engin Chem Res 45: 6953– 6959, 2006. Nephron 22: 423– 431, 1978.
68. Deen WM. Cellular contributions to glomerular size-selectivity. 92. Fujihara CK, Arcos-Fajardo M, Brandao De Almeida Prado E,
Kidney Int 69: 1295–1297, 2006. Jose Brandao De Almeida Prado M, Sesso A, Zatz R. Enhanced
69. Deen WM. Hindered transport of large molecules in liquid-filled glomerular permeability to macromolecules in the Nagase analbu-
pores. AIChE J 33: 1409 –1425, 1987. minemic rat. Am J Physiol Renal Physiol 282: F45–F50, 2002.
70. Deen WM. What determines glomerular capillary permeability? 93. Gekle D, von Bruchhausen F, Fuchs G. On the size of the pore
J Clin Invest 114: 1412–1414, 2004. equivalent in isolated basement membrane of the rat kidney.
71. Deen WM, Bohrer MP, Epstein NB. Effects of molecular size and Pflügers Arch 289: 180 –190, 1966.
configuration on diffusion in microporous membranes. AIChE J 27:
94. Ghitescu L, Desjardins M, Bendayan M. Immunocytochemical
952–959, 1981.
study of glomerular permeability to anionic, neutral and cationic
72. Deen WM, Bridges CR, Brenner BM, Myers BD. Heteroporous
albumins. Kidney Int 42: 25–32, 1992.
model of glomerular size selectivity: application to normal and
95. Giddings JC, Kucera E, Russel CP, Myers MN. Statistical theory
nephrotic humans. Am J Physiol Renal Fluid Electrolyte Physiol
for the equilibrium distribution of rigid molecules in inert porous
249: F374 –F389, 1985.
networks. Exclusion chromatography. J Phys Chem 72: 4397– 4408,
73. Deen WM, Lazzara MJ, Myers BD. Structural determinants of
1968.
glomerular permeability. Am J Physiol Renal Physiol 281: F579 –
96. Gitay-Goren H, Soker S, Vlodavsky I, Neufeld G. The binding
F596, 2001.
of vascular endothelial growth factor to its receptors is dependent
74. Deen WM, Robertson CR, Brenner BM. A model of glomerular
ultrafiltration in the rat. Am J Physiol 223: 1178 –1183, 1972. on cell surface-associated heparin-like molecules. J Biol Chem 267:
75. Deen WM, Robertson CR, Brenner BM. Concentration polariza- 6093– 6098, 1992.
tion in an ultrafiltering capillary. Biophys J 14: 412– 431, 1974. 97. Glandt ED. Distribution equilibrium between a bulk phase and
76. Deen WM, Satvat B, Jamieson JM. Theoretical model for glo- small pores. AlChE J 27: 51–59, 1981.
merular filtration of charged solutes. Am J Physiol Renal Fluid 98. Glassock RJ. Circulating permeability factors in the nephrotic
Electrolyte Physiol 238: F126 –F139, 1980. syndrome: a fresh look at an old problem. J Am Soc Nephrol 14:
77. Desjardins C, Duling BR. Microvessel hematocrit: measurement 541–543, 2003.
and implications for capillary oxygen transport. American 99. Green DF, Hwang KH, Ryan US, Bourgoignie JJ. Culture of
J Physiol Heart Circ Physiol 252: H494 –H503, 1987. endothelial cells from baboon and human glomeruli. Kidney Int 41:
78. Du Bois R, Decoodt P, Gassée JP, Verniory A, Lambert PP. 1506 –1516, 1992.
Determination of glomerular intracapillary and transcapillary pres- 100. Greive KA, Osicka TM, Russo LM, Comper WD. Fragmentation
sure gradients from sieving data. Pflügers Arch 356: 299 –316, 1974. of filtered proteins and implications for glomerular protein sieving
79. Durvasula RV, Shankland SJ. Podocyte injury and targeting ther- in Fanconi syndrome. Kidney Int 61: 1549 –1550, 2002.
apy: an update. Curr Opin Nephrol Hypertens 15: 1–7, 2006. 101. Groggel GC, Stevenson J, Hovingh P, Linker A, Border WA.
80. Casassa EF. Equilibrium distribution of flexible polymer chains Changes in heparan sulfate correlate with increased glomerular
between a macroscopic solution phase and small voids. J Polym permeability. Kidney Int 33: 517–523, 1988.
Sci Polym Lett 5: 773–778, 1967. 102. Groop PH, Forsblom C, Thomas MC. Mechanisms of disease:
81. Edwards A, Daniels BS, Deen WM. Hindered transport of mac- pathway-selective insulin resistance and microvascular complica-
romolecules in isolated glomeruli. II. Convection and pressure tions of diabetes. Nat Clin Pract Endocrinol Metab 1: 100 –110,
effects in basement membrane. Biophys J 72: 214 –222, 1997. 2005.
82. Edwards A, Deen WM, Daniels BS. Hindered transport of mac- 103. Guan F, Villegas G, Teichman J, Mundel P, Tufro A. Autocrine
romolecules in isolated glomeruli. I. Diffusion across intact and VEGF-A system in podocytes regulates podocin and its interaction
cell-free capillaries. Biophys J 72: 204 –213, 1997. with CD2AP. Am J Physiol Renal Physiol 291: F422–F428, 2006.
83. El-Aouni C, Herbach N, Blattner SM, Henger A, Rastaldi MP, 104. Guan N, Ding J, Zhang J, Yang J. Expression of nephrin, podocin,
Jarad G, Miner JH, Moeller MJ, St-Arnaud R, Dedhar S, alpha-actinin, WT1 in children with nephrotic syndrome. Pediatr
Holzman LB, Wanke R, Kretzler M. Podocyte-specific deletion Nephrol 18: 1122–1127, 2003.
of integrin-linked kinase results in severe glomerular basement 105. Guasch A, Deen WM, Myers BD. Charge selectivity of the glo-
membrane alterations and progressive glomerulosclerosis. J Am merular filtration barrier in healthy and nephrotic humans. J Clin
Soc Nephrol 17: 1334 –1344, 2006. Invest 92: 2274 –2282, 1993.
84. Eppel GA, Osicka TM, Pratt LM, Jablonski P, Howden B, 106. Guasch A, Hashimoto H, Sibley RK, Deen WM, Myers BD.
Glasgow EF, Comper WD. The return of glomerular filtered Glomerular dysfunction in nephrotic humans with minimal

changes or focal glomerulosclerosis. Am J Physiol Renal Fluid 127. Huber TB, Reinhardt HC, Exner M, Burger JA, Kerjaschki D,
Electrolyte Physiol 260: F728 –F737, 1991. Saleem MA, Pavenstädt H. Expression of functional CCR and
107. Guimaraes MA, Nikolovski J, Pratt LM, Greive K, Comper CXCR chemokine receptors in podocytes. J Immunol 168: 6244 –
WD. Anomalous fractional clearance of negatively charged Ficoll 6252, 2002.
relative to uncharged Ficoll. Am J Physiol Renal Physiol 285: 128. Hutchison AS, O’Reilly DS, MacCuish AC. Albumin excretion
F1118 –F1124, 2003. rate, albumin concentration, albumin/creatinine ratio compared for
108. Gyenge CC, Tenstad O, Wiig H. In vivo determination of steric screening diabetics for slight albuminuria. Clin Chem 34: 2019 –
and electrostatic exclusion of albumin in rat skin and skeletal 2021, 1988.
muscle. J Physiol 552: 907–916, 2003. 129. Huxley VH, Williams DA. Role of a glycocalyx on coronary
109. Ha TS, Duraisamy S, Faulkner JL, Kasinath BS. Regulation of arteriole permeability to proteins: evidence from enzyme treat-
glomerular endothelial cell proteoglycans by glucose. J Korean ments. Am J Physiol Heart Circ Physiol 278: H1177–H1185, 2000.
Med Sci 19: 245–252, 2004. 130. Igarashi S, Nagase M, Oda T, Honda N. Molecular sieving by
110. Hagen SG, Michael AF, Butkowski RJ. Immunochemical and glomerular basement membrane isolated from normal and ne-
biochemical evidence for distinct basement membrane heparan phrotic rabbits. Clin Chim Acta 68: 255–258, 1976.
sulfate proteoglycans. J Biol Chem 268: 7261–7269, 1993. 131. Iozzo RV, Cohen IR, Grassel S, Murdoch AD. The biology of
111. Halbesma N, Kuiken DS, Brantsma AH, Bakker SJ, Wetzels perlecan: the multifaceted heparan sulphate proteoglycan of base-
JF, De Zeeuw D, De Jong PE, Gansevoort RT. Macroalbumin-
ment membranes and pericellular matrices. Biochem J 302: 625–
uria is a better risk marker than low estimated GFR to identify
639, 1994.
individuals at risk for accelerated GFR loss in population screen-
132. Jarad G, Cunningham J, Shaw AS, Miner JH. Proteinuria pre-
ing. J Am Soc Nephrol 17: 2582–2590, 2006.
cedes podocyte abnormalities inLamb2⫺/⫺ mice, implicating the
112. Haldenby KA, Chappell DC, Winlove CP, Parker KH, Firth JA.
Focal and regional variations in the composition of the glycocalyx glomerular basement membrane as an albumin barrier. J Clin
of large vessel endothelium. J Vasc Res 31: 2–9, 1994. Invest 116: 2272–2279, 2006.
113. Halfter W, Dong S, Schurer B, Cole GJ. Collagen XVIII is a 133. Jeansson M, Granqvist AB, Nyström JS, Haraldsson B. Func-
basement membrane heparan sulfate proteoglycan. J Biol Chem tional and molecular alterations of the glomerular barrier in long-
273: 25404 –25412, 1998. term diabetes in mice. Diabetologia 49: 2200 –2209, 2006.
114. Haraldsson B, Rippe B. Orosomucoid as one of the serum com- 134. Jeansson M, Haraldsson B. Glomerular size and charge selectiv-
ponents contributing to normal capillary permselectivity in rat ity in the mouse after exposure to glucosaminoglycan-degrading
skeletal muscle. Acta Physiol Scand 129: 127–135, 1987. enzymes. J Am Soc Nephrol 14: 1756 –1765, 2003.
115. Haraldsson B, Rippe B. Serum factors other than albumin are 135. Jeansson M, Haraldsson B. Morphological and functional evi-
needed for the maintenance of normal capillary permeability in rat dence for an important role of the endothelial cell glycocalyx in the
hindlimb muscle. Acta Physiol Scand 123: 427– 436, 1985. glomerular barrier. Am J Physiol Renal Physiol 290: F111–F116,
116. Haraldsson B, Sörensson J. Why do we not all have proteinuria? 2006.
An update of our current understanding of the glomerular barrier. 136. Johnson EM, Deen WM. Electrostatic effects on the equilibrium
News Physiol Sci 19: 7–10, 2004. partitioning of spherical colloids in random fibrous media.. J Col-
117. Haraldsson BS, Johnsson EK, Rippe B. Glomerular permselec- loid Interface Sci 178: 749 –756, 1996.
tivity is dependent on adequate serum concentrations of orosomu- 137. Johnson EM, Deen WM. Errata: electrostatic effects on the equi-
coid. Kidney Int 41: 310 –316, 1992. librium partitioning of spherical colloids in random fibrous media.
118. Hassell JR, Robey PG, Barrach HJ, Wilczek J, Rennard SI, J Colloid Interface Sci 195: 268 –268, 1997.
Martin GR. Isolation of a heparan sulfate-containing proteoglycan 138. Johnson EM, Deen WM. Hydraulic permeability of agarose gels.
from basement membrane. Proc Natl Acad Sci USA 77: 4494 – 4498, AIChE J 42: 1220 –1224, 1996.
1980. 139. Johnsson E, Haraldsson B. Addition of purified orosomucoid
119. Hillege HL, Janssen WM, Bak AA, Diercks GF, Grobbee DE, preserves the glomerular permeability for albumin in isolated per-
Crijns HJ, Van Gilst WH, De Zeeuw D, De Jong PE. Microalbu- fused rat kidneys. Acta Physiol Scand 147: 1– 8, 1993.
minuria is common, also in a nondiabetic, nonhypertensive popu- 140. Johnsson E, Haraldsson B. An isolated perfused rat kidney prep-
lation, an independent indicator of cardiovascular risk factors and aration designed for assessment of glomerular permeability char-
cardiovascular morbidity. J Internal Med 249: 519 –526, 2001. acteristics. Acta Physiol Scand 144: 65–73, 1992.
120. Hilliard LM, Osicka TM, Clavant SP, Robinson PJ, Nikolic- 141. Johnston ST, Deen WM. Hindered convection of Ficoll and pro-
Paterson DJ, Comper WD. Characterization of the urinary albu- teins in agarose gels. Ind Eng Chem Res 41: 340 –346, 2002.
min degradation pathway in the isolated perfused rat kidney. J Lab 142. Johnstone DB, Holzman LB. Clinical impact of research on the
Clin Med 147: 36 – 44, 2006. podocyte slit diaphragm. Nature Clin Pract 2: 271–282, 2006.
121. Hjalmarsson C, Johansson BR, Haraldsson B. Electron micro- 143. Julian BA, Novak J. IgA nephropathy: an update. Curr Opin
scopic evaluation of the endothelial surface layer of glomerular
Nephrol Hypertens 13: 171–179, 2004.
capillaries. Microvasc Res 67: 9 –17, 2004.
144. Kalluri R. Proteinuria with and without renal glomerular podocyte
122. Hjalmarsson C, Lidell ME, Haraldsson B. Beneficial effects of
effacement. J Am Soc Nephrol 17: 2383–2389, 2006.
orosomucoid on the glomerular barrier in puromycin aminonucleo-
145. Kanwar YS, Linker A, Farquhar MG. Increased permeability of
side-induced nephrosis. Nephrol Dial Transplant 21: 1223–1230,
2006. the glomerular basement membrane to ferritin after removal of
123. Hjalmarsson C, Ohlson M, Haraldsson B. Puromycin amino- glycosaminoglycans (heparan sulfate) by enzyme digestion. J Cell
nucleoside damages the glomerular size barrier with minimal ef- Biol 86: 688 – 693, 1980.
fects on charge density. Am J Physiol Renal Physiol 281: F503– 146. Kanwar YS, Rosenzweig LJ. Clogging of the glomerular base-
F512, 2001. ment membrane. J Cell Biol 93: 489 – 494, 1982.
124. Hryciw DH, Ekberg J, Pollock CA, Poronnik P. ClC-5: a chlo- 147. Karumanchi SA, Maynard SE, Stillman IE, Epstein FH,
ride channel with multiple roles in renal tubular albumin uptake. Sukhatme VP. Preeclampsia: a renal perspective. Kidney Int 67:
Int J Biochem Cell Biol 38: 1036 –1042, 2006. 2101–2113, 2005.
125. Huber TB, Benzing T. The slit diaphragm: a signaling platform to 148. Kasinath BS. Glomerular endothelial cell proteoglycans—regula-
regulate podocyte function. Curr Opin Nephrol Hypertens 14: tion by TGF-beta 1. Arch Biochem Biophys 305: 370 –377, 1993.
211–216, 2005. 149. Kelley VE, Cavallo T. Glomerular permeability: transfer of native
126. Huber TB, Kwoh C, Wu H, Asanuma K, Godel M, Hartleben B, ferritin in glomeruli with decreased anionic sites. Lab Invest 39:
Blumer KJ, Miner JH, Mundel P, Shaw AS. Bigenic mouse 547–553, 1978.
models of focal segmental glomerulosclerosis involving pairwise 150. Kerjaschki D. Dysfunctions of cell biological mechanisms of vis-
interaction of CD2AP, Fyn, synaptopodin. J Clin Invest 116: 1337– ceral epithelial cell (podocytes) in glomerular diseases. Kidney Int
1345, 2006. 45: 300 –313, 1994.

151. Kerjaschki D, Sharkey DJ, Farquhar MG. Identification and SA. Soluble endoglin and other circulating antiangiogenic factors
characterization of podocalyxin—the major sialoprotein of the re- in preeclampsia. N Engl J Med 355: 992–1005, 2006.
nal glomerular epithelial cell. J Cell Biol 98: 1591–1596, 1984. 172. Levine RJ, Maynard SE, Qian C, Lim KH, England LJ, Yu KF,
152. Kerjaschki D, Vernillo AT, Farquhar MG. Reduced sialylation Schisterman EF, Thadhani R, Sachs BP, Epstein FH, Sibai
of podocalyxin–the major sialoprotein of the rat kidney glomeru- BM, Sukhatme VP, Karumanchi SA. Circulating angiogenic fac-
lus–in aminonucleoside nephrosis. Am J Pathol 118: 343–349, 1985. tors and the risk of preeclampsia. N Engl J Med 350: 672– 683, 2004.
153. Kestila M, Lenkkeri U, Mannikko M, Lamerdin J, McCready 173. Lieberthal W, Rennke HG, Sandock KM, Valeri CR, Levinsky
P, Putaala H, Ruotsalainen V, Morita T, Nissinen M, Herva R, NG. Ischemia in the isolated erythrocyte-perfused rat kidney. Pro-
Kashtan CE, Peltonen L, Holmberg C, Olsen A, Tryggvason K. tective effect of hypothermia. Renal Physiol Biochem 11: 60 – 69,
Positionally cloned gene for a novel glomerular protein–nephrin–is 1988.
mutated in congenital nephrotic syndrome. Mol Cell 1: 575–582, 174. Lijnen HR, Collen D. Endothelium in hemostasis and thrombosis.
1998. Prog Cardiovasc Dis 39: 343–350, 1997.
154. Kestila M, Mannikko M, Holmberg C, Gyapay G, Weissenbach 175. Lindström KE, Blom A, Johnsson E, Haraldsson B, Fries E.
J, Savolainen ER, Peltonen L, Tryggvason K. Congenital ne- High glomerular permeability of bikunin despite similarity in
phrotic syndrome of the Finnish type maps to the long arm of charge and hydrodynamic size to serum albumin. Kidney Int 51:
chromosome 19. Am J Hum Genet 54: 757–764, 1994. 1053–1058, 1997.
155. Limbach KW, Nitsche JM, Wei J. Partitioning of nonspherical 176. Lindström KE, Johnsson E, Haraldsson B. Glomerular charge
molecules between bulk solution and porous solids. AIChE J 35: selectivity for proteins larger than serum albumin as revealed by
42–52, 1989. lactate dehydrogenase isoforms. Acta Physiol Scand 162: 481– 488,
156. Kitahara T, Hiromura K, Ikeuchi H, Yamashita S, Kobayashi 1998.
S, Kuroiwa T, Kaneko Y, Ueki K, Nojima Y. Mesangial cells 177. Lindström KE, Rönnstedt L, Jaremko G, Haraldsson B. Phys-
stimulate differentiation of endothelial cells to form capillary-like iological and morphological effects of perfusing isolated rat kid-
networks in a three-dimensional culture system. Nephrol Dial neys with hyperosmolal mannitol solutions. Acta Physiol Scand
Transplant 20: 42– 49, 2005. 166: 231–238, 1999.
157. Klassen RB, Crenshaw K, Kozyraki R, Verroust PJ, Tio L, 178. Luft JH. Fine structures of capillary and endocapillary layer as
Atrian S, Allen PL, Hammond TG. Megalin mediates renal up- revealed by ruthenium red. Federation Proc 25: 1773–1783, 1966.
take of heavy metal metallothionein complexes. Am J Physiol 179. Lund U, Rippe A, Venturoli D, Tenstad O, Grubb A, Rippe B.
Renal Physiol 287: F393–F403, 2004. Glomerular filtration rate dependence of sieving of albumin and
158. Kosto KB, Deen WM. Diffusivities of macromolecules in compos- some neutral proteins in rat kidneys. Am J Physiol Renal Physiol
ite hydrogels. AIChE J 50: 2648 –2658, 2004. 284: F1226 –F1234, 2003.
159. Kosto KB, Deen WM. Hindered convection of macromolecules in 180. Maack T. Physiological evaluation of the isolated perfused rat
hydrogels. Biophys J 88: 277–286, 2005. kidney. Am J Physiol Renal Fluid Electrolyte Physiol 238: F71–
160. Kosto KB, Panuganti S, Deen WM. Equilibrium partitioning of F78, 1980.
Ficoll in composite hydrogels. J Colloid Interface Sci 277: 404 – 409, 181. Maddox DA, Deen WM, Brenner BM. Glomerular filtration. In:
2004. Handbook of Physiology. Renal Physiology.Bethesda, MD: Am.
161. Kramer A, van den Hoven M, Rops A, Wijnhoven T, van den Physiol. Soc., 1992, sect. 8, vol. I, chapt. 13, p. 545– 638.
Heuvel L, Lensen J, van Kuppevelt T, van Goor H, van der 182. Marrero MB, Banes-Berceli AK, Stern DM, Eaton DC. Role of
Vlag J, Navis G, Berden JH. Induction of glomerular heparanase the JAK/STAT signaling pathway in diabetic nephropathy. Am J
expression in rats with adriamycin nephropathy is regulated by Physiol Renal Physiol 290: F762–F768, 2006.
reactive oxygen species and the renin-angiotensin system. J Am 183. Mattix HJ, Hsu CY, Shaykevich S, Curhan G. Use of the albu-
Soc Nephrol 17: 2513–2520, 2006. min/creatinine ratio to detect microalbuminuria: implications of
162. Kreisberg JI, Venkatachalam M, Troyer D. Contractile proper- sex and race. J Am Soc Nephrol 13: 1034 –1039, 2002.
ties of cultured glomerular mesangial cells. Am J Physiol Renal 184. Maunsbach AB. Absorption of 125I-labeled homologous albumin
Physiol 249: F457–F463, 1985. by rat kidney proximal tubule cells. A study of microperfused
163. Kriz W, Kaissling B. Structural organization of the mammalian single proximal tubules by electron microscopic autoradiography
kidney. In: The Kidney Physiology and Pathophysiology, edited by and histochemistry 1966. J Am Soc Nephrol 8: 323–351, 1997.
Seldin GW, Giebisch G. New York: Raven, 1992, p. 707–777. 185. Maunsbach AB. Absorption of I-125-labeled homologous albumin
164. Lagrue G, Xheneumont S, Branellec A, Hirbec G, Weil B. A by rat kidney proximal tubule cells. A study of microperfused
vascular permeability factor elaborated from lymphocytes. I. Dem- single proximal tubules by electron microscopic autoradiography
onstration in patients with nephrotic syndrome. Biomedicine 23: and histochemistry. J Ultrastruct Res 15: 197–241, 1966.
37– 40, 1975. 186. Maynard SE, Min JY, Merchan J, Lim KH, Li J, Mondal S,
165. Lai AS, Lai KN. Molecular basis of IgA nephropathy. Curr Mol Libermann TA, Morgan JP, Sellke FW, Stillman IE, Epstein
Med 5: 475– 487, 2005. FH, Sukhatme VP, Karumanchi SA. Excess placental soluble
166. Lazzara MJ, Blankschtein D, Deen WM. Effects of multisolute fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial
steric interactions on membrane partition coefficients. J Colloid dysfunction, hypertension, proteinuria in preeclampsia. J Clin
Interface Sci 226: 112–122, 2000. Invest 111: 649 – 658, 2003.
167. Lazzara MJ, Deen WM. Effects of concentration on the partition- 187. Maynard SE, Venkatesha S, Thadhani R, Karumanchi SA.
ing of macromolecule mixtures in agarose gels. J Colloid Interface Soluble Fms-like tyrosine kinase 1 and endothelial dysfunction in
Sci 272: 288 –297, 2004. the pathogenesis of preeclampsia. Pediatr Res 57: 1R–7R, 2005.
168. Lazzara MJ, Deen WM. Effects of plasma proteins on sieving of 188. McCarthy ET, Sharma R, Sharma M, Li JZ, Ge XL, Dileezpan
tracer macromolecules in glomerular basement membrane. Am J KN, Savin VJ. TNF-alpha increases albumin permeability of iso-
Physiol Renal Physiol 281: F860 –F868, 2001. lated rat glomeruli through the generation of superoxide. J Am Soc
169. Lazzara MJ, Deen WM. Model of albumin reabsorption in the Nephrol 9: 433– 438, 1998.
proximal tubule. Am J Physiol Renal Physiol 292: F430 –F439, 189. Mene P, Simonson MS, Dunn MJ. Physiology of the mesangial
2007. cell. Physiol Rev 69: 1347–1424, 1989.
170. Lemmink HH, Mochizuki T, van den Heuvel LP, Schroder CH, 190. Meyrier A. Mechanisms of disease: focal segmental glomeruloscle-
Barrientos A, Monnens LA, van Oost BA, Brunner HG, Reed- rosis. Nature Clin Pract 1: 44 –54, 2005.
ers ST, Smeets HJ. Mutations in the type IV collagen alpha 3 191. Michel CC, Phillips ME. The effect of Ficoll 70 and bovine serum
(COL4A3) gene in autosomal recessive Alport syndrome. Hum Mol albumin on the permeability properties of individually perfused
Genet 3: 1269 –1273, 1994. frog mesenteric capillaries. J Physiol 291: 39P, 1979.
171. Levine RJ, Lam C, Qian C, Yu KF, Maynard SE, Sachs BP, 192. Miner JH, Morello R, Andrews KL, Li C, Antignac C, Shaw AS,
Sibai BM, Epstein FH, Romero R, Thadhani R, Karumanchi Lee B. Transcriptional induction of slit diaphragm genes by Lmx1b

is required in podocyte differentiation. J Clin Invest 109: 1065– 211. Ogston AG. The spaces in a uniform random suspension of fibres.
1072, 2002. Trans Faraday Soc 54: 1754 –1757, 1958.
193. Miner JH, Sanes JR. Collagen IV alpha 3, alpha 4, alpha 5 chains 212. Ohlson M, Sörensson J, Haraldsson B. A gel-membrane model
in rodent basal laminae: sequence, distribution, association with of glomerular charge and size selectivity in series. Am J Physiol
laminins, developmental switches. J Cell Biol 127: 879 – 891, 1994. Renal Physiol 280: F396 –F405, 2001.
194. Mitchell BD, Deen WM. Theoretical effects of macromolecule 213. Ohlson M, Sörensson J, Haraldsson B. Glomerular size and
concentration and charge on membrane rejection coefficients. J charge selectivity in the rat as revealed by FITC-ficoll and albumin.
Membr Sci 19: 75–100, 1984. Am J Physiol Renal Physiol 279: F84 –F91, 2000.
195. Mochizuki T, Lemmink HH, Mariyama M, Antignac C, Gubler 214. Ohlson M, Sörensson J, Lindström K, Blom AM, Fries E,
MC, Pirson Y, Verellen-Dumoulin C, Chan B, Schroder CH, Haraldsson B. Effects of filtration rate on the glomerular barrier
Smeets HJ. Identification of mutations in the alpha 3(IV) and and clearance of four differently shaped molecules. Am J Physiol
alpha 4(IV) collagen genes in autosomal recessive Alport syn- Renal Physiol 281: F103–F113, 2001.
drome. Nature Genet 8: 77– 81, 1994. 215. Ohyama K, Seyer JM, Raghow R, Kang AH. Extracellular matrix
196. Morigi M, Zoja C, Figliuzzi M, Remuzzi G, Remuzzi A. Super- phenotype of rat mesangial cells in culture. Biosynthesis of colla-
natant of endothelial cells exposed to laminar flow inhibits mesan- gen types I, III, IV, V and a low molecular weight collagenous
gial cell proliferation. Am J Physiol Cell Physiol 264: C1080 –C1083, component and their regulation by dexamethasone. J Lab Clin Med
1993. 116: 219 –227, 1990.
197. Morita H, Yoshimura A, Inui K, Ideura T, Watanabe H, Wang 216. Oken DE, Flamenbaum W. Micropuncture studies of proximal
L, Soininen R, Tryggvason K. Heparan sulfate of perlecan is tubule albumin concentrations in normal and nephrotic rats. J Clin
involved in glomerular filtration. J Am Soc Nephrol 16: 1703–1710, Invest 50: 1498 –1505, 1971.
2005. 217. Oliver JD 3rd, Anderson S, Troy JL, Brenner BM, Deen WM.
198. Mulivor AW, Lipowsky HH. Inflammation- and ischemia-induced Determination of glomerular size-selectivity in the normal rat with
shedding of venular glycocalyx. Am J Physiol Heart Circ Physiol Ficoll. J Am Soc Nephrol 3: 214 –228, 1992.
286: H1672–H1680, 2004. 218. Oliver JD 3rd, Deen WM. Random-coil model for glomerular
199. Mundel P, Kriz W. Cell culture of podocytes. Exp Nephrol 4: sieving of dextran. Bull Math Biol 56: 369 –389, 1994.
263–266, 1996. 219. Oliver JD 3rd, Simons JL, Troy JL, Provoost AP, Brenner BM,
200. Mundel P, Reiser J, Kriz W. Induction of differentiation in cul- Deen WM. Proteinuria and impaired glomerular permselectivity in
tured rat and human podocytes. J Am Soc Nephrol 8: 697–705, 1997. uninephrectomized fawn-hooded rats. Am J Physiol Renal Fluid
201. Mundel P, Reiser J, Zuniga Mejia Borja A, Pavenstädt H, Electrolyte Physiol 267: F917–F925, 1994.
Davidson GR, Kriz W, Zeller R. Rearrangements of the cytoskel- 220. Olson JL, Rennke HG, Venkatachalam MA. Alterations in the
eton and cell contacts induce process formation during differenti- charge and size selectivity barrier of the glomerular filter in ami-
ation of conditionally immortalized mouse podocyte cell lines. Exp nonucleoside nephrosis in rats. Lab Invest 44: 271–279, 1981.
221. Osicka TM, Comper WD. Characterization of immunochemically
Cell Res 236: 248 –258, 1997.
nonreactive urinary albumin. Clin Chem 50: 2286 –2291, 2004.
202. Muragaki Y, Timmons S, Griffith CM, Oh SP, Fadel B, Quer-
222. Osicka TM, Comper WD. Glomerular charge selectivity for an-
termous T, Olsen BR. Mouse Col18a1 is expressed in a tissue-
ionic and neutral horseradish peroxidase. Kidney Int 47: 1630 –
specific manner as three alternative variants and is localized in
1637, 1995.
basement membrane zones. Proc Natl Acad Sci USA 92: 8763– 8767,
223. Osicka TM, Comper WD. Protein degradation during renal pas-
1995.
sage in normal kidneys is inhibited in experimental albuminuria.
203. Myers BD, Guasch A. Selectivity of the glomerular filtration bar-
Clin Sci 93: 65–72, 1997.
rier in healthy and nephrotic humans. Am J Nephrol 13: 311–317,
224. Osicka TM, Comper WD. Tubular inhibition destroys charge se-
1993.
lectivity for anionic and neutral horseradish peroxidase. Biochim
204. Nieuwdorp M, van Haeften TW, Gouverneur MC, Mooij HL, Biophys Acta 1381: 170 –178, 1998.
van Lieshout MH, Levi M, Meijers JC, Holleman F, Hoekstra 225. Osicka TM, Hankin AR, Comper WD. Puromycin aminonucleo-
JB, Vink H, Kastelein JJ, Stroes ES. Loss of endothelial glyco- side nephrosis results in a marked increase in fractional clearance
calyx during acute hyperglycemia coincides with endothelial dys- of albumin. Am J Physiol Renal Physiol 277: F139 –F145, 1999.
function and coagulation activation in vivo. Diabetes 55: 480 – 486, 226. Osicka TM, Houlihan CA, Chan JG, Jerums G, Comper WD.
2006. Albuminuria in patients with type 1 diabetes is directly linked to
205. Noakes PG, Miner JH, Gautam M, Cunningham JM, Sanes JR, changes in the lysosome-mediated degradation of albumin during
Merlie JP. The renal glomerulus of mice lacking s-laminin/laminin renal passage. Diabetes 49: 1579 –1584, 2000.
beta 2: nephrosis despite molecular compensation by laminin beta 227. Osicka TM, Panagiotopoulos S, Jerums G, Comper WD. Frac-
1. Nature Genet 10: 400 – 406, 1995. tional clearance of albumin is influenced by its degradation during
206. Norden AG, Lapsley M, Igarashi T, Kelleher CL, Lee PJ, renal passage. Clin Sci 93: 557–564, 1997.
Matsuyama T, Scheinman SJ, Shiraga H, Sundin DP, Thakker 228. Osicka TM, Pratt LM, Comper WD. Glomerular capillary wall
RV, Unwin RJ, Verroust P, Moestrup SK. Urinary megalin defi- permeability to albumin and horseradish peroxidase. Nephrology
ciency implicates abnormal tubular endocytic function in Fanconi 199 –212, 1996.
syndrome. J Am Soc Nephrol 13: 125–133, 2002. 229. Osicka TM, Yu Y, Panagiotopoulos S, Clavant SP, Kiriazis Z,
207. Norden AG, Lapsley M, Lee PJ, Pusey CD, Scheinman SJ, Tam Pike RN, Pratt LM, Russo LM, Kemp BE, Comper WD, Jerums
FW, Thakker RV, Unwin RJ, Wrong O. Fragmentation of filtered G. Prevention of albuminuria by aminoguanidine or ramipril in
proteins and implications for glomerular protein sieving in Fanconi streptozotocin-induced diabetic rats is associated with the normal-
syndrome. Kidney Int 62: 349, 2002. ization of glomerular protein kinase C. Diabetes 49: 87–93, 2000.
208. Norden AG, Lapsley M, Lee PJ, Pusey CD, Scheinman SJ, Tam 230. Pappenheimer JR, Renkin EM, Borrero LM. Filtration,diffusion
FW, Thakker RV, Unwin RJ, Wrong O. Glomerular protein siev- and molecular sieving through peripheral capillary membranes. A
ing and implications for renal failure in Fanconi syndrome. Kidney contribution to the pore theory of capillary permeability. Am J
Int 60: 1885–1892, 2001. Physiol 167: 13– 46, 1951.
209. Norden AG, Sharratt P, Cutillas PR, Cramer R, Gardner SC, 231. Park CH, Maack T. Albumin absorption and catabolism by iso-
Unwin RJ. Quantitative amino acid and proteomic analysis: very lated perfused proximal convoluted tubules of the rabbit. J Clin
low excretion of polypeptides ⬎750 Da in normal urine. Kidney Int Invest 73: 767–777, 1984.
66: 1994 –2003, 2004. 232. Pavenstadt H. The charge for going by foot: modifying the surface
210. Ogamo A, Nagai A, Nagasawa K. Binding of heparin fractions and of podocytes. Exp Nephrol 6: 98 –103, 1998.
other polysulfated polysaccharides to plasma fibronectin: effects of 233. Pavenstädt H, Henger A, Briner V, Greger R, Schollmeyer P.
molecular size and degree of sulfation of polysaccharides. Biochim Extracellular ATP regulates glomerular endothelial cell function. J
Biophys Acta 841: 30 – 41, 1985. Auton Pharmacol 16: 389 –391, 1996.

234. Pavenstädt H, Kriz W, Kretzler M. Cell biology of the glomerular 256. Rosenzweig LJ, Kanwar YS. Removal of sulfated (heparan sul-
podocyte. Physiol Rev 83: 253–307, 2003. fate) or nonsulfated (hyaluronic acid) glycosaminoglycans results
235. Phillips RJ. A hydrodynamic model for hindered diffusion of in increased permeability of the glomerular basement membrane to
125
proteins and micelles in hydrogels. Biophys J 79: 3350 –3353, 2000. I-bovine serum albumin. Lab Invest 47: 177–184, 1982.
236. Platts SH, Linden J, Duling BR. Rapid modification of the gly- 257. Ross BD. The isolated perfused rat kidney. Clin Sci Mol Med Suppl
cocalyx caused by ischemia-reperfusion is inhibited by adenosine 55: 513–521, 1978.
A2A receptor activation. Am J Physiol Heart Circ Physiol 284: 258. Rossi M, Morita H, Sormunen R, Airenne S, Kreivi M, Wang L,
H2360 –H2367, 2003. Fukai N, Olsen BR, Tryggvason K, Soininen R. Heparan sulfate
237. Pollak MR. The genetic basis of FSGS and steroid-resistant ne- chains of perlecan are indispensable in the lens capsule but not in
phrosis. Semin Nephrol 23: 141–146, 2003. the kidney. EMBO J 22: 236 –245, 2003.
238. Pyke C, Kristensen P, Ostergaard PB, Oturai PS, Romer J. 259. Rostgaard J, Qvortrup K. Electron microscopic demonstrations
Proteoglycan expression in the normal rat kidney. Nephron 77: of filamentous molecular sieve plugs in capillary fenestrae. Micro-
461– 470, 1997. vasc Res 53: 1–13, 1997.
239. Raats CJ, van den Born J, Bakker MA, Oppers-Walgreen B, 260. Rostgaard J, Qvortrup K. Sieve plugs in fenestrae of glomerular
Pisa BJ, Dijkman HB, Assmann KJ, Berden JH. Expression of capillaries–site of the filtration barrier? Cells Tissues Organs 170:
agrin, dystroglycan, utrophin in normal renal tissue and in experi- 132–138, 2002.
mental glomerulopathies. Am J Pathol 156: 1749 –1765, 2000. 261. Roth J, Brown D, Orci L. Regional distribution of N-acetyl-D-
240. Rabkin R, Kitabchi AE. Factors influencing the handling of insu- galactosamine residues in the glycocalyx of glomerular podocytes.
lin by the isolated rat kidney. J Clin Invest 62: 169 –175, 1978. J Cell Biol 96: 1189 –1196, 1983.
241. Reiser J, von Gersdorff G, Loos M, Oh J, Asanuma K, Gi- 262. Rubio-Gayosso I, Platts SH, Duling BR. Reactive oxygen spe-
ardino L, Rastaldi MP, Calvaresi N, Watanabe H, Schwarz K, cies mediate modification of glycocalyx during ischemia-reperfu-
Faul C, Kretzler M, Davidson A, Sugimoto H, Kalluri R, sion injury. Am J Physiol Heart Circ Physiol 290: H2247–H2256,
Sharpe AH, Kreidberg JA, Mundel P. Induction of B7-1 in podo- 2006.
cytes is associated with nephrotic syndrome. J Clin Invest 113: 263. Ruggenenti P, Mosconi L, Sangalli F, Casiraghi F, Gambara V,
1390 –1397, 2004. Remuzzi G, Remuzzi A. Glomerular size-selective dysfunction in
242. Remuzzi A, Perico N, Sangalli F, Vendramin G, Moriggi M, NIDDM is not ameliorated by ACE inhibition or by calcium channel
Ruggenenti P, Remuzzi G. ACE inhibition and ANG II receptor blockade. Kidney Int 55: 984 –994, 1999.
blockade improve glomerular size-selectivity in IgA nephropathy. 264. Russo LM, Bakris GL, Comper WD. Renal handling of albumin:
Am J Physiol Renal Physiol 276: F457–F466, 1999. a critical review of basic concepts and perspective. Am J Kidney
243. Remuzzi A, Puntorieri S, Battaglia C, Bertani T, Remuzzi G. Dis 39: 899 –919, 2002.
Angiotensin converting enzyme inhibition ameliorates glomerular 265. Russo LM, Comper WD, Osicka TM. Mechanism of albuminuria
filtration of macromolecules and water and lessens glomerular associated with cardiovascular disease and kidney disease. Kidney
injury in the rat. J Clin Invest 85: 541–549, 1990. Int Suppl S67–S68, 2004.
244. Remuzzi A, Viberti G, Ruggenenti P, Battaglia C, Pagni R, 266. Russo LM, Sandoval RM, McKee M, Osicka TM, Collins AB,
Remuzzi G. Glomerular response to hyperglycemia in human dia- Brown D, Molitoris BA, Comper WD. The normal kidney filters
betic nephropathy. Am J Physiol Renal Fluid Electrolyte Physiol nephrotic levels of albumin retrieved by proximal tubule cells:
259: F545–F552, 1990. retrieval is disrupted in nephrotic states. Kidney Int 71: 504 –513,
245. Renkin EM, Robinson RR. Glomerular filtration. N Engl J Med 2007.
290: 785–792, 1974. 267. Ryan GB, Hein SJ, Karnovsky MJ. Glomerular permeability to
246. Rennke HG, Patel Y, Venkatachalam MA. Glomerular filtration proteins. Effects of hemodynamic factors on the distribution of
of proteins: clearance of anionic, neutral, cationic horseradish endogenous immunoglobulin G and exogenous catalase in the rat
peroxidase in the rat. Kidney Int 13: 278 –288, 1978. glomerulus. Lab Invest 34: 415– 427, 1976.
247. Rennke HG, Venkatachalam MA. Glomerular permeability of 268. Ryan GB, Karnovsky MJ. Distribution of endogenous albumin in
macromolecules. Effect of molecular configuration on the frac- the rat glomerulus: role of hemodynamic factors in glomerular
tional clearance of uncharged dextran and neutral horseradish barrier function. Kidney Int 9: 36 – 45, 1976.
peroxidase in the rat. J Clin Invest 63: 713–717, 1979. 269. Salant DJ. The structural biology of glomerular epithelial cells in
248. Rijkmans BG, Buurman WA, Kootstra G. Six-day canine kidney proteinuric diseases. Curr Opin Nephrol Hypertens 3: 569 –574,
preservation. Hypothermic perfusion combined with isolated blood 1994.
perfusion. Transplantation 37: 130 –134, 1984. 270. Saleem MA, O’Hare MJ, Reiser J, Coward RJ, Inward CD,
249. Rinta-Valkama J, Palmen T, Lassila M, Holthofer H. Podocyte- Farren T, Xing CY, Ni L, Mathieson PW, Mundel P. A condi-
associated proteins FAT, alpha-actinin-4 and filtrin are expressed in tionally immortalized human podocyte cell line demonstrating
Langerhans islets of the pancreas. Mol Cell Biochem 294: 117–125, nephrin and podocin expression. J Am Soc Nephrol 13: 630 – 638,
2007. 2002.
250. Rippe B, Haraldsson B. Transport of macromolecules across 271. Sanes JR, Engvall E, Butkowski R, Hunter DD. Molecular
microvascular walls: the two-pore theory. Physiol Rev 74: 163–219, heterogeneity of basal laminae: isoforms of laminin and collagen IV
1994. at the neuromuscular junction and elsewhere. J Cell Biol 111:
251. Rippe C, Asgeirsson D, Venturoli D, Rippe A, Rippe B. Effects 1685–1699, 1990.
of glomerular filtration rate on Ficoll sieving coefficients (theta) in 272. Satchell SC, Tasman CH, Singh A, Ni L, Geelen J, von Ruh-
rats. Kidney Int 69: 1326 –1332, 2006. land CJ, O’Hare MJ, Saleem MA, van den Heuvel LP, Mathie-
252. Rippe C, Rippe A, Larsson A, Asgeirsson D, Rippe B. Nature of son PW. Conditionally immortalized human glomerular endothelial
glomerular capillary permeability changes following acute renal cells expressing fenestrations in response to VEGF. Kidney Int 69:
ischemia/reperfusion (I/R) injury in rats. Am J Physiol Renal 1633–1640, 2006.
Physiol 292: F1362–F1368, 2006. 273. Savage CO. The biology of the glomerulus: endothelial cells. Kid-
253. Robinson GB, Walton HA. Glomerular basement membrane as a ney Int 45: 314 –319, 1994.
compressible ultrafilter. Microvasc Res 38: 36 – 48, 1989. 274. Savin VJ, Sharma R, Lovell HB, Welling DJ. Measurement of
254. Rodewald R, Karnovsky MJ. Porous substructure of the glomer- albumin reflection coefficient with isolated rat glomeruli. J Am Soc
ular slit diaphragm in the rat and mouse. J Cell Biol 60: 423– 433, Nephrol 3: 1260 –1269, 1992.
1974. 275. Scheinman JI, Fish AJ, Brown DM, Michael AJ. Human glo-
255. Rops AL, van der Vlag J, Lensen JF, Wijnhoven TJ, van den merular smooth muscle (mesangial) cells in culture. Lab Invest 34:
Heuvel LP, van Kuppevelt TH, Berden JH. Heparan sulfate 150 –158, 1976.
proteoglycans in glomerular inflammation. Kidney Int 65: 768 –785, 276. Schena FP, Gesualdo L. Pathogenetic mechanisms of diabetic
2004. nephropathy. J Am Soc Nephrol 16 Suppl 1: S30 –S33, 2005.

277. Schlondorff D. The glomerular mesangial cell: an expanding role 298. Sörensson J, Björnson A, Ohlson M, Ballermann BJ, Haralds-
for a specialized pericyte. FASEB J 1: 272–281, 1987. son B. Synthesis of sulfated proteoglycans by bovine glomerular
278. Schnitzer JE, Carley WW, Palade GE. Specific albumin binding endothelial cells in culture. Am J Physiol Renal Physiol 284: F373–
to microvascular endothelium in culture. Am J Physiol Heart Circ F380, 2003.
Physiol 254: H425–H437, 1988. 299. Sörensson J, Matejka GL, Ohlson M, Haraldsson B. Human
279. Schnitzer JE, Pinney E. Quantitation of specific binding of oroso- endothelial cells produce orosomucoid, an important component
mucoid to cultured microvascular endothelium: role in capillary of the capillary barrier. Am J Physiol Heart Circ Physiol 276:
permeability. Am J Physiol Heart Circ Physiol 263: H48 –H55, H530 –H534, 1999.
1992. 300. Sörensson J, Ohlson M, Björnson A, Haraldsson B. Orosomu-
280. Sengoelge G, Luo W, Fine D, Perschl AM, Fierlbeck W, coid has a cAMP-dependent effect on human endothelial cells and
Haririan A, Sorensson J, Rehman TU, Hauser P, Trevick JS, inhibits the action of histamine. Am J Physiol Heart Circ Physiol
Kulak SC, Wegner B, Ballermann BJ. A SAGE-based compari- 278: H1725–H1731, 2000.
son between glomerular and aortic endothelial cells. Am J Physiol 301. Sörensson J, Ohlson M, Haraldsson B. A quantitative analysis of
Renal Physiol 288: F1290 –F1300, 2005. the glomerular charge barrier in the rat. Am J Physiol Renal
281. Sharma M, Sharma R, Ge XL, Fish BL, McCarthy ET, Savin VJ, Physiol 280: F646 –F656, 2001.
Cohen EP, Moulder JE. Early detection of radiation-induced 302. Sörensson J, Ohlson M, Lindström K, Haraldsson B. Glomer-
glomerular injury by albumin permeability assay. Radiat Res 155: ular charge selectivity for horseradish peroxidase and albumin at
474 – 480, 2001. low and normal ionic strengths. Acta Physiol Scand 163: 83–91,
282. Sharma M, Sharma R, McCarthy ET, Savin VJ. The focal seg- 1998.
mental glomerulosclerosis permeability factor: biochemical char- 303. Takami H, Naramoto A, Shigematsu H, Ohno S. Ultrastructure
acteristics and biological effects. Exp Biol Med 229: 85–98, 2004. of glomerular basement membrane by quick-freeze and deep-etch
283. Sharma M, Sharma R, McCarthy ET, Savin VJ. “The FSGS methods. Kidney Int 39: 659 – 664, 1991.
factor”: enrichment and in vivo effect of activity from focal seg- 304. Takeda T. Podocyte cytoskeleton is connected to the integral
mental glomerulosclerosis plasma. J Am Soc Nephrol 10: 552–561, membrane protein podocalyxin through Na⫹/H⫹-exchanger regu-
1999. latory factor 2 and ezrin. Clin Exp Nephrol 7: 260 –269, 2003.
284. Sharma M, Sharma R, Reddy SR, McCarthy ET, Savin VJ. 305. Takemoto M, Asker N, Gerhardt H, Lundkvist A, Johansson
Proteinuria after injection of human focal segmental glomerulo- BR, Saito Y, Betsholtz C. A new method for large scale isolation
sclerosis factor. Transplantation 73: 366 –372, 2002. of kidney glomeruli from mice. Am J Pathol 161: 799 – 805, 2002.
285. Sharma R, Khanna A, Sharma M, Savin VJ. Transforming 306. Tay M, Comper WD, Singh AK. Charge selectivity in kidney
growth factor-beta1 increases albumin permeability of isolated rat ultrafiltration is associated with glomerular uptake of transport
glomeruli via hydroxyl radicals. Kidney Int 58: 131–136, 2000. probes. Am J Physiol Renal Fluid Electrolyte Physiol 260: F549 –
286. Sharma R, Suzuki K, Nagase H, Savin VJ. Matrix metallopro- F554, 1991.
teinase (stromelysin-1) increases the albumin permeability of iso- 307. Taylor AE, Granger DN. Exchange of macromolecules across the
lated rat glomeruli. J Lab Clin Med 128: 297–303, 1996. microcirculation. In: Handbook of Physiology. The Cardiovascular
287. Shih NY, Li J, Karpitskii V, Nguyen A, Dustin ML, Kanagawa System. Microcirculation.Bethesda, MD: Am. Physiol. Soc., 1984,
O, Miner JH, Shaw AS. Congenital nephrotic syndrome in mice sect. 2, vol. IV, pt. 2, chapt. 11, p. 467–520.
lacking CD2-associated protein. Science 286: 312–315, 1999. 308. Taylor GM, Neuhaus TJ, Shah V, Dillon S, Barratt TM. Charge
288. Simionescu M, Simionescu N. Ultrastructure of the microvascu- and size selectivity of proteinuria in children with idiopathic ne-
lar wall: functional correlations. In: Handbook of Physiology. The phrotic syndrome. Pediatr Nephrol 11: 404 – 410, 1997.
Cardiovascular System. Microcirculation.Bethesda, MD: Am. 309. Ten Dam MA, ten Kate RW, Werter CJ, van Kamp GJ, Meu-
Physiol. Soc., 1984, sect. 2, vol. IV, pt. 2, chapt. 3, p. 41–101. wissen SG, Donker AJ. Measurement of the glomerular sieving
289. Sims DE, Westfall JA, Kiorpes AL, Horne MM. Preservation of coefficient of endogenous proteins in man. Contr Nephrol 83: 47–
tracheal mucus by nonaqueous fixative. Biotech Histochem 66: 52, 1990.
173–180, 1991. 310. Tencer J, Frick IM, Öquist BW, Alm P, Rippe B. Size-selectivity
290. Sjöberg EM, Blom A, Larsson BS, Alston-Smith J, Sjöquist M, of the glomerular barrier to high molecular weight proteins: upper
Fries E. Plasma clearance of rat bikunin: evidence for receptor- size limitations of shunt pathways. Kidney Int 53: 709 –715, 1998.
mediated uptake. Biochemical J 308: 881– 887, 1995. 311. Tencer J, Torffvit O, Thysell H, Rippe B, Grubb A. Urine
291. Smith FG III, Deen WM. Electrostatic effects on the partitioning IgG2/IgG4-ratio indicates the significance of the charge selective
of spherical colloids between dilute bulk solution and cylindrical properties of the glomerular capillary wall for the macromolecular
pores. J Colloid Interface Sci 91: 571–590, 1983. transport in glomerular diseases. Nephrol Dial Transplant 14:
292. Smithies O. Why the kidney glomerulus does not clog: a gel 1425–1429, 1999.
permeation/diffusion hypothesis of renal function. Proc Natl Acad 312. Tenstad O, Roald AB, Grubb A, Aukland K. Renal handling of
Sci USA 100: 4108 – 4113, 2003. radiolabelled human cystatin C in the rat. Scand J Clin Lab Invest
293. Spiegler KS, Kedem O. Thermodynamics of hyperfiltration 56: 409 – 414, 1996.
(reverse osmosis): criteria for efficient membranes. Desalination 313. Tenstad O, Williamson HE, Clausen G, Oien AH, Aukland K.
1: 311–326, 1966. Glomerular filtration and tubular absorption of the basic polypep-
294. Stefanidis I, Heintz B, Stocker G, Mrowka C, Sieberth HG, tide aprotinin. Acta Physiol Scand 152: 33–50, 1994.
Haubeck HD. Association between heparan sulfate proteoglycan 314. Thelle K, Christensen EI, Vorum H, Orskov H, Birn H. Char-
excretion and proteinuria after renal transplantation. J Am Soc acterization of proteinuria and tubular protein uptake in a new
Nephrol 7: 2670 –2676, 1996. model of oral L-lysine administration in rats. Kidney Int 69: 1333–
295. Strong KJ, Osicka TM, Comper WD. Urinary-peptide excretion 1340, 2006.
by patients with and volunteers without diabetes. J Lab Clin Med 315. Tisher CC, Brenner BM. Structure and Function of the Glomer-
145: 239 –246, 2005. ulus. Philadelphia, PA: Lippincott, 1994.
296. Sugimoto H, Hamano Y, Charytan D, Cosgrove D, Kieran M, 316. Tisher CC, Madsen KM. Anatomy of the Kidney. In: The Kidney,
Sudhakar A, Kalluri R. Neutralization of circulating vascular edited by Brenner B, Rector F. Philadelphia, PA: Saunders, 1991,
endothelial growth factor (VEGF) by anti-VEGF antibodies and p. 3–75.
soluble VEGF receptor 1 (sFlt-1) induces proteinuria. J Biol Chem 317. Tojo A, Endou H. Intrarenal handling of proteins in rats using
278: 12605–12608, 2003. fractional micropuncture technique. Am J Physiol Renal Fluid
297. Sumpio BE, Chaudry IH, Baue AE. Reduction of the drug-in- Electrolyte Physiol 263: F601–F606, 1992.
duced nephrotoxicity by ATP-MgCl2. 1. Effects on the cis-diam- 318. Trachtman H, Greenbaum LA, McCarthy ET, Sharma M, Gau-
minedichloroplatinum-treated isolated perfused kidneys. J Surg thier BG, Frank R, Warady B, Savin VJ. Glomerular permeabil-
Res 38: 429 – 437, 1985. ity activity: prevalence and prognostic value in pediatric patients

with idiopathic nephrotic syndrome. Am J Kidney Dis 44: 604 – 610, with urinary albumin excretion in men and women. J Am Soc
2004. Nephrol 14: 1330 –1335, 2003.
319. Tryggvason K. Unraveling the mechanisms of glomerular ultrafil- 337. Verroust PJ. Pathophysiology of cubilin: of rats, dogs and men.
tration: nephrin, a key component of the slit diaphragm. J Am Soc Nephrol Dial Transplant 17 Suppl 9: 55–56, 2002.
Nephrol 10: 2440 –2445, 1999. 338. Vink H, Duling BR. Capillary endothelial surface layer selectively
320. Tryggvason K, Patrakka J. Thin basement membrane nephrop- reduces plasma solute distribution volume. Am J Physiol Heart
athy. J Am Soc Nephrol 17: 813– 822, 2006. Circ Physiol 278: H285–H289, 2000.
321. Tryggvason K, Patrakka J, Wartiovaara J. Hereditary protein- 339. Vink H, Duling BR. Identification of distinct luminal domains for
uria syndromes and mechanisms of proteinuria. N Engl J Med 354: macromolecules, erythrocytes, leukocytes within mammalian cap-
1387–1401, 2006. illaries. Circ Res 79: 581–589, 1996.
322. Tryggvason K, Pikkarainen T, Patrakka J. Nck links nephrin to 340. Vyas SV, Comper WD. Dextran sulfate binding to isolated rat
actin in kidney podocytes. Cell 125: 221–224, 2006. glomeruli and glomerular basement membrane. Biochim Biophys
323. Tryggvason K, Wartiovaara J. How does the kidney filter Acta 1201: 367–372, 1994.
plasma? Physiology 20: 96 –101, 2005. 341. Vyas SV, Parker JA, Comper WD. Uptake of dextran sulphate by
324. Uchida K, Uchida S, Nitta K, Yumura W, Marumo F, Nihei H. glomerular intracellular vesicles during kidney ultrafiltration. Kid-
Glomerular endothelial cells in culture express and secrete vascu- ney Int 47: 945–950, 1995.
lar endothelial growth factor. Am J Physiol Renal Fluid Electrolyte 342. Wartiovaara J, Ofverstedt LG, Khoshnoodi J, Zhang J,
Physiol 266: F81–F88, 1994. Makela E, Sandin S, Ruotsalainen V, Cheng RH, Jalanko H,
325. Waga I, Yamamoto J, Sasai H, Munger WE, Hogan SL, Preston Skoglund U, Tryggvason K. Nephrin strands contribute to a
GA, Sun HW, Jennette JC, Falk RJ, Alcorta DA. Altered mRNA porous slit diaphragm scaffold as revealed by electron tomography.
expression in renal biopsy tissue from patients with IgA nephrop- J Clin Invest 114: 1475–1483, 2004.
athy. Kidney Int 64: 1253–1264, 2003. 343. Weiss C, Passow H, Rothstein A. Autoregulation of flow in
326. Walz G. Slit or pore? A mutation of the ion channel TRPC6 causes isolated rat kidney in absence of red cells. Am J Physiol 196:
FSGS. Nephrol Dial Transplant 20: 1777–1779, 2005. 1115–1118, 1959.
327. Van Damme M, Prevost M. Transport of charged macromolecules 344. White JA, Deen WM. Equilibrium partitioning of flexible macro-
across a biological charged membrane. Comput Prog Biomed 19: molecules in fibrous membranes and gels. Macromolecules 33:
107–117, 1985. 8504 – 8511, 2000.
328. Van den Berg B, Vink H. Glycocalyx perturbation: cause or con- 345. Wolf G, Stahl RA. CD2-associated protein and glomerular disease.
sequence of damage to the vasculature? Am J Physiol Heart Circ Lancet 362: 1746 –1748, 2003.
Physiol 290: H2174 –H2175, 2006. 346. Wolgast M, Källskog Ö, Wahlström H. Characteristics of the
329. Van den Berg BM, Vink H, Spaan JA. The endothelial glycocalyx glomerular capillary membrane of the rat kidney as a hydrated gel.
protects against myocardial edema. Circ Res 92: 592–594, 2003. I. Hypothetical structure. Acta Physiol Scand 158: 213–224, 1996.
330. Van den Berg JG, van den Bergh Weerman MA, Assmann KJ, 347. Wolgast M, Källskog Ö, Wahlström H. Characteristics of the
Weening JJ, Florquin S. Podocyte foot process effacement is not glomerular capillary membrane of the rat kidney as a hydrated gel.
correlated with the level of proteinuria in human glomerulopathies. II. On the validity of the model. Acta Physiol Scand 158: 225–232,
Kidney Int 66: 1901–1906, 2004. 1996.
331. Van Setten PA, van Hinsbergh VW, van der Velden TJ, van de 348. Wolgast M, Öjteg G. Electrophysiology of renal capillary mem-
Kar NC, Vermeer M, Mahan JD, Assmann KJ, van den Heuvel branes: gel concept applied and Starling model challenged. Am J
LP, Monnens LA. Effects of TNF alpha on verocytotoxin cytotox- Physiol Renal Fluid Electrolyte Physiol 254: F364 –F373, 1988.
icity in purified human glomerular microvascular endothelial cells. 349. Zeman LJ, Zydney AL. Microfiltration and Ultrafiltration. New
Kidney Int 51: 1245–1256, 1997. York: Dekker, 1996, p. 350 –352.
332. VanTeeffelen JW, Dekker S, Fokkema DS, Siebes M, Vink H, 350. Zenker M, Aigner T, Wendler O, Tralau T, Muntefering H,
Spaan JA. Hyaluronidase treatment of coronary glycocalyx in- Fenski R, Pitz S, Schumacher V, Royer-Pokora B, Wuhl E,
creases reactive hyperemia but not adenosine hyperemia in dog Cochat P, Bouvier R, Kraus C, Mark K, Madlon H, Dotsch J,
hearts. Am J Physiol Heart Circ Physiol 289: H2508 –H2513, 2005. Rascher W, Maruniak-Chudek I, Lennert T, Neumann LM,
333. Vassiliou P, Tay M, Comper WD. Partial ischemia and protein- Reis A. Human laminin beta2 deficiency causes congenital nephro-
uria during isolated kidney perfusion is accompanied by the release sis with mesangial sclerosis and distinct eye abnormalities. Hum
of vascular [35S]heparan sulfate. Biochem Int 19: 1241–1251, 1989. Mol Genet 13: 2625–2632, 2004.
334. Vehaskari VM, Chang CT, Stevens JK, Robson AM. The effects 351. Zhang X, Curry FR, Weinbaum S. Mechanism of osmotic flow in
of polycations on vascular permeability in the rat. A proposed role a periodic fiber array. Am J Physiol Heart Circ Physiol 290:
for charge sites. J Clin Invest 73: 1053–1061, 1984. H844 –H852, 2006.
335. Venturoli D, Rippe B. Ficoll and dextran vs. globular proteins as 352. Zoja C, Wang JM, Bettoni S, Sironi M, Renzi D, Chiaffarino F,
probes for testing glomerular permselectivity: effects of molecular Abboud HE, Van Damme J, Mantovani A, Remuzzi G. Interleu-
size, shape, charge, deformability. Am J Physiol Renal Physiol 288: kin-1 beta and tumor necrosis factor-alpha induce gene expression
F605–F613, 2005. and production of leukocyte chemotactic factors, colony-stimulat-
336. Verhave JC, Hillege HL, Burgerhof JG, Navis G, de Zeeuw D, ing factors, and interleukin-6 in human mesangial cells. Am J
de Jong PE. Cardiovascular risk factors are differently associated Pathol 138: 991–1003, 1991.

Properties of The Glomerular Barrier and Mechanisms

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Properties of The Glomerular Barrier and Mechanisms

Uploaded by

Copyright:

Available Formats

Physiol Rev 88: 451– 487, 2008;

Properties of the Glomerular Barrier and Mechanisms

BÖRJE HARALDSSON, JENNY NYSTRÖM, AND WILLIAM M. DEEN

Department of Molecular and Clinical Medicine/Nephrology, Institute of Medicine, Göteborg University,

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

The intricate properties of the glomerular barrier

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

allow high permeability to water and small solutes in the

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

FIG. 4. Electron microcraphs show-

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

FIG. 6. Human podocytes in culture.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

FIG. 7. Human glomerular endothelial

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

FIG. 8. Glomerular cells produce

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Species Albumin Sieving Coefficient Technique Reference Nos.

Rat 0.00033 In vivo ⫹ lysine 310

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

TABLE 2. Sieving coefficients for various proteins in selected studies

Type of Protein SE Radius Sieving Coefficient Charge References

Neutral or slightly cationic proteins

C. Dextran Polymers D. Ficoll Polymers

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

FIG. 9. Glomerular size selectivity is illus-

FIG. 10. Glomerular sieving coeffi-

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

B. Sieving in Homogeneous Membranes

The experimental measure of membrane selectivity

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

C共0兲 C共L) Of course, the diffusional hindrance factor (H),filtrate

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org

Downloaded from journals.physiology.org/journal/physrev (182.001.175.191) on June 13, 2021.

Physiol Rev • VOL 88 • APRIL 2008 • www.prv.org