Clinical Science (1997) 93,65-72 (Printed in Great Britain)
65
Protein degradation during renal passage in normal kidneys is
inhibited in experimental albuminuria
Tanya M. OSlCKA and Wayne D.COMPER*
Department of Biochemistry and Molecular Biology, Monash University, Clayton, Victoria, Australia 3 I68
(Received 31 October 1996/26 February 1997; accepted 26 February 1997)
1. lkitium labelled proteins, namely bovine serum
albumin ([3H]BSA), rat serum albumin ([3H]RSA),
anionic horseradish peroxidase ( [3H]aHRP) and
immunoglobulin present in urine fractions from rat
filtration studies in vivo and isolated perfused rat
kidneys (IPKs) have been shown by gel chromatographic analysis to be severely degraded to small
peptides. The degradation of RSA and BSA in vivo
has been shown to be similar.
2. Degradation of proteins in the urine from IPK
experiments was inhibited by including 150 mmol/l
lysine in the perfusate. Similarly, [3H]BSA and
[3H]aHRP excreted from rats with puromycin
aminonucleoside nephrosis was again essentially
intact for both IPK and in vivo experiments.
3. It appears that the degradation of proteins
observed in urine obtained from control kidneys is
due, in part, to proteolytic activity associated with
the proximal tubule. Inhibition of proximal tubule
function, which occurs for both lysine and puromycin aminonucleoside treatments (as calibrated by
lysozyme uptake), results in inhibition of the degradation observed. Glomerular epithelial cells could
also contribute to the degradation.
4. There was no generation of low-molecular-weight
material in the perfusate or plasma arising from
breakdown of circulating proteins or recycling of
potential degradation products from the tubules.
INTRODUCTION
The glomerular capillary wall (GCW) severely
restricts the passage of albumin and other plasma
proteins during filtration. However, in normal kidneys, some protein transport across the capillary
wall occurs which is then subject to endocytosis by
tubular cells. Therefore, the appearance of protein
in urine is the result of a dominant filtration rejection at the capillary wall and post-glomerular scavenging by tubules.
The glomerular barrier offers little hindrance to
the filtration of low-molecular-weight proteins such
as lysozyme, growth hormone and insulin (all
<20 000 molecular weight). There is extensive mor-
phological and biochemical data showing that lowmolecular-weight proteins are segregated into
endocytic vesicles of tubular cells which consequently fuse with lysosomes, providing strong qualitative evidence that the post-glomerular processing
of these proteins involves intracellular catabolism
[l-31 in normal kidneys. The integrity of highmolecular-weight proteins ( >40 000) which are
excreted in the urine has not been widely investigated, although recent studies have demonstrated
fragmentation during renal passage in rats of
anionic horseradish peroxidase (aHRP) [4], albumin
[5] and bikunin [6], as well as the fragmentation of
albumin in diabetic patients [7].
In this study we examine the structural integrity of
filtered tritium-labelled high-molecular-weight proteins, namely bovine serum albumin ([3H]BSA), rat
serum albumin ([3H]RSA), [3H]aHRP and immunoglobulin ([3H]IgG) excreted in both the isolated perfused kidney and in vivo. This is achieved by
analysing urine which is fractionated by size-exclusion chromatography. These experiments are also
performed in rats where tubular uptake has been
inhibited through the use of either lysine or puromycin aminonucleoside (PA).
METHODS
Materials
Male Sprague-Dawley rats (300-350 g) were
obtained from the Monash University Central
Animal House. PA (6-dimethylamino-9[3’ amino-3‘
deoxyribosyl purine), type VI peroxidase (EC
1.11.1.7) (288 units/mg of solid from horseradish),
catalase (EC 1.11.1.6) (1600 units/mg of solid from
bovine liver), lysozyme (EC 3.2.1.17) (30000 units/
mg of solid from chicken egg white), benzoylated
dialysis tubing (molecular weight cut-off of 2000)
and the amino acids lysine, tyrosine, serine, cysteine,
aspartate, glutamate, asparagine and glutamine were
from Sigma Chemical Co. (St. Louis, MO, U.S.A.).
BSA (fraction V), immunoglobulin (bovine), Micrococcus Zuteus (ATCC 4698) and superoxide dismutase (EC 1.15.1.1) (5000 units/mg of solid from
Key words: albumin, aminonucleoside nephrosis, degradation, horseradish peroxidase, immunoglobulin, isolated perfused kidney lysine, puromycin.
Abbreviations: aHRP, anionic horseradish peroxidase; GCW, glomerular capillary wall: GFR, glomerular filtration rate; GSC, glomerular sieving coefficient; IPK, isolated perfused
kidney; PA, puromycin aminonucleoside; PAN, puromycin aminonucleoside nephrosis; RSA, rat serum albumin.
Correspondence: Dr Wayne D. Comper.
66
T.
M. Osicka and W. D. Comper
bovine erythrocytes) were purchased from
Boehringer Mannheim GmbH Biochemica (Mannheim, Germany). Sephadex G-100, Sephacryl S-200,
Sephadex G-25 in PD-10 columns and Blue Dextran
T2000 were purchased from Pharmacia Fine Chemicals (Uppsala, Sweden). Nembutal (60 mg/ml) was
from Cera Chemicals Australia Pty. Ltd (Hornsby,
New South Wales, Australia). Synthamin (a source
of amino acids) was from Travenol Laboratories
(New South Wales, Australia). Sodium heparin was
from Commonwealth Serum Laboratories (Melbourne, Australia). Mannitol was from CSR Chemicals Ltd (Rhodes, New South Wales, Australia).
Tritiated water (0.25 mCi/g) and [~arboxyl-'~C]inulin (2.7 mCi/g) were obtained from Du Pont (Wilmington, Detroit, MI, U.S.A.) and sodium
b~ro-[~H]hydride
(132 mCi/mg) was from Amersham International (Buckinghamshire, U.K.).
Solutions and buffers
PBS, pH 7.4, contained (mmol/l) 136.9 NaC1, 2.68
KCl, 8.1 Na2HP04 and 1.5 KH2P04. Krebs-Henseleit buffer, pH 7.4, contained (mmol/l) 122 NaCl,
4.6 KC1, 0.115 KH2Po4, 0.115 MgS04, 24.9
NaHC03 and 0.1 CaC12.H20. Kidney perfusate
solution was 5% BSA in JSrebs-Henseleit buffer
that contained 5 mmol/l glucose and 36 mg of mannitol, 10600 units of superoxide dismutase and
110000 units of catalase (all per 200 ml of perfusate) and amino acids (mmol/l) tyrosine 0.2, serine
1.0, cysteine 0.5, aspartate 0.2, glutamate 0.5,
asparagine 0.2, glutamine 2.0, leucine 0.4, phenylalanine 0.32, methionine 0.33, lysine 1.0, isoleucine
0.3, valine 0.33, histidine 0.24, threonine 0.24, tryptophan 0.07, alanine 2.0, glycine 2.3, arginine 0.5
and proline 0.31.
Inductionof puromycin aminonucleoside nephrosis
(PAN)
Male Sprague-Dawley rats were injected intravenously with 15 mg/100 g body weight PA given as
a 3.5% solution in PBS in the tail vein. Age- and
weight-matched control rats were injected with the
same dosage of PBS for protein urinary excretion
measurements. Urine was collected over 24 h in a
metabolic cage and total protein excretion was
determined by Biuret assay (using BSA as standard)
[8] at days 0, 5 and 9 after PA and PBS administration. Animals were given free access to food and
water.
Tritium labelling
BSA, rat serum albumin (RSA) and IgG were
labelled with tritium by the reductive methylation
procedure of Tack et al. [9]. HRP was also tritium
labelled and then succinylated to prepare an anionic
derivative (aHRP) as previously described [4, 101.
The tritium-labelling reaction involves a brief exposure to formaldehyde and sodium b~ro-[~H]hydride.
Tritium labelling is specific for the a-amino groups
of the amino terminal residues and the &-amino
groups of lysyl residues. The purity of the tritiumlabelled preparations is ensured in that the samples
(-lo9 d.p.m./ml; 3.5 ml) are extensively dialysed in
dialysis tubing with a molecular-weight cut-off of
2000 until there is no tritium in the dialysate
(volume = 4 litres). Assuming that there is less than
100 d.p.m./ml in the dialysate then this corresponds
to less than 0.0025% of the original sample that
could be contaminated with free tritium or lowmolecular-weight labelled material. The dialysed
preparation was applied to a PD-10 column
immediately before use. The specific activity of
[3H]BSAwas 2.31 x lo8 d.p.m./mg, 3.02 x lo8 d.p.m./
mg for [3H]RSA, 7 . 5 4 ~ 1 0 ~d.p.m./mg for
[3H]aHRP and 1.99 x lo8 d.p.m./mg for PHIIgG.
Kidney perfusion
Male Sprague-Dawley rats were anaesthetized by
a 1 ml intraperitoneal injection of Nembutal (18
mg/ml). A 1 ml volume containing 10% mannitol
and 200 units of sodium heparin was injected into
the femoral vein. A laparotomy was performed and
the right ureter was cannulated with polyethylene
tubing (PE-10, Dural Plastics and Engineering,
Auburn, New South Wales, Australia). The right
renal artery was cannulated via the superior mesenteric artery and the kidney was removed by en bloc
dissection. This whole procedure took no longer
than 10 min. The perfusion pressure was maintained
at 90-100 mmHg with a peristaltic pump monitored
by a calibrated aneroid manometer, whilst the flow
rate was monitored by a ball flowmeter. Kidneys
were perfused with 160 ml of recirculated filtered
5% BSA in Krebs-Henseleit buffer containing
glucose, essential amino acids [ll-131 and oxygenradical scavengers to prevent partial ischaemia [ 131.
The system was maintained at 37°C and the perfusate was continually gassed with 95% 02/5% C02.
The kidney was allowed to equilibrate for 10 min
and urine and perfusate samples were collected
after 40 and 60 min of perfusion. The perfusate contained either
1x lo6 d.p.m./ml [3H]BSA, or
[3H]aHRP or [3H]IgG. Perfusions were also performed using [3H]BSA on rats 5 days after the
administration of PA. Measurement of glomerular
filtration rate (GFR) was made with [carboxyl14C]inulin. Perfusions were also performed in the
presence of 150 mmol/l lysine, a lysosomotropic
agent [14, 151.
Urine samples obtained between 40 and 60 min of
perfusion and perfusate samples obtained 60 min
after perfusion were analysed by size-exclusion gel
chromatography. Sephadex G-100 was used for
experiments involving [3H]BSA and [3H]aHRP,
whereas Sephacryl S-200 was used for experiments
involving [3H]IgG. Samples (1 ml) were loaded onto
-
Renal degradation of proteins
67
Table 1. Renal parameters in control, lysine-treated and PA-treated (day 5) IPKs. Variation of the urine flow rate (UFR), GFR,
and protein excretion rate as measured by the Biuret assay as a function of perfusion time (min).
~~
UFR (ml/min)
Time (min)
Contiol (n= 6)
Lysine (n = 4)
PAN (n = 6)
0-40
0.12210.071
0.151 50.041
0.0058+0.0046
40-60
0.14850.065
0.113+0.017
0.0159+0.0122
each column and 95 fractions of 1.65 ml were eluted
with PBS (pH 7.4) at 20 ml/h at 4°C. For routine
analysis, approximately 120000 d.p.m. of perfusate
was applied to the column, whereas for urine
samples 30 000-100 000 d.p.m. samples were applied.
The void volume (VO)was determined with Blue
Dextran T2000 and the total volume (Vt) with tritiated water.
Processing of proteins in vivo
Experiments in vivo were performed by injecting
-namely
1 lo8 d.p.m./ml of each tritium-labelled protein,
[3H]BSA, [3H]RSA, [3H]aHRP and [3H]IgG,
x
into the tail vein of Sprague-Dawley rats which
were maintained in a metabolic cage for 2 h with
free access to food and water. The urine collected
after this time was again analysed by column
chromatography as described above.
GFR (ml/min)
0-40
0.86750.224
0.167k0.041
0.085k0.103
Protein excretion rate @g/min)
40-60
0.828k0.166
0.168+0.023
0.121 k0.082
0-40
282574
428+92
221 5 9 7
40-60
2485 I37
554+144
273 rt I02
Calculations
All quantitative data are expressed as means fSD
where n represents the number of determinations.
RESULTS
The isolated perfused kidney preparation when
studied over a 1h period has been shown by this
laboratory to be a stable, steady-state preparation
(Table 1) [5]. Kidneys from rats were perfused for
60 min with various tritium-labelled proteins. The
gel chromatographic profile on Sephadex G-100 of
rH]BSA from a 60 min perfusate sample (Fig. la)
appears to be essentially the same as the original
[3H]BSA sample (results not shown) as it elutes as a
single major peak with a Kay of 0.190. The small
15000
Lysozyrne assay
This assay is based on the lysis of M. luteus cells
using the method described by Harrison and Barnes
[16] with adaptations from Litwack [17]. Sodium
phosphate buffer, 0.067 mol/l (pH 6.20), was used
for the substrate, standards and dilutions. The substrate was a freshly prepared suspension consisting
of dried M. luteus cells (25 mg/100 ml). The stock
standard lysozyme solution contained 400 pg/ml and
was kept at 4°C. Fresh working standards contained
2, 5, 10, 15 and 20 pg/ml lysozyme and were prepared from the stock solution. The assay was performed in an Ultrospec I11 spectrophotometer
(Pharmacia LKB Biotechnology, Cambridge, U.K.)
at room temperature. To 0.9 ml of bacterial suspension was added 0.1 ml of sample or standard. Urine
and perfusate samples were routinely diluted
between 1/10 and 1/20. The rate of change in transmission was recorded at 450 nm upon the addition
of sample to the bacterial suspension. The change in
transmission between 30 and 60 s was linear and was
plotted against lysozyme concentration.
Counting of radioactivity
Tritium and 14C radioactivity was determined in
1ml aqueous samples with 3 ml of scintillant [18]
and recorded on a Wallac 1410 liquid scintillation
counter (Wallac Oy, Turku, Finland).
w
10000
c
3
z
I
5000
K
w
n
g
v)
z
I-
0
8000
6000
z
v,
0
4000
2000
fl
0 10 20 30 40 50 60 70 80 90
FRACTION NO.
Fig. I.Representative profiles of the size-exclusion chromatography
on Sephadex G-100 of [3H]BSA in (a) the perfusate (135000 d.p.m.)
sampled after 60 min perfusion of Sprague-Dawley rat kidneys
(n = 6), and (b) urine (75000 d.p.m.) collected between 40 and 60
min of perfusion (n = 6). The open squares show the elution profile for
['Hllysine. Vo is the void volume of the column as determined with Blue
Dextran T2000, and Vr is the total column volume as determined with
['HIHzO.
T. M. Osicka and W. D. Comper
6a
leading peak is the albumin dimer but there is no
low-molecular-weight material in the perfusate. The
elution profile of the urine obtained from isolated
rat kidneys perfused with [3H]BSA, shown in Fig.
lb, reveals that most of the [3H]BSA in the urine is
degraded to small peptides (88.9 f7.1%). This lowmolecular-weight material was not free tritium,
which would elute at the Vt of the column, nor was
it free labelled amino acid as it too elutes at the Vt
(Fig. lb). This demonstrates that the labelled
material is macromolecular with a molecular weight
>2500.
Further experiments were performed to eliminate
any possibility of contamination by any low-molecular-weight peptides in the [3H]BSA, which would be
preferentially filtered during the perfusion giving
rise to the urine profile observed. These peptides
may have been present in the original BSA preparation or be generated during the course of the perfusion. The original [3H]BSA stock solution ( 3 x lo9
d.p.m.) was fractionated on Sephadex G-100 and the
peak fraction only (fraction 41 from a similar profile
to that shown in Fig. l a ) was then used in a subsequent perfusion. Again, as shown in Fig. 2b, degradation was still observed. No low-molecular-weight
material was observed in the perfusate after the 60
min perfusion (Fig. 2a); any undetected contamination would be less than 0.03% of the perfusate
sample. This confirms that the low-molecular-weight
material shown in Figs l b and 2b is in fact material
derived de novo from the degradation of albumin
during its post-glomerular basement membrane
transport.
The presence of relatively low-molecular-weight
material in the urine profile is not the result of
degradation of albumin due to proteases in the
urine, as incubation for 2 h of [3H]BSA with urine
from an isolated perfused kidney (IPK) at 37°C
resulted in a profile similar to that of the original
material [5].
Degradation too has been observed with the
excretion of [3H]aHRP (61 f3% degradation; Fig.
3b) [4]. It was apparent that for [3H]IgG the perfusate sample contained a number of minor fractions,
yet there appeared to be significant low-molecularweight material excreted in the urine (Fig. 4b).
To eliminate the possibility that the IPK system
contributed to the degradation observed for the protein molecules, studies were performed in vivo. The
renal parameters for the studies in vivo are shown in
Table 2. In rats injected intravenously with [3H]BSA
there was the appearance of 90% low-molecularweight material in urine collected (Fig. 5a). A similar pattern was also seen in urine when [3H]RSA
was injected intravenously into rats, shown in Fig.
5b. Using RSA, as opposed to BSA, eliminates the
possibility that BSA is processed differently in rats.
Degradation was also observed in vivo for both
[3H]aHRP (Fig. 5c) and [3H]IgG (Fig. 5d).
Ultrapure [3H]BSA prepared by gel chromatography as described above was also injected intra-
-
venously into rats, yielding similar results (Fig. 2c).
Again, the plasma from these rats (Fig. 2d) did not
show the presence of any low-molecular-weight
material, and in this case any undetected contamination would be less than 0.003% of the plasma
sample. These results therefore confirm that the
appearance of tritium-labelled low-molecular-weight
material seen in the IPK is also evident for proteins
filtered in vivo. It also confirms that there was no
detectable generation of low-molecular-weight
material in the plasma that may arise through breakdown of circulating albumin or through the recycling
of filtered albumin through the tubules.
In order to examine the influence of tubular
uptake of proteins we used 150 mmol/l lysine (which
results in 100% inhibition of lysozyme reabsorption
[5]) as a tubular inhibitor. The inclusion of lysine in
50000
30000L
40000
20000
I
I
11
I
,
10000
0
3e+5 2e+5 -
4e+5
d
I
8
I,
41
I
FRACTION NO.
Fig. 2. Size-exclusion chromatography on Sephadex G-I00 of (a) the
perfusate (300000 d.p.m.) sampled after 60 min perfusion of Sprague-Dawley kidneys, (b) urine (I7000 d.p.m.) collected between 40
and 60 min of perfusion, (c) urine (20000 d.p.m.) collected from rats
2 h after a bolus intravenous injection of pH]BSA, and (d) plasma
(2.8 x lo6 d.p.m.) from the rat 2 h after a bolus intravenous injection
of fH]BSA. Note that specially purified PH]BSA was used in these experiments to eliminate any possibility of contamination in the [3H]BSA stock
solution.
69
Renal degradation of proteins
the perfusion medium clearly has an influence on
the urine profile of all proteins studied, as shown in
Fig. 6. In the urine profile of t3H]BSA, shown in
Fig. 6a, the major portion of radioactivity elutes
near the monomer peak and there is far less relatively low-molecular-weight material (37.6 & 6.8%)
present as compared with normal urine from the
perfused kidney. Again, this was evident for
t3H]aHRP (Fig. 6b) and [3H]IgG (Fig. 6c) in the
presence of lysine.
A feature of the results for albumin is that while
lysine treatment results in a marked change in the
ratio of intact albumin to low-molecular-weight
material in the urine there is no such marked difference in the protein excretion rates of lysine-treated
compared with the control in IPKs (Table 1). The
comparison is valid, as we have previously demon-
strated that protein excretion measured by the
Biuret assay is the same as that measured by radioactivity [5]. These results would then demonstrate
that lysine has a genuine inhibitory effect on the
process of degradation of albumin during its renal
passage.
PA has been widely used to produce marked albuminuria in experimental animals. The intravenous
administration of PA to rats in a single dose of 15
mg/100 g body weight resulted in a significant timedependent increase in proteinuria, as shown in
Table 2. These results show a similar trend to those
found by others [19-211. Similar to lysine treatment,
PA administration resulted in 100% inhibition of
protein reabsorption as determined through the use
of lysozyme.
0 10 20 30 40 50 60 70 80 90
0 10 20 30 40 50 60 70 80 90
FRACTION NO.
FRACTION NO.
Fig. 4. Representative profiles of the size-exclusion chromatography
on Sephacryl S-200 of [IHIlgG in (a) the perfusate (I20000 d.p.m.)
sampled after a 60 min perfusion, and (b) urine (II0000 d.p.m.) collected between 40 and 60 min of perfusion (n = 6). Vo is the void
volume of the column as determined with Blue DextranT2000, and V, is the
total column volume as determined with rH]H20.
Fig. 3. Representative profiles of the size-exclusion chromatography
on Sephadex G-100 of [)H]aHRP in (a) the perfusate (I25000
d.p.m.) sampled after 60 min perfusion of Sprague-Dawley rat kidneys (n = 4), and (b) urine (28000 d.p.m.) collected between 40 and
60 min of perfusion (n = 4)
Table 2. Renal parameters in control and PA-treated rats in vivo. GFR was measured using the creatinine assay. Protein excretion
may vary depending on the assay. Protein excretion using the sulphosalicyclic acid turbidity assay gave 13.5 k 3. I pg/min for the control
and 128 f 10.6 & n i n for day 5 PAN rats and I52 k 15.8 pgglmin for day 9 PAN rats. UFR urine flow rate: n.d., not determined.
Control (n = 4)
PAN (n = 3) day 5
PAN (n = 3) day 9
UFR (mlhin)
GFR (mlglmin)
Protein excretion rate (,ug/min)
0.0059 k0.0017
0.0066+0.0026
0.02 I 3 5 0.008I
1.44k0.39
n.d.
104+39
769k76
IOlOJr 123
n.d.
70
T. M. Osicka and W. D. Comper
Kidneys perfused 9 days after PA administration
failed to function successfully, therefore kidneys
were subsequently perfused 5 days after PA administration. Steady state was achieved within the 1h
perfusion period. The renal parameters associated
with the perfusion of kidneys isolated from PAN
rats are shown in Table 1. These results are similar
to studies of PAN rats in vivo [19].
A striking feature of the results presented here is
that the degradation for both [3H]BSA(Fig. 7b) and
[3H]aHRP (Fig. 7c) is totally absent in urine collected from day 9 PAN rats, whereas degradation is
severely reduced (24.11 k5.2%; Figure 7a) in urine
collected from perfused PAN kidneys 5 days after
PA administration. For albumin, in particular, the
equivalent protein excretion rate of the PAN rats
compared with the control in the IPK studies (Table
1) would demonstrate that PA treatment, like lysine,
has a bona fide inhibitory effect on the degradation
of albumin.
DISCUSSION
Tritium labelling of proteins by reductive methylation [9] is specific for the a-amino groups of the
amino terminal residues and the &-aminogroups of
lysyl residues. Therefore, it is likely that most of the
peptides resulting from endopeptidase action, as
seen in the chromatography profiles of urine, would
be detected through the presence of their tritium
label. This has been confirmed by the good agreement between the Biuret assay and radioactivity
measurements.
The gel chromatographic profiles reveal that proteins present in normal urine from the IPK are
degraded to peptides, as compared with the original
material. Degradation was not a result of the IPK
system, as similar degradation patterns were also
observed in vivo, nor was it the result of any con-
I
a
3000
5000
2000
4000
3000
2000
1000
I000
0
10000
E3
z
8000
6000
3
0
'B
8000
$
6000
F9
4000
n
4000
xg
-
2000
C
6000
4000
2000
0
0
0 102030405060708090
FRACTION NO.
Fig. 5. Representative profiles of the sizeexclusion chromatography
of urine collected from rats 2 h after a bolus intravenous (I x lo8
d.p.m.) injection of (a) ['HIBSA (45000 d.p.m.) (n = 4), (b) ['HIRSA
(70000 d.p.m.) (n = 4), (c) fH]aHRP (90000 d.p.m.) (n = 3), and (d)
['HIlgG (75000 d.p.m.) (n = 4). Sephadex G-100 was used for separation
of all molecules except for rH]lgG, where Sephactyl S-200 was used.
0 10 20 30 40 50 60 70 80 90
FRACTION NO.
Fig. 6. Representative profiles of the size-exclusion chromatography
of urine collected from kidneys perfused in the presence of 150
.mmol/l lysine and (a) TH]BSA (40000 d.p.m.) (n = 4), (b) [)H]aHRP
(I25000 d.p.m.) (n =4), and (c) ['HIlgG (I20000 d.p.m.) (n =4).
Sephadex G-100 was used for separation of all molecules except for
rH]lgG, where Sephacryl S-200 was used.
Renal degradation of proteins
tamination of the protein probes used, as similar
degradation patterns were observed when a specially
purified sample of [3H]BSA was used in experiments. However, in the presence of 150 mmol/l
lysine or in PAN rats where tubular reabsorption of
proteins was inhibited, degradation was inhibited
significantly. These studies establish that in normal
filtering kidneys there is a relatively rapid tubular
uptake of protein and release into the tubular lumen
of degraded protein, which is then excreted into the
urine. Glomerular epithelial cells may also contribute to the degradation observed.
The lysine or PAN effect could be viewed alternatively as the onset of large non-selective pores,
allowing the preferential flux of intact albumin
across the capillary wall [22], i.e. glomerular protei-
4000
3000
2000
1000
w
0
5
20000
5
15000
=
10000
za
n
rn
s
5000
5
0
F
z
n
C
4000
3000
200c
IOOC
C
0 10 20 30 40 50 60 70 80 90
FRACTION NO.
Fig. 7. Representative profiles of the size-exclusion chromatography
on Sephadex G-100 of (a) urine collected from day 5 PAN kidneys
perfused with [3H]BSA (62000 d.p.m.) (n = 5), (b) urine collected
from day 9 PAN rats Z h after a bolus intravenous injection of
[3H]BSA (ZOO000 d.p.m.) (n = 3), and (c) urine collected from day 9
PAN rats 2 h after a bolus intravenous injection of [3H]aHRP (45000
d.p.m.) (n = 3)
71
nuria. Recent studies, described below, would reject
this possibility. The concept of large non-selective
pores has arisen from' two features associated with
the transcapillary transport of albumin, i.e. (1) that
albumin appears to be a relatively larger molecule
than its normal hydrodynamic radius of 36 A, on the
basis of its glomerular sieving properties in normal
kidneys [23] (thought to be due to charge repulsion
of the molecule by the GCW), and (2) that the disproportionate increase in albumin transport as compared with other molecules, like dextran in
albuminuric states, can be explained in part by the
increased number of these non-selective pores
(which are not charge selective) [24].
There are two major biophysical factors governing
transport across the capillary wall [25], i.e. (1) the
excluded-volume partitioning at the interface
between the capillary wall and its bathing fluid
(blood or urine), and (2) the drag on the molecule
as it is being transported through the capillary wall.
Charge effects associated with either of these factors
have now been demonstrated to be negligible [4, 5,
23, 26-28]. The excluded-volume partitioning of the
capillary wall has been shown to be very similar for
dextran, which has the same hydrodynamic radius as
albumin [5]. Similarly, the excluded-volume partitioning of dextran as compared with globular Ficoll
with the same radius has been demonstrated to be
almost identical in gel chromatographic material
[29]. Therefore, flexible dextran is equivalent to the
more globular albumin or Ficoll in terms of
excluded-volume partitioning. The drag coefficients
also tend to be similar, as evidenced by the fractional clearance values of albumin compared with
dextran in a tubular inhibited system [5], and by
comparison of the fractional clearance of dextran
and Ficoll [30, 311. These studies demonstrate that
albumin is like any other molecule, such as dextran,
and that in tubular inhibited systems it behaves as if
it is merely size selected on the basis of a molecule
with a radius of approximately 36 A [5].
In albuminuric states it is apparent that while
fractional clearance of dextrans with radii >40 A
may be increased (as a result of the increased flow
through large npn-selective pores) the clearance for
dextrans of 36 A may either not be affected or may
even decrease. In the same manner, albumin clearance would also be expected to follow the same
behaviour. Under these circumstances, and recognizing that charge selectivity is not a factor determining
albumin transport, the putative increased role of
non-selective pores in albuminuric states could not
account for the vast increase in albumin clearance.
In the case of lysine-induced albuminuria, the
increase in the fractional clearance of albumin
(including its degradation products in urine) is
almost 10-fold, whereas there was no significant
change in dextran fractional clearance for molecules
with molecular radii in the range (26-50 A) studied
[5]. Similar arguments would apply to albuminuria
in PAN, where size selectivity as determined by dex-
T. M. Osicka and W. D. Comper
72
tran fractional clearance has been shown to be
unchanged for molecules with radii of 36 A [19].
Recent studies performed by us [5] have demonstrated that albumin is merely size selected at the
GCW and has a glomerular sieving coefficient
(GSC) in the range 0.04-0.07 in normal kidneys.
The relatively large amounts of albumin which are
transported across the GCW are then taken up by a
specific tubular cell pathway that eventually returns
the albumin to the blood supply undegraded. Albuminuria induced by either lysine or PA would be
consistent with the inhibition of this high-capacity
tubular-uptake system. Some caution should prevail
concerning the apparent contradiction provided by
micropuncture analysis of fluids from the proximal
tubule, which gives low concentrations of albumin
that are assumed to be directly related to the GSC.
The assumption does not take into account the highcapacity rapid removal of albumin from the tubular
lumen by the tubular cells. It is only in situations
where tubular uptake is inhibited, as shown in this
study with PA, that micropuncture albumin concentration measurements reflect the true GSC. Under
such conditions, Lewy and Pesce [32] have measured
albumin concentrations of 234 mg/100 ml, corresponding to a GSC of -0.06.
-
ACKNOWLEDGMENTS
This work was supported by grants from the
National Health and Medical Research Council of
Australia and the Australian Research Council.
REFERENCES
I. Maack T. Renal handling of low molecular weight protein. Am J Med 1975; 58:
57-64.
2. Maack T, JohnsonV, Kau ST, Figueiredo J, Sigulem D. Renal filtration, transport,
and metabolism of low-molecular weight proteins: a review. Kidney Int 1979; 1 6
25 1-70.
3. Maack T, Park CH, Camargo MJF. Renal filtration, transport, and metabolism of
proteins. In: Seldin DW, Giebisch G, eds. The Kidney: Physiology and
Pathophysiology. 2nd edn. New York: Raven Press, 1992 3005-38.
4. Osicka TM, Comper WD. Glomerular charge selectivity for anionic and neutral
horseradish peroxidase. Kidney Int 1995; 47: 1630-7.
5. Osicka TM, Pratt LM, Comper WD. Glomerular capillary wall permeability to
albumin and horseradish peroxidase. Nephrology 1996: 2 199-212.
6. Lindstrom K, Blom A, Johnson E, Haraldsson B, Fries E. High glomerular
permeability of bikunin despite size and charge similar to that of serum albumin.
1 Am Soc Nephrol 1995; 6: 681.
7. Yagame M, Suzuki K, Jinde K, Yano N, Naka R Abe Y, Nomoto Y, Sakai H,
Suzuki H, Ohashi. Urinary albumin fragments as a new clinical parameter for the
early detection of diabetic nephropathy. Int Med 1995; 3 4 463-8.
8. lahaki RF, Gill DM. A micro-biuret method for estimating proteins. Anal
Biochem 196% 9 40 I - 10.
9. Tack BF, Dean J, Eilat D, Lorenz PE, Schecter AN. Tritium labelling of proteins
to high specific radioactivity by reductive methylation.J Biol Chem 1980 255
8842-7.
10. Klotz IM. Succinylation. Methods Enzymol 1967; II: 576-80.
I I. Vassiliou P, Tay M. Comper WD. Partial ischemiaand proteinuriaduring isolated
kidney perfusion is accompanied by the release of vascular [3SS]heparansulfate.
Biochem Int 1989; 19: 1241-51.
12. Tay M, Comper WD, Singh AK. Charge selectivity in kidney ultrafiltrationis
associated with glomerular uptake of transport probes. Am J Physiol 199 I ; 260
F549-54.
13. Tay M, Comper WD, Vassiliou P, Glasgow EF, Baker MS, Pratt L. The inhibitory
action of oxygen radical scavengers on proteinuriaand glomerular heparan sulfate
loss in the isolated perfused kidney. Biochem Int 1990 20: 767-78.
14. Solling K, Mogenson CE. Studies on the mechanism of renal tubular protein
reabsorption. Proc Eur Dial Transplant Assoc 1977; 14 543-9.
15. Tucker BJ, Rasch R, Blantz RC. Glomerular filtration and tubular reabsorption of
albumin in preproteinuricand proteinuric diabetic rats. J Clin Invest 1993; 9 2
686-94.
16. Harrison IF, Barnes AD. The urinary excretion of lysozyme in dogs. Clin Sci
1970 38: 533-47.
17. Litwack G. Photometric determination of lysozyme activity. Proc SOC Exp Biol
Med 1955; 8 9 401-3.
18. Fox 5. Application of Triton X- I00 colloid scintillation counting in biochemistry.
Int J Radiat lsot 1968; 1 9 717-30.
19. Olson JL, Rennke HG, Venkatachalam M A Alterations in the charge and size
selectivity barrier of the glomerular filter in aminonucleoside nephrosis in rats.
Lab Invest 1981; 4 4 271-9.
20. Ryan GB. Karnovsky MJ. An ultrastructural study of the mechanisms of
proteinuriain aminonucleoside nephrosis. Kidney Int 1975; 8: 219-32.
21. lnokuchi S, Sakai T, Shirato I, Tomino Y, Koide H. Ultrastructuralchanges in the
glomerular epithelial cells in acute puromycin aminonucleoside nephrosis: a study
by high-resolutionscanning electron microscopy. Virchows Arch A: Pathol Anat
1993; 423: I I 1-19.
22. Deen WM, Bridges CR, Brenner BM, Myers BD. Heteroporousmodel of
glomerular size selectivity: application to normal and nephrotic humans. Am J
Physiol 1985; 249 F374-89.
23. Comper WD. Glasgow EF. Charge selectivity in kidney ultrafiltration. Kidney Int
1995; 47: 1242-5 I.
24. Myers BD, Guasch A. Mechanisms of proteinuria in nephrotic humans. Pediatr
Nephrol 1994; 8: 107- 12.
25. Paine PL, Scherr P. Drag coefficientsfor the movement of rigid spheres through
liquid-filled cylindrical pores. BiophysJ 1975; 1 5 1087-91.
26. Vyas SV, Burne MJ, Pratt LM, Comper WD. Glomerular processing of dextran
sulfate during transcapillary transport. Arch Biochem Biophys 1996; 332:
205- 12.
27. Burne MJ, Vyas SV, Smit MF, Pratt LM, Comper WD. Competition between
polyanions in glomerular binding and renal clearance. Arch Biochem Biophys
1997; 340: 257-64.
28. Smit MF, Comper WD. A codiffusingsystem as a novel model for capillary wall
charge selectivity. Biophys Chem 1996; 6 2 73-80.
29. Oliver ID, Anderson S, Troy JL, Brenner BM, Deen WM. Determination of
glomerular size selectivity in normal rat with Ficoll. J Am Soc Nephrol 1992; 3
214-28.
30. Bohrer MP, Deen WM, Robertson CR, Troy JL, Brenner BM. Influence of
molecular configuration on the passage of macromolecules across the glomerular
capillary wall. J Gen Physiol 1979; 74: 583-93.
3 I. Smit MF, Pratt LM, Comper WD. Conformational dependence of the fractional
clearance of renal transport probes. Nephrology 1996 2: 355-60.
32. L e y JE, Pesce A. Micropuncture study of albumin transfer in aminonucleoside
nephrosis in the rat. Pediatr Res 1973; 7: 553-9.