1Renal Research Group, Institute of Medicine, 2Department of Physiology, University of Bergen and 3Department of Pathology, Haukeland University Hospital, Bergen, Norway
Correspondence and offprint requests to: Rolf E. F. Christiansen, Institute of Medicine, N-5021 Haukeland University Hospital, Bergen, Norway. Email: rolf.christiansen{at}meda.uib.no
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods. The glomerular charge barrier was evaluated in 80-week-old SHR and age-matched normotensive WistarKyoto rats (WKY) by measuring the glomerular clearance of radiolabelled cationic and anionic chymotrypsinogen (Chym and aChym, MW 25 000) accumulated by the proximal tubular cells in the outer, middle and inner cortex following intravenous injection. The glomerular filtration rates (GFR) in the cortical zones were obtained by aprotinin (MW 6500) and the sieving coefficient () of the protein tracers calculated as their glomerular clearance/GFR.
Results. The aChym was similar in SHR and WKY except in the inner cortex where it was 35% higher in SHR (0.65±0.05) than in WKY (0.48±0.03) (P = 0.01). The ratio of
aChym to
Chym was increased from 0.55±0.06 to 0.77±0.05 (P<0.02) in the inner cortex of SHR, whereas
Chym remained the same in all cortical zones. Finally, the percentage of glomeruli with adsorption droplets in podocytes quantified by light microscopy was higher in the inner than the outer cortex of SHR (P<0.05).
Conclusions. The study supports the theory of a functioning glomerular charge barrier. An increased relative clearance of aChym in the inner cortex of SHR indicates impairment of the charge barrier, which, at least in part, could explain the increased protein excretion in SHR with hypertensive nephropathy.
Keywords: glomerular charge selectivity; hypertensive nephropathy; proteinuria
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The spontaneously hypertensive rat (SHR), which is recognized as a model of essential hypertension, develops hypertensive renal disease at high age and the morphological features are comparable with the findings in human kidneys exposed to long-lasting essential hypertension. The typical morphological findings are glomerulosclerosis, increased thickening of the vascular wall and interstitial fibrosis in the juxtamedullary cortex rendering the superficial parts of the renal cortex nearly normal [13].
During glomerular filtration, solutes are believed to be restricted in their passage into Bowman's space on basis of their size (size selectivity), shape (shape selectivity) and charge (charge selectivity). The property of glomerular selectivity to filtration of solutes is thought to be caused by functioning glomerular pores and a net negative charge (negative charge density) in the glomerulus. The negative glomerular charge density is thought to repel anionic plasma proteins, and hence major serum proteins such as albumin are excluded from the tubular filtrate mainly by means of charge repulsion. The charge selective barrier has been documented in earlier work [4], where the filterability of anionic dextrans was studied demonstrating a lower fractional clearance of anionic compared with neutral or cationic dextrans of same molecular size. In recent years the validity of the classical model of glomerular charge selectivity has been questioned and the possibility of a physiological relatively high albumin clearance in conjunction with tubular retrieval has been proposed [5]. However, at the same time, experimental evidence supporting the concept of a glomerular charge barrier is newly presented [6].
In SHR the glomerular capillary pressure in the deep cortex is higher than in the superficial cortex, an observation that is different from the normotensive animals where the glomerular capillary pressure is similar in the deep and superficial renal cortex [3]. Increased glomerular capillary pressure is found in the deep cortex of 10-week-old SHR and increases with age [3]. Proteinuria usually develops at 3040 weeks of age, and the increase in glomerular capillary pressure consequently precedes development of proteinuria and morphological changes in the deep cortex of SHR.
The present study was designed to investigate the role of the glomerular charge barrier on protein filtration in hypertensive renal disease in SHR. To accomplish this, chymotrypsinogen A (Sigma C 4879, MW 25 000, Stokes radius 21 Å), a cationic polypeptide filtered with a sieving coefficient () of 0.85±0.02, was anionized, resulting in an unchanged hydrodynamic radius but a reduction of
to 0.53±0.02 confirming the existence of a glomerular charge barrier. Measuring the clearance ratio of anionic to cationic chymotrypsinogen (aChym to Chym) and GFR in different cortical zones using aprotinin (Sigma A 4529, MW 6500) [7,8], we found a reduction of GFR and an attenuated glomerular charge barrier in the juxtamedullary cortex of old SHR.
![]() |
Subjects and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The experiments were performed with the approval of the Norwegian State Board for Biological Experiments with Living Animals.
The animals were anaesthetized with intraperitoneal pentobarbital (50 mg/kg body weight) and were placed on a servocontrolled heating pad to keep body temperature stable at 3738°C. Infusion of 5% bovine serum albumin in a 0.9% sodium chloride solution was given at a rate of 1.0 ml/100 g body weight per hour through a polyethylene catheter, which was inserted into the right jugular vein. The same catheter was used for bolus injection of radioactive tracers.
Haemodynamic measurements
A polyethylene catheter was inserted into the right carotid artery for measurement of arterial blood pressure and blood sampling. The arterial blood pressure was measured with a Hewlett-Packard pressure transducer connected to a Gould-4000 recorder. Urine was collected from the left urether and the bladder.
After surgery, the rats were allowed to recover for 30 min before the experiment commenced.
Tracers
Chymotrypsinogen was acetylated and thereby anionized by blocking its free amino groups [9]. The net molecular charge of aChym was confirmed by isoelectric focusing (IEF) and ionic exchange chromatography (Figure 1). The hydrodynamic radius of aChym was 21.3 Å and of Chym 21.0 Å and was measured by size exclusion chromatography using low molecular weight calibration kit proteins as standards (Superdex 75-10/30 HR; Amersham Biosciences). Aprotinin, aChym and Chym were each dissolved (1 mg/ml) in a 0.05 M Na-phosphate buffer at pH 7.5 and labelled with 131I or 125I, respectively, by using an IODO-GEN technique (Pierce Chemical Company, Rockford, IL, USA). Iodination did not alter the hydrodynamic radius of the probes or their isoelectric points as measured by size exclusion chromatography and IEF.
|
Clearance of aChym, Chym and measurement of zonal and total GFR
At zero time, 131I-labelled aprotinin (131I-Ap) and 125I-labelled anionic chymotrypsinogen (125I-aChym) or chymotrypsinogen (125I-Chym) in a volume of 100 µl of saline was injected as a bolus into the right jugular vein. Arterial blood samples (0.1 ml) were taken at time 15 s, 1, 3, 5 and 7 min after the start of the experiment. A 0.1 ml volume of 5% bovine serum albumin was injected after each blood sample to compensate for the blood loss. Total and zonal GFR was obtained by measuring tubular uptake of 131I-Ap as described by Tenstad et al. [7,8]. Similarly, total and zonal clearance of 125I-aChym and 125I-Chym was obtained by dividing the total 125I activity in the kidney or tissue samples from the outer, middle and inner cortex by the time-integrated plasma concentration during the 7 min clearance period.
The blood samples were centrifuged in a Hettich Rotanta Centrifuge at 2700 r.p.m. for 5 min, and 20 µl of plasma was taken from each sample and radioactivity was counted in a gamma counter (Cobra 2; Auto Gamma) for 5 min. Ten microlitres of plasma was also taken from each sample for measurement of total protein concentration with a refractometer (American Optical) together with 15 µl of plasma from each sample for HPLC (Superdex 75-10/30 HR; Amersham Biosciences), performed to measure protein binding of the tracers.
At the end of the experiment, 1 ml of 2% Alcian blue was injected intra-arterially and the renal arteries were ligated. The kidneys were removed and frozen in isopenthane prechilled to 80°C.
The left kidney was counted in a gamma counter and thereafter placed in a constantly cooled dish containing isopenthane at 10°C. The kidney was first divided in three 34 mm transverse slices, sectors of these slices were cut using a scalpel, and then five tissue samples of similar thickness from the outer, middle and inner cortex were prepared from each of these sectors under the microscope at 16x magnification. The corticomedullary border was determined by the use of Alcian blue and by the position of the arcuate vessels.
Each piece of cortical tissue was put into pre-weighed counting tubes, reweighed and radioactivity was counted for 5 min. The urine was collected in a test tube and radioactivity counted.
Clearance of aChym and Chym was corrected for their protein binding of 20%, as measured by HPLC in each experiment. Aprotinin clearance was corrected for its protein binding of
5% and GFR was calculated by correcting for a glomerular GibbsDonnan distribution according to an empirical formula developed by Tenstad et al. [8]:
![]() |
Measurements of urinary protein excretion
Urine was collected for 24 h while the rats were in metabolic cages with free access to food and water during the collection period. Urinary protein and albumin concentration was measured by agarose gel electrophoresis on a Hydrasys (Sebia) semi-automated analyser. The normal rats excreted <20 mg of protein/24 h.
Light and electron microscopy
Kidneys from both WKY and SHR were fixed in 10% formalin and embedded in paraffin by standard procedures. Frontal paraffin sections were cut (4 µm) and stained with periodic acid-Schiff (PAS). All analyses were performed in a blinded manner. 100 glomeruli from the outer cortex and at least 70 glomeruli from the inner cortex were analysed regarding podocytes with adsorption droplets (magnification x400) and development of synechiae/adherences. The degree of tubulointerstitial damage was assessed using a semi-quantitative scoring system (04). In PAS-stained sections, at least 20 fields from the inner or outer cortex were randomly sampled. At a magnification of x200 they were graded as follows: grade 0, no changes; grade 1, lesions involving <25% of the area; grade 2, lesions involving 2550% of the area; grade 3, lesions involving >50% of the area; grade 4 involving (almost) the entire area [10]. Tubulointerstitial damage was defined as tubular dilatation, atrophy, cast formation or interstitial expansion with inflammation or fibrosis.
For ultrastructural investigation pieces of superficial and deep cortical tissue from both WKY and SHR were fixed in McDowells solution. The pieces were post-fixed in 1% osmium tetroxide, dehydrated and embedded in Epon 812. Semi-thin sections were stained with toluidine blue. Ultra-thin sections were stained with uranylacetate and lead citrate and studied in a Philips 400 electron microscope.
Statistical analysis
The results are presented as means±SE. Differences between groups were assessed by non-parametric ANOVA on Ranks followed by one-way ANOVA in groups to be compared. Where significant differences were found, the groups were compared by unpaired t-test. Differences between cortical layers within the same strain in the histological analysis were assessed by paired comparisons. A value of P<0.05 was accepted as statistically significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
GFR was calculated from the aprotinin clearance. As shown in Figure 2, total GFR was numerically 17% lower in SHR (0.69±0.07) than in WKY (0.83± 0.08 ml min1 g1), although not a statistically significant finding (P = 0.2). GFR in different layers of the renal cortex were, however, different in WKY and SHR. Outer cortex GFR was numerically 11% lower in SHR (1.09±0.11) than in WKY (1.22±0.11 ml min1 g1), but failed to reach significance (P = 0.4). GFR in the middle part of the renal cortex was 31% lower in SHR (0.74±0.07) than in WKY (1.07±0.09 ml min1 g1) (P<0.05), while GFR of the inner part of the renal cortex was 55% lower in SHR (0.33±0.04) than in WKY (0.73±0.09 ml min1 g1) (P<0.01).
|
The sieving coefficient of aChym ( aChym) and Chym (
Chym) was calculated as their clearances divided by GFR as measured by the aprotinin method. As shown in Figure 3A,
aChym was similar in SHR and WKY for the whole kidney and in the outer and middle cortical zones, while
aChym in the inner cortex was 35% higher in SHR (0.65±0.05) than in WKY (0.48±0.03) (P = 0.01). The
aChym of
0.85 was similar in all cortical zones in SHR (P = 0.36), and the ratio of
aChym to
aChym was increased from 0.55±0.06 to 0.77±0.05 (P<0.02) in the inner cortex of SHR as shown in Figure 3B.
|
|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
SHR is a commonly accepted experimental model for essential hypertension in man and the morphological picture of human hypertensive renal disease is comparable with renal histological changes observed in old SHR. Focal and segmental glomerulosclerosis associated with tubular atrophy and interstitial fibrosis fits well with the morphological picture in man, described as decompensated benign nephrosclerosis by Ratschek et al. [11]. As in SHR, the glomerular and tubulointerstitial changes begin and are most severe in the juxtamedullary cortex.
In SHR proteinuria usually develops at the age of 3040 weeks and GFR starts to decline at higher age [3], although in the present study reduced GFR was only found in the middle and inner renal cortex. Other investigators have demonstrated that the urinary protein excretion in old SHR originated from juxtamedullary glomeruli [1,2], protein casts in tubuli were most abundant in juxtamedullary nephrons and no protein excretion was found in superficial glomeruli of male SHR using a micropuncture technique. The observations correlate with the well known morphological changes in SHR where glomerulosclerosis, increased interstitial cell infiltration and vascular wall thickness were found in deep cortex, while the more superficial part of the cortex showed minor changes [13].
The glomerular capillary filtration barrier consists of the glomerular basement membrane (GBM) with the fenestrated endothelium of the capillaries lining the inside while the visceral epithelium covers the outside. The epithelial cells are in contact with the GBM through the numerous epithelial podocyte foot processes. Proteoglycans, especially heparan sulphate proteoglycan (HSPG), in the form of perlecan and agrin are molecular constituents of the GBM which constitute parts of the negative glomerular charge density and are produced mainly in the epithelial cells [12]. There are also other negative charges in the GBM and important negatively charged glycoproteins on the endothelium, podocytes, podocyte foot processes and slit diaphragms. The precise localization of the anionic sites which are most important in conferring charge selectivity to the glomerular barrier is disputable [13]. Evidence suggest that anionic sites in the GBM are not crucial for the charge selectivity since knockout mice without glomerular basal membrane HSPG do not have proteinuria [14] and clearance values from studies of isolated basal membranes do not indicate a functioning charge barrier [15].
Although most studies show that the glomerulus has a net negative charge density, the hypothesis of a functioning glomerular charge barrier as postulated in earlier studies is debated [16] and a model of a negligible glomerular charge barrier with a normally relatively high albumin clearance together with tubular albumin reabsorption and degradation is presented [5]. There are, however, several problems in accepting a crucial role of tubular reabsorption and cellular processing in the control of albuminuria. Dysfunctions of the receptors for tubular albumin uptake (cubulin and megalin receptors) for example, are not followed by heavy proteinuria and acute tubular necrosis is not associated with marked albuminuria [5]. In addition, evidence supporting the classical theory of a glomerular charge barrier normally restricting the passage of albumin has been recently published [6]. Sorensson et al. [6] studying different proteins and Ficoll changed the ionic strength of a perfusate to alter glomerular charge interactions concluding that glomerular negative charges are important determinants of the glomerular permselectivity.
Methodological considerations
In the present study we have examined the glomerular charge barrier in different cortical zones of SHR and WKY by a new approach. The use of aprotinin, aChym and Chym is methodologically a technique where glomerular clearance can be measured in the outer, middle and inner cortex. It should be noted that even though the single glomerular filtration rate (SNGFR) is highest in the larger glomeruli of the deep cortex, the GFR per gram wet weight as estimated by the aprotinin method is lower due to the fewer number of glomeruli in the deep cortex than in the superficial cortex.
Chym is filtered with a of 0.85, whereas aChym with a practically identical hydrodynamic radius has a restricted clearance (
= 0.53), which is evidence of an existing glomerular charge barrier. Since the fractional Chym clearance is normally high (
= 0.85), this marker has limited capability to detect a potential charge- independent increase in filterability. Nevertheless, the clearance ratio between aChym and Chym provides a direct assessment of the charge barrier in the different cortical layers and the finding of a clear increase in the aChym/Chym clearance ratio only in the inner cortex of SHR without any change in Chym sieving suggests that the charge barrier in the juxtamedullary cortex of SHR is impaired.
The circulatory pattern in the juxtamedullary cortex may interfere with the measurements of CaChym and CChym. Increasing the time of exposure of the tracers to the glomerular filter would probably increase the clearance values due to increased diffusive transport through the glomerular membrane. However, in this part of the renal cortex in SHR, the glomeruli are enlarged [3], SNGFR are probably increased and the glomeruli hyperperfused. Taking this into account, the values for CaChym and CChym may be underestimated rather than overestimated in the juxtamedullary cortex, however, a disproportionate change in the clearance of the tracers would not be expected.
Using the aprotinin method to measure zonal GFR, we found unchanged GFR in outer cortex (numerically 11% lower) and a significantly lower GFR in the middle (31%) and inner cortex (55%) of SHR compared with WKY, whereas the total GFR was not significantly reduced (numerically 17% lower). Due to the spherical shape of the kidney and the method of dividing the kidney cortex into three different zones, the GFR in the outer, middle and inner cortex may represent 60,
30 and
10% of the total GFR. This intracortical distribution of GFR implies a 22% reduction of total GFR in SHR compared with WKY, in fairly good agreement with the observed value of a 17% reduction.
It is well known [17] that anionic peptides are less readily reabsorbed in proximal tubular cells than cationic peptides. This could explain the observed urinary excretion of aChym compared with Chym and aprotinin (10 vs <1%). The handling of aChym is, however, similar in SHR and WKY and should therefore not interfere with our overall conclusion. However, one fundamental reservation should be stated. In spite of detailed studies in our group on the proximal tubular reabsorption of peptides, we cannot yet totally exclude the possibility of a different outer, middle and inner proximal tubular reabsorption of aChym in old WKY and SHR. Increased inner cortical reabsorption of aChym in SHR could theoretically be a pitfall that could explain the increased inner cortical sieving coefficient in these animals. However, the finding of unaltered urinary excretion of aChym and unaltered reabsorption of native chymotrypsinogen in the outer, middle and inner cortical zones makes this explanation unlikely.
Our finding of an increased sieving coefficient for the anionic peptide in the inner cortex of old SHR with a selective proteinuria (65% albumin) is strongly supported by the observations of Feld et al. [2]. In their study using female SHR with proteinuria and hypertensive nephropathy, the glomeruli were examined by the use of polyanionic staining. Marked decreased staining affinity for polyanions was found in the juxtamedullary cortex compared with the superficial glomeruli in the rats with hypertensive nephropathy. However, to the best of our knowledge, there are little available data on the pathophysiological importance of the glomerular charge barrier in relation to increased urinary protein excretion in human hypertensive renal disease. Heintz et al. [18] demonstrated a decreased content of HSPG in GBMs of hypertensive subjects. In addition, a diminished post exercise urinary excretion of HSPG was observed in hypertensive subjects. This could suggest that the HSPG production is reduced or the turnover is changed in hypertensive individuals. Abnormal proteoglycan metabolism is involved in the onset of the functional derangement in diabetic nephropathy and this determines the loss of glomerular membrane anionic charges, which may lead to proteinuria. Hertzan-Levy et al. [19] observed reduced membrane polyanions such as HSPG in salt-sensitive hypertension in rats. Taken together, these data support our observation of a functional effect of reduced negative glomerular charge density in the kidney damaged by hypertension.
The function of the podocytes and their slit diaphragms in the barrier function of the filter has gained increased interest [20]. Lack of podocyte expression of a slit diaphragm peptide named nephrin is associated with congenital nephrotic syndrome of the Finnish type and heavy proteinuria [21], hence emphasizing the importance of the slit diaphragms. The slit pores were not studied in the present paper. We observed podocyte hypertrophy and loss of the normal cell shape with fusion of the foot processes and prominent cytoskeleton in the inner cortex. These changes are commonly seen in renal diseases with proteinuria [22]. They are considered adaptive to different injuries and thought to be reversible. Podocytes neighbouring areas with damage of the filtration barrier reabsorb filtered proteins and store them in lysosomes [20]. These adsorption droplets are easily detected as cytoplasmatic PAS positive droplets using light microscopy under high magnification and serve as a parameter of a distorted glomerular filter. Increased numbers of glomeruli in the inner cortex of SHR showed adsorption droplets compared with glomeruli in the outer cortex or compared with WKY. This observation fits nicely with the results of our functional studies. Podocyte pseudocysts, a sign of more severe podocyte damage, were also seen in the inner cortex. It is thought that damaged podocytes detach from the GBM and that bare basement membranes get attached to the parietal layer of Bowman's capsule, thus forming a synechia. Synechiae are considered the earliest lesions in the development of segmental sclerosis. They have a tendency to grow, and established segmental sclerosis usually shows broad adherences with the Bowman's capsule [20]. Both synechiae and adherences represent different stages in the progress of irreversible sclerotic lesions in glomeruli and were found significantly more often in the inner than in the outer cortex of SHR in the present study. The possible interplay between foot process fusion and effacement, detachment from the basement membrane, decreased negative glomerular charge density and proteinuria is not clear, but most observations indicate that the glomerular anionic sites play an important role.
A common feature in hypertensive as well as diabetic nephropathy is the increased glomerular capillary pressure. We have shown earlier that the increase in capillary pressure is mainly localized to the juxtamedullary cortex and the more superficial glomeruli have a normal capillary pressure [3]. How increased glomerular capillary pressure increases protein excretion and possibly interferes with glomerular anionic charge is not known. One may speculate that stretching of the glomerular capillary loops due to increased glomerular capillary pressure may interfere with podocyte attachment to the basal membrane [20] and synthesis of HSPG and thus reduce the negative glomerular charge density.
Our data support the notion of a functioning glomerular charge barrier, and the finding of an increased aChym to
Chym ratio in the inner cortex of old SHR suggests an impairment of the glomerular charge barrier, which could be of pathogenetic importance for the proteinuria observed in SHR hypertensive nephropathy.
![]() |
Acknowledgments |
---|
Conflict of interest statement. None declared.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|