Glomerular filtration is normal in the absence of both ... - Oxford Journals

Download PDF
5 downloads 0 Views 693KB Size Report
Comment
Jan 14, 2009 - of Nox4, Cbfa1, ALP and osteopontin in HASMCs. A. NADPH oxidase inhibitor and antioxidants inhibited IS- induced ROS production and ...
2044 20. Saleem MA, O’Hare MJ, Reiser J et al. A conditionally immortalized human podocyte cell line demonstrating nephrin and podocin expression. J Am Soc Nephrol 2002; 13: 630–638 21. Hultenby K, Reinholt FP, Oldberg A et al. Ultrastructural immunolocalization of osteopontin in metaphyseal and cortical bone. Matrix 1991; 11: 206–213 22. Wernerson A, Duner F, Pettersson E et al. Altered ultrastructural distribution of nephrin in minimal change nephrotic syndrome. Nephrol Dial Transplant 2003; 18: 70–76 23. Weibel E. Stereological methods. In: Practical Methods for Biological Morphometry. London: Academic, 1979 24. De Preter K, Speleman F, Combaret V et al. Quantification of MYCN, DDX1, and NAG gene copy number in neuroblastoma using a realtime quantitative PCR assay. Mod Pathol 2002; 15: 159–166 25. Rodewald R, Karnovsky MJ. Porous substructure of the glomerular slit diaphragm in the rat and mouse. J Cell Biol 1974; 60: 423–433 26. Haraldsson B, Sorensson J. Why do we not all have proteinuria? An update of our current understanding of the glomerular barrier. News Physiol Sci 2004; 19: 7–10 27. Mundel P, Shankland SJ. Podocyte biology and response to injury. J Am Soc Nephrol 2002; 13: 3005–3015 28. Benzing T. Signaling at the slit diaphragm. J Am Soc Nephrol 2004; 15: 1382–1391 29. Akhtar M, Al Mana H. Molecular basis of proteinuria. Adv Anat Pathol 2004; 11: 304–309 30. Barisoni L, Kopp JB. Update in podocyte biology: putting one’s best foot forward. Curr Opin Nephrol Hypertens 2003; 12: 251–258 31. Chugh SS, Kaw B, Kanwar YS. Molecular structure-function relationship in the slit diaphragm. Semin Nephrol 2003; 23: 544–555 32. Tryggvason K. Unraveling the mechanisms of glomerular ultrafiltration: nephrin, a key component of the slit diaphragm. J Am Soc Nephrol 1999; 10: 2440–2445 33. Godin RE, Robertson EJ, Dudley AT. Role of BMP family members during kidney development. Int J Dev Biol 1999; 43: 405–411

S. Goldberg et al. 34. Winnier G, Blessing M, Labosky PA et al. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev 1995; 9: 2105–2116 35. Zhang H, Bradley A. Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 1996; 122: 2977–2986 36. Dudley AT, Lyons KM, Robertson EJ. A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev 1995; 9: 2795–2807 37. Dudley AT, Robertson EJ. Overlapping expression domains of bone morphogenetic protein family members potentially account for limited tissue defects in BMP7 deficient embryos. Dev Dyn 1997; 208: 349– 362 38. Ozkaynak E, Schnegelsberg PN, Jin DF et al. Osteogenic protein-2. A new member of the transforming growth factor-beta superfamily expressed early in embryogenesis. J Biol Chem 1992; 267: 25220– 25227 39. Tufro A. VEGF spatially directs angiogenesis during metanephric development in vitro. Dev Biol 2000; 227: 558–566 40. Sugimoto H, Hamano Y, Charytan D et al. Neutralization of circulating vascular endothelial growth factor (VEGF) by anti-VEGF antibodies and soluble VEGF receptor 1 (sFlt-1) induces proteinuria. J Biol Chem 2003; 278: 12605–12608 41. Guan F, Villegas G, Teichman J et al. Autocrine VEGF-A system in podocytes regulates podocin and its interaction with CD2AP. Am J Physiol Renal Physiol 2006; 291: F422–F428 42. Simons M, Schwarz K, Kriz W et al. Involvement of lipid rafts in nephrin phosphorylation and organization of the glomerular slit diaphragm. Am J Pathol 2001; 159: 1069–1077 43. Simons K, Toomre D. Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 2000; 1: 31–39 44. Cherukuri A, Dykstra M, Pierce SK. Floating the raft hypothesis: lipid rafts play a role in immune cell activation. Immunity 2001; 14: 657–660 Received for publication: 14.5.08; Accepted in revised form: 11.12.08

Nephrol Dial Transplant (2009) 24: 2044–2051 doi: 10.1093/ndt/gfn758 Advance Access publication 14 January 2009

Glomerular filtration is normal in the absence of both agrin and perlecan–heparan sulfate from the glomerular basement membrane Seth Goldberg1 , Scott J. Harvey1 , Jeanette Cunningham1 , Karl Tryggvason2 and Jeffrey H. Miner1 1 2

Renal Division, Department of Internal Medicine, Washington University School of Medicine, St Louis, MO 63110, USA and Division of Matrix Biology, Department of Medical Biochemistry and Biophysics, Karolinska Institute, 171 77 Stockholm, Sweden

Correspondence and offprint requests to: Jeffrey H. Miner; E-mail: [email protected]

Abstract Background. For several decades, it has been thought that the glomerular basement membrane (GBM) provides a charge-selective barrier for glomerular filtration. However, recent evidence has presented challenges to this concept: selective removal of heparan sulfate (HS) moieties that impart a negative charge to the GBM causes little if any increase in proteinuria. Removal of agrin, the major GBM HS-proteoglycan (HSPG), from the GBM causes a pro-

found reduction in the glomerular anionic charge without changing the excretion of a negatively charged tracer. Perlecan is another HSPG present in the GBM, as well as in the mesangium and Bowman’s capsule, that could potentially contribute to a charge barrier in the absence of agrin. Methods. Here we studied the nature of the glomerular filtration barrier to albumin in mice lacking the HS chains of perlecan either alone or in combination with podocytespecific loss of agrin.

 C The Author [2009]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: [email protected]

Agrin, perlecan and glomerular filtration

Results. The results show significant reductions in anionic sites within the GBM in perlecan-HS and in perlecanHS/agrin double mutants. Podocyte and overall glomerular architecture were normal, and renal function was normal up to 15 months of age with no measurable proteinuria. Moreover, excretion of a negatively charged Ficoll tracer was unchanged as compared to control mice. Conclusions. These findings cast further doubt upon a critical role for the GBM in charge selectivity. Keywords: charge selectivity; Ficoll; glomerular basement membrane; glomerular filtration; perlecan

Introduction The glomerular capillary wall serves as a semi-permeable filtration barrier. Classic experiments from the 1970s suggested the existence of a size-selective component based on the differential clearance of dextrans with radii between ˚ [1]. However, it was noted that proteins such 18 and 60 A as albumin possessed filtration characteristics that could not be wholly accounted for by their size, with fractional clearances lower than what would be predicted from studies of dextrans of equal size [2–5]. This phenomenon was explained in part by conformational differences between the random coil of dextrans and the prolate ellipsoid of albumin [6]. Ficoll, a spherical polysaccharide more rigid than its dextran counterpart, also exhibited a fractional clearance less than would be expected based upon size alone [7,8]. The concept of charge selectivity arose from these early observations [9]. Electrostatic forces were thought to also account for the diminished clearance of albumin, which is a polar protein with an overall net charge between −12 and −18 at physiologic pH [6]. Furthermore, studies with differentially charged dextrans showed that, at a given molecular radius, neutral dextrans were filtered more freely than negatively charged dextran sulfate, but with a lower clearance than polycationic DEAE dextran [10]. Similar results were obtained with differentially charged variants of horseradish peroxidase [11]. It was postulated that electrostatic repulsion at the glomerular filtration barrier accounted for these observations. The components of the glomerular capillary wall are known to possess an overall net negative charge, provided at anatomically distinct sites. Podocalyxin, a sialylated transmembrane protein, largely provides the anionic charge found within the glycocalyx of endothelial cells and podocytes [12]. In the latter, podocalyxin is thought to play a critical role in the maintenance of foot process architecture due in part to its extracellular charge but perhaps more so to its linkage to the actin cytoskeleton via ezrin [13,14]. The glomerular basement membrane (GBM) likewise possesses a net negative charge, with sulfated glycosaminoglycan side chains of its constituent proteoglycans providing the majority of the anionic charge. Heparan sulfate (HS) proteoglycans (HSPGs) have been described along the lamina rara externa and to a lesser extent along the lamina rara interna [15–17].

2045

Three distinct HSPGs have been identified within the GBM: agrin, perlecan and collagen XVIII [18]. Agrin is produced primarily by the podocytes, and it is the predominant proteoglycan of adult GBM in all species studied [19]. It exists as a core protein of 212 kDa with at least two HS side chains [20]. It is also present at nerve–muscle synapses where it plays an important developmental role, with Agrn knockout mice dying at birth due to paralysis caused by the absence of neuromuscular junctions [21]. Perlecan, in contrast, is produced primarily by glomerular endothelial cells and is present in the GBM during development. Later, perlecan is found predominantly in the mesangial matrix and Bowman’s capsule, as is collagen XVIII [22,23]. Perlecan’s core protein of 467 kDa is linked to three HS glycosaminoglycan side chains via N-terminal domain I attachment sites [24,25]. Homozygous perlecan-null (Hspg2−/−) mice die in utero or shortly after birth with disordered cartilage matrix and basement membrane disruption in regions of increased mechanical stress [26,27], but animals homozygous for an exon-specific mutation that removes the HS attachment sites (Hspg23/3 ) exhibit only mild extrarenal defects and are otherwise viable and fertile [28]. Our prior work [29] examined mice with a podocytespecific knockout of agrin. These mice were developmentally normal, viable as adults and fertile. The glomerular filtration barrier was not compromised despite the absence of a major GBM component (agrin) and a significant reduction in the subepithelial GBM anionic charge. In addition, the fractional clearance of an anionic tracer was unchanged. These results suggest that GBM charge is not a major contributor to glomerular charge selectivity. The present study examines the potential role of perlecan’s HS side chains by investigating whether their loss, either alone or in combination with podocyte-specific agrin deficiency, results in a change in glomerular charge selectivity. Methods Agrin and perlecan mutant mice Mice denoted herein as agrin mutants possessed two nonfunctional Agrn alleles in podocytes. Mutants carried either two conditional ‘floxed’ Agrn alleles (Agrnfl/fl ) or one floxed and one null (Agrndel/fl ) allele in the presence of the 2.5 PCre transgene, which uses the human podocin (NPHS2) promoter to drive Cre expression specifically in podocytes [30]. The Agrnfl allele harbours loxP sites within introns 6 and 33, and the intervening genomic sequence is removed upon Cre-mediated recombination to generate a null allele [29]. In the Agrndel allele, the genomic sequence from within exon 6 to intron 33 is replaced by a PGK-neo cassette [21]. Mice denoted as perlecan mutants possessed two copies of the mutated perlecan gene (Hspg2). The allele (Hspg23 ) contains a deletion of exon 3, removing the sites for HS attachment, as previously described [28]. Agrin/perlecan double mutants were generated by intercrossing. Both Agrn and Hspg2 are on chromosome 4 and are separated by 8.6 centimorgans, which facilitated the production of double mutants once the targeted alleles became linked via meiotic recombination. Control mice had at least one wild-type allele for both Agrn and Hspg2 and/or lacked the 2.5 PCre transgene. All mice were studied on a mixed 129 × C57BL/6J × CBA/J background. Animal experiments were approved by the Washington University Animal Studies Committee. Histology and immunostaining Kidneys were formalin fixed and paraffin embedded prior to sectioning and staining with haematoxylin and eosin (H&E) or periodic acid-Schiff (PAS) reagent for light microscopy. Kidneys were also embedded in

2046 optimal cutting temperature compound (Sakura finetek, Torrance, CA, USA) and frozen in 2-methylbutane cooled in a dry-ice/ethanol bath. Unfixed cryosections were immunostained after being blocked with 1.5% goat serum in PBS. For agrin staining, the rabbit anti-mouse agrin LG (COOH-terminus) antibody (T. Sasaki, Oregon Health & Sciences University, Portland, OR, USA) was used at a 1:25 000 dilution [19]. For HS detection, the sections were fixed for 10 min in acetone fixing at 4◦ C and blocked using the Mouse-on-Mouse kit (Vector Laboratories, Burlingame, CA, USA) prior to staining with the antibody JM403 (Seikagaku Corporation, Tokyo, Japan) at a 1:100 dilution [31]. The secondary fluorophore-labelled antibodies were Cy3-conjugated goat anti-rabbit IgG (1:500) for agrin and FITC-conjugated goat anti-mouse IgM (1:100) for HS. Electron microscopy and polyethyleneimine labelling For routine electron microscopy, kidney cortices were immersion fixed in 2% paraformaldehyde and 2% glutaraldehyde in a 0.1 M sodium cacodylate buffer, pH 7.2. Tissue was then post-fixed in 1% OsO4 and serially dehydrated prior to embedding in Polybed (Polysciences, Warrington, PA, USA). Counterstaining was performed with uranyl acetate and lead citrate. Thin sections were examined with a CX-100 electron microscope (JEOL, Tokyo, Japan). The GBM thickness was measured along 15 evenly spaced points within each photograph. Comparisons were made after the thickest and thinnest measurements in each photograph were excluded. For polyethyleneimine (PEI) labelling, a minced kidney cortex was incubated in 0.5% PEI (1.8 kDa; Sigma, St Louis, MO, USA) in 0.9% NaCl, pH 7.3 for 30 min. The specimens were then washed in a 0.1 M sodium cacodylate buffer and incubated in 2% phosphotungstic acid and 2.5% glutaraldehyde in a 0.1 M sodium cacodylate buffer for 1 h. Tissue was then post-fixed and embedded as above. Glomerular capillary loops were photographed at 14 000× magnification, and the number of PEI aggregates per micrometre was counted along each aspect of the GBM (subepithelial and subendothelial). Several litters of adult mice between the ages of 9 and 15 months were analysed using this technique, with totals of two controls, one agrin mutant, four perlecan mutants and three agrin/perlecan double mutants. Clinical chemistry and Ficoll clearance studies Urinary protein concentrations were measured using Biuret reactions, and creatinine concentrations were measured using Jaffe reactions, each on a Cobas Mira Plus analyser (Roche Diagnostics, Indianapolis, IN, USA). Levels of total serum protein and blood urea nitrogen were also analysed. For Ficoll clearance studies, the mice underwent tail injection with a bolus of FITC-labelled carboxymethyl Ficoll-70 in 0.9% NaCl (TdB Consultancy, Uppsala, Sweden) at a dose of 250 µg/g of body weight. The mice were housed for 24 h in metabolic cages (Hatteras Instruments, Cary, NC, USA), and their urine samples were collected and assayed. The concentration of tracer present in the urine was determined on a QuantaMaster fluorimeter (Photon Technology International, Lawrenceville, NJ, USA) using an excitation wavelength of 488 nm and an emission wavelength of 529 nm. The experimental samples were compared with standards after dilution in 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.5. Clearance of the tracer was expressed as the percentage of the total amount injected that was excreted in 24 h. Statistical analyses The two-sample t-test was used to analyse the PEI, urinalysis and serum data, as well as the Ficoll fractional clearances. Differences were considered to be significant at P < 0.05.

Results Disruption of GBM–HSPGs alone or in combination does not disrupt renal function All mutants displayed normal development and survival into adulthood through at least 15 months of age. Testing for underlying renal dysfunction by analysing blood urea nitrogen levels demonstrated no significant differences between

S. Goldberg et al. Table 1. Clinical chemistry assays for serum urea nitrogen, serum protein and urine protein-to-creatinine ratio showed no statistically significant differences among the wild-type, perlecan mutant and perlecan/agrin double-mutant groups (all P-values were >0.05)

Assay Serum urea nitrogen Serum protein Urine protein-tocreatinine ratio (mg/mg)

Wild-type (n = 7)

Perlecan mutant (n = 5)

Double mutant (n = 7)

19.1 ± 5.6 mg/dL 26.2 ± 1.8 mg/dL 21.9 ± 9.3 mg/dL 5.04 ± 0.6 g/dL 20.8 ± 12.5

4.26 ± 1.3 g/dL 31.46 ± 16.7

4.54 ± 1.6 g/dL 20.9 ± 9.1

the mutants (perlecan mutants or agrin/perlecan double mutants) and controls (Table 1). The wild-type average blood urea nitrogen of 19.1 ± 5.6 mg/dL did not differ statistically from that of the perlecan mutant group (26.2 ± 1.8 mg/ dL, P = 0.07) or the double-mutant group (21.9 ± 9.3 mg/ dL, P = 0.20). Likewise, there was no statistical difference between the serum protein measurements (Table 1). The wild-type group had an average serum protein level of 5.04 ± 0.6 g/dL, which was not significantly different from the perlecan mutant group (4.26 ± 1.3 g/dL, P = 0.21) or the double-mutant group (4.54 ± 1.6 g/dL, P = 0.45). Urine samples from all groups were analysed for protein levels and standardized to creatinine concentration. The urinary protein-to-creatinine ratios (mg/mg) demonstrated no significant differences among the groups (Table 1). The double-mutant group (20.9 ± 9.1) showed no difference as compared to the control group (20.8 ± 12.5, P = 0.98). Furthermore, no statistical difference was observed when the double-mutant group was compared to the perlecan mutant group (31.46 ± 16.7, P = 0.19). Immunofluorescence reveals selective loss of agrin in the GBM of agrin and double mutants As we previously showed, mice lacking a functional agrin allele in podocytes (Agrnfl/fl ; 2.5 PCre and Agrndel/fl ; 2.5 PCre mice) demonstrated a dramatic reduction in GBM agrin (Figure 1B and D). Whether agrin was eliminated alone or in combination with perlecan-HS, the agrin mutants showed glomerular agrin deposition that was confined primarily to the mesangial matrix. In contrast, the control and perlecan mutant groups showed bright, linear staining for agrin along the GBM (Figure 1A and C). Differential reduction of glomerular heparan sulfate in perlecan, agrin and double mutants We used the anti-HS antibody JM403 to label HS in adult kidneys. The glomeruli of the control mice revealed staining in the mesangium as well as linearly along the GBM (Figure 2A). The selective elimination of podocyte agrin synthesis led to an absence of HS staining in the GBM, but mesangial deposition was preserved (Figure 2B). The reverse was found in the perlecan mutants: mesangial staining was diminished, whereas the GBM maintained its

Agrin, perlecan and glomerular filtration

2047

Fig. 3. Light microscopic analysis (H&E stain) of representative glomeruli from adult wild-type (A), agrin-only mutant (B), perlecan-only mutant (C) and double-mutant (D) kidneys. No structural abnormalities were observed. Magnification 400×. Fig. 1. Immunofluorescence with the agrin LG antibody shows normal GBM deposition in the wild-type (A) and perlecan-HS mutants (C). The agrin-only mutant (B) and double mutant (D) show reduced GBM deposition. Magnification 400×.

Fig. 4. Light microscopic analysis (H&E stain) of representative tubules from adult wild-type (A), agrin-only mutant (B), perlecan-only mutant (C) and double-mutant (D) kidneys. No structural abnormalities were observed. Magnification 400×.

Fig. 2. Immunofluorescence with the HS antibody JM403 shows linear GBM and mesangial staining in wild-type mice (A). The podocyte-specific agrin mutant (B) shows reduced GBM staining with preserved mesangial staining. The perlecan-HS mutation results in the elimination of most of the mesangial staining (C), but the linear GBM staining is preserved. Doublemutant glomeruli (D) also showed reduced HS staining. Magnification 400×.

distinctive linear staining pattern (Figure 2C). The double mutants showed an overall decrease in staining for HS, although this effect was variable throughout the kidney (Figure 2D). These staining patterns suggest that while there may be some overlap, agrin was primarily localized to the GBM, whereas perlecan was present mostly in the mesangium.

Light and electron microscopy reveals no renal structural defects in mutants When compared to wild-type controls, no histological abnormalities were noted in the kidneys in the singleor double-mutant animals. Figure 3 shows the normal glomerular architecture, with open and intact capillary loops, no obvious GBM thickening and no mesangial expansion. Likewise, the tubular architecture showed no differences between the wild-type and mutant mice by both H&E and PAS staining (Figure 4). Glomerular ultrastructure in mutants was indistinguishable from the controls (Figure 5). All genotypes demonstrated the maintenance of organized podocyte foot process architecture. Irregularities of the GBM were present in all groups in comparable numbers, including the control

2048

Fig. 5. Electron microscopic analysis of wild-type (A), agrin-only mutant (B), perlecan-only mutant (C) and double-mutant (D) glomerular capillary wall. No significant differences among the groups were observed. Magnification 14 000×; scale bar 1 µm.

S. Goldberg et al.

mutants (4.34 ± 0.3 versus 11.21 ± 1.0, P < 0.001). A similar reduction was also found in the agrin/perlecan double mutants as compared to littermate controls (4.73 ± 2.3 versus 11.15 ± 1.2, P < 0.001) and to the perlecan-only mutants (4.73 ± 2.3 versus 11.21 ± 1.0, P < 0.001). There was no difference between the agrin-only mutant group and the double mutants (4.34 ± 0.3 versus 4.73 ± 2.3, P = 0.72). Of note, the double-mutant kidneys showed stretches of GBM with PEI staining comparable to the control group; these accounted for 20% of the length of the GBM that was photographed and measured. These could represent either agrin-positive segments stemming from inefficient Cre activity in a subset of podocytes or upregulated deposition of an alternative anionic protein, such as collagen XVIII. These regions were included in the calculations, demonstrating statistical significance nonetheless. Subendothelial PEI labelling differed between some groups. Although no difference was found between agrinonly mutants and littermate controls (4.80 ± 0.5 versus 5.03 ± 0.7, P = 0.53), the perlecan-only mutants demonstrated a reduction in anionic sites as compared both to littermate controls (3.83 ± 0.8 versus 5.03 ± 0.7, P < 0.001) and to the agrin-only mutant group (3.83 ± 0.8 versus 4.80 ± 0.5, P = 0.02). Likewise, a similar reduction was found between the agrin/perlecan mutants as compared to the littermate controls (3.93 ± 0.8 versus 5.03 ± 0.7, P = 0.002) as well as the agrin-only mutants (3.93 ± 0.8 versus 4.80 ± 0.5, P = 0.03). There was no difference between the perlecan-only mutant group and the double mutants (3.83 ± 0.8 versus 3.93 ± 0.8, P = 0.70). Ficoll clearance is unaffected in double mutants

Fig. 6. Measurements of the glomerular basement membrane (GBM) thickness revealed no differences among the groups (all P-values >0.05).

kidneys and likely were plane-of-section artefacts. GBM thicknesses were measured and are shown in Figure 6. There were no statistically significant differences among any of the groups. Polyethyleneimine labelling demonstrates significant GBM charge alterations PEI labelling was used to visualize anionic sites along the GBM (Figure 7). Quantitation of the number of subepithelial anionic sites per micrometre revealed no difference between the perlecan mutant and the wild-type littermate control (11.21 ± 1.0 versus 11.15 ± 1.2, P = 0.89). However, agrin mutants showed a significant reduction in charged sites compared with littermate controls (4.34 ± 0.3 versus 11.15 ± 1.2, P < 0.001) and with the perlecan-only

Next we investigated whether the reduction in GBM anionic sites increased the rate of clearance of a negatively charged tracer, which would be expected if these sites contribute to the charge barrier. Urinary excretion of intravenously injected, fluorescently labelled carboxymethy-Ficoll-70 was measured, and the overall fractional clearance was determined and expressed as the percentage of the total amount injected that was excreted over a 24-h period. These data are shown in Table 2. There were no statistically significant differences among the groups. The control group averaged a Ficoll clearance of 24.10% ± 8.3%, and this did not differ from the perlecan group (26.25% ± 17.7%, P = 0.78) or the double-mutant group (27.36% ± 5.9%, P = 0.59).

Discussion This study tested the hypothesis that the observed charge selectivity of the glomerular filter is provided by the intrinsic negative charge of the GBM. However, our results challenge this theory by showing that removal of components known to provide the GBM with its negative charge did not lead to a measurable difference in the fractional clearance of a negatively charged tracer. These findings are in agreement with and extend our earlier work showing that the selective loss of podocyte-derived agrin had no effect on glomerular filtration despite the significant loss of ∼50% of the anionic sites along the GBM [29]. Here we

Agrin, perlecan and glomerular filtration

2049

Fig. 7. Disruption of anionic sites in the GBM are manifested as a reduction in PEI labelling. In control mice, the punctuate PEI labelling is distributed in a semi-regular pattern along the subepithelial and subendothelial laminae rarae (A). Agrin mutants (B) show a reduction in the subepithelial labelling only, whereas perlecan mutants (C) show a decrease in the subendothelial layer only. Double mutants (D) show diminished labelling along both the subepithelial and subendothelial aspects. Quantitative analyses (E and F) reveal a significant reduction in subepithelial labelling in agrin mutants and double mutants when compared to littermate controls and perlecan mutants (P < 0.001). A significant reduction in subendothelial labelling was found in perlecan mutants and double mutants when compared to littermate controls and agrin mutants (P < 0.03). Magnification 14 000×; scale bar, 1 µm. Table 2. The fractional clearances of carboxymethylated FITC-Ficoll after 24 h showed no statistically significant differences between any of the groups (all P-values were >0.05)

Genotype

Ficoll fractional clearance (% of amount injected)

Wild-type (n = 7) Perlecan mutant (n = 5) Double mutant (n = 7)

24.10 ± 8.3 26.25 ± 17.7 26.90 ± 11.3

demonstrated a similar reduction in the subepithelial charge density when agrin was eliminated along with perlecan-HS (double mutants). Perlecan, which is present in the adult GBM, albeit to a much lesser degree than agrin, was shown to contribute to the overall negative charge. The mice lacking the HS attachment sites on perlecan demonstrated a statistically significant reduction in anionic sites along the GBM, though only along the subendothelial aspect. This was not unexpected, given that perlecan is produced primarily by the endothelial cells. The immunofluorescent staining for HS supports this pattern of expression, with a primarily mesangial loss in perlecan mutants, but preserved linear staining in the GBM. However, as with the Agrn mutant mice, disruption in this intrinsic charge did not result in any measurable difference in the fractional clearance of polyanionic Ficoll. To show that a subtle compensatory upregulation of either agrin or perlecan expression in the absence of the other did not account for the lack of alterations in the glomerular barrier, we eliminated both components. These double mutants, like their agrin-only and perlecan-only mutant counterparts, displayed no observable histologic or ultrastructural abnormalities in glomeruli. Also, while they demonstrated diminished staining for HS and a decrease in

subepithelial and subendothelial anionic sites, they showed no difference from control mice in the fractional clearance of negatively charged Ficoll. The evaluation of both urine and serum samples showed no defects in kidney function. Urinary protein excretion, when standardized to creatinine, showed no evidence of proteinuria in any of the groups (control, perlecanHS mutant and agrin/perlecan-HS double mutant). Also, measurements of blood urea nitrogen levels as well as protein concentrations revealed no significant differences. The mice used in this study were at least 9 months old, and through at least 15 months of age they demonstrated no survival disadvantages. Taken together, these results suggest that an electrostatic charge barrier to negatively charged compounds is not provided by the proteoglycan components of the GBM. Other work has shown that an isolated GBM loses much of its charge-selective properties [32]. Moreover, HS, while comprising the majority of the anionic charge within the GBM, makes up only 1% of the overall dry weight of the rat GBM matrix [33]. Type IV collagen comprises a majority of the GBM scaffolding and is known to possess carbohydrate side-chain moieties, including negatively charged sialic acid residues [34]. Thus, a contribution of type IV collagen to the GBM’s negative charge, which should be unaffected by the selective knockout of agrin and perlecanHS, cannot be fully excluded. More recent evidence also agrees with our findings. In one study, overexpression of heparanase similarly resulted in neither alteration of glomerular architecture nor decline in renal function, despite a small but statistically significant increase in albuminuria [35]. In another study, podocytespecific knockout of Ext1, whose gene product is required for HS synthesis, did not lead to significant albuminuria [36].

2050

The question regarding the source and location of the charge barrier, if such a barrier exists, remains unanswered. Focus has shifted from the extracellular matrix towards the cellular constituents of the glomerular filtration barrier. Such candidates would include the negatively charged glycocalyx of both endothelial cells and podocytes. On the capillary side, endothelial cells have been noted to desulfate exogenously administered polysaccharides [37,38]. Such modification and removal of negative charges may result in an effective decrease in their concentration at the filtration interface, and this, as well as cellular uptake time, may thus reduce their rate of passage across the barrier [39]. In podocytes, loss of podocalyxin has been shown to lead to perinatal lethality with effacement of the foot processes [13]. Distortion of this delicate architecture has made it difficult to distinguish the specific effect of the loss of this negatively charged compound, although it is reasonable to suggest that it is the charge itself that allows for the maintenance of normal podocyte structure and function. Alternatively, there is evidence that podocalyxin may contribute to organizing the podocyte cytoskeleton independent of its anionic charge [40]. Acknowledgements. We thank Jennifer Richardson for mouse genotyping, the Mouse Genetics Core for animal care and the Pulmonary Morphology Core (supported by P01HL029594) for histology services. This work was supported by an American Heart Association Established Investigator Award and NIH grants R01DK064687, R01GM060432 and R01DK078314 to J.H.M., and by grants from the Swedish Medical Research Council, the Knut and Alice Wallenberg Foundation and the Novo Nordisk Foundation to K.T. S.G. was supported by NIH training grant T32DK007126. Valuable technical support was provided by the Washington University Center for Kidney Disease Research (P30DK079333). Conflict of interest statement. None declared.

References 1. Chang RL, Ueki IF, Troy JL et al. Permselectivity of the glomerular capillary wall to macromolecules: II. Experimental studies in rats using neutral dextran. Biophys J 1975; 15: 887–906 2. Bertolatus JA, Hunsicker LG. Glomerular sieving of anionic and neutral bovine albumins in proteinuric rats. Kidney Int 1985; 28: 467– 476 3. Purtell JN, Pesce AJ, Clyne DH et al. Isoelectric point of albumin: effect on renal handling of albumin. Kidney Int 1979; 16: 366–376 4. Assel E, Neumann KH, Schurek HJ et al. Glomerular albumin leakage and morphology after neutralization of polyanions: I. Albumin clearance and sieving coefficient in the isolated perfused rat kidney. Ren Physiol 1984; 7: 357–364 5. Bertolatus JA, Abuyousef M, Hunsicker LG. Glomerular sieving of high molecular weight proteins in proteinuric rats. Kidney Int 1987; 31: 1257–1266 6. Peters T Jr. Serum albumin. Adv Protein Chem 1985; 37: 161–245 7. Bohrer MP, Deen WM, Robertson CR et al. Influence of molecular configuration on the passage of macromolecules across the glomerular capillary wall. J Gen Physiol 1979; 74: 583–593 8. Oliver JD 3rd, Anderson S, Troy JL et al. Determination of glomerular size-selectivity in the normal rat with Ficoll. J Am Soc Nephrol 1992; 3: 214–228 9. Comper WD, Glasgow EF. Charge selectivity in kidney ultrafiltration. Kidney Int 1995; 47: 1242–1251 10. Bohrer MP, Baylis C, Humes HD et al. Permselectivity of the glomerular capillary wall. Facilitated filtration of circulating polycations. J Clin Invest 1978; 61: 72–78

S. Goldberg et al. 11. Rennke HG, Patel Y, Venkatachalam MA. Glomerular filtration of proteins: clearance of anionic, neutral, and cationic horseradish peroxidase in the rat. Kidney Int 1978; 13: 278–288 12. Kerjaschki D, Sharkey DJ, Farquhar MG. Identification and characterization of podocalyxin—the major sialoprotein of the renal glomerular epithelial cell. J Cell Biol 1984; 98: 1591–1596 13. Doyonnas R, Kershaw DB, Duhme C et al. Anuria, omphalocele, and perinatal lethality in mice lacking the CD34-related protein podocalyxin. J Exp Med 2001; 194: 13–27 14. Orlando RA, Takeda T, Zak B et al. The glomerular epithelial cell anti-adhesin podocalyxin associates with the actin cytoskeleton through interactions with ezrin. J Am Soc Nephrol 2001; 12: 1589– 1598 15. Caulfield JP, Farquhar MG. Distribution of annionic sites in glomerular basement membranes: their possible role in filtration and attachment. Proc Natl Acad Sci U S A 1976; 73: 1646–1650 16. Kanwar YS, Farquhar MG. Anionic sites in the glomerular basement membrane. In vivo and in vitro localization to the laminae rarae by cationic probes. J Cell Biol 1979; 81: 137–153 17. Kanwar YS, Linker A, Farquhar MG. Increased permeability of the glomerular basement membrane to ferritin after removal of glycosaminoglycans (heparan sulfate) by enzyme digestion. J Cell Biol 1980; 86: 688–693 18. Harvey SJ, Miner JH. Revisiting the glomerular charge barrier in the molecular era. Curr Opin Nephrol Hypertens 2008; 17: 393–398 19. Groffen AJ, Ruegg MA, Dijkman H et al. Agrin is a major heparan sulfate proteoglycan in the human glomerular basement membrane. J Histochem Cytochem 1998; 46: 19–27 20. Groffen AJ, Buskens CA, van Kuppevelt TH et al. Primary structure and high expression of human agrin in basement membranes of adult lung and kidney. Eur J Biochem 1998; 254: 123–128 21. Lin W, Burgess RW, Dominguez B et al. Distinct roles of nerve and muscle in postsynaptic differentiation of the neuromuscular synapse. Nature 2001; 410: 1057–1064 22. Groffen AJ, Veerkamp JH, Monnens LA et al. Recent insights into the structure and functions of heparan sulfate proteoglycans in the human glomerular basement membrane. Nephrol Dial Transplant 1999; 14: 2119–2129 23. Saarela J, Rehn M, Oikarinen A et al. The short and long forms of type XVIII collagen show clear tissue specificities in their expression and location in basement membrane zones in humans. Am J Pathol 1998; 153: 611–626 24. Murdoch AD, Dodge GR, Cohen I et al. Primary structure of the human heparan sulfate proteoglycan from basement membrane (HSPG2/perlecan). A chimeric molecule with multiple domains homologous to the low density lipoprotein receptor, laminin, neural cell adhesion molecules, and epidermal growth factor. J Biol Chem 1992; 267: 8544–8557 25. Kallunki P, Tryggvason K. Human basement membrane heparan sulfate proteoglycan core protein: a 467-kD protein containing multiple domains resembling elements of the low density lipoprotein receptor, laminin, neural cell adhesion molecules, and epidermal growth factor. J Cell Biol 1992; 116: 559–571 26. Arikawa-Hirasawa E, Watanabe H, Takami H et al. Perlecan is essential for cartilage and cephalic development. Nat Genet 1999; 23: 354–358 27. Costell M, Gustafsson E, Aszodi A et al. Perlecan maintains the integrity of cartilage and some basement membranes. J Cell Biol 1999; 147: 1109–1122 28. Rossi M, Morita H, Sormunen R et al. Heparan sulfate chains of perlecan are indispensable in the lens capsule but not in the kidney. EMBO J 2003; 22: 236–245 29. Harvey SJ, Jarad G, Cunningham J et al. Disruption of glomerular basement membrane charge through podocyte-specific mutation of agrin does not alter glomerular permselectivity. Am J Pathol 2007; 171: 139–152 30. Moeller MJ, Sanden SK, Soofi A et al. Podocyte-specific expression of cre recombinase in transgenic mice. Genesis 2003; 35: 39– 42

IS induces oxidative stress in VSMCs 31. Van Den Born J, Van Den Heuvel LP, Bakker MA et al. A monoclonal antibody against GBM heparan sulfate induces an acute selective proteinuria in rats. Kidney Int 1992; 41: 115–123 32. Bertolatus JA, Klinzman D. Macromolecular sieving by glomerular basement membrane in vitro: effect of polycation or biochemical modifications. Microvasc Res 1991; 41: 311–327 33. Kanwar YS, Farquhar MG. Isolation of glycosaminoglycans (heparan sulfate) from glomerular basement membranes. Proc Natl Acad Sci U S A 1979; 76: 4493–4497 34. Nayak BR, Spiro RG. Localization and structure of the asparaginelinked oligosaccharides of type IV collagen from glomerular basement membrane and lens capsule. J Biol Chem 1991; 266: 13978–13987 35. Van Den Hoven MJ, Wijnhoven TJ, Li JP et al. Reduction of anionic sites in the glomerular basement membrane by heparanase does not lead to proteinuria. Kidney Int 2008; 73: 278–287

2051 36. Chen S, Wassenhove-McCarthy DJ, Yamaguchi Y et al. Loss of heparan sulfate glycosaminoglycan assembly in podocytes does not lead to proteinuria. Kidney Int 2008; 74: 289–299 37. Dawes J, Pepper DS. Human vascular endothelial cells catabolise exogenous glycosaminoglycans by a novel route. Thromb Haemost 1992; 67: 468–472 38. Comper WD, Tay M, Wells X et al. Desulphation of dextran sulphate during kidney ultrafiltration. Biochem J 1994; 297(Pt 1): 31– 34 39. Tay M, Comper WD, Singh AK. Charge selectivity in kidney ultrafiltration is associated with glomerular uptake of transport probes. Am J Physiol 1991; 260: F549–F554 40. Takeda T, McQuistan T, Orlando RA et al. Loss of glomerular foot processes is associated with uncoupling of podocalyxin from the actin cytoskeleton. J Clin Invest 2001; 108: 289–301 Received for publication: 27.10.08; Accepted in revised form: 17.12.08

Nephrol Dial Transplant (2009) 24: 2051–2058 doi: 10.1093/ndt/gfn757 Advance Access publication 22 January 2009

Indoxyl sulphate induces oxidative stress and the expression of osteoblast-specific proteins in vascular smooth muscle cells Gulinuer Muteliefu1 , Atsushi Enomoto2,3 , Ping Jiang3 , Masahide Takahashi3 and Toshimitsu Niwa1 1 3

Department of Clinical Preventive Medicine, Nagoya University Hospital, 2 Institute for Advanced Research, Nagoya University and Department of Pathology, Nagoya University Graduate School of Medicine, Nagoya, Japan

Correspondence and offprint requests to: Toshimitsu Niwa; E-mail: [email protected]

Abstract Background. Previously, we demonstrated that indoxyl sulphate (IS), a uraemic toxin, induced aortic calcification in hypertensive rats. This study aimed to determine if IS induces the production of reactive oxygen species (ROS) and the expression of osteoblast-specific proteins in human aortic smooth muscle cells (HASMCs). Methods. In order to achieve these goals, HASMCs were incubated with IS. ROS were detected using probes with a fluorescence detector. The expression of alkaline phosphatase (ALP), osteopontin and organic anion transporters (OAT1, OAT3) was studied by western blotting. The expression of core binding factor 1 (Cbfa1), ALP, osteopontin and NADPH oxidases (Nox1, Nox2 and Nox4) was analysed by reverse transcription-polymerase chain reaction (RT-PCR). Knockdown of Nox4 was performed by RNA interference (RNAi). Results. IS induced ROS generation and the expression of Nox4, Cbfa1, ALP and osteopontin in HASMCs. A NADPH oxidase inhibitor and antioxidants inhibited ISinduced ROS production and mRNA expression of Cbfa1 and ALP. Knockdown of Nox4 using small interfering RNA (siRNA) inhibited IS-induced ROS production and mRNA expression of Cbfa1, ALP and osteopontin. OAT3 was expressed in HASMCs.

Conclusions. IS induces ROS generation by upregulating Nox4, and the expression of osteoblast-specific proteins such as Cbfa1, ALP and osteopontin in HASMCs. Keywords: indoxyl sulphate; NADPH oxidase Nox4; osteoblast-specific proteins; reactive oxygen species; vascular smooth muscle cells

Introduction Cardiovascular disease accounts for premature death in more than 50% of patients undergoing regular dialysis [1]. Cardiovascular disease mortality in dialysis patients is much higher especially in younger age categories than age- and sex-matched controls without chronic kidney disease (CKD) [1,2]. Vascular calcification plays a pivotal role in the development of cardiovascular morbidity and subsequent increased mortality. Vascular calcification affects both vascular intima and media layers, and its mechanism remains poorly understood. In addition to traditional cardiovascular risk factors, hyperphosphataemia, calcium overload, increased oxidized low-density lipoprotein cholesterol, uraemic toxins, increased oxidative stress, hyperhomocysteinaemia, haemodynamic overload and

 C The Author [2009]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: [email protected]