- Email: [email protected]
Abolishment of proximal tubule albumin endocytosis does not affect plasma albumin during nephrotic syndrome in mice
basic research
www.kidney-international.org
Abolishment of proximal tubule albumin endocytosis does not affect plasma albumin during nephrotic syndrome in mice Kathrin Weyer1, Pia K. Andersen1, Kasper Schmidt1, Geraldine Mollet2,3, Corinne Antignac2,3,4, Henrik Birn1,5, Rikke Nielsen1 and Erik I. Christensen1 1
Department of Biomedicine, Aarhus University, Aarhus, Denmark; 2Inserm U1163, Imagine Institute, Laboratory of Hereditary Kidney Diseases, Paris, France; 3Paris Descartes-Sorbonne Paris Cité University, Paris, France; 4Department of Genetics, Necker Hospital, Assistance Publique-Hôpitaux de Paris, Paris, France; and 5Department of Renal Medicine, Aarhus University Hospital, Aarhus, Denmark
The megalin/cubilin receptor complex is required for proximal tubular endocytosis and degradation of filtered albumin. An additional high-capacity retrieval pathway of intact albumin for the recovery of large amounts of filtered albumin has been proposed, possibly involving cooperation between megalin/cubilin and the neonatal Fc receptor. To clarify the potential role of such a pathway, we examined the effects of megalin/cubilin gene inactivation on tubular albumin uptake and plasma albumin levels in nephrotic, podocin knockout mice. Immunofluorescence microscopy of megalin/cubilin/podocin knockout mouse kidneys demonstrated abolishment of proximal tubule albumin uptake, in contrast to the excessive albumin accumulation observed in podocin knockout mice compared to controls. Correspondingly, urinary albumin excretion was increased 1.4 fold in megalin/cubilin/podocin compared to podocin knockout mice (albumin/creatinine: 226 vs. 157 mg/mg). However, no difference in plasma albumin levels was observed between megalin/cubilin/ podocin and podocin knockout mice, as both were reduced to approximately 40% of controls. There were no differences in liver albumin synthesis by mRNA levels and protein abundance. Thus, megalin/cubilin knockout efficiently blocks proximal tubular albumin uptake in nephrotic mice but plasma albumin levels did not differ as a result of megalin/cubilin-deficiency, suggesting no significance of the megalin/cubilin-pathway for albumin homeostasis by retrieval of intact albumin. Kidney International (2017) j.kint.2017.07.024
-, -–-;
http://dx.doi.org/10.1016/
KEYWORDS: albuminuria; endocytosis; nephrotic syndrome; proteinuria; proximal tubule Copyright ª 2017, International Society of Nephrology. Published by Elsevier Inc. All rights reserved.
Correspondence: Kathrin Weyer, Department of Biomedicine, Aarhus University, Wilhelm Meyers Allé 3, Building 1234, DK-8000, Aarhus C, Denmark. E-mail: [email protected] Received 11 June 2017; revised 11 July 2017; accepted 27 July 2017
Kidney International (2017) -, -–-
T
he renal handling of albumin, the quantitatively most important plasma protein, has been rigorously investigated. Increased urinary excretion of albumin, i.e., albuminuria, is an early manifestation and important prognostic marker in various forms of kidney disease.1 Nevertheless, several aspects of renal albumin handling remain controversial, in particular with regard to the amount of albumin filtered by the normal glomeruli and its route of reabsorption, degradation, and potential transcytosis in the proximal tubule cells. Understanding the contribution of these pathways is important for the elucidation of the pathophysiological mechanisms of albuminuria and its potentially harmful role in the progression of kidney diseases.2 The prevailing view is that normal glomerular filtration of albumin is limited by the properties of the glomerular filtration barrier to 0.5 to 2 g/d in humans and that the filtered albumin is efficiently reabsorbed and degraded by the proximal tubule.3 This perception has been challenged by the proposal that large amounts of albumin are normally filtered in the kidney (w225 g/d) and that a high-capacity retrieval pathway in the proximal tubule returns albumin to the circulation to maintain plasma albumin concentrations.4,5 The proximal tubule cell is responsible for renal reabsorption of all filtered proteins, including albumin, primarily mediated by the megalin/cubilin receptor complex.6–12 Following megalin/cubilin-mediated endocytosis, albumin is targeted for lysosomal degradation, whereas the receptors recycle to the apical plasma membrane. It has been suggested that albumin can also be rescued by the neonatal Fc receptor (FcRn), leading to transcytosis and the return of intact albumin to the circulation.13,14 As FcRn knockout (KO) mice develop hypoalbuminemia,15 FcRn-mediated transcytosis of albumin has been claimed to have a crucial role in maintaining normal plasma albumin levels.16 FcRn binds albumin with higher affinity at low pH (pH 6.1),17 allowing it to bind albumin in the acidic environment of the endosome and releasing it at a higher pH in the extracellular environment. Because the pH of the proximal tubule lumen does not facilitate albumin binding to FcRn, it was further suggested that the receptor pathways cooperate in the transcellular 1
basic research
transport of albumin.13 This implies that the megalin/cubilin receptor pathway is responsible for the initial binding, endocytosis, and transport of albumin to the early endosomes. The acidic pH of this compartment causes dissociation of albumin from the megalin/cubilin receptors and binding to the FcRn, mediating transport across the proximal tubule cells and release at the basolateral surface. This hypothesis implies that megalin/cubilin-mediated albumin uptake contributes to the retrieval of intact albumin and thereby preservation of plasma albumin levels. To address this, we investigated proximal tubular albumin uptake and plasma albumin levels when megalin/cubilin-mediated uptake of
K Weyer et al.: Albumin uptake in nephrotic mice
albumin was abolished in normal and nephrotic mice using novel, inducible KO mouse models. RESULTS Generation of inducible nephrotic megalin/cubilin receptor KO mice
Tamoxifen treatment of inducible megalin/cubilin, megalin/ cubilin/podocin, and podocin KO mice resulted in efficient (w85%–92%) inactivation of all the respective genes in the kidney (Figure 1a–c). This was confirmed by immunohistochemistry, showing the almost complete absence of megalin and cubilin staining in proximal tubule cells and podocin
Figure 1 | Megalin/cubilin/podocin gene inactivation in the kidney by quantitative real-time PCR (qRT-PCR) and immunohistochemistry. (a–c) The relationship of megalin, cubilin, and podocin to the actin mRNA ratio from inducible megalin/cubilin knockout (KO) mice, megalin/cubilin/podocin KO mice, and podocin KO mice compared with control mice, determined by qRT-PCR using whole kidney samples. The graphs show the mean values from all mice in each group (N ¼ 8), and error bars indicate SEM. (d–i) Immunohistochemical labeling for megalin, cubilin, and podocin in kidney sections from control, inducible megalin/cubilin KO, or inducible podocin KO mice, as indicated above the images. (g,h) Arrows point out residual expression of megalin and cubilin in a few mosaic tubules in serial sections from kidneys of inducible megalin/cubilin KO mice. Note that the megalin/cubilin receptors appear to be knocked out together in the proximal tubule cells. (d,e,g,h) Bars ¼ 50 mm. (f,i) Bars ¼ 25 mm. To optimize viewing of this image, please see the online version of this article at www. kidney-international.org. 2
Kidney International (2017) -, -–-
K Weyer et al.: Albumin uptake in nephrotic mice
staining in podocytes in sections from the relevant KO mice compared with controls (Figure 1d–i). The expression levels and distribution of megalin/cubilin in podocin KO mice were comparable to those of controls (Supplementary Figure S1). Podocin-deficient mice develop nephrotic-range albuminuria 3 to 4 weeks post-induction.18,19 Thus, we successfully generated a mouse model with an inducible, massively increased filtration of albumin and a simultaneous efficient knockdown of the proximal tubular albumin receptors megalin and cubilin. Gene KO of the megalin/cubilin receptors blocks albumin uptake in nephrotic mice
The crucial role of the megalin/cubilin receptors in the proximal tubule uptake of albumin was confirmed by immunostaining of kidney sections for albumin. Compared with control mice (Figure 2a), no albumin staining was seen in the proximal tubules of megalin/cubilin KO mice (Figure 2b). Concurrent with the development of albuminuria, podocin KO mice displayed a strongly enhanced albumin accumulation compared with controls (Figure 2c vs. 2a). This was completely abolished in megalin/cubilin/podocin KO mice (Figure 2d) despite the increased load of filtered albumin. Analysis of mosaic KO proximal tubules further emphasized
basic research
the requirement of the megalin/cubilin receptors for albumin uptake in megalin/cubilin/podocin KO mice (Figure 2e). Importantly, the urinary albumin excretion was significantly increased in megalin/cubilin/podocin KO mice compared with podocin KO mice (albumin/creatinine: 225.9 8.7 vs. 157.0 13.8 mg/mg) (Figure 3), reflecting the augmented urinary loss of albumin in the absence of megalin/cubilinmediated albumin uptake. Low-range albuminuria developed in the inducible megalin/cubilin KO mice compared with controls (albumin/creatinine: 1.09 0.08 vs. 0.05 0.02 mg/mg), equivalent to previously reported values using conditional megalin/cubilin KO mice12 and inducible megalin KO mice.20 Deficiency of the megalin/cubilin receptors does not affect plasma albumin levels in nephrotic mice
Consistent with the nephrotic state, plasma albumin levels were reduced both in megalin/cubilin/podocin KO and podocin KO mice compared with controls (24.8 1.0 g/l) as well as with the inducible megalin/cubilin KO mice (24.1 1.3 g/l) (Figure 4). Notably, no difference was observed in plasma albumin levels between megalin/cubilin/podocin KO mice (11.0 0.9 g/l) and podocin KO mice and (9.1 1.0 g/l) (Figure 4). This finding was confirmed by Coomassie
Figure 2 | Tubular albumin uptake is abolished in proximal tubules of megalin (Meg)/cubilin (Cub), and Meg/Cub/podocin (Pod) knockout (KO) mice. Coimmunostaining for megalin (red) and albumin (green) in kidney sections from control mice (a), inducible Meg/Cub KO mice (b), inducible Pod KO mice (c), and inducible megalin/cubilin/podocin KO mice (d). (e) A mosaic megalin-expressing proximal tubule from Meg/Cub/Pod KO mice reveals albumin uptake only in cells expressing megalin. The arrow and white lines indicate the localization of proximal tubule cells lacking megalin. Confocal images were taken at 63 magnification; bars ¼ 10 mm. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org. Kidney International (2017) -, -–-
3
basic research
K Weyer et al.: Albumin uptake in nephrotic mice
levels were increased in both megalin/cubilin/podocin KO and podocin KO mice compared with controls (Figure 5d). Liver albumin synthesis is unaltered in megalin/cubilin deficiency
Figure 3 | Increased urinary albumin excretion in megalin/cubilin/ podocin knockout (KO) mice compared with podocin KO mice. Urinary albumin excretion (albumin/creatinine, mg/mg) in control, megalin/cubilin KO, podocin KO, and megalin/cubilin/podocin KO mice (N ¼ 8 in each group). Error bars indicate SEM. **P < 0.01; ****P < 0.0001.
As the liver is the main site of albumin synthesis, we further measured liver albumin mRNA and protein abundance to exclude any effects of differences in albumin production on the plasma albumin levels. No differences in albumin mRNA levels were observed between megalin/cubilin/podocin KO and podocin KO mice, whereas both were increased by w5-fold compared with megalin/cubilin KO and controls (Figure 6a). Also, we found no differences in liver albumin protein levels between controls, podocin KO, megalin/cubilin/ podocin KO mice (Figure 6b), or megalin/cubilin KO (not shown). These data suggest that the liver albumin synthesis is not affected by the megalin/cubilin deficiency in neither nonnephrotic (megalin/cubilin KO vs. control mice) nor in nephrotic mice (megalin/cubilin/podocin KO vs. podocin KO mice). DISCUSSION
staining of plasma proteins, showing equally reduced albumin levels in podocin KO and megalin/cubilin/podocin KO mice (Figure 5a). The total amount of protein in plasma was nevertheless maintained in the podocin KO mice (49.4 2.4 g/l) and megalin/cubilin/podocin KO mice (50.8 3.2 g/l) compared with controls (51.0 3.5 g/l) (Figure 5b), and the Coomassie stain showed increased levels of distinct protein bands (Figure 5a), suggesting a compensatory increase in several other plasma proteins. Examining the plasma Ig abundance by immunoblotting revealed substantially reduced IgG levels in megalin/cubilin/podocin and podocin KO mice compared with controls (Figure 5c), whereas IgM plasma
Figure 4 | Unaltered plasma albumin levels in megalin/cubilin/ podocin knockout (KO) mice compared with podocin KO mice. Plasma albumin levels (g/l) in control, megalin/cubilin KO mice, podocin KO, and megalin/cubilin/podocin KO mice (N ¼ 8 in each group). Error bars indicate SEM. Not significant (ns) indicates P > 0.05. 4
The mechanisms of renal albumin handling, in particular with regard to the amount of albumin filtered by normal glomeruli and the proposed existence of a high-capacity retrieval pathway in the proximal tubules responsible for maintenance of plasma albumin levels, have been highly debated.21–25 Using new inducible KO mouse models, we identified a 1.4-fold increase in albumin excretion in nephrotic megalin/cubilin KO mice compared with nephrotic podocin KO with preserved tubular endocytosis. The total filtered amount of albumin may be estimated based on the urinary excretion in megalin/cubilin KO mice, assuming no tubular reabsorption in these mice. Based on this assumption, we calculate a glomerular sieving coefficient (GSC) for albumin of 7.5 105, which is increased to 0.08 in the nephrotic mice. Mori et al.20 recently made similar calculations of total albumin filtration by studying experimental, diabetic nephropathy in megalin(lox/lox);Ndrg1-CreERT2 KO mice. In their study, the development of diabetes in megalindeficient mice was associated with a 1.9-fold increase in total albumin filtration compared with diabetic controls. They further estimated the GSC of albumin to be 1.7 105, which was approximately doubled in the diabetic mice.20 Also, the albumin GSC in Dent disease patients, characterized by defective proximal tubular protein reabsorption, has been estimated to be 7.7 105.26 Although multiple factors may affect the calculation of the estimated GSCs (e.g., the contribution of reabsorption by the distal nephron, the stability of albumin in urine, and residual proximal tubular uptake in animal models), these estimated albumin GSCs are consistent with micropuncture measures showing an albumin GSC of 0.0006.27 A study based on 2-photon confocal microscopy found a much higher GSC of 0.0344; however, more recent 2-photon studies, refining several of the technical challenges with this technique, have reported low albumin Kidney International (2017) -, -–-
K Weyer et al.: Albumin uptake in nephrotic mice
basic research
Figure 5 | Analysis of plasma protein levels in megalin (Meg)/cubilin (Cub)/podocin (Pod) knockout (KO) mice compared with Pod KO mice. (a) Visualization of plasma proteins by Coomassie stain of plasma samples (N ¼ 5) from control, Pod KO mice, and Meg/Cub/Pod KO mice. The asterisk indicates the albumin-protein band. (b) Total plasma protein levels (g/l) from control, Meg/Cub KO, Meg/Cub/Pod KO, and Pod KO mice (N ¼ 8 in each group); error bars indicate SEM. ns (not significant) indicates P > 0.05. (c) Western blotting identifying IgG in plasma samples (N ¼ 5) from control, Pod KO, and Meg/Cub/Pod KO mice. HC and LC indicate the IgG heavy and light chain, respectively. The graph shows the mean densities of the IgG HC bands normalized to actin in arbitrary units (AU), relative to the control groups mean (defined as 1). Error bars indicate SEM. **P < 0.01; ****P < 0.0001. (d) Western blotting against IgM in plasma samples (N ¼ 5) from control, inducible Pod KO, and inducible Meg/Cub/Pod KO mice. The graph below shows the mean density of the IgM heavy chain normalized to actin in AU, relative to the mean of the control groups (defined as 1). ns indicates P > 0.05; *P < 0.05. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org.
GSC values of 0.000728 and in the range 0.005 to 0.008.29,30 These findings are in accordance with the notion of restricted albumin filtration in the normal kidney and that glomerular damage induces a dramatic increase in albumin filtration. Our findings further suggest that the capacity of the megalin/cubilin-mediated albumin-reabsorption pathway in nephrotic mouse kidney proximal tubules is w1.9 g/d per kilogram of body weight, estimated from the increase in albumin excretion of megalin/cubilin/podocin KO mice compared with podocin KO mice. Kidney International (2017) -, -–-
A previous study suggested that the megalin/cubilin receptor pathway contributes to the preservation of plasma albumin levels.31 However, we find no evidence in support of this view in our current observations, demonstrating similar plasma albumin levels in podocin KO mice and megalin/ cubilin/podocin KO mice despite complete blockage of albumin reclamation in the latter. The increase in liver albumin synthesis was similar in megalin/cubilin/podocin KO mice compared with podocin KO mice. This suggests that the stimulus for increased albumin synthesis is independent of 5
basic research
K Weyer et al.: Albumin uptake in nephrotic mice
Figure 6 | Liver albumin synthesis is similar in inducible megalin (Meg)/cubilin (Cub)/podocin (Pod) knockout (KO) mice compared with Pod KO mice. (a) The albumin-to-actin mRNA ratio in liver samples from control, Meg/Cub KO, Pod KO, and Meg/Cub/Pod KO mice as determined by quantitative real-time polymerase chain reaction. Values from all mice in each group (N ¼ 8) are shown. Error bars indicate SEM; ns (not significant) indicates P > 0.05. (b) Western blotting identifying albumin and actin in liver samples (N ¼ 5) from control, Pod KO, and Meg/Cub/Pod KO mice. The graph below shows the ratio of the albumin-to-actin band densities in arbitrary units (AU), relative to the mean of the control group (defined as 1). Error bars indicate SEM; ns indicates P > 0.05 versus controls by a 2-tailed unpaired Student t test.
tubular albumin uptake, consistent with the notion that megalin/cubilin-mediated uptake leads to lysosomal degradation rather than retrieval of intact albumin. Extrarenal albumin clearance was not estimated in the study; however, because the KO of megalin and cubilin was essentially kidney specific, it is highly unlikely that the extrarenal clearance of albumin would differ between megalin/cubilin/podocin KO mice and podocin KO mice. In our model, the nephrotic state did not significantly affect megalin and cubilin expression, which has been suggested in other studies.32 Our findings are consistent with a previous analysis of megalin/cubilin levels in podocin KO mice, demonstrating a normal expression level of both receptors by immunoblotting, quantitative real-time polymerase chain reaction (qRT-PCR), and immunohistochemistry at 3 to 5 weeks after induced gene deletion.19 Thus, at the time points studied, megalin and cubilin expression is maintained in the podocin KO mice and appear functional as supported by the heavy tubular accumulation of albumin observed in this study and the increased renal uptake of several megalin/cubilin ligands, including radiolabeled albumin, previously demonstrated in podocin KO mice.19 In addition to albumin, plasma IgG levels were also reduced in the nephrotic mice, consistent with increased urinary excretion due to disruption of the glomerular permselectivity as a result of podocin deficiency. The decrease in IgG levels was slightly greater in podocin KO mice compared with megalin/ cubilin/podocin KO mice. Similar to albumin, this suggests that the proximal tubule megalin/cubilin complex is not essential for maintaining IgG levels during the nephrotic syndrome. In contrast, plasma IgM was markedly increased in both megalin/ cubilin/podocin KO and podocin KO mice to counteract the loss of albumin, which corresponds to observations made in nephrotic syndrome patients.33 In conclusion, this study demonstrates that the albumin uptake by the proximal tubule is effectively abolished by 6
inactivation of the megalin/cubilin receptor complex in an inducible nephrotic mouse model. Despite heavy albuminuria, plasma albumin levels are not affected by the megalin/ cubilin KO, suggesting no significant megalin/cubilindependent tubular retrieval pathway for intact albumin. METHODS Animal model Meglox/lox,12 Cubnlox/lox,12 and Nph2lox/lox,18 mice were crossed with tamoxifen-inducible UBC-cre/ERT2 transgenic mice (The Jackson Laboratory, Bar Harbor, ME). Tail DNA was analyzed by PCR for genotyping.12,19 All mice were of a mixed C57BL/6-129/Svj background. At age 8 to 12 weeks, female inducible megalin/cubilin KO (Meglox/lox,Cubnlox/lox;Creþ), podocin KO (Nph2lox,lox;Creþ), megalin/cubilin/podocin KO (Meglox/lox,Cubnlox/lox,Nph2lox,lox;Creþ), and littermate control mice (Meglox/lox,Cubnlox/lox, Nph2lox,lox;Cre-), were induced by i.p. injection of tamoxifen (Sigma-Aldrich, St. Louis, MO) at a dose of 33 mg/kg body weight for 5 consecutive days. The knockdown of the target genes in the kidney was assessed by quantitative real-time PCR and immunohistochemistry (see the following) at weeks 3 and 5 of the experiment (counting from the first day of tamoxifen induction), with similar results. Mouse breeding and experiments were carried out in a certified animal facility according to the Danish Animal Experiments Inspectorate. Plasma, urine, and tissue samples Urine was collected at week 5 of the experiment by housing the mice in metabolic cages for 24 hours. Urine samples were collected in tubes containing a protease inhibitor mix (Complete; Roche, Basel, Switzerland). The urine was centrifuged at 1000g for 5 minutes and stored at 80 C. The day after urine collection, the mice were anesthetized using isoflurane, and a blood sample was collected from the inferior cava vein in heparin-containing tubes, followed by the removal of 1 kidney and a section of liver tissue, which were snap frozen in liquid nitrogen and stored at 80 C for preparation of tissue homogenates. The remaining kidney was fixed by retrograde perfusion through the abdominal aorta with 2% paraformaldehyde Kidney International (2017) -, -–-
basic research
K Weyer et al.: Albumin uptake in nephrotic mice
in 0.1 mmol/l sodium cacodylate buffer, pH 7.4. The tissue was subsequently dehydrated and embedded in paraffin by standard methods. Blood plasma was prepared by centrifugation at 2000g for 10 minutes at 4 C. Plasma samples were analyzed for creatinine, albumin, and total protein at MRC Harwell (Oxfordshire, UK). The kidney and liver tissues were homogenized in ice-cold buffer containing 0.3 mmol/l sucrose, 25 mmol/l imidazole, and 1 mmol/l ethylenediamine tetraacetic acid (pH 7.2) with a protease inhibitor mix (complete Roche, Sigma-Aldrich, St Louis, MO) for 30 s with an IKA Ultra-turrax T8 homogenizer (IKA Werke, Staufen, Germany), centrifuged at 4000g for 15 minutes at 4 C, and the supernatants were stored at 80 C. Antibodies Primary antibodies used were sheep anti–rat megalin,34 rabbit anti–rat cubilin,35 rabbit anti–podocin,36 rabbit anti–human albumin (A0001, DAKO, Glostrup, Denmark), rabbit anti–actin (A5060, SigmaAldrich), chicken anti–b-actin (AB13822, Abcam, Cambridge, MA), and goat anti–murine IgM (AB9167, Abcam). Secondary antibodies used for immunoblotting were IRDye 800CW goat anti–rabbit IgG (926-32211, LI-COR, Biosciences, Cambridge, UK), IRDye 680CW donkey anti-chicken IgG (926-68075, LI-COR), and IRDye 800CW donkey anti–goat (926-32214, LI-COR). IgG was visualized by immunoblotting using IRDye 680CW goat anti–mouse IgG (92632220, LI-COR). Secondary antibodies used for immunohistochemistry were donkey anti–rabbit Alexa Fluor 488 (21206, Invitrogen, Carlsbad, CA), donkey anti–sheep Alexa Fluor 546 (21098, Invitrogen), horseradish peroxidase–conjugated goat anti–rabbit (P448, DAKO), and rabbit anti–sheep Ig-horseradish peroxidase (P163, DAKO). The specificity of the rabbit anti–human albumin antibody (A0001, DAKO) was validated by immunoblotting using purified mouse serum albumin (Sigma-Aldrich A3559) (not shown). Urinary albumin/creatinine measurements Urine albumin concentrations were determined using a mouse albumin ELISA Quantification kit (Bethyl Labs, Montgomery, TX), according to the manufacturer’s guidelines. Urine samples were analyzed for creatinine at MRC Harwell. Calculation of albumin GSC estimates Based on the assumption that the urinary excretion in megalin/ cubilin-deficient mice reflects the total filtered amount of albumin, the GSCs of albumin in normal and nephrotic mice were estimated as the urinary albumin-to-creatinine ratio divided by plasma albumin-to-creatinine ratio,20,26 in megalin/cubilin KO and megalin/ cubilin/podocin KO mice, respectively. Sodium dodecylsulfate–polyacrylamide gel electrophoresis and immunoblotting Liver samples were prepared by mixing homogenate (10 mg) with 4 sodium dodecylsulfate buffer (Invitrogen) containing 20 mM dithiothreitol (final concentration). Plasma samples were prepared by mixing 0.125 ml plasma with 4 sodium dodecylsulfate buffer, containing 20 mM dithiothreitol (final). Samples were separated on 12% Bis-Tris NuPage Gels (Invitrogen) in 1 MOPS buffer (Invitrogen). For protein-staining, gels were incubated with Coomassie dye (GelCode Blue Stain, Thermo Fisher Scientific, Rockford, IL), according to the manufacturer’s guidelines. For immunoblotting, the proteins were transferred to a polyvinylidene defluoride membrane (Immobilon-P, Merck Millipore, Darmstadt, Germany) and blocked in Odyssey blocking buffer (LI-COR) for 1 hour at room Kidney International (2017) -, -–-
temperature. The membrane was incubated overnight at 4 C with primary antibody and subsequently washed 3 times in phosphatebuffered saline with Tween. Incubation with a secondary antibody was performed at room temperature for 1 hour. Before exposure on the LI-COR Odyssey Infrared Imaging System (LI-COR, Biosciences, Cambridge, UK), the membrane was washed 3 times in phosphatebuffered saline with Tween and once in phosphate-buffered saline. Band intensities were quantified using Odyssey software (version 1.2; LI-COR). Immunohistochemistry Tissue preparation, sectioning, and labeling were performed as described previously.19 For immunoperoxidase labeling, renal paraffin sections were incubated with primary antibodies at 4 C overnight, followed by incubation with horseradish peroxidase– conjugated secondary antibody for 1 hour at room temperature. Peroxidase labeling was visualized by incubation with diaminobenzidine and 0.03% H2O2 for 10 minutes, and sections were counterstained with Mayer’s hematoxylin (Sigma-Aldrich). Immunoperoxidase-labeled images were acquired with a Leica DMR microscope equipped with a Leica DFC320 camera (Leica, Wetzlar, Germany). For immunofluorescent labeling, renal paraffin sections were incubated with primary antibodies at 4 C overnight, followed by incubation with secondary antibodies for 1 hour at room temperature. Immunofluorescent images were taken using a confocal laser-scanning microscope (LSM 510-META, Carl Zeiss GmbH, Jena, Germany) and processed using Zeiss Zen software (2009, Light Edition). All microscope and camera settings were kept identical when obtaining images from different groups. Quantitative real-time polymerase chain reaction Total RNA was purified from liver or kidney tissue using RNeasy Mini Kit (Qiagen, Hilden, Germany). The purity and concentration were determined by optical density at wavelengths of 260 and 280 nm, respectively. Purified RNA (20–100 ng) was incubated in a reverse transcriptase reaction mixture as previously described.19 A reaction without a reverse transcriptase enzyme was made in parallel to ascertain that the amplification occurred on RNA and not DNA. Realtime PCR was performed using a LightCycler480 (Roche) and SYBR Green I Master (Roche), according to the manufacturer. The primers used are shown in Supplementary Table S1. A standard curve, a calibration control, and negative controls were included in each run. All mRNA analyzed was normalized according to actin mRNA in the same sample and calibrated according to the calibrator (calibrator normalized relative quantification). All samples were run in duplicate. Data analysis and statistics All data were expressed as mean SEM. The Kolmogorov-Smirnov test (when N ¼ 5 per group) or the d’Agostino-Pearson test (when N ¼ 8 per group) was used to analyze for normality of the data. Data were compared between groups using an unpaired, 2-tailed, Student t test. Values of P < 0.05 were considered significant. All analyses were performed using GraphPad Prism 7.0 for Windows (GraphPad Software, San Diego, CA). DISCLOSURE
All the authors declared no competing interests. ACKNOWLEDGMENTS
We thank Inger Blenker Kristoffersen, Helle Salling Gittins, Hanne Sidelmann, and Pia Kamuk Nielsen for excellent technical assistance. 7
basic research
Sources of support for this work included grants from the Danish Medical Research Council, The Lundbeck Foundation, and the Novo Nordisk Foundation. SUPPLEMENTARY MATERIAL Figure S1. Megalin and cubilin localization and mRNA levels in the kidney of podocin knockout mice. Representative images showing the immunohistochemical labeling for megalin (A) and cubilin (B) in kidney sections from podocin knockout (KO) mice. Bars ¼ 50 mm. The relative megalin (C) and cubilin (D) to actin mRNA ratio in whole kidney samples from podocin KO compared with control (Cre-) mice, determined by quantitative real-time polymerase chain reaction. The graphs represent mean values from all mice in each group (N ¼ 8), and error bars indicate SEM. ns, not significant (P > 0.05) versus controls. Table S1. Primers used for quantitative real-time polymerase chain reaction. Supplementary material is linked to the online version of the paper at www.kidney-international.org. REFERENCES 1. Lambers Heerspink HJ, Brinkman JW, Bakker SJ, et al. Update on microalbuminuria as a biomarker in renal and cardiovascular disease. Curr Opin Nephrol Hypertens. 2006;15:631–636. 2. Remuzzi G, Ruggenenti P, Benigni A. Understanding the nature of renal disease progression. Kidney Int. 1997;51:2–15. 3. Haraldsson B, Nystrom J, Deen WM. Properties of the glomerular barrier and mechanisms of proteinuria. Physiol Rev. 2008;88:451–487. 4. Russo LM, Sandoval RM, McKee M, et al. The normal kidney filters nephrotic levels of albumin retrieved by proximal tubule cells: retrieval is disrupted in nephrotic states. Kidney Int. 2007;71:504–513. 5. Comper WD. Albuminuria is controlled primarily by proximal tubules. Nat Rev Nephrol. 2014;10:180. 6. Cui S, Verroust PJ, Moestrup SK, et al. Megalin/gp330 mediates uptake of albumin in renal proximal tubule. Am J Physiol. 1996;271:F900–F907. 7. Amsellem S, Gburek J, Hamard G, et al. Cubilin is essential for albumin reabsorption in the renal proximal tubule. J Am Soc Nephrol. 2010;21: 1859–1867. 8. Birn H, Fyfe JC, Jacobsen C, et al. Cubilin is an albumin binding protein important for renal tubular albumin reabsorption. J Clin Invest. 2000;105: 1353–1361. 9. Storm T, Emma F, Verroust PJ, et al. A patient with cubilin deficiency. N Engl J Med. 2011;364:89–91. 10. Storm T, Tranebjaerg L, Frykholm C, et al. Renal phenotypic investigations of megalin-deficient patients: novel insights into tubular proteinuria and albumin filtration. Nephrol Dial Transplant. 2013;28:585–591. 11. Weyer K, Nielsen R, Christensen EI, et al. Generation of urinary albumin fragments does not require proximal tubular uptake. J Am Soc Nephrol. 2012;23:591–596. 12. Weyer K, Storm T, Shan J, et al. Mouse model of proximal tubule endocytic dysfunction. Nephrol Dial Transplant. 2011;26:3446–3451. 13. Tenten V, Menzel S, Kunter U, et al. Albumin is recycled from the primary urine by tubular transcytosis. J Am Soc Nephrol. 2013;24:1966–1980. 14. Sandoval RM, Wagner MC, Patel M, et al. Multiple factors influence glomerular albumin permeability in rats. J Am Soc Nephrol. 2012;23:447–457.
8
K Weyer et al.: Albumin uptake in nephrotic mice
15. Kim J, Bronson CL, Hayton WL, et al. Albumin turnover: FcRn-mediated recycling saves as much albumin from degradation as the liver produces. Am J Physiol Gastrointest Liver Physiol. 2006;290:G352–G360. 16. Sarav M, Wang Y, Hack BK, et al. Renal FcRn reclaims albumin but facilitates elimination of IgG. J Am Soc Nephrol. 2009;20:1941–1952. 17. Chaudhury C, Brooks CL, Carter DC, et al. Albumin binding to FcRn: distinct from the FcRn-IgG interaction. Biochemistry. 2006;45: 4983–4990. 18. Mollet G, Ratelade J, Boyer O, et al. Podocin inactivation in mature kidneys causes focal segmental glomerulosclerosis and nephrotic syndrome. J Am Soc Nephrol. 2009;20:2181–2189. 19. Nielsen R, Mollet G, Esquivel EL, et al. Increased lysosomal proteolysis counteracts protein accumulation in the proximal tubule during focal segmental glomerulosclerosis. Kidney Int. 2013;84:902–910. 20. Mori KP, Yokoi H, Kasahara M, et al. Increase of Total Nephron Albumin Filtration and Reabsorption in Diabetic Nephropathy. J Am Soc Nephrol. 2017;28:278–289. 21. Russo LM, Sandoval RM, Brown D, et al. Controversies in nephrology: response to ’renal albumin handling, facts, and artifacts’. Kidney Int. 2007;72:1195–1197. 22. Christensen EI, Birn H, Rippe B, et al. Controversies in nephrology: renal albumin handling, facts, and artifacts! Kidney Int. 2007;72:1192–1194. 23. Remuzzi A, Sangalli F, Fassi A, et al. Albumin concentration in the Bowman’s capsule: multiphoton microscopy vs micropuncture technique. Kidney Int. 2007;72:1410–1411. author reply, 1411. 24. Norden AG, Unwin RJ. Is the albumin retrieval hypothesis a paradigm shift for nephrology? J Am Soc Nephrol. 2012;23:569–571. 25. Dickson LE, Wagner MC, Sandoval RM, et al. The proximal tubule and albuminuria: really! J Am Soc Nephrol. 2014;25:443–453. 26. Norden AG, Lapsley M, Lee PJ, et al. Glomerular protein sieving and implications for renal failure in Fanconi syndrome. Kidney Int. 2001;60: 1885–1892. 27. Tojo A, Endou H. Intrarenal handling of proteins in rats using fractional micropuncture technique. Am J Physiol. 1992;263:F601–F606. 28. Schiessl IM, Hammer A, Kattler V, et al. Intravital Imaging Reveals Angiotensin II-Induced Transcytosis of Albumin by Podocytes. J Am Soc Nephrol. 2016;27:731–744. 29. Nakano D, Kobori H, Burford JL, et al. Multiphoton imaging of the glomerular permeability of angiotensinogen. J Am Soc Nephrol. 2012;23: 1847–1856. 30. Wagner MC, Campos-Bilderback SB, Chowdhury M, et al. Proximal Tubules Have the Capacity to Regulate Uptake of Albumin. J Am Soc Nephrol. 2016;27:482–494. 31. Aseem O, Smith BT, Cooley MA, et al. Cubilin maintains blood levels of HDL and albumin. J Am Soc Nephrol. 2014;25:1028–1036. 32. Caruso-Neves C, Pinheiro AA, Cai H, et al. PKB and megalin determine the survival or death of renal proximal tubule cells. Proc Natl Acad Sci U S A. 2006;103:18810–18815. 33. Kaysen GA, al Bander H. Metabolism of albumin and immunoglobulins in the nephrotic syndrome. Am J Nephrol. 1990;10(Suppl 1):36–42. 34. Moestrup SK, Nielsen S, Andreasen P, et al. Epithelial glycoprotein-330 mediates endocytosis of plasminogen activator-plasminogen activator inhibitor type-1 complexes. J Biol Chem. 1993;268:16564–16570. 35. Sahali D, Mulliez N, Chatelet F, et al. Comparative immunochemistry and ontogeny of two closely related coated pit proteins. The 280-kd target of teratogenic antibodies and the 330-kd target of nephritogenic antibodies. Am J Pathol. 1993;142:1654–1667. 36. Roselli S, Gribouval O, Boute N, et al. Podocin localizes in the kidney to the slit diaphragm area. Am J Pathol. 2002;160:131–139.
Kidney International (2017) -, -–-
www.kidney-international.org
Abolishment of proximal tubule albumin endocytosis does not affect plasma albumin during nephrotic syndrome in mice Kathrin Weyer1, Pia K. Andersen1, Kasper Schmidt1, Geraldine Mollet2,3, Corinne Antignac2,3,4, Henrik Birn1,5, Rikke Nielsen1 and Erik I. Christensen1 1
Department of Biomedicine, Aarhus University, Aarhus, Denmark; 2Inserm U1163, Imagine Institute, Laboratory of Hereditary Kidney Diseases, Paris, France; 3Paris Descartes-Sorbonne Paris Cité University, Paris, France; 4Department of Genetics, Necker Hospital, Assistance Publique-Hôpitaux de Paris, Paris, France; and 5Department of Renal Medicine, Aarhus University Hospital, Aarhus, Denmark
The megalin/cubilin receptor complex is required for proximal tubular endocytosis and degradation of filtered albumin. An additional high-capacity retrieval pathway of intact albumin for the recovery of large amounts of filtered albumin has been proposed, possibly involving cooperation between megalin/cubilin and the neonatal Fc receptor. To clarify the potential role of such a pathway, we examined the effects of megalin/cubilin gene inactivation on tubular albumin uptake and plasma albumin levels in nephrotic, podocin knockout mice. Immunofluorescence microscopy of megalin/cubilin/podocin knockout mouse kidneys demonstrated abolishment of proximal tubule albumin uptake, in contrast to the excessive albumin accumulation observed in podocin knockout mice compared to controls. Correspondingly, urinary albumin excretion was increased 1.4 fold in megalin/cubilin/podocin compared to podocin knockout mice (albumin/creatinine: 226 vs. 157 mg/mg). However, no difference in plasma albumin levels was observed between megalin/cubilin/ podocin and podocin knockout mice, as both were reduced to approximately 40% of controls. There were no differences in liver albumin synthesis by mRNA levels and protein abundance. Thus, megalin/cubilin knockout efficiently blocks proximal tubular albumin uptake in nephrotic mice but plasma albumin levels did not differ as a result of megalin/cubilin-deficiency, suggesting no significance of the megalin/cubilin-pathway for albumin homeostasis by retrieval of intact albumin. Kidney International (2017) j.kint.2017.07.024
-, -–-;
http://dx.doi.org/10.1016/
KEYWORDS: albuminuria; endocytosis; nephrotic syndrome; proteinuria; proximal tubule Copyright ª 2017, International Society of Nephrology. Published by Elsevier Inc. All rights reserved.
Correspondence: Kathrin Weyer, Department of Biomedicine, Aarhus University, Wilhelm Meyers Allé 3, Building 1234, DK-8000, Aarhus C, Denmark. E-mail: [email protected] Received 11 June 2017; revised 11 July 2017; accepted 27 July 2017
Kidney International (2017) -, -–-
T
he renal handling of albumin, the quantitatively most important plasma protein, has been rigorously investigated. Increased urinary excretion of albumin, i.e., albuminuria, is an early manifestation and important prognostic marker in various forms of kidney disease.1 Nevertheless, several aspects of renal albumin handling remain controversial, in particular with regard to the amount of albumin filtered by the normal glomeruli and its route of reabsorption, degradation, and potential transcytosis in the proximal tubule cells. Understanding the contribution of these pathways is important for the elucidation of the pathophysiological mechanisms of albuminuria and its potentially harmful role in the progression of kidney diseases.2 The prevailing view is that normal glomerular filtration of albumin is limited by the properties of the glomerular filtration barrier to 0.5 to 2 g/d in humans and that the filtered albumin is efficiently reabsorbed and degraded by the proximal tubule.3 This perception has been challenged by the proposal that large amounts of albumin are normally filtered in the kidney (w225 g/d) and that a high-capacity retrieval pathway in the proximal tubule returns albumin to the circulation to maintain plasma albumin concentrations.4,5 The proximal tubule cell is responsible for renal reabsorption of all filtered proteins, including albumin, primarily mediated by the megalin/cubilin receptor complex.6–12 Following megalin/cubilin-mediated endocytosis, albumin is targeted for lysosomal degradation, whereas the receptors recycle to the apical plasma membrane. It has been suggested that albumin can also be rescued by the neonatal Fc receptor (FcRn), leading to transcytosis and the return of intact albumin to the circulation.13,14 As FcRn knockout (KO) mice develop hypoalbuminemia,15 FcRn-mediated transcytosis of albumin has been claimed to have a crucial role in maintaining normal plasma albumin levels.16 FcRn binds albumin with higher affinity at low pH (pH 6.1),17 allowing it to bind albumin in the acidic environment of the endosome and releasing it at a higher pH in the extracellular environment. Because the pH of the proximal tubule lumen does not facilitate albumin binding to FcRn, it was further suggested that the receptor pathways cooperate in the transcellular 1
basic research
transport of albumin.13 This implies that the megalin/cubilin receptor pathway is responsible for the initial binding, endocytosis, and transport of albumin to the early endosomes. The acidic pH of this compartment causes dissociation of albumin from the megalin/cubilin receptors and binding to the FcRn, mediating transport across the proximal tubule cells and release at the basolateral surface. This hypothesis implies that megalin/cubilin-mediated albumin uptake contributes to the retrieval of intact albumin and thereby preservation of plasma albumin levels. To address this, we investigated proximal tubular albumin uptake and plasma albumin levels when megalin/cubilin-mediated uptake of
K Weyer et al.: Albumin uptake in nephrotic mice
albumin was abolished in normal and nephrotic mice using novel, inducible KO mouse models. RESULTS Generation of inducible nephrotic megalin/cubilin receptor KO mice
Tamoxifen treatment of inducible megalin/cubilin, megalin/ cubilin/podocin, and podocin KO mice resulted in efficient (w85%–92%) inactivation of all the respective genes in the kidney (Figure 1a–c). This was confirmed by immunohistochemistry, showing the almost complete absence of megalin and cubilin staining in proximal tubule cells and podocin
Figure 1 | Megalin/cubilin/podocin gene inactivation in the kidney by quantitative real-time PCR (qRT-PCR) and immunohistochemistry. (a–c) The relationship of megalin, cubilin, and podocin to the actin mRNA ratio from inducible megalin/cubilin knockout (KO) mice, megalin/cubilin/podocin KO mice, and podocin KO mice compared with control mice, determined by qRT-PCR using whole kidney samples. The graphs show the mean values from all mice in each group (N ¼ 8), and error bars indicate SEM. (d–i) Immunohistochemical labeling for megalin, cubilin, and podocin in kidney sections from control, inducible megalin/cubilin KO, or inducible podocin KO mice, as indicated above the images. (g,h) Arrows point out residual expression of megalin and cubilin in a few mosaic tubules in serial sections from kidneys of inducible megalin/cubilin KO mice. Note that the megalin/cubilin receptors appear to be knocked out together in the proximal tubule cells. (d,e,g,h) Bars ¼ 50 mm. (f,i) Bars ¼ 25 mm. To optimize viewing of this image, please see the online version of this article at www. kidney-international.org. 2
Kidney International (2017) -, -–-
K Weyer et al.: Albumin uptake in nephrotic mice
staining in podocytes in sections from the relevant KO mice compared with controls (Figure 1d–i). The expression levels and distribution of megalin/cubilin in podocin KO mice were comparable to those of controls (Supplementary Figure S1). Podocin-deficient mice develop nephrotic-range albuminuria 3 to 4 weeks post-induction.18,19 Thus, we successfully generated a mouse model with an inducible, massively increased filtration of albumin and a simultaneous efficient knockdown of the proximal tubular albumin receptors megalin and cubilin. Gene KO of the megalin/cubilin receptors blocks albumin uptake in nephrotic mice
The crucial role of the megalin/cubilin receptors in the proximal tubule uptake of albumin was confirmed by immunostaining of kidney sections for albumin. Compared with control mice (Figure 2a), no albumin staining was seen in the proximal tubules of megalin/cubilin KO mice (Figure 2b). Concurrent with the development of albuminuria, podocin KO mice displayed a strongly enhanced albumin accumulation compared with controls (Figure 2c vs. 2a). This was completely abolished in megalin/cubilin/podocin KO mice (Figure 2d) despite the increased load of filtered albumin. Analysis of mosaic KO proximal tubules further emphasized
basic research
the requirement of the megalin/cubilin receptors for albumin uptake in megalin/cubilin/podocin KO mice (Figure 2e). Importantly, the urinary albumin excretion was significantly increased in megalin/cubilin/podocin KO mice compared with podocin KO mice (albumin/creatinine: 225.9 8.7 vs. 157.0 13.8 mg/mg) (Figure 3), reflecting the augmented urinary loss of albumin in the absence of megalin/cubilinmediated albumin uptake. Low-range albuminuria developed in the inducible megalin/cubilin KO mice compared with controls (albumin/creatinine: 1.09 0.08 vs. 0.05 0.02 mg/mg), equivalent to previously reported values using conditional megalin/cubilin KO mice12 and inducible megalin KO mice.20 Deficiency of the megalin/cubilin receptors does not affect plasma albumin levels in nephrotic mice
Consistent with the nephrotic state, plasma albumin levels were reduced both in megalin/cubilin/podocin KO and podocin KO mice compared with controls (24.8 1.0 g/l) as well as with the inducible megalin/cubilin KO mice (24.1 1.3 g/l) (Figure 4). Notably, no difference was observed in plasma albumin levels between megalin/cubilin/podocin KO mice (11.0 0.9 g/l) and podocin KO mice and (9.1 1.0 g/l) (Figure 4). This finding was confirmed by Coomassie
Figure 2 | Tubular albumin uptake is abolished in proximal tubules of megalin (Meg)/cubilin (Cub), and Meg/Cub/podocin (Pod) knockout (KO) mice. Coimmunostaining for megalin (red) and albumin (green) in kidney sections from control mice (a), inducible Meg/Cub KO mice (b), inducible Pod KO mice (c), and inducible megalin/cubilin/podocin KO mice (d). (e) A mosaic megalin-expressing proximal tubule from Meg/Cub/Pod KO mice reveals albumin uptake only in cells expressing megalin. The arrow and white lines indicate the localization of proximal tubule cells lacking megalin. Confocal images were taken at 63 magnification; bars ¼ 10 mm. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org. Kidney International (2017) -, -–-
3
basic research
K Weyer et al.: Albumin uptake in nephrotic mice
levels were increased in both megalin/cubilin/podocin KO and podocin KO mice compared with controls (Figure 5d). Liver albumin synthesis is unaltered in megalin/cubilin deficiency
Figure 3 | Increased urinary albumin excretion in megalin/cubilin/ podocin knockout (KO) mice compared with podocin KO mice. Urinary albumin excretion (albumin/creatinine, mg/mg) in control, megalin/cubilin KO, podocin KO, and megalin/cubilin/podocin KO mice (N ¼ 8 in each group). Error bars indicate SEM. **P < 0.01; ****P < 0.0001.
As the liver is the main site of albumin synthesis, we further measured liver albumin mRNA and protein abundance to exclude any effects of differences in albumin production on the plasma albumin levels. No differences in albumin mRNA levels were observed between megalin/cubilin/podocin KO and podocin KO mice, whereas both were increased by w5-fold compared with megalin/cubilin KO and controls (Figure 6a). Also, we found no differences in liver albumin protein levels between controls, podocin KO, megalin/cubilin/ podocin KO mice (Figure 6b), or megalin/cubilin KO (not shown). These data suggest that the liver albumin synthesis is not affected by the megalin/cubilin deficiency in neither nonnephrotic (megalin/cubilin KO vs. control mice) nor in nephrotic mice (megalin/cubilin/podocin KO vs. podocin KO mice). DISCUSSION
staining of plasma proteins, showing equally reduced albumin levels in podocin KO and megalin/cubilin/podocin KO mice (Figure 5a). The total amount of protein in plasma was nevertheless maintained in the podocin KO mice (49.4 2.4 g/l) and megalin/cubilin/podocin KO mice (50.8 3.2 g/l) compared with controls (51.0 3.5 g/l) (Figure 5b), and the Coomassie stain showed increased levels of distinct protein bands (Figure 5a), suggesting a compensatory increase in several other plasma proteins. Examining the plasma Ig abundance by immunoblotting revealed substantially reduced IgG levels in megalin/cubilin/podocin and podocin KO mice compared with controls (Figure 5c), whereas IgM plasma
Figure 4 | Unaltered plasma albumin levels in megalin/cubilin/ podocin knockout (KO) mice compared with podocin KO mice. Plasma albumin levels (g/l) in control, megalin/cubilin KO mice, podocin KO, and megalin/cubilin/podocin KO mice (N ¼ 8 in each group). Error bars indicate SEM. Not significant (ns) indicates P > 0.05. 4
The mechanisms of renal albumin handling, in particular with regard to the amount of albumin filtered by normal glomeruli and the proposed existence of a high-capacity retrieval pathway in the proximal tubules responsible for maintenance of plasma albumin levels, have been highly debated.21–25 Using new inducible KO mouse models, we identified a 1.4-fold increase in albumin excretion in nephrotic megalin/cubilin KO mice compared with nephrotic podocin KO with preserved tubular endocytosis. The total filtered amount of albumin may be estimated based on the urinary excretion in megalin/cubilin KO mice, assuming no tubular reabsorption in these mice. Based on this assumption, we calculate a glomerular sieving coefficient (GSC) for albumin of 7.5 105, which is increased to 0.08 in the nephrotic mice. Mori et al.20 recently made similar calculations of total albumin filtration by studying experimental, diabetic nephropathy in megalin(lox/lox);Ndrg1-CreERT2 KO mice. In their study, the development of diabetes in megalindeficient mice was associated with a 1.9-fold increase in total albumin filtration compared with diabetic controls. They further estimated the GSC of albumin to be 1.7 105, which was approximately doubled in the diabetic mice.20 Also, the albumin GSC in Dent disease patients, characterized by defective proximal tubular protein reabsorption, has been estimated to be 7.7 105.26 Although multiple factors may affect the calculation of the estimated GSCs (e.g., the contribution of reabsorption by the distal nephron, the stability of albumin in urine, and residual proximal tubular uptake in animal models), these estimated albumin GSCs are consistent with micropuncture measures showing an albumin GSC of 0.0006.27 A study based on 2-photon confocal microscopy found a much higher GSC of 0.0344; however, more recent 2-photon studies, refining several of the technical challenges with this technique, have reported low albumin Kidney International (2017) -, -–-
K Weyer et al.: Albumin uptake in nephrotic mice
basic research
Figure 5 | Analysis of plasma protein levels in megalin (Meg)/cubilin (Cub)/podocin (Pod) knockout (KO) mice compared with Pod KO mice. (a) Visualization of plasma proteins by Coomassie stain of plasma samples (N ¼ 5) from control, Pod KO mice, and Meg/Cub/Pod KO mice. The asterisk indicates the albumin-protein band. (b) Total plasma protein levels (g/l) from control, Meg/Cub KO, Meg/Cub/Pod KO, and Pod KO mice (N ¼ 8 in each group); error bars indicate SEM. ns (not significant) indicates P > 0.05. (c) Western blotting identifying IgG in plasma samples (N ¼ 5) from control, Pod KO, and Meg/Cub/Pod KO mice. HC and LC indicate the IgG heavy and light chain, respectively. The graph shows the mean densities of the IgG HC bands normalized to actin in arbitrary units (AU), relative to the control groups mean (defined as 1). Error bars indicate SEM. **P < 0.01; ****P < 0.0001. (d) Western blotting against IgM in plasma samples (N ¼ 5) from control, inducible Pod KO, and inducible Meg/Cub/Pod KO mice. The graph below shows the mean density of the IgM heavy chain normalized to actin in AU, relative to the mean of the control groups (defined as 1). ns indicates P > 0.05; *P < 0.05. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org.
GSC values of 0.000728 and in the range 0.005 to 0.008.29,30 These findings are in accordance with the notion of restricted albumin filtration in the normal kidney and that glomerular damage induces a dramatic increase in albumin filtration. Our findings further suggest that the capacity of the megalin/cubilin-mediated albumin-reabsorption pathway in nephrotic mouse kidney proximal tubules is w1.9 g/d per kilogram of body weight, estimated from the increase in albumin excretion of megalin/cubilin/podocin KO mice compared with podocin KO mice. Kidney International (2017) -, -–-
A previous study suggested that the megalin/cubilin receptor pathway contributes to the preservation of plasma albumin levels.31 However, we find no evidence in support of this view in our current observations, demonstrating similar plasma albumin levels in podocin KO mice and megalin/ cubilin/podocin KO mice despite complete blockage of albumin reclamation in the latter. The increase in liver albumin synthesis was similar in megalin/cubilin/podocin KO mice compared with podocin KO mice. This suggests that the stimulus for increased albumin synthesis is independent of 5
basic research
K Weyer et al.: Albumin uptake in nephrotic mice
Figure 6 | Liver albumin synthesis is similar in inducible megalin (Meg)/cubilin (Cub)/podocin (Pod) knockout (KO) mice compared with Pod KO mice. (a) The albumin-to-actin mRNA ratio in liver samples from control, Meg/Cub KO, Pod KO, and Meg/Cub/Pod KO mice as determined by quantitative real-time polymerase chain reaction. Values from all mice in each group (N ¼ 8) are shown. Error bars indicate SEM; ns (not significant) indicates P > 0.05. (b) Western blotting identifying albumin and actin in liver samples (N ¼ 5) from control, Pod KO, and Meg/Cub/Pod KO mice. The graph below shows the ratio of the albumin-to-actin band densities in arbitrary units (AU), relative to the mean of the control group (defined as 1). Error bars indicate SEM; ns indicates P > 0.05 versus controls by a 2-tailed unpaired Student t test.
tubular albumin uptake, consistent with the notion that megalin/cubilin-mediated uptake leads to lysosomal degradation rather than retrieval of intact albumin. Extrarenal albumin clearance was not estimated in the study; however, because the KO of megalin and cubilin was essentially kidney specific, it is highly unlikely that the extrarenal clearance of albumin would differ between megalin/cubilin/podocin KO mice and podocin KO mice. In our model, the nephrotic state did not significantly affect megalin and cubilin expression, which has been suggested in other studies.32 Our findings are consistent with a previous analysis of megalin/cubilin levels in podocin KO mice, demonstrating a normal expression level of both receptors by immunoblotting, quantitative real-time polymerase chain reaction (qRT-PCR), and immunohistochemistry at 3 to 5 weeks after induced gene deletion.19 Thus, at the time points studied, megalin and cubilin expression is maintained in the podocin KO mice and appear functional as supported by the heavy tubular accumulation of albumin observed in this study and the increased renal uptake of several megalin/cubilin ligands, including radiolabeled albumin, previously demonstrated in podocin KO mice.19 In addition to albumin, plasma IgG levels were also reduced in the nephrotic mice, consistent with increased urinary excretion due to disruption of the glomerular permselectivity as a result of podocin deficiency. The decrease in IgG levels was slightly greater in podocin KO mice compared with megalin/ cubilin/podocin KO mice. Similar to albumin, this suggests that the proximal tubule megalin/cubilin complex is not essential for maintaining IgG levels during the nephrotic syndrome. In contrast, plasma IgM was markedly increased in both megalin/ cubilin/podocin KO and podocin KO mice to counteract the loss of albumin, which corresponds to observations made in nephrotic syndrome patients.33 In conclusion, this study demonstrates that the albumin uptake by the proximal tubule is effectively abolished by 6
inactivation of the megalin/cubilin receptor complex in an inducible nephrotic mouse model. Despite heavy albuminuria, plasma albumin levels are not affected by the megalin/ cubilin KO, suggesting no significant megalin/cubilindependent tubular retrieval pathway for intact albumin. METHODS Animal model Meglox/lox,12 Cubnlox/lox,12 and Nph2lox/lox,18 mice were crossed with tamoxifen-inducible UBC-cre/ERT2 transgenic mice (The Jackson Laboratory, Bar Harbor, ME). Tail DNA was analyzed by PCR for genotyping.12,19 All mice were of a mixed C57BL/6-129/Svj background. At age 8 to 12 weeks, female inducible megalin/cubilin KO (Meglox/lox,Cubnlox/lox;Creþ), podocin KO (Nph2lox,lox;Creþ), megalin/cubilin/podocin KO (Meglox/lox,Cubnlox/lox,Nph2lox,lox;Creþ), and littermate control mice (Meglox/lox,Cubnlox/lox, Nph2lox,lox;Cre-), were induced by i.p. injection of tamoxifen (Sigma-Aldrich, St. Louis, MO) at a dose of 33 mg/kg body weight for 5 consecutive days. The knockdown of the target genes in the kidney was assessed by quantitative real-time PCR and immunohistochemistry (see the following) at weeks 3 and 5 of the experiment (counting from the first day of tamoxifen induction), with similar results. Mouse breeding and experiments were carried out in a certified animal facility according to the Danish Animal Experiments Inspectorate. Plasma, urine, and tissue samples Urine was collected at week 5 of the experiment by housing the mice in metabolic cages for 24 hours. Urine samples were collected in tubes containing a protease inhibitor mix (Complete; Roche, Basel, Switzerland). The urine was centrifuged at 1000g for 5 minutes and stored at 80 C. The day after urine collection, the mice were anesthetized using isoflurane, and a blood sample was collected from the inferior cava vein in heparin-containing tubes, followed by the removal of 1 kidney and a section of liver tissue, which were snap frozen in liquid nitrogen and stored at 80 C for preparation of tissue homogenates. The remaining kidney was fixed by retrograde perfusion through the abdominal aorta with 2% paraformaldehyde Kidney International (2017) -, -–-
basic research
K Weyer et al.: Albumin uptake in nephrotic mice
in 0.1 mmol/l sodium cacodylate buffer, pH 7.4. The tissue was subsequently dehydrated and embedded in paraffin by standard methods. Blood plasma was prepared by centrifugation at 2000g for 10 minutes at 4 C. Plasma samples were analyzed for creatinine, albumin, and total protein at MRC Harwell (Oxfordshire, UK). The kidney and liver tissues were homogenized in ice-cold buffer containing 0.3 mmol/l sucrose, 25 mmol/l imidazole, and 1 mmol/l ethylenediamine tetraacetic acid (pH 7.2) with a protease inhibitor mix (complete Roche, Sigma-Aldrich, St Louis, MO) for 30 s with an IKA Ultra-turrax T8 homogenizer (IKA Werke, Staufen, Germany), centrifuged at 4000g for 15 minutes at 4 C, and the supernatants were stored at 80 C. Antibodies Primary antibodies used were sheep anti–rat megalin,34 rabbit anti–rat cubilin,35 rabbit anti–podocin,36 rabbit anti–human albumin (A0001, DAKO, Glostrup, Denmark), rabbit anti–actin (A5060, SigmaAldrich), chicken anti–b-actin (AB13822, Abcam, Cambridge, MA), and goat anti–murine IgM (AB9167, Abcam). Secondary antibodies used for immunoblotting were IRDye 800CW goat anti–rabbit IgG (926-32211, LI-COR, Biosciences, Cambridge, UK), IRDye 680CW donkey anti-chicken IgG (926-68075, LI-COR), and IRDye 800CW donkey anti–goat (926-32214, LI-COR). IgG was visualized by immunoblotting using IRDye 680CW goat anti–mouse IgG (92632220, LI-COR). Secondary antibodies used for immunohistochemistry were donkey anti–rabbit Alexa Fluor 488 (21206, Invitrogen, Carlsbad, CA), donkey anti–sheep Alexa Fluor 546 (21098, Invitrogen), horseradish peroxidase–conjugated goat anti–rabbit (P448, DAKO), and rabbit anti–sheep Ig-horseradish peroxidase (P163, DAKO). The specificity of the rabbit anti–human albumin antibody (A0001, DAKO) was validated by immunoblotting using purified mouse serum albumin (Sigma-Aldrich A3559) (not shown). Urinary albumin/creatinine measurements Urine albumin concentrations were determined using a mouse albumin ELISA Quantification kit (Bethyl Labs, Montgomery, TX), according to the manufacturer’s guidelines. Urine samples were analyzed for creatinine at MRC Harwell. Calculation of albumin GSC estimates Based on the assumption that the urinary excretion in megalin/ cubilin-deficient mice reflects the total filtered amount of albumin, the GSCs of albumin in normal and nephrotic mice were estimated as the urinary albumin-to-creatinine ratio divided by plasma albumin-to-creatinine ratio,20,26 in megalin/cubilin KO and megalin/ cubilin/podocin KO mice, respectively. Sodium dodecylsulfate–polyacrylamide gel electrophoresis and immunoblotting Liver samples were prepared by mixing homogenate (10 mg) with 4 sodium dodecylsulfate buffer (Invitrogen) containing 20 mM dithiothreitol (final concentration). Plasma samples were prepared by mixing 0.125 ml plasma with 4 sodium dodecylsulfate buffer, containing 20 mM dithiothreitol (final). Samples were separated on 12% Bis-Tris NuPage Gels (Invitrogen) in 1 MOPS buffer (Invitrogen). For protein-staining, gels were incubated with Coomassie dye (GelCode Blue Stain, Thermo Fisher Scientific, Rockford, IL), according to the manufacturer’s guidelines. For immunoblotting, the proteins were transferred to a polyvinylidene defluoride membrane (Immobilon-P, Merck Millipore, Darmstadt, Germany) and blocked in Odyssey blocking buffer (LI-COR) for 1 hour at room Kidney International (2017) -, -–-
temperature. The membrane was incubated overnight at 4 C with primary antibody and subsequently washed 3 times in phosphatebuffered saline with Tween. Incubation with a secondary antibody was performed at room temperature for 1 hour. Before exposure on the LI-COR Odyssey Infrared Imaging System (LI-COR, Biosciences, Cambridge, UK), the membrane was washed 3 times in phosphatebuffered saline with Tween and once in phosphate-buffered saline. Band intensities were quantified using Odyssey software (version 1.2; LI-COR). Immunohistochemistry Tissue preparation, sectioning, and labeling were performed as described previously.19 For immunoperoxidase labeling, renal paraffin sections were incubated with primary antibodies at 4 C overnight, followed by incubation with horseradish peroxidase– conjugated secondary antibody for 1 hour at room temperature. Peroxidase labeling was visualized by incubation with diaminobenzidine and 0.03% H2O2 for 10 minutes, and sections were counterstained with Mayer’s hematoxylin (Sigma-Aldrich). Immunoperoxidase-labeled images were acquired with a Leica DMR microscope equipped with a Leica DFC320 camera (Leica, Wetzlar, Germany). For immunofluorescent labeling, renal paraffin sections were incubated with primary antibodies at 4 C overnight, followed by incubation with secondary antibodies for 1 hour at room temperature. Immunofluorescent images were taken using a confocal laser-scanning microscope (LSM 510-META, Carl Zeiss GmbH, Jena, Germany) and processed using Zeiss Zen software (2009, Light Edition). All microscope and camera settings were kept identical when obtaining images from different groups. Quantitative real-time polymerase chain reaction Total RNA was purified from liver or kidney tissue using RNeasy Mini Kit (Qiagen, Hilden, Germany). The purity and concentration were determined by optical density at wavelengths of 260 and 280 nm, respectively. Purified RNA (20–100 ng) was incubated in a reverse transcriptase reaction mixture as previously described.19 A reaction without a reverse transcriptase enzyme was made in parallel to ascertain that the amplification occurred on RNA and not DNA. Realtime PCR was performed using a LightCycler480 (Roche) and SYBR Green I Master (Roche), according to the manufacturer. The primers used are shown in Supplementary Table S1. A standard curve, a calibration control, and negative controls were included in each run. All mRNA analyzed was normalized according to actin mRNA in the same sample and calibrated according to the calibrator (calibrator normalized relative quantification). All samples were run in duplicate. Data analysis and statistics All data were expressed as mean SEM. The Kolmogorov-Smirnov test (when N ¼ 5 per group) or the d’Agostino-Pearson test (when N ¼ 8 per group) was used to analyze for normality of the data. Data were compared between groups using an unpaired, 2-tailed, Student t test. Values of P < 0.05 were considered significant. All analyses were performed using GraphPad Prism 7.0 for Windows (GraphPad Software, San Diego, CA). DISCLOSURE
All the authors declared no competing interests. ACKNOWLEDGMENTS
We thank Inger Blenker Kristoffersen, Helle Salling Gittins, Hanne Sidelmann, and Pia Kamuk Nielsen for excellent technical assistance. 7
basic research
Sources of support for this work included grants from the Danish Medical Research Council, The Lundbeck Foundation, and the Novo Nordisk Foundation. SUPPLEMENTARY MATERIAL Figure S1. Megalin and cubilin localization and mRNA levels in the kidney of podocin knockout mice. Representative images showing the immunohistochemical labeling for megalin (A) and cubilin (B) in kidney sections from podocin knockout (KO) mice. Bars ¼ 50 mm. The relative megalin (C) and cubilin (D) to actin mRNA ratio in whole kidney samples from podocin KO compared with control (Cre-) mice, determined by quantitative real-time polymerase chain reaction. The graphs represent mean values from all mice in each group (N ¼ 8), and error bars indicate SEM. ns, not significant (P > 0.05) versus controls. Table S1. Primers used for quantitative real-time polymerase chain reaction. Supplementary material is linked to the online version of the paper at www.kidney-international.org. REFERENCES 1. Lambers Heerspink HJ, Brinkman JW, Bakker SJ, et al. Update on microalbuminuria as a biomarker in renal and cardiovascular disease. Curr Opin Nephrol Hypertens. 2006;15:631–636. 2. Remuzzi G, Ruggenenti P, Benigni A. Understanding the nature of renal disease progression. Kidney Int. 1997;51:2–15. 3. Haraldsson B, Nystrom J, Deen WM. Properties of the glomerular barrier and mechanisms of proteinuria. Physiol Rev. 2008;88:451–487. 4. Russo LM, Sandoval RM, McKee M, et al. The normal kidney filters nephrotic levels of albumin retrieved by proximal tubule cells: retrieval is disrupted in nephrotic states. Kidney Int. 2007;71:504–513. 5. Comper WD. Albuminuria is controlled primarily by proximal tubules. Nat Rev Nephrol. 2014;10:180. 6. Cui S, Verroust PJ, Moestrup SK, et al. Megalin/gp330 mediates uptake of albumin in renal proximal tubule. Am J Physiol. 1996;271:F900–F907. 7. Amsellem S, Gburek J, Hamard G, et al. Cubilin is essential for albumin reabsorption in the renal proximal tubule. J Am Soc Nephrol. 2010;21: 1859–1867. 8. Birn H, Fyfe JC, Jacobsen C, et al. Cubilin is an albumin binding protein important for renal tubular albumin reabsorption. J Clin Invest. 2000;105: 1353–1361. 9. Storm T, Emma F, Verroust PJ, et al. A patient with cubilin deficiency. N Engl J Med. 2011;364:89–91. 10. Storm T, Tranebjaerg L, Frykholm C, et al. Renal phenotypic investigations of megalin-deficient patients: novel insights into tubular proteinuria and albumin filtration. Nephrol Dial Transplant. 2013;28:585–591. 11. Weyer K, Nielsen R, Christensen EI, et al. Generation of urinary albumin fragments does not require proximal tubular uptake. J Am Soc Nephrol. 2012;23:591–596. 12. Weyer K, Storm T, Shan J, et al. Mouse model of proximal tubule endocytic dysfunction. Nephrol Dial Transplant. 2011;26:3446–3451. 13. Tenten V, Menzel S, Kunter U, et al. Albumin is recycled from the primary urine by tubular transcytosis. J Am Soc Nephrol. 2013;24:1966–1980. 14. Sandoval RM, Wagner MC, Patel M, et al. Multiple factors influence glomerular albumin permeability in rats. J Am Soc Nephrol. 2012;23:447–457.
8
K Weyer et al.: Albumin uptake in nephrotic mice
15. Kim J, Bronson CL, Hayton WL, et al. Albumin turnover: FcRn-mediated recycling saves as much albumin from degradation as the liver produces. Am J Physiol Gastrointest Liver Physiol. 2006;290:G352–G360. 16. Sarav M, Wang Y, Hack BK, et al. Renal FcRn reclaims albumin but facilitates elimination of IgG. J Am Soc Nephrol. 2009;20:1941–1952. 17. Chaudhury C, Brooks CL, Carter DC, et al. Albumin binding to FcRn: distinct from the FcRn-IgG interaction. Biochemistry. 2006;45: 4983–4990. 18. Mollet G, Ratelade J, Boyer O, et al. Podocin inactivation in mature kidneys causes focal segmental glomerulosclerosis and nephrotic syndrome. J Am Soc Nephrol. 2009;20:2181–2189. 19. Nielsen R, Mollet G, Esquivel EL, et al. Increased lysosomal proteolysis counteracts protein accumulation in the proximal tubule during focal segmental glomerulosclerosis. Kidney Int. 2013;84:902–910. 20. Mori KP, Yokoi H, Kasahara M, et al. Increase of Total Nephron Albumin Filtration and Reabsorption in Diabetic Nephropathy. J Am Soc Nephrol. 2017;28:278–289. 21. Russo LM, Sandoval RM, Brown D, et al. Controversies in nephrology: response to ’renal albumin handling, facts, and artifacts’. Kidney Int. 2007;72:1195–1197. 22. Christensen EI, Birn H, Rippe B, et al. Controversies in nephrology: renal albumin handling, facts, and artifacts! Kidney Int. 2007;72:1192–1194. 23. Remuzzi A, Sangalli F, Fassi A, et al. Albumin concentration in the Bowman’s capsule: multiphoton microscopy vs micropuncture technique. Kidney Int. 2007;72:1410–1411. author reply, 1411. 24. Norden AG, Unwin RJ. Is the albumin retrieval hypothesis a paradigm shift for nephrology? J Am Soc Nephrol. 2012;23:569–571. 25. Dickson LE, Wagner MC, Sandoval RM, et al. The proximal tubule and albuminuria: really! J Am Soc Nephrol. 2014;25:443–453. 26. Norden AG, Lapsley M, Lee PJ, et al. Glomerular protein sieving and implications for renal failure in Fanconi syndrome. Kidney Int. 2001;60: 1885–1892. 27. Tojo A, Endou H. Intrarenal handling of proteins in rats using fractional micropuncture technique. Am J Physiol. 1992;263:F601–F606. 28. Schiessl IM, Hammer A, Kattler V, et al. Intravital Imaging Reveals Angiotensin II-Induced Transcytosis of Albumin by Podocytes. J Am Soc Nephrol. 2016;27:731–744. 29. Nakano D, Kobori H, Burford JL, et al. Multiphoton imaging of the glomerular permeability of angiotensinogen. J Am Soc Nephrol. 2012;23: 1847–1856. 30. Wagner MC, Campos-Bilderback SB, Chowdhury M, et al. Proximal Tubules Have the Capacity to Regulate Uptake of Albumin. J Am Soc Nephrol. 2016;27:482–494. 31. Aseem O, Smith BT, Cooley MA, et al. Cubilin maintains blood levels of HDL and albumin. J Am Soc Nephrol. 2014;25:1028–1036. 32. Caruso-Neves C, Pinheiro AA, Cai H, et al. PKB and megalin determine the survival or death of renal proximal tubule cells. Proc Natl Acad Sci U S A. 2006;103:18810–18815. 33. Kaysen GA, al Bander H. Metabolism of albumin and immunoglobulins in the nephrotic syndrome. Am J Nephrol. 1990;10(Suppl 1):36–42. 34. Moestrup SK, Nielsen S, Andreasen P, et al. Epithelial glycoprotein-330 mediates endocytosis of plasminogen activator-plasminogen activator inhibitor type-1 complexes. J Biol Chem. 1993;268:16564–16570. 35. Sahali D, Mulliez N, Chatelet F, et al. Comparative immunochemistry and ontogeny of two closely related coated pit proteins. The 280-kd target of teratogenic antibodies and the 330-kd target of nephritogenic antibodies. Am J Pathol. 1993;142:1654–1667. 36. Roselli S, Gribouval O, Boute N, et al. Podocin localizes in the kidney to the slit diaphragm area. Am J Pathol. 2002;160:131–139.
Kidney International (2017) -, -–-