Brief Communication

Murine Double Minute-2 Prevents p53-Overactivation-Related Cell Death (Podoptosis) of Podocytes

Dana Thomasova, Hauke A. Bruns, Victoria Kretschmer, Martrez Ebrahim, Simone Romoli, Helen Liapis, Ahmed M. Kotb, Nicole Endlich and Hans-Joachim Anders
JASN July 2015, 26 (7) 1513-1523; DOI: https://doi.org/10.1681/ASN.2014040345

This article requires a subscription to view the full text. If you have a subscription you may use the login form below to view the article. Access to this article can also be purchased.

Abstract

Murine double minute-2 (MDM2), an E3 ligase that regulates the cell cycle and inflammation, is highly expressed in podocytes. In podocyte injury, MDM2 drives podocyte loss by mitotic catastrophe, but the function of MDM2 in resting podocytes has not been explored. Here, we investigated the effects of podocyte MDM2 deletion in vitro and in vivo. In vitro, MDM2 knockdown by siRNA caused increased expression of p53 and podocyte death, which was completely rescued by coknockdown of p53. Apoptosis, pyroptosis, pyronecrosis, necroptosis, ferroptosis, and parthanatos were excluded as modes of occurrence for this p53-overactivation-related cell death (here referred to as podoptosis). Podoptosis was associated with cytoplasmic vacuolization, endoplasmic reticulum stress, and dysregulated autophagy (previously described as paraptosis). MDM2 knockdown caused podocyte loss and proteinuria in a zebrafish model, which was consistent with the phenotype of podocyte-specific MDM2-knockout mice that also showed the aforementioned ultrastructual podocyte abnormalities before and during progressive glomerulosclerosis. The phenotype of both animal models was entirely rescued by codeletion of p53. We conclude that MDM2 maintains homeostasis and long-term survival in podocytes by preventing podoptosis, a p53-regulated form of cell death with unspecific features previously classified as paraptosis.

Podocytes are terminally differentiated epithelial cells situated at the outer surface of the glomerular capillary loops and a component of the glomerular filtration barrier (GFB). Significant podocyte loss leads to glomerular basement membrane denudation and adhesion to Bowman’s capsule with, initially, focal segmental scarring and subsequently, global glomerulosclerosis, a condition that accounts for 90% of CKD.1–4 It is rather remarkable that podocytes are able to withstand the glomerular filtration pressure as well as toxic or immunologic stress over an individual’s lifespan, but the mechanisms are not well understood. The E3 ubiquitin ligase murine double minute-2 (MDM2) regulates proteasomal degradation of the tumor suppressor gene and cell cycle regulator p53 but also, facilitates NF-κB signaling.5,6 MDM2 overexpression promotes tumor growth, whereas MDM2 antagonists are effective in cancer.7,8 We have previously shown that MDM2 drives NF-κB signaling-related inflammation in acute tubular injury, whereas its proregenerative effect in the healing phase was exclusively p53-dependent.9 In podocyte injury, MDM2 promotes podocyte loss by bypassing cell cycle G2/M arrest and entering mitosis (i.e., mitotic catastrophe).10 However, podocytes also constitutively express high levels of MDM2,10 but what for is unknown. We speculated that resting podocytes need MDM2 to maintain homeostasis and sought to define in which manner. To test our hypothesis, we deleted MDM2 or MDM2/p53 in cultured mouse podocytes by small interfering RNA (siRNA), in zebrafish by using morpholino (MO) oligonucleotides, and by generating podocyte-specific knockout mice.

Transfection of mouse podocytes with MDM2 siRNA (Supplemental Figure 1A) resulted in prominent vacuolization (Supplemental Figure 1B) and decreased podocyte viability compared with transfection with control siRNA (Figure 1A). MDM2 suppression induced an increase of p53 mRNA and numerous p53 target genes known to regulate the cell cycle, cell death, and autophagy, such as p21, Fas, and ATG5 (Figure 1B). Simultaneous knockdown of MDM2 and p53 (Supplemental Figure 1A) completely restored podocyte viability just like the p53 inhibitor pifithrin-α (Figure 1C, Supplemental Figure 2). This implies that MDM2 is needed to prevent p53-overactivation-related cell death (here referred to as podoptosis). Podoptosis is thought to represent apoptosis, but pancaspase inhibition with Z-VAD FMK did not abrogate death of MDM2-depleted podocytes (Figure 1D, Supplemental Figure 2), which excludes not only apoptosis but also, caspase-1–dependent pyroptosis. Also, necrostatin-1, ferrostatin, cyclosporin A, and CA-074Me could not abrogate death of MDM2-depleted podocytes, which excludes all other known forms of regulated cell death (i.e., necroptosis, ferroptosis, parthanatos, and pyroptosis).11 Because the autophagy inhibitor chloroquine rather enhanced MDM2 siRNA-induced podocyte death (Supplemental Figure 2), we further examined the distribution of markers for autophagosomes and lysosomes in a transgenic GFP-LC3 human podocyte cell line treated with MDM2 siRNA. These cells showed cytoplasmic increase of LC3-GFP punctate and redistribution and accumulation of lysosomal LAMP1 in perinuclear and perivacuolar distribution (Figure 1E), suggesting aberrations in autophagic flux. Also, p62 accumulation was apparent, indicating insufficient autophagic protein aggregate clearance as another unspecific indicator of cell stress (Figure 1E). Additionally, densitometric analysis of Western blots from MDM2 knockdown mouse podocytes incubated with chloroquine showed a slower conversion of LC3-I to LC3-II as well as accumulation of p62, indicating a low rate of autophagic flux (Supplemental Figure 1C). Furthermore, we investigated how the MDM2 deficiency in podocytes alters p53 ubiquitination and degradation. Knockdown of endogenous MDM2 decreased the levels of ubiquitinated p53, but the levels of the endogenious p53 were increased (Supplemental Figure 1D), confirming that MDM2 promotes ubiquitination and proteasome-mediated degradation of p53.

Figure 1.
Figure 1.

MDM2 knockdown in murine and human podocyte cell line results in decreased viability and dysregulated autophagy in podocytes. (A) Murine podocytes were incubated for 24 hours with MDM2 siRNA or control scrambled siRNA, and the cell death was measured by LDH release. Podocytes treated with 2% Triton-X were used as positive controls. (B) RNA was isolated from murine podocytes treated for 48 hours with MDM2 siRNA or control siRNA and transcribed into cDNA. Quantitative RT-PCR was performed, and mRNA expression was determined using 18S as a housekeeping gene. Expression levels are illustrated in the bar graph as fold induction compared with control siRNA-transfected podocytes, which served as controls. (C) Murine podocytes were transiently transfected with siRNAs for MDM2, p53, and scrambled or cotransfected with MDM2+p53 siRNAs; 24 hours later, the cell death and cell viability were determined by LDH and MTT assays, respectively. (D) p53-overactivation-related cell death (podoptosis) schema. (E) The representative confocal microscopy images of human podocytes expressing LC3-GFP protein transfected with MDM2 and control siRNAs. MDM2 knockdown podocytes compared with the control cells exhibited accumulation of autophagosomal LC3 and lysosomal LAMP1 proteins with occasional colocalization; p62, a marker of dysfunctional autophagy, was also accumulated in MDM2 knockdown podocytes. All experiments were performed in triplicate. Data are means±SEMs. CTRL, control. Scale bar, 10 µm. **P<0.01; ***P<0.005.

To validate the nonredundant role of MDM2 in podocytes in vivo, MDM2 was knocked down in the zebrafish model by the use of morpholinos (MOs). MOs generated against zMdm2 and a control MO were injected into fertilized eggs of the zebrafish strain ET expressing enhanced green fluorescent protein (EGFP) specifically in podocytes.12 The zMdm2 knockdown was confirmed by real-time PCR (RT-PCR) (Figure 2A). zMdm2 MOs treated larvae showed a strong pericardial edema compared with control MO larvae at 3 days postfertilizaton (Figure 2B). Histologic sections revealed the changed morphology of the glomerular tuft and the significant reduction of nephrin expression in zMdm2 knockdown larvae glomeruli compared with control MO larvae (Figure 2B). To investigate the integrity of the GFB, we used a zebrafish strain Tg(Casper; l-fabp:DBP-EGFP) (CADE) that expresses the vitamin D-binding protein (DBP) coupled to EGFP in blood13 (A.M. Kotb et al., unpublished data) and does not pass the intact GFB. If the GFB becomes leaky, DBP-EGFP passes the barrier, and the fluorescence in the blood vessels decreases over time. In 4-days postfertilizaton-old zMdm2 MO larvae, the intensity of DBP-EGFP was significantly diminished, indicating a loss of DBP-EGFP through the filtration barrier, whereas the fluorescence in control larvae vessels was unchanged (Figure 2C). To verify that the EGFP loss is because of damaged GFB, we injected Alexa Fluor 647-conjugated 10-kD dextran into CADE larvae and monitored its filtration into the proximal tubules. The tubular epithelial cells of control larvae showed only the dextran uptake. In contrast, zMdm2 MO larvae accumulated dextran together with DBP-EGFP, indicating disintegrated GFB (Supplemental Figure 2). Interestingly, a coinjection of MOs against zMdm2 and zp53 (1:1) to induce simultaneously zMdm2 and zp53 deficiency (Figure 2A) completely rescued the structural and functional phenotypes of the zMdm2 knockdown (Figure 2, B and C, Supplemental Figure 3). This suggests that Mdm2 is essential for podocytes to avoid p53-dependent podocyte dysfunction and loss in cultured podocytes as well as the zebrafish model.

Figure 2.
Figure 2.

zMdm2 knockdown in the zebrafish larvae causes malformation of the GFB and loss of nephrin. (A) After the treatment of fertilized zebrafish eggs with zMdm2 MO or zMDM2+zp53 MO, mRNA expressions of Mdm2 or both Mdm2 and p53 were significantly reduced in contrast to larvae treated with control MO. The target mRNA expression levels were determined by real-time PCR and expressed as the mean of the ratio versus the respective CTRB1 mRNA level±SEM. Data are means±SEMs. (B) Representative images of zebrafish larvae targeted with control MO, zMdm2 MO, and zMdm2+zp53 MO. The figure (row 1) shows a zebrafish larva (3 days postfertilizaton) that was treated with control MO. The glomerulus was well developed, and the podocytes (green) expressed nephrin (red) that is shown as a merged pictured (yellow). In contrast, zMdm2 MO-treated larva developed a pericardial edema (arrow). The glomerular tuft is not properly formed, and the podocytes (green) have reduced nephrin expression (red). Coinjection of zMdm2 and zp53 MO (row 4) completely rescued this phenotype, and the zebrafish larva (4 days postfertilizaton) showed a normal developed glomerulus. The podocytes expressed nephrin comparable with the control MO-treated larva (4 days postfertilizaton). (C) Upper panel shows a loss of fluorescent DBP-EGFP protein from the vessels of Mdm2 MO-treated larvae caused by a leakage in the filtration barrier. Lower panel indicates the rescue of the functional Mdm2 knockdown phenotype in zebrafish larvae that are double deficient for zMmd2 and zp53. CTRL, control. Scale bar, 10 µm. *P<0.05; ***P<0.005.

To validate these results in the mammalian kidney, we generated mice with podocyte-specific MDM2 deletion (MDM2Δpodocyte) (Figure 3A). These mice were born at expected Mendelian ratios. MDM2 immunostaining was selectively diminished in podocytes of MDM2Δpodocyte mice compared with control littermates (Figure 3A), and MDM2 mRNA levels from purified glomeruli were also diminished (Supplemental Figure 4A). Up to 3 weeks of age, the MDM2Δpodocyte mice and control littermates were indistinguishable concerning weight, proteinuria, and glomerular histology. At 3–5 weeks of age, the MDM2Δpodocyte mice developed significant proteinuria, which increased with age (Figure 3B) and was associated with a reduced lifespan compared with control littermates up to 7 months of life (Figure 3C). MDM2Δpodocyte mouse kidneys were normal on light microscopy at 2–3 weeks, but at 8 weeks, collapsing FSGS was evident (Figure 3D). Glomerular lesions were associated with respective tubular atrophy and intraluminal protein casts (Figure 3D). Ultrastructural analysis showed glomerular hyalinosis and marked abnormalities in the podocytes compared with control glomeruli. There was profound podocyte accumulation of vacuoles, enlarged lysosomes and mitochondria, endoplasmic reticulum (ER) abnormalities, and clumping of the actin cytoskeleton, suggesting significant podocyte stress (Figure 3E). At 14 weeks of age, the MDM2Δpodocyte mice displayed global glomerulosclerosis, segmental scarring, and profound podocyte vacuolization in most glomeruli and massive tubular dilation with flattened tubular epithelium and multifocal tubular casts (Figure 3D). The vacuolated podocytes stained strongly with calnexin and GRP78 at 8 weeks of age, indicating high level of ER stress (Supplemental Figure 4B). Electron microscopy identified degenerated podocytes with their cytoplasm filled with huge single membrane-bound vacuoles (Figure 3E, Supplemental Figure 5D); 3-week-old mice in both MDM2Δpodocyte mice and the control group had the same count of Wilms' tumor 1 (WT-1)/nephrin-positive podocytes per glomerulus as well as the same number of nephrons corrected to kidney planar surface (Supplemental Figure 5A). Podocyte counts declined with age in MDM2Δpodocyte mice, whereas that ratio increased in their control littermates (Figure 3F, Supplemental Figure 5B). Lack of MDM2 in podocytes was associated with an age-dependent increase of p53 positivity (Figure 3G, Supplemental Figure 5C). To investigate whether dysregulated autophagy contributes to the deleterious renal phenotype in MDM2Δpodocyte mice, we examined the kidney sections of 8- and 14-week-old mice by staining with markers of dysregulated autophagy. Confocal microscopy showed accumulation of LC3-positive autophagosomes, LAMP1- and LAMP2-positive lysosomes, and p62 protein aggregates in affected podocytes (Supplemental Figure 4C). To validate the functional role of uncontrolled p53 activity as a cause of the MDM2Δpodocyte mouse phenotype, we generated podocyte-specific double-knockout MDM2/p53Δpod mice and littermate controls with one p53 allele intact MDM2Δpod/p53f/wt. Lack of p53 completely rescued the severe structural and functional phenotype of MDM2Δpodocyte mice (Figure 4, Supplemental Figure 6).

Figure 3.
Figure 3.

Podocyte-specific MDM2 knockout in our mouse model results in p53-mediated podocyte loss and FSGS. (A) Mice expressing cre recombinase under control of podocin Nphs2 promoter were crossed with MDM2fl/fl mice to generate podocyte-specific MDM2 knockout mice. Immunohistochemical staining confirmed MDM2 deletion in podocytes, whereas in tubular epithelial cells, the MDM2 remained unchanged. Scale bar, 20 µm. (B) MDM2Δpodocyte mice developed increasing proteinuria by 4 weeks of age (n=11). Light, alb/cre<5; normal, alb/cre<1; severe, alb/cre>5. (C) Kaplan–Meyer survival curve of MDM2fl/fl control and MDM2Δpodocyte mice (n=14 per group). (D) Representative images of kidney sections from periodic acid–Schiff (PAS) and silver stains of 8- and 14-week-old MDM2Δpodocyte mice and 14-week-old littermate controls. By 8 weeks of age, the kidney shows FSGS and glomerular hyalinosis; by 14 weeks, the glomeruli are globally sclerosed with prominent vacuolization. Scale bars, 100 µm in top panel; 20 µm in middle and bottom panels. (E) Electron microscopy at 8 weeks of age reveals foot process effacement and podocyte vacuolization in MDM2Δpodocyte mice in contrast to intact foot processes and normal-appearing podocytes in controls. By 14 weeks, the podocyte cytoplasm is filled with huge vacuoles, accumulated enlarged lysosomes, and lipofuscin granules (electron microscopy in successive low to medium to high 1500–12,000 magnification). The vacuoles are clearly in visceral podocytes; peripheral podocytes appear fibroblast-like (elongated) at the periphery of the glomerulus with no cytoplasmic vacuoles. Scale bars, 10 µm in left and center panels; 2 µm in right panel. (F) WT-1/nephrin costaining was used to quantify differentiated podocytes in kidneys of control and MDM2Δpodocyte mice. Numbers of nephrin/WT-1 double-positive cells were quantified in both groups at 3 and 14 weeks of age. The data are means±SEMs. Original magnification, ×400. Scale bar, 20 µm. **P<0.01 versus control group. (G) Representative images of p53 immunohistochemistry showing the increased p53 stain in MDM2Δpodocyte glomeruli at 14 weeks of age. CTRL, control. Scale bar, 20 µm.

Figure 4.
Figure 4.

Concomitant podocyte-specific MDM2 and p53 deletion rescues the podocytes in the MDM2/p53Δpod mouse model. (A) Results from 4 months of follow-up of MDM2/p53Δpod mice for proteinuria (8-week-old mice: n=17 control MDM2Δpod/p53f/wt and n=14 MDM2/p53Δpod; 14-week-old mice: n=10 control MDM2Δpod/p53f/wt and n=8 MDM2/p53Δpod). ***P<0.005. (B) Representative images of kidney sections of 14-week-old MDM2fl/fl control, MDM2Δpod/p53f/wt, and MDM2/p53 Δpod mice. Scale bars, 100 µm in top panel; 20 µm in middle panel. The bottom panel shows the corresponding electron micrographs in control and experimental mice, showing complete reversal of the vacuolated phenotype in MDM2/p53Δpod mice (magnifications of 4000–12,000). Scale bars, 10 µm in bottom left and bottom center; 2 µm in bottom right. (C) WT-1/nephrin costaining was used to quantify differentiated podocytes in kidneys of control MDM2Δpod/p53f/wt and MDM2/p53Δpod mice kidneys. Numbers of nephrin/WT-1 double-positive cells were quantified in both groups at 3 and 14 weeks of age. The data are means±SEMs. CTRL, control; EM, electron microscopy; PAS, periodic acid–Schiff. **P<0.01 versus control group.

MDM2 ubiquitinates p53 for proteasomal degradation, suppresses p53 transcription, and fosters nuclear export of p53 back to the cytoplasm.6 This way, MDM2 prevents p53-mediated cell cycle arrest, cell senescence, and premature death.8,14 For example, gain-of-function mutations in the MDM2 gene promote tumorigenesis, and specific MDM2 induction triggers lymphoma or lymphoproliferation in systemic lupus erythematosus.15,16 The disruption of MDM2 and p53 binding is being tested as a potential cancer therapy in patients who retain wild-type p53. Several MDM2 antagonists have been developed and are undergoing clinical trials.7 However, our data and the results from other studies17,18 indicate that p53 activation could have deleterious consequences in kidney, especially in podocytes and other healthy tissues. Genetic deletion of the MDM2 gene in mice results in embryonic lethality from massive p53-dependent cell death, whereas concomitant p53 deficiency completely rescues this phenotype.19,20 Specific deletion of MDM2 from the embryonic kidney compromises stem/progenitor cell renewal and differentiation.21 Our results are in line with these data and document the same nonredundant role of MDM2 in controlling p53 in terminally differentiated podocytes during kidney homeostasis. However, our results differ from those observed in other epithelial tissues. For example, MDM2 depletion in intestinal epithelial cells does not cause a persistent phenotype because of rapid clonal overgrowth of cells with insufficient Cre recombinase activity and MDM2 depletion.22 This could not occur in postmitotic podocytes that have a low turnover rate.23 Like neurons, the terminally differentiated podocytes did not retain the ability to divide, and after lost, they hardly can be replaced. Through MDM2-controlled p53 suppression, the podocytes can be driven into mitosis, which results in podocyte death by mitotic catastrophe.10 However, as our data show, the uncontrolled p53 overexpression leads to podocyte demise. Because of their complex structure, podocytes cannot be renewed by mitosis of the surviving, intact podocytes, and their permanent loss results in irreversible glomerular damage and kidney injury. This indicates that keeping p53 in balance is absolutely essential for lifelong survival of podocytes and kidney homeostasis. However, which form of podocyte death is prevented by MDM2?24 Our in vitro data exclude apoptosis, which is consistent with a previous study that excluded p53-mediated apoptotic cell death in MDM2-depleted epithelial cells.18 Our in vitro data also exclude all other known regulated cell death pathways, such as pyroptosis, pyronecrosis, necroptosis, ferroptosis, and parthanatos.11 Necroptosis was also excluded to mediate adriamycin-induced podocyte death.25 In our search for a cell death mode, we noticed dysregulated autophagy in MDM2-depleted podocytes, but this seems to be a rather unspecific phenomenon during cell stress that does not cause cell death on its own, at least not within a few weeks of life.26 Oncosis could not be formally excluded, because no inhibitors are available for this passive form of injury-related cell necrosis. However, the occurrence of podocyte death under homeostatic conditions basically excludes oncosis as an explanation for spontaneous podocyte loss. Because nowadays, cell death is no longer defined by morphologic features but by signaling pathways, the precise way of podocyte podoptosis other than apoptosis remains unknown.27 On the basis of the morphologic signature of MDM2-deficient podocytes (massive cytoplasmic vacuolization and signs of ER stress), it may be currently classified as what has been called paraptosis-like,28,29 but the associated signaling events remain to be identified. We conclude that podocytes need MDM2 to maintain homeostasis and long-term survival. MDM2 prevents podoptosis, a p53-overactivation-related form of cell death, which in cultured podocytes, was not apoptosis or any of the known modes of regulated necrosis but showed unspecific features previously named paraptosis.

Concise Methods

Cell Culture Experiments

Human podocytes expressing the GFP-LC3 transgene were provided by Tobias Huber (University of Freiburg, Freiburg, Germany). Mouse podocytes were cultured as previously described.30,31 For the viability assays, the mouse podocytes were seeded in 96-well plates (5000 cells/well). For the RNA extraction, the murine or human podocytes were cultured in six-well plates (60,000 cells/well) and kept under nonpermissive conditions (37°C in growth medium without INF-y) for 12 days to induce the differentiation of podocytes. The podocytes were then incubated with 25 nM siRNA for 24–48 hours using Lipofectamine RNAiMAX transfection reagent (Life Technologies, Darmstadt, Germany) for transient transfection. Specific siRNA to silence selectively murine and human MDM2 or murine p53 as well as appropriate control siRNA (negative control no. 1) were purchased from Ambion (Life Technologies).

At 48–72 hours post-transfection, the podocytes were collected for RNA isolation, or cytotoxicity/viability assays were performed; 16 hours after the transfection, the cells were treated with Z-VAD FMK (25 µM), necrostatin-1 (10 µM), cyclosporin A (1 µM; Enzo Life Sciences, Lörrach, Germany), chloroquine (25 µM; Invivogen, San Diego, CA), pifithrin-α (10 µM; Sigma-Aldrich, Munich, Germany), ferrostatin (1 µM), and CA-074Me (500 nM; Calbiochem; EMD Millipore, Billerica, MA) for 24 hours, and viability assay was performed. The podocyte cell death and cell viability were determined using LDH assay (Cytotoxicity Detection Kit [LDH]; Roche, Mannheim, Germany) and MTT assay (CellTiter 96 Non-Radioactive Cell Proliferation Assay; Promega, Madison, WI) according to the manufacturer’s instructions. Experiments were performed in triplicate. For immunofluorescence and confocal microscopy analysis, the human LC3-GFP podocytes were grown, differentiated (8000 cells/well) for 12 days, and transfected with 25 nM siRNA specific for human MDM2 and negative scrambled control siRNA in eight-well chamber slides coated with collagen IV (ibidi GmbH, Planegg, Germany); 48 hours after transfection, the cells were fixed in ice-cold acetone, stained using primary antibodies against LAMP-1 (Abcam, Inc., Cambridge, United Kingdom) or p62 (Progen Biotechnik, Heidelberg, Germany), and evaluated by confocal microscopy.

Zebrafish Experiments

Zebrafish were grown, mated, and maintained at 28.5°C according to standard protocols and as previously described.32 Embryos were kept and handled in E3 solution. Two transgene zebrafish strains were used (ET and CADE). Imaging of the zebrafish pronephros is greatly improved in transparent zebrafish mutants.12 Therefore, we crossed the transparent zebrafish line Casper33 (provided by Uwe Strähle, Karlruhe, Germany) with the wt1a:EGFP strain (provided by Christoph Englert, Jena, Germany) that expresses EGFP in podocytes.34 Furthermore, we crossed the transgenic l-fabp:DBP-EGFP zebrafish line (provided by Anand-Apte) expressing DBP coupled to EGFP in the blood plasma35 with the zebrafish strain Casper. This zebrafish line is referred to as CADE. The DBP-EGFP has the size and charge similar to albumin and cannot pass the intact filtration barrier.

Dextran Microinjection and Confocal Microscopy

Tricaine (Sigma-Aldrich) at a concentration of 0.1%–0.5% (in E3 solution) was used for anesthetizing larvae before microinjection and microscopy. The CADE larvae were injected with 20 mg/ml Alexa Fluor 647-conjugated 10-kD dextran solution (Invitrogen, San Diego, CA) into the caudal vein using Femtotips micropipettes (Eppendorf, Hamburg, Germany). After recovery in E3 medium for 40 minutes, the larvae were anesthetized again. The dextran filtration into the proximal tubules was monitored in the living zebrafish larvae by confocal microscopy 4 hours after injection. For confocal microscopy, a Leica TCS SP5 (Leica Microsystems, Wetzlar, Germany) was used, and the larvae were positioned on their sides in a glass-bottomed petri dish and covered with 8% methyl cellulose in PBS. The vasculature and the pronephric tubule were imaged over time to study the distribution and the uptake of injected dextran. We used a 20× water immersion objective (HCX PL APO 20×/0.7).

MO-Mediated Knockdown

MOs were designed by Gene Tools LCC (Philomath, OR). To target the 5′ untranslated region including the start codon of zMdm2, we have used the following MO sequence: 5′-AACAACTCTCTGTTGCCATTTTGGT-3′. As a second MO directed against a splicing site, the sequence 5′-CTCTGTTGCCATTTTGGTAGTTATC-3′ was used. To target the 5′ untranslated region including the start codon of zp53, we have used the MO sequence 5′-GCGCCATTGCTTTGCAAGAATTG-3′. For control, we have used a standard control MO provided by Gene Tools LCC with the sequence 5′-CCTCTTACCTCAGTTACAATTTATA-3′. MOs were diluted to a concentration of 1 mM. A volume of approximately 3 nl per zebrafish egg was injected into the yolk at the two- to four-cell stage using a microinjector (Transjector 5246; Eppendorf).

Zebrafish Histology and Immunohistochemistry

For fluorescence microscopy of zebrafish cryosections, larvae were fixed in 2% paraformaldehyde in 1× PBS (pH 7.4) for 3 hours at room temperature and incubated in 30% sucrose at 4°C overnight. After snap freezing in liquid nitrogen using Tissue-Tek (Sakura, Staufen, Germany), cryostat sections (60 μm) were cut and permeabilized with 0.3% Triton X-100 (Merck, Darmstadt, Germany) for 30 seconds. The cryosections were stained with phalloidin (Alexa Fluor 546-conjugated phalloidin; Invitrogen) and 1 mg/100 ml Hoechst 33342 (Sigma-Aldrich) for 30 minutes. For the nephrin staining, the anti-rabbit nephrin antibody (gift from A. Majumdar, Uppsala, Sweden) was incubated overnight at 4°C. After the washing steps, the Cy3-conjugated anti-rabbit secondary antibody (1:250; Dianova, Hamburg, Germany) was applied for 30 minutes.

Mouse Experiments

The MDM2flox/flox;Nphs2-cre+ (MDM2Δpodocyte ) mice were generated by breeding the MDM2flox/flox mice, in which loxP sites flanked exons 4 and 5 of the MDM2 gene, with Nphs2-cre+ mice, expressing the Cre recombinase under control of the podocin promoter. The MDM2flox/flox mice (/B6 mixed background) were a gift from Guillermina Lozano (University of Texas, Houston, TX), and the Nphs2-cre+ mice (C56Bl/6 background) were provided by Tobias Huber. Both mice strains were previously described.36,37 The MDM2flox/flox littermates lacking the podo-cre transgene were used as control mice. To generate the podocyte-specific double-knockout MDM2flox/flox p53flox/flox; Nphs2-cre+ (MDM2/p53Δpod ) mice, we bred the MDM2flox/flox; Nphs2-cre+ mice with p53flox/flox mice, in which the loxP sites were inserted in introns 1 and 10 to ensure Cre-mediated removal of nearly all coding sequences.38 The p53flox/flox mice were a gift from Karl Lenhard Rudolph (University of Ulm, Ulm, Germany). The MDM2flox/flox p53flox/wt; Nphs2-cre+ (MDM2Δpod/p53f/wt) littermates were used as positive controls. All animal studies were approved by the Committee on Research Animal Care, Regierungspräsidium Oberbayern.

Renal Function Measurement

The mice urine samples were collected every week from week 2 to 17. The urine albumin was measured by the albumine ELISA kit (Bethyl Laboratories, Montgomery, TX). The creatinine kit (DiaSys, Holzheim, Germany) was used for detection of urine creatinine according to the manufacturer’s instructions. Albumin-to-creatinine ratio was calculated as milligrams per deciliter albumin per milligrams per deciliter creatinine.

Renal Histology, Immunohistochemistry, Confocal Microscopy, and Electron Microscopy

Kidney tissues were fixed in 4% neutral-buffered formalin, dehydrated in graded alcohols, and embedded in paraffin. For routine histology, the 4-µm sections were stained with periodic acid–Schiff reagent. For immunohistochemistry, sections were deparaffinized, rehydrated, transferred into citrate buffer, either autoclaved or microwave-treated for antigen retrieval, and processed as described.9 The following primary antibodies were used: guinea pig anti-mouse nephrin (1:100; Acris Antibodies, Herford, Germany), rabbit anti-mouse WT1 (1:25; Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti-mouse MDM2 (1:100; Abcam, Inc.), rabbit anti-mouse p53 (1:500; Vector Laboratories, Burlingame, CA), rabbit anti-mouse LAMP1 (1:200; Abcam, Inc.), rabbit anti-mouse LAMP2 (1:250; Abcam, Inc.), guinea pig anti-mouse p62 (1:50; Progen Biotechnik), rabbit anti-mouse LC3B (1:50; Cell Signaling Technology, Danvers, MA), rabbit anti-mouse calnexin (1:100; Enzo Life Sciences, Farmingdale, NY), and rabbit anti-mouse GRP78 (1:100; Abcam, Inc.). Immunofluorescence staining was evaluated using a LSM 510 confocal microscope and LSM software (Carl Zeiss AG). For transmission electron microscopy, the kidney cortex was sectioned into 1×1-mm cubes and immediately immersed in fixative containing 3% glutaraldehyde and 1% paraformaldehyde in PBS. Postfixation kidneys were immersed in cold fixative containing 2% glutaraldehyde and 2% paraformaldehyde in sodium cacodylate buffer (pH 7.4). Kidneys were postfixed in phosphate cacodylate-buffered 2% OsO4 for 1 hour, dehydrated in graded ethanols with a final dehydration in propylene oxide, and embedded in Embed-812 (Electron Microscopy Sciences, Hatfield, PA). Ultrathin sections (approximately 90-nm thick) were stained with uranyl acetate and Venable’s lead citrate and viewed with a JEOL model 1200EX electron microscope (JEOL, Tokyo, Japan).

RNA Preparation and Real-Time Quantitative PCR

Total RNA was extracted from mouse glomeruli and isolated using Dynabead perfusion or the murine podocyte cell line using Ambion RNA extraction kit (Invitrogen, Carlsbad, CA) following the manufacturer’s instructions. After quantification, RNA quality was assessed using agarose gels before reverse transcription with Superscript II (Invitrogen) as described.39 Real-time RT-PCR was performed using SYBRGreen PCR master mix and analyzed with a Light Cycler 480 (Roche). All gene expression values were normalized using 18s RNA as a housekeeping gene. All primers used for amplification were from Metabion (Martinsried, Germany). Primers used are listed in the Table 1.

View this table:
Table 1.

Primers used for real-time RT-PCR

Western Blotting and Immunoprecipitation

Cells were lysed in radioimmunoprecipitation assay buffer supplemented with protease inhibitors (Roche) and 20 mmol/L NEM (Calbiochem) followed by centrifugation at 12,000 rpm for 20 minutes.

Protein concentration was determined by Bio-Rad protein assay. For immunoprecipitation reactions, samples were incubated with 2 µg primary antibody overnight at 4°C with Dynabead-conjugated protein G (Dynal). Proteins were separated by SDS-PAGE and then transferred to a polyvinylidene difluoride membrane. The following antibodies were used: anti-mouse MDM2, anti-rabbit β-actin, rabbit p53 anti-mouse (Abcam, Inc.), goat anti-mouse p53 (Santa Cruz Biotechnology), anti-mouse Ub (FK2; Enzo Life Sciences), rabbit anti-mouse LC3B (Cell Signaling Technology), and guinea pig anti-mouse p62 (Progen Biotechnik). The signals were visualized by an enhanced chemiluminescence system (Amersham, Bukinghamshire, United Kingdom).

Statistical Analyses

Values are expressed as means±SEMs. Statistical analyses were performed using Graphpad Prism5 software. Significance of differences was determined by the appropriate two-sided t test for single comparisons. ANOVA with post hoc Bonferroni’s correction was used for multiple comparisons. P values<0.05 were considered statistically significant.

Disclosures

None.

Acknowledgments

We thank Henny Wegner, Regina Maciejewski, Dan Draganovici, and Janina Mandelbaum for their expert technical assistance and Jaclyn Lett.

The work was supported by Deutsche Forschungsgemeinschaft Grants TH1836/1-2 and AN372/11-2 and European Union FP7-REGPOT-2010-1 EnVision Grant 264143. This work is also part of the research project Greifswald Approach to Individualized Medicine (GANI_MED). The GANI_MED Consortium is funded by the Federal Ministry of Education and Research and the Ministry of Cultural Affairs of the Federal State of Mecklenburg-West Pomerania (Grant 03IS2061A). The Washington University in St. Louis Research Core Laboratory for Auditory and Vestibular Studies was funded by National Institutes of Health/National Institute on Deafness and Other Communication Disorders Grant P30-DC04665.

Footnotes

References

  1. ↵
    1. Wiggins RC
    : The spectrum of podocytopathies: A unifying view of glomerular diseases. Kidney Int 71: 1205–1214, 2007pmid:17410103
    OpenUrlCrossRefPubMed
    1. Kriz W,
    2. Lemley KV
    : The role of the podocyte in glomerulosclerosis. Curr Opin Nephrol Hypertens 8: 489–497, 1999pmid:10491745
    OpenUrlCrossRefPubMed
    1. Remuzzi G,
    2. Benigni A,
    3. Remuzzi A
    : Mechanisms of progression and regression of renal lesions of chronic nephropathies and diabetes. J Clin Invest 116: 288–296, 2006pmid:16453013
    OpenUrlCrossRefPubMed
  4. ↵
    1. Wharram BL,
    2. Goyal M,
    3. Wiggins JE,
    4. Sanden SK,
    5. Hussain S,
    6. Filipiak WE,
    7. Saunders TL,
    8. Dysko RC,
    9. Kohno K,
    10. Holzman LB,
    11. Wiggins RC
    : Podocyte depletion causes glomerulosclerosis: Diphtheria toxin-induced podocyte depletion in rats expressing human diphtheria toxin receptor transgene. J Am Soc Nephrol 16: 2941–2952, 2005pmid:16107576
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Thomasova D,
    2. Mulay SR,
    3. Bruns H,
    4. Anders HJ
    : p53-independent roles of MDM2 in NF-κB signaling: Implications for cancer therapy, wound healing, and autoimmune diseases. Neoplasia 14: 1097–1101, 2012pmid:23308042
    OpenUrlCrossRefPubMed
  6. ↵
    1. Bond GL,
    2. Hu W,
    3. Levine AJ
    : MDM2 is a central node in the p53 pathway: 12 years and counting. Curr Cancer Drug Targets 5: 3–8, 2005pmid:15720184
    OpenUrlCrossRefPubMed
  7. ↵
    1. Vassilev LT,
    2. Vu BT,
    3. Graves B,
    4. Carvajal D,
    5. Podlaski F,
    6. Filipovic Z,
    7. Kong N,
    8. Kammlott U,
    9. Lukacs C,
    10. Klein C,
    11. Fotouhi N,
    12. Liu EA
    : In vivo activation of the p53 pathway by small-molecule antagonists of MDM2. Science 303: 844–848, 2004pmid:14704432
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Eischen CM,
    2. Lozano G
    : p53 and MDM2: Antagonists or partners in crime? Cancer Cell 15: 161–162, 2009pmid:19249672
    OpenUrlCrossRefPubMed
  9. ↵
    1. Mulay SR,
    2. Thomasova D,
    3. Ryu M,
    4. Anders HJ
    : MDM2 (murine double minute-2) links inflammation and tubular cell healing during acute kidney injury in mice. Kidney Int 81: 1199–1211, 2012pmid:22297670
    OpenUrlCrossRefPubMed
  10. ↵
    1. Mulay SR,
    2. Thomasova D,
    3. Ryu M,
    4. Kulkarni OP,
    5. Migliorini A,
    6. Bruns H,
    7. Gröbmayr R,
    8. Lazzeri E,
    9. Lasagni L,
    10. Liapis H,
    11. Romagnani P,
    12. Anders HJ
    : Podocyte loss involves MDM2-driven mitotic catastrophe. J Pathol 230: 322–335, 2013pmid:23749457
    OpenUrlCrossRefPubMed
  11. ↵
    1. Vanden Berghe T,
    2. Linkermann A,
    3. Jouan-Lanhouet S,
    4. Walczak H,
    5. Vandenabeele P
    : Regulated necrosis: The expanding network of non-apoptotic cell death pathways. Nat Rev Mol Cell Biol 15: 135–147, 2014pmid:24452471
    OpenUrlCrossRefPubMed
  12. ↵
    1. Endlich N,
    2. Simon O,
    3. Göpferich A,
    4. Wegner H,
    5. Moeller MJ,
    6. Rumpel E,
    7. Kotb AM,
    8. Endlich K
    : Two-photon microscopy reveals stationary podocytes in living zebrafish larvae. J Am Soc Nephrol 25: 681–686, 2014pmid:24309184
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Zhou W,
    2. Hildebrandt F
    : Inducible podocyte injury and proteinuria in transgenic zebrafish. J Am Soc Nephrol 23: 1039–1047, 2012pmid:22440901
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Marine JC,
    2. Lozano G
    : Mdm2-mediated ubiquitylation: p53 and beyond. Cell Death Differ 17: 93–102, 2010pmid:19498444
    OpenUrlCrossRefPubMed
  15. ↵
    1. Watanabe T,
    2. Ichikawa A,
    3. Saito H,
    4. Hotta T
    : Overexpression of the MDM2 oncogene in leukemia and lymphoma. Leuk Lymphoma 21: 391–397, 1996
    OpenUrlCrossRefPubMed
  16. ↵
    1. Allam R,
    2. Sayyed SG,
    3. Kulkarni OP,
    4. Lichtnekert J,
    5. Anders HJ
    : Mdm2 promotes systemic lupus erythematosus and lupus nephritis. J Am Soc Nephrol 22: 2016–2027, 2011pmid:21949095
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Jaramillo-Rangel G,
    2. Ortega-Martínez M,
    3. Sepúlveda-Saavedra J,
    4. Saucedo-Cárdenas O,
    5. Montes-de-Oca-Luna R
    : p53 E3 ubiquitin protein ligase homolog regulates p53 in vivo in the adult mouse eye lens. Mol Vis 19: 2468–2476, 2013pmid:24339722
    OpenUrlPubMed
  18. ↵
    1. Liu G,
    2. Terzian T,
    3. Xiong S,
    4. Van Pelt CS,
    5. Audiffred A,
    6. Box NF,
    7. Lozano G
    : The p53-Mdm2 network in progenitor cell expansion during mouse postnatal development. J Pathol 213: 360–368, 2007pmid:17893884
    OpenUrlCrossRefPubMed
  19. ↵
    1. Montes de Oca Luna R,
    2. Wagner DS,
    3. Lozano G
    : Rescue of early embryonic lethality in mdm2-deficient mice by deletion of p53. Nature 378: 203–206, 1995pmid:7477326
    OpenUrlCrossRefPubMed
  20. ↵
    1. Jones SN,
    2. Roe AE,
    3. Donehower LA,
    4. Bradley A
    : Rescue of embryonic lethality in Mdm2-deficient mice by absence of p53. Nature 378: 206–208, 1995pmid:7477327
    OpenUrlCrossRefPubMed
  21. ↵
    1. Hilliard SA,
    2. Yao X,
    3. El-Dahr SS
    : Mdm2 is required for maintenance of the nephrogenic niche. Dev Biol 387: 1–14, 2014pmid:24440154
    OpenUrlCrossRefPubMed
  22. ↵
    1. Valentin-Vega YA,
    2. Okano H,
    3. Lozano G
    : The intestinal epithelium compensates for p53-mediated cell death and guarantees organismal survival. Cell Death Differ 15: 1772–1781, 2008pmid:18636077
    OpenUrlCrossRefPubMed
  23. ↵
    1. Wanner N,
    2. Hartleben B,
    3. Herbach N,
    4. Goedel M,
    5. Stickel N,
    6. Zeiser R,
    7. Walz G,
    8. Moeller MJ,
    9. Grahammer F,
    10. Huber TB
    : Unraveling the role of podocyte turnover in glomerular aging and injury. J Am Soc Nephrol 25: 707–716, 2014
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Tharaux PL,
    2. Huber TB
    : How many ways can a podocyte die? Semin Nephrol 32: 394–404, 2012pmid:22958494
    OpenUrlCrossRefPubMed
  25. ↵
    1. Hakroush S,
    2. Cebulla A,
    3. Schaldecker T,
    4. Behr D,
    5. Mundel P,
    6. Weins A
    : Extensive podocyte loss triggers a rapid parietal epithelial cell response. J Am Soc Nephrol 25: 927–938, 2013
    OpenUrlPubMed
  26. ↵
    1. Hartleben B,
    2. Gödel M,
    3. Meyer-Schwesinger C,
    4. Liu S,
    5. Ulrich T,
    6. Köbler S,
    7. Wiech T,
    8. Grahammer F,
    9. Arnold SJ,
    10. Lindenmeyer MT,
    11. Cohen CD,
    12. Pavenstädt H,
    13. Kerjaschki D,
    14. Mizushima N,
    15. Shaw AS,
    16. Walz G,
    17. Huber TB
    : Autophagy influences glomerular disease susceptibility and maintains podocyte homeostasis in aging mice. J Clin Invest 120: 1084–1096, 2010pmid:20200449
    OpenUrlCrossRefPubMed
  27. ↵
    1. Galluzzi L,
    2. Vitale I,
    3. Abrams JM,
    4. Alnemri ES,
    5. Baehrecke EH,
    6. Blagosklonny MV,
    7. Dawson TM,
    8. Dawson VL,
    9. El-Deiry WS,
    10. Fulda S,
    11. Gottlieb E,
    12. Green DR,
    13. Hengartner MO,
    14. Kepp O,
    15. Knight RA,
    16. Kumar S,
    17. Lipton SA,
    18. Lu X,
    19. Madeo F,
    20. Malorni W,
    21. Mehlen P,
    22. Nuñez G,
    23. Peter ME,
    24. Piacentini M,
    25. Rubinsztein DC,
    26. Shi Y,
    27. Simon HU,
    28. Vandenabeele P,
    29. White E,
    30. Yuan J,
    31. Zhivotovsky B,
    32. Melino G,
    33. Kroemer G
    : Molecular definitions of cell death subroutines: Recommendations of the Nomenclature Committee on Cell Death 2012. Cell Death Differ 19: 107–120, 2012pmid:21760595
    OpenUrlCrossRefPubMed
  28. ↵
    1. Sperandio S,
    2. de Belle I,
    3. Bredesen DE
    : An alternative, nonapoptotic form of programmed cell death. Proc Natl Acad Sci U S A 97: 14376–14381, 2000pmid:11121041
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Schneider D,
    2. Gerhardt E,
    3. Bock J,
    4. Müller MM,
    5. Wolburg H,
    6. Lang F,
    7. Schulz JB
    : Intracellular acidification by inhibition of the Na+/H+-exchanger leads to caspase-independent death of cerebellar granule neurons resembling paraptosis. Cell Death Differ 11: 760–770, 2004pmid:15017383
    OpenUrlCrossRefPubMed
  30. ↵
    1. Mundel P,
    2. Reiser J,
    3. Zúñiga Mejía Borja A,
    4. Pavenstädt H,
    5. Davidson GR,
    6. Kriz W,
    7. Zeller R
    : Rearrangements of the cytoskeleton and cell contacts induce process formation during differentiation of conditionally immortalized mouse podocyte cell lines. Exp Cell Res 236: 248–258, 1997pmid:9344605
    OpenUrlCrossRefPubMed
  31. ↵
    1. Saleem MA,
    2. O’Hare MJ,
    3. Reiser J,
    4. Coward RJ,
    5. Inward CD,
    6. Farren T,
    7. Xing CY,
    8. Ni L,
    9. Mathieson PW,
    10. Mundel P
    : A conditionally immortalized human podocyte cell line demonstrating nephrin and podocin expression. J Am Soc Nephrol 13: 630–638, 2002pmid:11856766
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Müller T,
    2. Rumpel E,
    3. Hradetzky S,
    4. Bollig F,
    5. Wegner H,
    6. Blumenthal A,
    7. Greinacher A,
    8. Endlich K,
    9. Endlich N
    : Non-muscle myosin IIA is required for the development of the zebrafish glomerulus. Kidney Int 80: 1055–1063, 2011pmid:21849970
    OpenUrlCrossRefPubMed
  33. ↵
    1. White RM,
    2. Sessa A,
    3. Burke C,
    4. Bowman T,
    5. LeBlanc J,
    6. Ceol C,
    7. Bourque C,
    8. Dovey M,
    9. Goessling W,
    10. Burns CE,
    11. Zon LI
    : Transparent adult zebrafish as a tool for in vivo transplantation analysis. Cell Stem Cell 2: 183–189, 2008pmid:18371439
    OpenUrlCrossRefPubMed
  34. ↵
    1. Bollig F,
    2. Perner B,
    3. Besenbeck B,
    4. Köthe S,
    5. Ebert C,
    6. Taudien S,
    7. Englert C
    : A highly conserved retinoic acid responsive element controls wt1a expression in the zebrafish pronephros. Development 136: 2883–2892, 2009pmid:19666820
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Xie J,
    2. Farage E,
    3. Sugimoto M,
    4. Anand-Apte B
    : A novel transgenic zebrafish model for blood-brain and blood-retinal barrier development. BMC Dev Biol 10: 76, 2010pmid:20653957
    OpenUrlCrossRefPubMed
  36. ↵
    1. Grier JD,
    2. Yan W,
    3. Lozano G
    : Conditional allele of mdm2 which encodes a p53 inhibitor. Genesis 32: 145–147, 2002pmid:11857803
    OpenUrlCrossRefPubMed
  37. ↵
    1. Moeller MJ,
    2. Sanden SK,
    3. Soofi A,
    4. Wiggins RC,
    5. Holzman LB
    : Podocyte-specific expression of cre recombinase in transgenic mice. Genesis 35: 39–42, 2003pmid:12481297
    OpenUrlCrossRefPubMed
  38. ↵
    1. Jonkers J,
    2. Meuwissen R,
    3. van der Gulden H,
    4. Peterse H,
    5. van der Valk M,
    6. Berns A
    : Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model for breast cancer. Nat Genet 29: 418–425, 2001pmid:11694875
    OpenUrlCrossRefPubMed
  39. ↵
    1. Allam R,
    2. Pawar RD,
    3. Kulkarni OP,
    4. Hornung V,
    5. Hartmann G,
    6. Segerer S,
    7. Akira S,
    8. Endres S,
    9. Anders HJ
    : Viral 5′-triphosphate RNA and non-CpG DNA aggravate autoimmunity and lupus nephritis via distinct TLR-independent immune responses. Eur J Immunol 38: 3487–3498, 2008pmid:19009528
    OpenUrlCrossRefPubMed
View Abstract
Back to top

In this issue

View Selected Citations (0)
Print
Download PDF
Article Alerts
Email Article
Citation Tools
Request Permissions
Murine Double Minute-2 Prevents p53-Overactivation-Related Cell Death (Podoptosis) of Podocytes
Dana Thomasova, Hauke A. Bruns, Victoria Kretschmer, Martrez Ebrahim, Simone Romoli, Helen Liapis, Ahmed M. Kotb, Nicole Endlich, Hans-Joachim Anders
JASN Jul 2015, 26 (7) 1513-1523; DOI: 10.1681/ASN.2014040345
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Keywords