Fig 1.
Schip1 is expressed by glomerular podocytes.
(A) RT-PCR shows SCHIP1 expression in both glomerulus and the kidney fraction lacking glomeruli. In the glomerulus, expression is mostly detected in FACS-sorted podocytes. As an internal control, expression levels of GAPDH were measured. Controls for markers of various fractions are presented in S3 Fig. (Pod-podocytes, ROG-rest of glomerulus, GLOM-glomerulus, ROK-rest of kidney). (B) Northern blotting on a mouse multiple tissue panel shows the presence of two SCHIP1 mRNA transcripts enriched in the brain, heart, testes and kidney tissues. (C) Northern blotting on mouse glomerulus and ROK tissue confirms stronger SCHIP1 expression in the glomerulus, and presence of two transcripts. (D) By radioactive in situ hybridization on newborn mouse kidney sections SCHIP1 mRNA is localized to developing podocytes of the capillary loop stage glomerulus. (E) By Western blotting, the mouse 55kDa Schip1 protein is detected mostly in the glomerulus. Podocin was used as a positive control for the glomerular fraction, β-actin as a loading control. (F) Immunofluorescence on mouse and human kidney sections shows Schip1 glomerulus signal that partially overlaps with a podocyte foot process marker synaptopodin (Synpo). (G) By immunoelectron microscopy, Schip1 localizes to the glomerular podocyte foot processes (FP) in human kidney sections (GBM-glomerular basement membrane).
Fig 2.
Additional evidence of Schip1 expression in podocytes.
SCHIP1 is significantly upregulated in microarrays from human glomeruli vs. tubuli comparison (A), in mouse podocytes vs. non-podocyte cells (B) and in mouse podocytes (visceral epithelial cells) vs. parietal epithelial cells (C).
Fig 3.
Schip1 is expressed in zebrafish pronephros and its inactivation leads to pericardial edema and loss of podocyte specific GFP-expression.
(A) To confirm the presence of Schip1 in zebrafish, pronephroi were microdissected from the zebrafish line expressing the GFP under podocin promoter. RT-PCR from pronephros (Glom) and rest-of-fish (ROF) fractions shows the signal for Schip1 in pronephros. Podocin was used as the positive control to validate the purity of the pronephros fraction and GAPDH as the loading control. (B) Morpholino injection in zebrafish resulted in the reduction of Schip1 protein as confirmed by Western blot. An increase in Schip1 protein is detected in fish rescued by coinjection with mouse full-length Schip1 mRNA. Zebrafish Schip1 protein band 50kDa. Actin-beta used as loading control. (C) Inactivation of Schip1 led to development of pericardial edema in 96 hpf morphant embryos (upper panels, arrowheads). In podocin-GFP zebrafish line, Schip1 injection caused loss of GFP signal in pronephros (lower panels, arrowheads). (D) Graphs showing quantification of the zebrafish phenotypes. Data presented as mean with SEM of several experiments. WT-wild type, MO-morpholino, PCE-pericardial edema, ctrl-control, GFP-green fluorescent protein.
Fig 4.
Schip1 inactivation in zebrafish causes distortion of podocyte foot processes, podocytopenia and leakage of the filtration barrier.
(A) Brightfield images of PAS stained histological sections showing dilated Bowman’s space and distortion of proper podocyte structures (arrowheads) in Schip1 morphant zebrafish with milder (second panel) and more severe phenotype (third panel). Higher magnifications (size bars) shown in lower panel, including a cell nuclei quantification graph. (B) Electron microscopic analyses of zebrafish morphants show effaced podocyte foot processes with various degrees of damage (middle, arrowheads and right). However, slit diaphragms are consistently present (right, arrowheads). (C) Dye filtration assay in control and Schip1 zebrafish morphants. Rhodamine conjugated 10kDa is freely filtered into the tubuli of both control and morphant fish, whereas FITC labeled 500kDa dye remains in the pronephros in control embryos. In Schip1 morphants 500kDa dye accumulates in the Bowman’s space and also leaks into the tubuli. STD-standard control morpholino, F-fins, T-nephric tubuli, S-somites, B-brain, N-notochord, P-pronephros, G-gut, Y-yolk.
Fig 5.
Schip1 localizes to cell lamellipodia and associates with the cortical actin cytoskeleton.
(A) Both endogenous (arrowheads, upper panel) and ectopic (arrowheads, lower panel) Schip1 localize to lamellipodia in cultured human podocytes. (B) To test Schip1 association to actin-rich lamellipodia regions, transiently transfected human podocytes were treated with the standard procedure (fixation and Triton X-100 permeabilization, upper panel), or incubated with saponin prior to fixation and staining for MycSchip1 (lower panel). Peripheral Schip1 expression is partially retained after saponin treatment, indicating association of the protein with detergent-insoluble cytoskeletal/plasma membrane structures. The same was not observed in control cells transfected with Stx8 (syntaxin 8). (C) Schip1 colocalizes with cortical F-actin in the podocyte lamellipodia. Both the Z- and XY-scanning indicate considerable signal overlap between Schip1 and F-actin along the plasma membrane in cells presenting well-developed lamellipodia. (D) Treatment with latrunculin A results in the dissolution of actin fibers in both control and Schip1-transfected podocytes. However, Schip1 signal remains associated with disturbed F-actin positive residues. In contrast, treatment with cytochalasin D results mostly in preservation of the cortical actin in Schip1 transfected podocytes.
Fig 6.
Schip1 overexpression promotes cortical F-actin accumulation, dissolution of stress fibers and motility in PDGF-BB-stimulated cells by attenuating actin depolymerisation.
(A) In control cells PDGF-BB treatment enhances development of lamellipodia. In Schip1-transfected cells PDGF-BB stimulation induces similar changes but also marked actin cytoskeleton rearrangement with cortical actin accumulation and dissolution of the actin stress fibers (box, zoom). Observe the neighboring non-Myc-Schip1 expressing cells, presenting with normal actin cytoskeleton and pronounced stress fibers. (B) Stable Schip1-expressing and control HEK293 cells were stimulated with 10% FCS or PDGF-BB, scratched, and left to migrate for 24 h (the wound healing assay). Schip1 transfected cells exhibit similar migration rate as controls in medium supplemented with 10% FCS, but migrate faster when induced with PDGF-BB (graph). Microscopic images of control and Schip1-expressing cell monolayers 18 and 24 h after wound scratching (left). (C) In vitro actin polymerization (right panel) and depolymerization (left panel) assays with lysates from GFP–Schip1-expressing HEK293 cells and controls show that Schip1-overexpression slows down actin depolymerization in presence of PDGFBB in comparison to cells treated with 10% FCS (p<0.0001). Results are representative of three separate experiments. RFU-relative fluorescence units.
Fig 7.
Schip1 interacts and colocalizes with Nherf2 and ezrin in vitro.
(A) Myc-tagged Schip1 and Flag-tagged Nherf2 coimmunoprecipitate from lysates of cotransfected HEK293 cells (upper panel). Negative control (MycSchip1 transfection, Flag IP, anti-Myc blot) is shown in lane 4 and positive control (MycSchip transfection, Myc IP, anti-Myc blot) in lane 1. MycSchip1 also interacts with ezrin, a protein known to be in complex with Nherf2, in cotransfected HEK293 (middle panel). Negative control (Ezrin transfection, Myc IP, anti-Ezrin blot) is here in lane 1 and positive control (Ezrin transfection, Ezrin IP, anti-Ezrin blot) in lane 2. Endogenous Schip1 and ezrin interact in pig glomerular lysates as shown by coimmunoprecipitation (lower panel). Whole blots are presented in S4 Fig. OE-overexpression by transfection, IP-immunoprecipitation, WB-Western blot. (B) Schip1 colocalizes with ezrin and Nherf2 in cultured human podocytes to cortical actin zones and lamellipodia. Schip1 and ezrin show very close overlap (arrowheads, boxed area, zoom), whereas Schip1 and Nherf2 colocalize partially (boxed area, zoom). (C) Interaction of Schip1 with Nherf2 and ezrin was confirmed by FRET in cotransfected, fixed podocytes. MycSchip1 signal intensity increases upon Alexa568 bleaching (boxed area, arrowheads) as a result of FRET between the two fluorophores (Alexa568 and Alexa488) suggesting associations of Schip1 with ezrin and Nherf2 proteins (middle and bottom panels). As a positive control, we used GFP-Schip1 stained with anti-GFP- and Alexa Fluor 568-conjugated secondary antibodies (upper panel). We detected about 40% FRET signal increase between Schip1 and ezrin. Lower FRET signal increase of about 10% was detected between Schip1 and Nherf2 (n = 20 ROIs tested in each experiment).
Fig 8.
Schip1, ezrin and Nherf2 colocalize in the podocyte foot processes of the human kidney glomerulus.
(A) Immunofluorescence on human kidney sections shows partial colocalization of Schip/ezrin and Schip/Nherf2 in the glomerulus (arrowheads). (B) By IEM, Schip1 is localized to the podocyte foot processes (FP), often to the apical but also to the basolateral side. Similar localization is seen for ezrin and Nherf2 (arrowheads). (C) Double IEM for Schip1 (10 nm gold particle) and ezrin or Nherf2 (5 nm gold particle) indicates that the proteins colocalize at the same subcellular area in the foot processes (arrowheads, zoom).