Crim1KST264/KST264 Mice Implicate Crim1 in the Regulation of Vascular Endothelial Growth Factor-A Activity during Glomerular Vascular Development
Lorine Wilkinson*,
Thierry Gilbert*,
Genevieve Kinna*,
Leah-Anne Ruta,
David Pennisi*,
Michelle Kett and
Melissa H. Little*
* Institute for Molecular Bioscience, University of Queensland, Brisbane, and Department of Physiology, Monash University, Melbourne, Australia
Address correspondence to: Prof. Melissa H. Little, Institute for Molecular Bioscience, University of Queensland, Brisbane, Australia 4072. Phone: +61-7-3346-2054; Fax: +61-7-3346-2101; E-mail: m.little{at}imb.uq.edu.au
Received for publication September 15, 2006.
Accepted for publication March 15, 2007.
Crim1, a transmembrane cysteine-rich repeat–containingprotein that is related to chordin, plays a role in the tetheringof growth factors at the cell surface. Crim1 is expressed inthe developing kidney; in parietal cells, podocytes, and mesangialcells of the glomerulus; and in pericytes that surround thearterial vasculature. A gene-trap mouse line with an insertionin the Crim1 gene (Crim1KST264/KST264) displayed perinatal lethalitywith defects in multiple organ systems. This study further analyzedthe defects that are present within the kidneys of these mice.Crim1KST264/KST264 mice displayed abnormal glomerular development,illustrated by enlarged capillary loops, podocyte effacement,and mesangiolysis. When outbred, homozygotes that reached birthdisplayed podocyte and glomerular endothelial cell defects andmarked albuminuria. The podocytic co-expression of Crim1 withvascular endothelial growth factor-A (VEGF-A) suggested a rolefor Crim1 in the regulation of VEGF-A action. Crim1 and VEGF-Awere shown to interact directly, providing evidence that cysteine-richrepeat–containing proteins can bind to non–TGF-superfamily ligands. Crim1KST264/KST264 mice display a mislocalizationof VEGF-A within the developing glomerulus, as assessed by immunogoldelectron microscopy and increased activation of VEGF receptor2 (Flk1) in the glomerular endothelial cells, suggesting thatCrim1 regulates the delivery of VEGF-A by the podocytes to theendothelial cells. This is the first in vivo demonstration ofregulation of VEGF-A delivery and supports the hypothesis thatCrim1 functions to regulate the release of growth factors fromthe cell of synthesis.
The critical role that vascular endothelial growth factor-A(VEGF-A) plays in renal vascular development is well established(1). Podocyte expression of VEGF-A is essential for the developmentof the glomerular filtration system (2). Expression from thepresumptive podocytes of the S-shaped body induces the inwardmigration of VEGF-A receptor 2 (flk1)-expressing endothelialcell precursors. The endothelial cells align in close appositionto the podocytes to become a capillary loop. This single loopsplits or branches to form a complex glomerular capillary bedthat is composed of many capillaries that are encased by podocytesand supported by mesangial cells. Within the glomerulus, thepodocytes and endothelial cells both contribute to the glomerularbasement membrane (GBM) that lies between these two cell types(3). Endothelial fenestrae that perforate the glomerular endotheliumand slit diaphragms that are formed along the GBM by foot processesfrom the podocytes allow filtration of small molecular weightmolecules from the blood. Podocyte expression of VEGF-A is thoughtto be necessary for the formation of endothelial fenestrationsin the developing glomerulus, and continued expression in adultkidney suggests that it may play a role in the maintenance ofboth the GBM and the glomerular capillary fenestrations (2,4).A series of transgenic mouse studies of podocyte-specific over-and underexpression or knockout of specific isoforms of VEGF-Asuggest that the availability of VEGF-A in the glomerulus istightly regulated and that small perturbations to levels ofVEGF-A during glomerular development have profound effects onglomerular architecture and function (1,2,5–7).
The Crim1 gene encodes a protein that is expressed in a spatiallyand temporally restricted manner during organogenesis of thelimbs, kidney, lens, pinna, erupting teeth, and testis (8–11).We previously demonstrated that Crim1 can bind to bone morphogeneticprotein-2 (BMP-2), -4, and -7; however, ligand binding occurswithin the cell and not in solution (12). Crim1 forms a complexat the cell surface with the preprotein forms of such ligands,retarding their secretion as mature active dimers (12). Althoughpossibly antagonistic, Crim1-ligand binding may act to regulatethe release of the ligand, with ligand release dependent onan external cue.
We previously characterized the phenotype of the KST264 gene-trapmouse line, which has a -Geo cassette inserted into intron 1of the Crim1 gene. This mouse line is predicted to be a Crim1hypomorph rather than a knockout as a result of the expressionof a minor alternately spliced isoform of Crim1 that seems tobe translated into a short form of the protein (11). The KST264insertion results in perinatal lethality in inbred homozygousanimals (Crim1KST264/KST264) but not in outbred homozygotesand defects in a variety of organ systems, including the limbs,eye, and kidneys (11). Our previous analysis of the Crim1KST264/KST264mouse kidney revealed Crim1 expression in the pericytes of therenal arterial vasculature and in afferent arterioles. Withinthe kidney, Crim1 is expressed in mesangial cells, parietalcells, and podocytes, with expression increasing in these sitesas the glomerulus matures (11). A role for Crim1 in vasculardevelopment was suggested by recent studies that suggested thatknockdown of Crim1 in human umbilical vein endothelial cellsprevented formation of vascular tubes (13), whereas in a zebrafishmodel, Crim1 morphants showed a disruption in hemangiogenesis(14).
In this study, we focused on characterizing the renal defectsin the KST264 gene-trap mouse line. In the kidneys of the Crim1KST264/KST264mice, there was evidence of a defect in the formation of theglomerular capillary loops, illustrated by disruption to endothelialcell adhesion to the GBM, disruption to the mesangial cells,and podocyte effacement. Adult outbred Crim1KST264/KST264 micepresented with albuminuria and a reduced GFR. We showed biochemicallythat the Crim1 protein could bind VEGF-A. Immunogold analysesrevealed a disrupted partitioning of VEGF-A between the podocytesand endothelial cells, and immunofluorescence analysis showedincreased Flk1 activation in the glomerular endothelial cells.These data provide evidence that Crim1 is involved in the tightlycontrolled regulation of VEGF-A delivery in the glomerulus andthat disruption to Crim1 can perturb development of the glomerularfiltration unit.
Maintenance of the KST264 Mouse Line
The maintenance and the production of mice that possess theKST264 gene-trap have been described previously (11). Embryonictissue was obtained from the C57/Bl6 inbred strain, whereasfor adult tissue, the C57/Bl6 were crossed with the outbredCD1 strain. Viable adult Crim1 homozygotes were obtained fromthis line. Use of animals in this study was done in accordancewith the animal ethics committee, University of Queensland (CMCB/535/04/NHMRC)and Monash University. Genotyping was performed as described(11).
Sample Preparation and X-gal Staining
Before X-gal staining, tissue samples were fixed in 4% paraformaldehydein PBS at 4°C for 1 h and washed in PBS at room temperature.Samples were incubated in stain buffer (5 mM potassium ferricyanide[Sigma, St. Louis, MO]), 5 mM potassium ferrocyanide, 2 mM magnesiumchloride, 0.01% sodium deoxycholate, 0.02% NP-40, and 1 mg/mlX-gal) at 37°C for 1 to 4 h. Samples were then washed inPBS before photography and processing for paraffin embedding.
Immunohistochemistry
Immunohistochemistry was performed on deparaffinized and rehydrated4-µm, paraformaldehyde (4%)-fixed, paraffin-embedded sections.Antigen retrieval was performed by treating sections with proteinaseK (0.01 mg/ml in buffer that contained 50 mM Tris [pH 8.0] and5 mM EDTA) for 10 min at 25°C, then endogenous peroxidaseactivity was quenched by incubation of sections in 3% H2O2 inwater for 30 min. For polyclonal antibodies, sections were blocked(2% sheep's serum in PBS) for 1 h and incubated for 2 h withprimary antibodies diluted in blocking serum. Secondary biotinylatedantibody and tertiary streptavidin incubations were performedusing an ABC kit (Vector Laboratories, Burlingame, CA), thenvisualized with the DAB-Plus Substrate kit (Zymed Laboratories,South San Francisco, CA) according to the manufacturer's instructions.For mAb, the MOM kit (Vector Laboratories) was used. Photographswere taken using an Olympus AX70 compound microscope with KodakElite Ektachrome 160T film. Antibodies that were used includedlaminin 5 (15), Wilms' tumor 1 (WT1; Dako, Botany, Australia),CD31 and phospho-Smad 2/3 (Chemicon Int., Temecula, CA), phospho-Smad1/5/8 (Cell Signaling Technology, Danvers, MA), and phospho-Flk1(Merck Pty. Ltd., Victoria, Australia). For immunofluorescence,secondary antibodies used were fluorophore-conjugated anti-mouseAlexa 488 or anti-rabbit Alexa 594 (Molecular Probes, Eugene,OR). Tissue was incubated with the secondary antibody for 45min, washed thoroughly in PBS, and mounted in Vector Shield(Vector Laboratories). Digital images were taken using an OlympusAX70 compound microscope.
Electron Microscopy and Immunogold Labeling of Embryonic Kidneys
Kidneys were dissected from embryos at day 17.5 and processedas already described (16). Briefly, samples for routine ultrastructuralanalysis were fixed in 2% glutaraldehyde in 0.1 M cacodylatebuffer and embedded in Epon resin. For immunoelectron microscopy,samples were fixed in 4% paraformaldehyde and 0.1% glutaraldehyde,partially dehydrated in acetone, and embedded in LR-Gold resinwith polymerization carried out at –25°C. Ultrathinsections were cut on a Leica Ultracut T microtome and viewedunder a JEOL 1010 electron microscope at 80 kV. For VEGF-A immunodetection,80-nm sections were mounted on 200-mesh thin-bar nickel gridspreviously coated with parlodion. Antibodies used were VEGF-A(Chemicon) and anti-mouse secondary coupled to 10 nm of goldparticles (Aurion, Wageningen, The Netherlands). Quantificationof gold particles associated with podocyte, GBM, and capillaryloop endothelium was performed at x50,000 magnification. Fortyareas of the glomerular filtration barrier of juxtamedullaryglomeruli were randomly selected in both wild-type littermatesand Crim1KST264/KST264 17.5 d post coitus (dpc) kidneys.
Cell Transfection
Cos-7 cells were seeded into six-well plates. After 24 h, mediumwas changed to OPTIMEM (Invitrogen, Mulgrave, Australia), andcells were transfected using Fugene (Roche) with constructsthat contained cDNA for the following receptors and ligands:mouse (m) Crim1, human growth hormone receptor (hGHr), mVEGF-A164, mnodal, mPDGF-B, mTGF-2, and mouse placental lactogen II(PLII). Medium or cells were harvested after 48 h.
Immunoprecipitation
For in vitro binding experiments, conditioned medium was clearedby centrifugation. Cell lysates were prepared by addition of600 µl of lysis buffer (0.1% Triton X100 and completeprotease inhibitors [Roche] in PBS) to wells and incubationfor 10 min. Immunoprecipitations were performed by additionof 4 µl of anti-myc or monoclonal VEGF-A antibody (Chemicon,Boronia, Australia) and incubation for 2 h at room temperature.A total of 20 µl of protein A Sepharose slurry (Sigma)was added and incubated for an additional hour. The beads werepelleted and washed three times with RIPA buffer (20 mM Tris[pH 7.4], 0.1% SDS, 1% Triton-X100, 1% deoxycholate, and 0.15M NaCl) and once with PBS. Protein was eluted by addition ofSDS-PAGE loading buffer (0.025 M Tris-HCl [pH 6.8], 4% SDS,10% -mercaptoethanol, 40% glycerol, and 0.05% bromophenol blue)and boiling for 5 min.
Immunofluorescence of Cultured Cells
For observation of cell-surface proteins, primary and secondaryantibody binding was performed on live cells that were incubatedon ice to prevent internalization as described previously (12).Antibody labeling was captured on an Olympus AX70 compound fluorescencemicroscope. Antibodies used were anti-myc (clone 9E10), anti-HA(HA.11; Chemicon), anti-rabbit cy3, and anti-mouse Alexa Fluor488 (Molecular Probes).
Renal Pathophysiology in Adult Crim1KST264/KST264 Mice
Twenty-four-hour conscious urinary excretory profile, anesthetizedrenal function, and renal histopathology were documented inoutbred male mice (Crim1+/+, n = 9; Crim1+/KST264, n = 14; Crim1KST264/KST264,n = 10). At 9 wk of age, trained mice were placed in individualmetabolic cages to obtain a 24-h urine collection. Urinary sodiumand potassium were analyzed using spectrophotometry (SynchronCX5CE Delta; Beckman Coulter, Sydney, Australia), and osmolalitywas measured by freezing point depression (Advanced Osmometer2020; Advanced Instruments, Norwood, MA). Urinary albumin concentrationwas determined using a murine-specific microalbuminuria ELISAkit (Exocell, Philadelphia, PA). Values for albumin were correctedfor urinary creatinine using a creatinine companion assay (Exocell).For anesthetized renal function, mice at 10 wk of age were anesthetizedintraperitoneally with thiobutabarbitone (Inactin 150 mg/kg;Sigma-Aldrich) and ketamine (10 mg/kg; Parnell Laboratories,Alexandria, Australia). Body temperature was maintained at 37.5°C.The trachea was catheterized (PE50 polyvinyl tubing; MicrotubeExtrusions, Melbourne, Australia), and a stream of oxygen wasblown over the end of the trachea tube. The right carotid arterywas catheterized (SV35) for measurement of mean arterial pressure,heart rate, and blood sampling. The right jugular vein was catheterized(PE10) for the infusion of 1% BSA (0.15 ml/h; Sigma-Aldrich)that contained 3H-Inulin (11.16 µCi/ml) and 14C-paraaminohippurate(0.74 µCi/ml; Perkin Elmer, Melbourne, Australia) to enablemeasurement of GFR and effective renal plasma flow, respectively,using standard clearance formulas. The bladder was catheterized(PE20) for the collection of urine. Mice were allowed to equilibratefor 1 h before two consecutive, timed urine collection periodsof approximately 15 min each. At the conclusion of the urinecollection periods, a blood sample was taken via the carotidartery catheter. For adult renal histology, a catheter (SV35)filled with 1 M phosphate buffer (pH 7.4) containing sodiumnitroprusside (Sigma-Aldrich) was placed into the aorta belowthe renal arteries. The kidneys were then perfused clear ofblood with the rinse solution before switching to a solutionthat contained Karnovsky's fixative (2% paraformaldehyde and,2% glutaraldehyde) in 1 M phosphate buffer for 10 min. All subsequentsections were coded such that analysis was performed in a blindmanner.
Fixed left kidneys were cut in half coronally, with one halfprocessed and embedded in paraffin and the remainder subjectedto electron microscopy (EM) as described. Blocks were sectionedat 4 µm and stained with Masson's trichrome method forexamination of gross renal pathology.
Statistical Analyses
Data are presented as means ± SEM. For comparison ofthe distribution of gold particles on the glomerular filtrationbarrier and for adult renal function tests, statistical analysiswas performed using a one-way ANOVA. Differences between groupswere determined by a Dunnett post hoc test.
Crim1 Expression in the Developing Kidney
We previously showed using X-gal staining of the KST264 gene-trapreporter that Crim1 is expressed in podocytes, parietal cells,mesangial cells, and vascular smooth muscle in both the embryonicand the adult kidney (11). In this study, we looked in moredetail at Crim1 expression during kidney development. X-galstaining first appeared at approximately 14 dpc in areas withinthe interstitium (Figure 1A). These cells aligned in cords withinthe interstitium (Figure 1B), ultimately forming the pericytesthat lie along the outside of the endothelial cells in the developingarterial vasculature (Figure 1C). X-gal–positive mesangialcells appeared in the developing glomerular tuft at the capillaryloop stage of renal corpuscle formation (Figure 1, B and D).In the mature embryonic glomerulus, staining was also presentin the podocytes and parietal cells (Figure 1E).
Figure 1. Analysis of Crim1KST264/KST264 kidneys reveals defects in glomerular capillary development. (A through G) LacZ analysis of embryonic Crim1KST264/KST264 kidney. (A) LacZ staining in the interstitium of 14 d post coitus (dpc) Crim1KST264/KST264 kidney marks presumptive pericytes (arrow). (B) LacZ staining in cells that form afferent arteriole (arrow) and in mesangial cells (arrowhead) of capillary loop stage glomerulus. (C) Strong lacZ expression along the walls of a developing afferent arteriole in 15.5 dpc kidney, suggesting expression in the pericytes (arrowhead) but not the endothelial cells of this structure. (D) LacZ staining was clearly evident in the parietal epithelial cells (pe) and mesangial cells (me) of a more mature capillary loop stage glomeruli and in mature glomeruli (17.5 dpc; E). Aberrant inflated glomerular capillary loops (*) in 15.5 dpc (F) showing lacZ staining of the podocytes, me, and pe. (G) LacZ staining of homozygous 15.5 dpc glomerulus showing severely affected glomerulus with a single dilated capillary filled with red blood cells. (H) A wild-type glomerulus showing podocytes (pd) and glomerular capillaries (*) is shown for comparison with Crim1KST264/KST264 glomeruli. (K through I) Thick sections of 17.5 dpc embryonic kidneys showing blood pooling (arrowheads) in Crim1KST264/264 (K) and Crim1+/KST264 (J) kidney compared with a wild-type kidney (I). (L) Hematoxylin and eosin section of 17.5 dpc glomerulus showing blood pooling in the dilated capillaries Bars = 10 µm (A, B, D, F through H, and L), 30 µm (C), 15 µm (E), and 300 µm (I through K).
Glomerular Capillary, Podocyte, and Mesangial Cell Defects in Embryonic Crim1KST264/KST264 Mice
Simplification and dilation of the glomerular capillaries wasevident from 15.5 dpc in glomeruli from both Crim1+/KST264 andCrim1KST264/KST264 mice (Figure 1F) and persisted in 17.5 dpcCrim1+/KST264 and Crim1KST264/KST264 glomeruli (Figure 1, Fand G) compared with wild-type mice (Figure 1H). At a macroscopiclevel, in comparison with wild-type E17.5 dpc kidney (Figure 1I),there was visible accumulation of blood within the glomeruliin Crim1+/KST264 (Figure 1J) and Crim1KST264/KST264 (Figure 1K)kidneys, and pooling of erythrocytes in some dilated capillarieswas evident in sections (Figure 1L). Thin section histologyand EM were used to characterize further the glomerular defectpresent. This analysis revealed that Crim1KST264/KST264 glomerulidisplayed multiple lesions of the glomerulus (Figure 2, D throughI) as compared with wild-type glomeruli (Figure 2, A throughC). Podocyte differentiation was severely impaired with defectsin foot process formation (Figure 2, F through H). Althoughthis feature was prominent, normal slit diaphragms were occasionallyobserved (Figure 2, F and G). The GBM was thin, as expectedfor this stage of development, and did not differ from thatof the wild-type littermates (Figure 2, B versus G). Nevertheless,irregular GBM was observed in the most damaged glomeruli (Figure 2H).Although light microscopy failed to reveal any changes in glomerularendothelial cells, electron micrographs showed that they werefrequently detached from the GBM (Figure 2F). In addition, somecapillaries exhibited signs of endothelium overgrowth with numerouscytoplasmic protrusions in the lumen (Figure 2D). In the glomerulartuft, mesangiolysis was present in the most mature glomeruliand was observed without any endothelial cell loss (Figure 2,E and I). Within the dilated lumens of these abnormal vasculartufts, there seemed to be many erythrocytes (Figure 2, A versusE). This may indicate that this structural anomaly was accompaniedby a reduction in pressure and a pooling of red blood cells.
Figure 2. Ultrastructural analysis of the kidney of Crim1KST264/KST264 mice. Light and electron micrographs were taken from embryonic day 17.5 control (A through C) and Crim1KST264/KST264 mice (D through I). In control glomeruli, mesangial cells (A and C, thick arrows) provide structural support to the developing capillary loops. Numerous endothelial cells are present (A, thin arrows). At higher magnification, the glomerular basement membrane (GBM; B, arrowheads) separates the capillary loop from the podocytes. At this stage, the glomerular endothelium is not fully fenestrated, and podocytes from the more mature glomeruli have developed numerous foot processes (B, arrows). In Crim1KST264/KST264 mice, glomeruli are enlarged and filled with numerous red blood cells (E). Endothelial cells that are focally detached from the GBM (F, arrows) and foot processes are obliterated (G, open arrows), although some slit diaphragm membranes can still be observed (F, arrowheads). In more severely affected glomeruli, the GBM is irregular and thickened (H, arrows). Mesangiolysis is prominent in the more mature glomeruli (E and I, *). cap, glomerular capillary; ec, endothelial cell; mc, mesangial cell; p, podocyte. Bars = 20 µm (A and E), 2 µm (B through D, H, and I), and 1 µm (F and G).
The presence of glomerular capillary abnormalities in Crim1KST264/KST264mice led us to examine more carefully the identity of the lacZ-positivecells within the developing renal corpuscle. With the use ofWT1 to mark the podocytes, immunohistochemistry confirmed thatthe podocytes of Crim1+/KST264 mice were lacZ positive (Figure 3A).Using antibodies for CD31/platelet-endothelial cell adhesionmolecule-1 to mark endothelial cells, we also confirmed thatthese cells were not lacZ positive (Figure 3B). WT1-positivepodocytes seemed to be reduced in number in Crim1KST264/KST264versus wild-type mice (Figure 3, C and D). This does not provea reduction in podocyte number per se but demonstrates a potentialloss of identity, which may correlate with the structural defectsthat were observed in the podocytes of these mice. CD31 immunohistochemistryshowed the clear presence of endothelial cells in both wild-typeand Crim1KST264/KST264 mice (Figure 3, E and F) but confirmedthe ultrastructural observations of an abnormal capillary networkin the mutants. Analysis of the Crim1 KST264/KST264 mice confirmedthe presence of desmin-positive mesangial cells (Figure 3, Gand H). Laminin 5, a component of the GBM that is importantfor interactions among the mesangial cells, podocytes, and endothelialcells (17,18), was also present but appeared disordered andthickened (Figure 3, I and J), suggestive of alterations tothe glomerular filtration barrier.
Figure 3. Immunohistochemical analysis of glomerular defects in Crim1KST264/KST264 mice. Sections were assessed by immunohistochemistry (A, B, and E through J) or immunofluorescence (C and D) using the antibodies indicated. A and B were stained for lacZ before sectioning to highlight expression of the Crim1 transgene. C through J are from 17.5 dpc embryos, whereas A and B are from 15.5 dpc embryos. (A) LacZ staining showed co-localization with the podocyte marker Wilms' tumor 1 (WT1). (B) CD31, which marks endothelial cells, was not co-localized with lacZ. WT1 immunofluorescence revealed fewer podocytes in the mature homozygous glomerulus compared with the wild-type (C and D). CD31 staining outlined the glomerular capillaries showing the gross dilation of the capillaries in the homozygous glomerulus (F) compared with the wild-type (E). Desmin staining (G and H) was equivalent in both wild-type and homozygous sections. Laminin 5 staining was present in both samples (I and J) but appeared abnormal in the Crim1KST264/KST264 mice (J), suggesting a GBM defect. Bars = 15 µm.
Evidence of Glomerular Defects in Adult Crim1KST264/KST264 Kidneys
As reported previously, Crim1KST264/KST264 mice on a C57/Bl6background are perinatal lethal. Outbreeding of the KST264 lineonto a CD1 background for one generation resulted in the birthof a proportion of homozygotes that survived to adulthood. Ratiosat birth for these litters reveal that only 15% of mice wereCrim1KST264/KST264, suggesting that a significant proportionof homozygotes still did not survive. We examined the pathologyand renal physiology of both Crim1+/KST264 and Crim1KST264/KST264outbred mice (Figure 4, A through F). A distinct feature ofadult kidneys of Crim1KST264/KST264 mice (Figure 4, C, D, F,and H through J) was the presence of small cysts and areas oftubular dilation (Figure 4, C and D). In addition, glomeruliand Bowman's capsule were markedly larger in Crim1KST264/KST264than in wild-type glomeruli and extended to the outer most regionsof the cortex (Figure 4, A versus C and D). Large glomerularcysts were also present (Figure 4E). The glomerular tuft inthese large cysts was abnormal and resembled collapsing glomerulopathy(Figure 4F). Crim1KST264/KST264 kidneys also showed areas ofinterstitial fibrosis and glomerular obsolescence (Figure 4D).Crim1+/KST264 kidneys seemed relatively normal, apart from occasionalareas of tubular dilation (Figure 4B) and interstitial fibrosis.
Figure 4. Evidence of glomeruli defects in adult Crim1KST264/KST264 kidneys. Light micrographs of the outer cortex of kidneys from Crim+/+ (A), Crim1+/KST264 (B), and Crim1KST264/KST264 (C and D) mice stained with Masson's trichrome. Crim1KST264/KST264 mice demonstrate numerous enlarged glomeruli extending into the outer cortex (C), glomerular obsolescence (D, *), tubular dilation, and areas of interstitial fibrosis (D, arrow). Cystic glomerulus showing enlarged Bowman's space (BS; E), and higher magnification showing collapse of the glomerular tuft (F). Electron micrographs of Crim+/+ (G) and Crim1KST264/KST264 (H) kidneys. Capillaries (C) from Crim1KST264/KST264 mice demonstrated areas of podocyte effacement and apparent loss of endothelial fenestrae. Albumin-to-creatinine excretion ratio was increased three-fold in Crim1KST264/KST264 adult mice compared with wild-type (I). Bars = 50 µm (A through F) and 2 µm (G and H).
Whereas some Crim1KST264/KST264 glomeruli demonstrated normalglomerular filtration barriers, many glomeruli demonstratedsegments of podocyte foot process flattening and effacement(Figure 4, H versus G). There was also evidence of loss of endothelialfenestrae and thickening of the endothelial cells (Figure 4H).The GBM did not seem to be different from that of wild-typemice (Figure 4, H versus G). The glomeruli of Crim1+/KST264mice seemed normal (data not shown).
Crim1KST264/KST264 Mice Are Albuminuric and Show a Reduced GFR
Adult Crim1+/KST264 and Crim1KST264/KST264 mice did not differfrom wild-type mice in body weight or 24-h urine, sodium, potassium,or osmolality excretion (Table 1). Crim1KST264/KST264 mice,however, showed a three-fold increase in 24-h albumin excretionand albumin-to-creatinine excretion ratio (Figure 4I). Crim1+/KST264mice showed no changes in 24-h albumin excretion. GFR of Crim1KST264/KST264mice was 40% lower than that in wild-type mice (P < 0.05;Table 2). Although not statistically significant, values foreffective renal plasma flow of Crim1KST264/KST264 mice werelower than those of wild-type mice (Table 2). Crim1KST264/KST264mice were not different from wild-type mice with respect toanesthetized mean arterial pressure, hematocrit, left ventricleweight, or left kidney weight, although the kidney-to-body weightratio of Crim1KST264/KST264 mice tended (P < 0.066) to begreater than that of wild-type mice (Table 2). Crim1+/KST264mice did not differ from wild-type mice (Table 2).
Table 2. Anesthetized renal function in adult Crim1KST264 mice
Crim1 Can Bind to Cystine-Knot–Containing Growth Factors Other than Members of the TGF- Superfamily, Including VEGF-A
The vascularization of the kidney is regulated, as in otherorgans, by the production of VEGF-A. VEGF-A is strongly expressedin the podocytes of the developing kidney (19). Although nota member of the TGF- superfamily, VEGF-A is structurally relatedto members of this family because all these proteins containa cystine-knot motif (20). Given the coexpression of Crim1 andVEGF-A, we used co-immunoprecipitation to demonstrate an interactionbetween Crim1 and VEGF-A (Figure 5A). To control for artifactualbinding, we showed that there was no interaction between VEGF-Aand another transmembrane protein, growth hormone receptor (Figure 5A).Immunoprecipitation showed Crim1 binds to a broad range of membersof the TGF- superfamily, including TGF-2 and nodal, as wellas another growth factor, PDGF, a protein that is more similarin structure to VEGF-A (Figure 5A). This suggests a broaderligand-binding capacity than previously described for Crim1.The commonality of these distinct protein families is the existenceof a cystine-knot motif (21). PLII, a growth factor from thegrowth hormone family that does not contain a cysteine-knotmotif, did not bind Crim1 (Figure 5A), supporting a model inwhich Crim1 binds to the growth factor via the cysteine-knotmotif. Using co-immunofluorescence, we showed that VEGF-A wastethered to the cell surface when coexpressed with Crim1 (Figure 5B).Using a series of Crim1 deletion constructs described previously(12) (Figure 5C), we determined that cysteine-rich repeat (CRR)domains in Crim1 are essential for VEGF-A binding. Deletionconstructs in which these domains were missing displayed nobinding to VEGF-A by co-immunoprecipitation (Figure 5D). Theseresults are similar to those found for Crim1 binding to BMP-4(12). However, whereas Crim1 maintained BMP-4 binding to a reducedextent when at least some of these repeats were present, VEGF-Aseems to require the full set of CRR for binding.
Figure 5. Evidence for an interaction between vascular endothelial growth factor-A (VEGF-A) and Crim1 in vivo. (A) Immunoprecipitation analyses of Cos-7 cells that were transfected with tagged Crim1 (CRIM1-HA), VEGF-A-myc (VEGF-A), TGF-2-myc (TGF2), Nodal-myc (Nodal), PDGF-B-myc (PDGFB), placental lactogen II (PLII), and growth hormone receptor (GHR-HA) either alone or in combination. Cell lysates and media were immunoprecipitated using an anti-myc antibody (IP-@myc). Western blots were then probed with anti-HA (IB-@HA). VEGF-A was co-immunoprecipitated with Crim1 but not with GHR. TGF-2-myc, Nodal-myc, and PDGFB-myc were co-immunoprecipitated with Crim1. Western analysis of aliquots of lysate before immunoprecipitation confirmed that all constructs were expressed (bottom panels, IB-@HA and @myc). (B) Immunofluorescence was performed on Cos-7 cells that were co-transfected with the same constructs that were used for immunoprecipitation and visualized using anti-HA and anti-myc antibodies. VEGF-A was visible on the cell surface when co-transfected with Crim1 but not when co-transfected with GHR. (C) Schematic of Crim1 deletion constructs used for immunoprecipitation in D. S, signal sequence; 1 to 6, cysteine-rich repeat (CRR) repeats 1 to 6; T, transmembrane domain; myc, site of introduced myc tag. (D) Immunoprecipitation analysis of Crim1 deletion constructs with VEGF-A. Cos-7 cells were co-transfected with a full-length Crim1 construct or the deletion constructs as indicated and VEGF-A-myc. Culture medium was immunoprecipitated (IP) with anti–VEGF-A, and Western analysis (IB) was performed using anti-myc. Cell lysates were subjected to Western analysis (IB) using anti-myc to ensure that transfection was successful. Only full-length Crim1 and deletions D1 and D3, which all contain the full six CRR, co-precipitated with VEGF-A. (E and F) Localization of VEGF-A expression in the wild-type (E) and Crim1KST264/KST264 (F) kidney by immunogold electron microscopy. Gold particles can be clearly seen in podocyte (p), GBM, and endothelial cell (ec) compartments (arrows). Bar = 0.5 µm. (G) Quantification of the distribution of VEGF-A over the glomerular filtration barrier of wild-type and Crim1KST264 mice. VEGF-A distribution is clearly more diffuse in the Crim1KST264/KST264 mice. *P < 0.01.
Loss of Crim1 Disrupts the Partitioning and Signaling of VEGF-A during Glomerular Development
Given this evidence for an interaction between VEGF-A and Crim1and their coexpression in the podocytes, we examined the possibilitythat the absence of full-length Crim1 protein in the Crim1KST264/KST264mice resulted in an impaired VEGF-A partitioning between thepodocytes and the endothelial cells in vivo. Using immunogoldEM, we analyzed the distribution of VEGF-A across the glomerularfiltration barrier in 17.5 dpc kidneys. In wild-type embryonickidneys, 65% of VEGF-A protein was present on the podocytes,8% within the GBM, and 27% over the endothelial cells (Figure 5,E and G). This is consistent with data that were gained frommature adult glomeruli (22). In contrast, immunogold localizationof VEGF-A in Crim1KST264/KST264 mice showed a reduction in podocyticlocalization (45%) and a concomitant increase in localizationof VEGF-A over the endothelial compartment (48%) with no changeover the GBM (6%; Figure 5, F and G).
Loss of Crim1 Results in Activation of Flk1 in Glomerular Endothelial Cells
VEGF-A primarily signals through the receptor Flk1, which isexpressed on endothelial cells in close proximity to VEGF-A–expressingcells (23). Upon VEGF-A binding, Flk1 dimerizes and autophosphorylates,beginning the signaling cascade that culminates in angiogenesisand vasculogenesis. Considering the altered distribution ofVEGF-A in the Crim1 KST264/KST264 glomerulus, we compared thelevels of Flk1 phosphorylation between the wild-type and Crim1KST264/KST26415.5 and 17.5 dpc embryo. Co-immunofluorescence of kidney sectionswas performed using an antibody against WT1 to mark podocytesand a phospho-specific Flk1 antibody. Results revealed the endothelialcell membrane localization of phospho-Flk1 and showed that itis markedly increased in the 17.5 dpc Crim1KST264/KST264 glomerulifrom capillary loop stage onward (Figure 6A) compared with wild-type(Figure 6B). Results for 15.5 dpc showed a similar increasein phospho-flk1, but levels overall for both wild-type and Crim1KST264/KST264glomeruli were lower (data not shown). Because Crim1 can bindmembers of the TGF- superfamily (12), we assessed whether therewas evidence for altered TGF- superfamily signaling in Crim1KST264/KST264mice using immunohistochemistry for phospho-Smad2/3 and phospho-Smad1/5/8. We could not detect any difference in nuclear phospho-Smad2/3or phospho-Smad 1/5/8 between wild-type and Crim1 KST264/KST264glomeruli at 15.5 dpc (data not shown) or at 17.5 dpc for phospho-Smad2/3 (Figure 6, C and D). However, at 17.5 dpc, there was a reductionin the levels of phospho-Smad 1/5/8 in the Crim1 KST264/KST264glomeruli (Figure 6E) compared with wild-type (Figure 6F). Thedecrease in phospho-Smad1/5/8 in the glomerulus was largelyconfined to the podocytes (Figure 6F) in mature glomeruli.
Figure 6. Loss of Crim1 results in activation of the VEGF-A pathway but not TGF- pathways. For assessment of which growth factor pathways were affected in the Crim1KST264/KST264 mutants, immunofluorescence (A and B) and immunohistochemistry (C through F) of 17.5 dpc glomeruli were performed using antibodies to the phosphorylated (activated) form of downstream affectors of the VEGF-A, TGF-1 and 2, and bone morphogenic protein pathways. Immunofluorescence using anti–phospho-Flk1(red) and anti-WT1 (green) to mark podocytes shows that the phosphorylated form of Flk1 was present to a greater extent in the Crim1KST264/KST264 glomerular endothelial cells (A) than in the wild-type glomerulus (B). In contrast, immunohistochemistry using antibodies to phospho-Smad 2/3 (C and D) showed no difference between wild-type (C) and Crim1KST264/KST264 (D) glomeruli. Phospho-Smad1/5/8 levels (E and F) were reduced in podocytes of the Crim1KST264/KST264 glomerulus compared with the wild-type.
Crim1 Acts to Regulate Presentation of Many Growth Factors at the Cell Surface
Crim1 was previously proposed to act via modulation of the TGF-superfamily, and we have verified that Crim1 can interact withBMP via its CRR motifs (12). In this study, we revealed a widerdegree of promiscuity with respect to Crim1 ligand binding,showing, for the first time, an interaction with ligands outsidethe TGF- superfamily, including an interaction between Crim1and VEGF-A. This is not surprising because, in elucidating thethree-dimensional structure of chordin-like repeats, it wasshown that these repeats contain a structurally equivalent foldto that of the VEGF-A binding domain of fibronectin type I (24).Furthermore, both the TGF-/BMP and the VEGF/PDGF contain a cystine-knotmotif (21), and connective tissue growth factor, a CRR-containingprotein, is known to bind VEGF-A (25). Crim1 may therefore acton multiple cystine-knot–containing ligands. In supportof this, we found that Crim1 does not bind to PLII, a growthfactor that does not contain a cystine-knot motif (26). Althougha number of different splice variants of VEGF-A are expressed,the region coding for the cystine knot (exons 2 to 5) is presentin all currently known variants (20,27), so Crim1 would be expectedto bind all VEGF-A isoforms.
Many growth factors are sequestered to the cell surface at somepoint after secretion, and the importance of cell-surface factorbinding for normal function is becoming increasingly apparent(6,28–30). Crim1 must be coexpressed with its bindingpartner and can act to tether the preprotein forms of BMP atthe cell surface (12). Although VEGF-A and PDGF are not producedas preproteins that require furin processing for activation,the longer isoforms of VEGF-A are known to be released fromthe cell surface via specific cell-surface proteases such asplasmin and urokinase type plasminogen activator, and VEGF-A189requires enzyme maturation for full receptor-binding activity(20). Hence, the tethering effect of Crim1 on VEGF-A may similarlyaffect rates of maturation and effective concentration gradients.Release of the active ligand may in part depend on cleavageof Crim1 from the cell surface (12), as with chordin and connectivetissue growth factor, which both require cleavage by metalloproteasesto release active ligand (25,31).
The gene-trap insertion that is present in the Crim1KST264/KST264mice is predicted to splice out exon 2. The loss of this regionalone would not affect the ability to bind to VEGF-A. Our previousanalysis of this mouse shows that the protein that is producedfrom the exon 2 deletion isoform also lacks the CRR region,resulting in a nonfunctional protein with respect to VEGF-Atethering. However, it remains possible that some aspects ofthe defect result more directly from the presence of this shorterversion of the protein.
Defects in Glomerular Capillaries Result from an Increase in VEGF-A Activity
The Crim1KST264/KST264 mouse presents with defects in glomerulardevelopment, including dilated capillaries, endothelial cellovergrowth, pooling of red cells in these capillaries, mesangiolysis,and podocyte effacement. Glomeruli in the adult mice also showpodocyte effacement, endothelial cell thickening, loss of endothelialcell fenestrae, and glomerular cysts. We have shown here thatthe loss of full-length Crim1 affects the regulation of VEGF-Adelivery to the glomerular endothelial cells, resulting in increasedactivation of the VEGF-A signaling pathway in these cells. Thisdysregulation would be expected to mimic most closely the mousemodel of podocyte-specific overexpression of VEGF-A (5) or themodel in which two of the cell surface binding isoforms of VEGF-A(VEGF-A164 and VEGF-A189) have been knocked out (6). The firstof these models shows a collapsing glomerular nephropathy, whereasthe latter shows glomerular sclerosis. Neither of these phenotypesexactly portrays the phenotype presented by the Crim1KST264/KST264mouse, although examples of collapsing glomerulopathy were seenin the adult homozygote. The extensive series of VEGF-A transgenicsdo show that regulation of VEGF-A is tightly controlled withonly minor changes resulting in profound effects on glomerulardevelopment. Our study would support this by representing anothersubtle variation in VEGF-A availability and hence activity.That the glomerular phenotype persists in the adult, in particularthe loss of endothelial fenestrae, suggests that regulationof VEGF-A continues to be important in the maintenance of thefiltration barrier. The mesangial defects may be secondary tochanges in endothelial cell architecture or GBM defects thatresult from the podocyte anomalies or may reflect a primarydefect given the expression of Crim1 in the mesangial cellsthemselves.
Crim1KST264/KST264 Mice Display a Complex Phenotype That Is Likely to Result from Disruptions to Multiple Growth Factor Pathways
The glomerular defects that are present in the embryonic Crim1KST264/KST264mice, in particular the dilated and simplified capillary network,resemble those that are found in other knockout models, suchas the integrin 3 (32), PDGF-B (33), and PDGF- receptor knockoutmice (34). Although it is possible that one or all of theseproteins directly interact with Crim1, their absence may simplyaffect the development of similar components of the glomerulusto elicit a similar phenotype. We have shown that Crim1 canbind PDGF-B, but PDGF-B is expressed in the endothelium andnot coexpressed with Crim1. We previously showed Crim1 bindingof BMP-7 (12). BMP-7 is expressed in the podocytes and is theonly BMP expressed in the mature glomerular tuft (35). Expressionlevels of BMP-7 have been reported to decrease in rat and mousemodels of ischemia/reperfusion and ureteral obstruction (36,37).Because Crim1 and BMP-7 are coexpressed in the podocyte, itis possible that BMP-7 signaling is also compromised in Crim1KST264/KST264kidneys. Smad1/5/8 are the intracellular signaling moleculeswithin the BMP pathway (32). Because the reduction in phospho-Smad1/5/8levels observed occurs only in the podocytes of the mature Crim1KST264/KST26417.5 dpc glomeruli, this most likely reflects late podocytedamage rather than being a direct result of Crim1 loss.
Our demonstration of a glomerular defect that results from adisruption to the VEGF-A gradient across the developing vasculartuft adds to the growing literature regarding the need for tightcontrol of VEGF-A in glomerular development. Crim1 seems tobe a promiscuous regulator of growth factor delivery, becausewe have shown binding of Crim1 to a variety of cystine-knot–containinggrowth factors. Hence, the many defects reported in these mice,including renal and nonrenal anomalies, may reflect disruptionsto gradients of more than one of these factors. This pleiotropicrole for Crim1 in the modulation of delivery and effective activityof a number of growth factors has far-reaching implicationsfor the biology of related proteins, including chordin. Thebiology of Crim1 will best be dissected using a Crim1 allelicseries of tissue-specific knockout mice as has been appliedto VEGF-A itself.
This work was funded by the National Health and Medical ResearchCouncil of Australia (grant 301056), and the Transgenic AnimalService of Queensland is supported by the Australian ResearchCouncil Special Research Centre for Functional and Applied Genomics.T.G. was funded by INSERM (Institut National de la Santéet de la Recherche Médicale) France. M.H.L. is an NationalHealth and Medical Research Council Principal Research Fellow.
We thank Marc Tessier-Levigne and Jeff Searle for scientificadvice and technical assistance and Margaret Cummings for adviceon pathology. The laminin 5 antibody was a gift from Jeff Miner.We acknowledge the support of staff in the School of BiomedicalSciences Animal Facility, the Transgenic Animal Service of Queensland,and the Queensland Histology Service, Queensland Institute ofMedical Research. We thank the Centre for Microscopy and Microanalysisof the University of Queensland.
We remember our colleague Toshiya Yamada.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Wakelin SJ, Marson L, Howie SE, Garden J, Lamb JR, Forsythe JL: The role of vascular endothelial growth factor in the kidney in health and disease.
Nephron Physiol 98
: 73
–79, 2004
Eremina V, Quaggin SE: The role of VEGF-A in glomerular development and function.
Curr Opin Nephrol Hypertens 13
: 9
–15, 2004[Medline]
Sariola H, Timpl R, von der Mark K, Mayne R, Fitch JM, Linsenmayer TF, Ekblom P: Dual origin of glomerular basement membrane.
Dev Biol 101
: 86
–96, 1984[CrossRef][Medline]
Eremina V, Sood M, Haigh J, Nagy A, Lajoie G, Ferrara N, Gerber HP, Kikkawa Y, Miner JH, Quaggin SE: Glomerular-specific alterations of VEGF-A expression lead to distinct congenital and acquired renal diseases.
J Clin Invest 111
: 707
–716, 2003[CrossRef][Medline]
Mattot V, Moons L, Lupu F, Chernavvsky D, Gomez RA, Collen D, Carmeliet P: Loss of the VEGF(164) and VEGF(188) isoforms impairs postnatal glomerular angiogenesis and renal arteriogenesis in mice.
J Am Soc Nephrol 13
: 1548
–1560, 2002[Abstract/Free Full Text]
Eremina V, Cui S, Gerber H, Ferrara N, Haigh J, Nagy A, Ema M, Rossant J, Jothy S, Miner JH, Quaggin SE: Vascular endothelial growth factor a signaling in the podocyte-endothelial compartment is required for mesangial cell migration and survival.
J Am Soc Nephrol 17
: 724
–735, 2006[Abstract/Free Full Text]
Kolle G, Georgas K, Holmes GP, Little MH, Yamada T: CRIM1, a novel gene encoding a cysteine-rich repeat protein, is developmentally regulated and implicated in vertebrate CNS development and organogenesis.
Mech Dev 90
: 181
–193, 2000[CrossRef][Medline]
Georgas K, Bowles J, Yamada T, Koopman P, Little MH: Characterisation of Crim1 expression in the developing mouse urogenital tract reveals a sexually dimorphic gonadal expression pattern.
Dev Dyn 219
: 582
–587, 2000[CrossRef][Medline]
Lovicu FJ, Kolle G, Yamada T, Little MH, McAvoy JW: Expression of Crim1 during murine ocular development.
Mech Dev 94
: 261
–265, 2000[CrossRef][Medline]
Pennisi DJ, Wilkinson L, Kolle G, Sohaskey ML, Gillinder K, Piper MJ, McAvoy JW, Lovicu FJ, Little MH: Crim1(KST264/KST264) mice display a disruption of the Crim1 gene resulting in perinatal lethality with defects in multiple organ systems.
Dev Dyn 236
: 502
–511, 2007[CrossRef][Medline]
Wilkinson L, Kolle G, Wen D, Piper M, Scott J, Little M: CRIM1 regulates the rate of processing and delivery of bone morphogenetic proteins to the cell surface.
J Biol Chem 278
: 34181
–34188, 2003[Abstract/Free Full Text]
Glienke J, Sturz A, Menrad A, Thierauch KH: CRIM1 is involved in endothelial cell capillary formation in vitro and is expressed in blood vessels in vivo.
Mech Dev 119
: 165
–175, 2002[CrossRef][Medline]
Kinna G, Kolle G, Carter A, Key B, Lieschke GJ, Perkins A, Little MH: Knockdown of zebrafish crim1 results in a bent tail phenotype with defects in somite and vascular development.
Mech Dev 123
: 277
–287, 2006[CrossRef][Medline]
Miner JH, Patton BL, Lentz SI, Gilbert DJ, Snider WD, Jenkins NA, Copeland NG, Sanes JR: The laminin alpha chains: Expression, developmental transitions, and chromosomal locations of alpha1–5, identification of heterotrimeric laminins 8–11, and cloning of a novel alpha3 isoform.
J Cell Biol 137
: 685
–701, 1997[Abstract/Free Full Text]
Gilbert T, Cibert C, Moreau E, Geraud G, Merlet-Benichou C: Early defect in branching morphogenesis of the ureteric bud in induced nephron deficit.
Kidney Int 50
: 783
–795, 1996[Medline]
Miner JH, Li C: Defective glomerulogenesis in the absence of laminin alpha5 demonstrates a developmental role for the kidney glomerular basement membrane.
Dev Biol 217
: 278
–289, 2000[CrossRef][Medline]
Miner JH, Cunningham J, Sanes JR: Roles of laminin in embryogenesis: Exencephaly, syndactyly, and placentopathy in mice lacking laminin alpha5 chain.
J Cell Biol 143
: 1713
–1723, 1998[Abstract/Free Full Text]
Brown LF, Berse B, Tognazzi K, Manseau EJ, Van de Water L, Senger DR, Dvorak HF, Rosen S: Vascular permeability factor mRNA and protein expression in human kidney.
Kidney Int 42
: 1457
–1461, 1992[Medline]
Robinson CJ, Stringer SE: The splice variants of vascular endothelial growth factor (VEGF) and their receptors.
J Cell Sci 114
: 853
–865, 2001[Abstract]
Vitt UA, Hsu SY, Hsueh AJ: Evolution and classification of cystine knot-containing hormones and related extracellular signaling molecules.
Mol Endocrinol 15
: 681
–694, 2001[Abstract/Free Full Text]
Foster RR, Hole R, Anderson K, Satchell SC, Coward RJ, Mathieson PW, Gillatt DA, Saleem MA, Bates DO, Harper SJ: Functional evidence that vascular endothelial growth factor may act as an autocrine factor on human podocytes.
Am J Physiol Renal Physiol 284
: F1263
–F1273, 2003[Abstract/Free Full Text]
O'Leary JM, Hamilton JM, Deane CM, Valeyev NV, Sandell LJ, Downing AK: Solution structure and dynamics of a prototypical chordin-like cysteine-rich repeat (von Willebrand factor type C module) from collagen IIA.
J Biol Chem 279
: 53857
–53866, 2004[Abstract/Free Full Text]
Hashimoto G, Inoki I, Fujii Y, Aoki T, Ikeda E, Okada Y: Matrix metalloproteinases cleave connective tissue growth factor and reactivate angiogenic activity of vascular endothelial growth factor 165.
J Biol Chem 277
: 36288
–36295, 2002[Abstract/Free Full Text]
Bole-Feysot C, Goffin V, Edery M, Binart N, Kelly PA: Prolactin (PRL) and its receptor: Actions, signal transduction pathways and phenotypes observed in PRL receptor knockout mice.
Endocr Rev 19
: 225
–268, 1998[Abstract/Free Full Text]
Muller YA, Li B, Christinger HW, Wells JA, Cunningham BC, de Vos AM: Vascular endothelial growth factor: Crystal structure and functional mapping of the kinase domain receptor binding site.
Proc Natl Acad Sci U S A 94
: 7192
–7197, 1997[Abstract/Free Full Text]
Ohkawara B, Iemura S, ten Dijke P, Ueno N: Action range of BMP is defined by its N-terminal basic amino acid core.
Curr Biol 12
: 205
–209, 2002[CrossRef][Medline]
Constam DB, Robertson EJ: Regulation of bone morphogenetic protein activity by pro domains and proprotein convertases.
J Cell Biol 144
: 139
–149, 1999[Abstract/Free Full Text]
Gregory KE, Ono RN, Charbonneau NL, Kuo CL, Keene DR, Bachinger HP, Sakai LY: The prodomain of BMP-7 targets the BMP-7 complex to the extracellular matrix.
J Biol Chem 280
: 27970
–27980, 2005[Abstract/Free Full Text]
Piccolo S, Agius E, Lu B, Goodman S, Dale L, De Robertis EM: Cleavage of chordin by xolloid metalloprotease suggests a role for proteolytic processing in the regulation of Spemann organizer activity.
Cell 91
: 407
–416, 1997[CrossRef][Medline]
Kreidberg JA, Donovan MJ, Goldstein SL, Rennke H, Shepherd K, Jones RC, Jaenisch R: Alpha 3 beta 1 integrin has a crucial role in kidney and lung organogenesis.
Development 122
: 3537
–3547, 1996[Abstract]
Leveen P, Pekny M, Gebre-Medhin S, Swolin B, Larsson E, Betsholtz C: Mice deficient for PDGF B show renal, cardiovascular, and hematological abnormalities.
Genes Dev 8
: 1875
–1887, 1994[Abstract/Free Full Text]
Soriano P: Abnormal kidney development and hematological disorders in PDGF beta-receptor mutant mice.
Genes Dev 8
: 1888
–1896, 1994[Abstract/Free Full Text]
Dudley AT, Robertson EJ: Overlapping expression domains of bone morphogenetic protein family members potentially account for limited tissue defects in BMP7 deficient embryos.
Dev Dyn 208
: 349
–362, 1997[CrossRef][Medline]
Vukicevic S, Basic V, Rogic D, Basic N, Shih MS, Shepard A, Jin D, Dattatreyamurty B, Jones W, Dorai H, Ryan S, Griffiths D, Maliakal J, Jelic M, Pastorcic M, Stavljenic A, Sampath TK: Osteogenic protein-1 (bone morphogenetic protein-7) reduces severity of injury after ischemic acute renal failure in rat.
J Clin Invest 102
: 202
–214, 1998[Medline]
Hruska KA, Guo G, Wozniak M, Martin D, Miller S, Liapis H, Loveday K, Klahr S, Sampath TK, Morrissey J: Osteogenic protein-1 prevents renal fibrogenesis associated with ureteral obstruction.
Am J Physiol Renal Physiol 279
: F130
–F143, 2000[Abstract/Free Full Text]
This article has been cited by other articles:
E. Bjorling, C. Lindskog, P. Oksvold, J. Linne, C. Kampf, S. Hober, M. Uhlen, and F. Ponten A Web-based Tool for in Silico Biomarker Discovery Based on Tissue-specific Protein Profiles in Normal and Cancer Tissues
Mol. Cell. Proteomics,
May 1, 2008;
7(5):
825 - 844.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
superfamily ligands. Crim1KST264/KST264 mice display a mislocalization of VEGF-A within the developing glomerulus, as assessed by immunogold electron microscopy and increased activation of VEGF receptor 2 (Flk1) in the glomerular endothelial cells, suggesting that Crim1 regulates the delivery of VEGF-A by the podocytes to the endothelial cells. This is the first in vivo demonstration of regulation of VEGF-A delivery and supports the hypothesis that Crim1 functions to regulate the release of growth factors from the cell of synthesis. 
5 (15), Wilms' tumor 1 (WT1; Dako, Botany, Australia), CD31 and phospho-Smad 2/3 (Chemicon Int., Temecula, CA), phospho-Smad 1/5/8 (Cell Signaling Technology, Danvers, MA), and phospho-Flk1 (Merck Pty. Ltd., Victoria, Australia). For immunofluorescence, secondary antibodies used were fluorophore-conjugated anti-mouse Alexa 488 or anti-rabbit Alexa 594 (Molecular Probes, Eugene, OR). Tissue was incubated with the secondary antibody for 45 min, washed thoroughly in PBS, and mounted in Vector Shield (Vector Laboratories). Digital images were taken using an Olympus AX70 compound microscope. 
