Β1 integrin expression by podocytes is required to maintain glomerular structural integrity

β1 integrin expression by podocytes is required to maintainglomerular structural integrity

Ambra Pozzi1,2,3, George Jarad6, Gilbert W. Moeckel5, Sergio Coffa2, Xi Zhang2,3, LeslieGewin2, Vera Eremina7, Billy G. Hudson2, Dorin-Bogdan Borza2, Raymond C. Harris1,2,4,Lawrence B. Holzman8, Carrie L. Phillips9, Reinhard Fassler10, Susan E. Quaggin7, JeffreyH. Miner6, and Roy Zent1,2,3,4

1Department of Medicine, Veterans Affairs Hospital, Vanderbilt University Medical Center, Nashville, 37232

2Division of Nephrology, Departments of Medicine, Vanderbilt University Medical Center, Nashville, 37232

3Department of Cancer Biology, Vanderbilt University Medical Center, Nashville, 37232

4Departments of Cell and Developmental Biology, Vanderbilt University Medical Center, Nashville, 37232

5Department of Pathology, Vanderbilt University Medical Center, Nashville, 37232

6Renal Division, Washington University School of Medicine, St. Louis, MO 63110

7Department of Maternal and Fetal Health, The Samuel Lunenfeld Research Institute, Mt. Sinai Hospital,University of Toronto, Toronto, Ontario, Canada

8Division of Nephrology, University of Michigan Medical School, Ann Arbor, Mich., USA

9Division of Nephrology and Department of Pathology, Indiana University School of Medicine, Indianapolis,Indiana, USA

10Max Planck Institute of Biochemistry, Department of Molecular Medicine, 82152 Martinsried, Germany

SummaryIntegrins are transmembrane heteromeric receptors that mediate interactions between cells andextracellular matrix (ECM). β1, the most abundantly expressed integrin subunit, binds at least 12 αsubunits. β1 containing integrins are highly expressed in the glomerulus of the kidney; however theirrole in glomerular morphogenesis and maintenance of glomerular filtration barrier integrity is poorlyunderstood. To study these questions we selectively deleted β1 integrin in the podocyte by crossingβ1flox/flox mice with podocyte specific podocin-cre mice (pod-Cre), which express cre at the time ofglomerular capillary formation. We demonstrate that podocyte abnormalities are visualized duringglomerulogenesis of the pod-Cre;β1flox/flox mice and proteinuria is present at birth, despite a grosslynormal glomerular basement membrane. Following the advent of glomerular filtration there isprogressive podocyte loss and the mice develop capillary loop and mesangium degeneration withlittle evidence of glomerulosclerosis. By three weeks of age the mice develop severe end stage renalfailure characterized by both tubulointerstitial and glomerular pathology. Thus, expression of β1containing integrins by the podocyte is critical for maintaining the structural integrity of theglomerulus.

Address correspondence to: Roy Zent, Room C3210 MCN, Vanderbilt University Medical Center, Nashville, TN, 37232, Email:[email protected], Telephone: (615) 322-4632.Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resultingproof before it is published in its final citable form. Please note that during the production process errors may be discovered which couldaffect the content, and all legal disclaimers that apply to the journal pertain.

NIH Public AccessAuthor ManuscriptDev Biol. Author manuscript; available in PMC 2009 April 15.

Published in final edited form as:Dev Biol. 2008 April 15; 316(2): 288–301.

NIH

-PA Author Manuscript

NIH


NIH


Keywordskidney; development; basement membrane; receptors

IntroductionIntegrins are heterodimeric transmembrane receptors composed of an α and a β subunit.Integrin β1, the most abundantly expressed subunit, heterodimerizes with at least 12 α subunits,forming dimers which are critical for cellular interactions with extracellular matrix (ECM)components (Hynes, 2002). Integrin α3β1 and α6β1 are major laminin binding receptors, whilethe predominant collagen receptors are integrins α1β1 and α2β1 (Hynes, 2002). These integrinsare highly expressed in multiple organs, including the kidneys, where they are found in boththe tubules and glomerulus (Kreidberg and Symons, 2000).

The glomerulus is the principal filtering unit of the kidney and its filtration barrier consists ofendothelial cells and podocytes separated by a glomerular basement membrane (GBM)comprised primarily of collagen IV and laminins (Miner, 2005). Both the cellular componentsand the GBM are required to maintain the integrity of the filtration barrier, and perturbationsof any of these components results in developmental or functional aberrations of theglomerulus. The importance of ECM components in glomerular development and integrity hasbeen demonstrated in genetically engineered mice (Cosgrove et al., 1996; Miner and Li,2000; Miner et al., 1997; Miner and Sanes, 1996; Noakes et al., 1995); however the role ofintegrins is far less clear.

α3β1 is the only integrin shown to play a significant role in glomerular development in vivo(Kreidberg et al., 1996). Mice lacking the integrin α3 subunit die in the neonatal period andhave aberrations in glomerular capillary loops, a disorganized GBM and podocyte foot processabnormalities (Kreidberg et al., 1996). More recently, the integrin α3 subunit was specificallydeleted in the podocyte (Sachs et al., 2006). These mice develop massive proteinuria in thefirst week of life and nephrotic syndrome by 5–6 weeks of age. The kidneys of the 6 week oldmice contained sclerosed glomeruli, a disorganized GBM and prominent protein casts in dilatedproximal tubules. Electron microscopy revealed complete effacement of podocyte footprocesses in newborn mice and widespread lamination and protrusions of the GBM in 6 week-old mice. In contrast to these mice, the total integrin α6-null mice do not have a renal phenotype(Georges-Labouesse et al., 1996), although integrin α6β1 is expressed with its major ligandslaminins-111, -511, and -521 (Aumailley et al., 2005) during various stages of glomerulardevelopment (Kreidberg and Symons, 2000; Miner, 1999). Furthermore, the total integrin α3/α6 double-deficient mice have a similar glomerular phenotype to that seen in the total α3-nullmice (De Arcangelis et al., 1999), suggesting that the α6 subunit is dispensable for glomerulardevelopment. Integrins α1β1 and α8β1, both highly expressed in the glomerulus, play a minorrole in glomerular development, as both integrin α1- and α8-null mice show subtle glomerularphenotypes characterized by increased glomerular collagen IV deposition and mesangialexpansion or mesangial hypercellularity and matrix deposition, respectively (Chen et al.,2004; Haas et al., 2003; Hartner et al., 2002). Glomeruli of integrin α2-null mice do not exhibitany overt glomerular phenotypes (A. Pozzi and R. Zent, unpublished data)

Although integrin α3β1 is the major ECM receptor required for glomerular development, wehypothesized that deleting β1 integrin in podocytes would result in a more profound phenotypethan that found in mice lacking theα3 integrin subunit in the podocytes as multiple αβ1 (αsβ1)integrins that interact with the GBM would not be expressed. In this context, ablation of theintegrin α3 subunit in the skin only generated microblistering (DiPersio et al., 1997; DiPersioet al., 2000), whereas specific integrin β1 deletion in the epidermis resulted in severe skin

Pozzi et al.

Dev Biol. Author manuscript; available in PMC 2009 April 15.

NIH


NIH


NIH


blistering, massive failure of BM assembly/organization, hemidesmosome instability, and afailure of keratinocytes within the hair follicles to remodel BM and invaginate into the dermis(Brakebusch et al., 2000; Raghavan et al., 2000). These results suggest that in addition toα3β1, multiple αsβ1 integrins contribute to keratinocyte proliferation/differentiation, ECMassembly and BM formation. In addition, although integrin α3 and α6 subunits are expressedin mammary glands, deleting these subunits in mice did not result in any functional ordevelopmental mammary gland abnormalities (Klinowska et al., 2001). However when β1 wasdeleted early in mammary gland development the alveoli were disorganized and containedclumps of epithelial cells bulging into what would normally be luminal space (Li et al.,2005; Naylor et al., 2005).

To determine the role of all the αsβ1 integrin heterodimers in glomerular development, wedeleted β1 integrin selectively in glomerular podocytes by crossing integrin β1flox/flox mice(Raghavan et al., 2000) with the same podocin-Cre (pod-Cre) mice (Moeller et al., 2003) usedto delete the integrin α3 subunit (Sachs et al., 2006). We provide evidence that pod-Cre;β1flox/flox mice develop end stage renal failure by 3 to 5 weeks of age due to glomerularabnormalities characterized by podocyte loss followed by degeneration of the glomerulus.Thus, mice in which integrin β1 was deleted from the podocytes have for the most part a similarbut worse phenotype than mice where only the α3 integrin subunit was deleted, suggesting thatα3β1 is the principal but not the only podocyte integrin required to maintain the glomerularfiltration barrier.

Material and MethodsGeneration of pod-Cre; β1flox/flox mice

All experiments were approved by the Vanderbilt University Institutional Animal Use and CareCommittee and are housed in a pathogen free environment. Integrin β1flox/flox mice (generousgift of Dr. E. Fuchs, Howard Hughes Medical Institute, The Rockefeller University) (Raghavanet al., 2000) or integrin β1flox/flox mice, in which a promoterless lacZ reporter gene wasintroduced after the downstream loxP site (Brakebusch et al., 2000) were crossed with thepodocin-Cre mice (pod-Cre) generated as previously described (Moeller et al., 2003). Micevaried between 4th and 6th generation C57BL6. Aged-matched littermates homozygous for thefloxed integrin β1 gene, but lacking Cre (β1flox/flox mice), were used as negative controls forpod-Cre; β1flox/flox mice. Mice were genotyped by PCR with the following primers: 5′-CGGCTCAAAGCAGAGTGTCAGTC-3′ and 5′-CCACAACTTTCCCAGTTAGCTCTC-3′for verification of the β1flox/flox (Raghavan et al., 2000); 5′-AGGTGCCCTTCCCTCTAGA-3′and 5′-GTGAAGTAGGTGAAAGGTAAC for verification of the integrin β1flox/flox mice withthe promoter-less lacZ reporter gene (Brakebusch et al., 2000); and 5′-GCATAACCAGTGAAACAGCATTGCTG-3′ and 5′-GGACATGTTCAGGGATCGCCAGGCG-3′ for verification of the podocin cre (Moeller etal., 2003).

Clinical parameters and morphologic analysisProteinuria was determined by analyzing 2 μl of urine per mouse on 10% SDS-PAGE gels thatwere subsequently stained by Coomassie Brilliant Blue.

For morphological and immunohistochemical analysis, kidneys at different stages ofdevelopment were removed immediately at sacrifice and fixed in 4% formaldehyde andembedded in paraffin, or embedded in OCT compound and stored at −80°C until use, or fixedin 2.5% glutaraldehyde, post-fixed in OsO4, dehydrated in ethanol and embedded in resin.Paraffin tissue sections were stained with either hematoxylin and eosin or Periodic AcidSchiff’s (PAS) for morphological evaluation by light microscopy.

Pozzi et al.


NIH


NIH


NIH


The mesangial cell number was evaluated in a blind fashion by a renal pathologist countingthe number of mesangial cells in 100 random glomerular sections per mouse. Glomeruli werecounted in 5 individual mice with a total of 500 glomeruli examined in each group. Themesangial cell number was expressed as a mean +/− standard deviation.

For electron microscopy, ultrastructural assessments of thin kidney sections were performedusing a Morgagni transmission electron microscope (FEI, Eindhoven, The Netherlands). GBMthickness was assessed by point to point measurements using a 2K × 2K camera (AdvancedMicroscopy Techniques Corp., Danvers, MA) with the associated digital imaging software.Ten different segments of GBM per mouse from 3 mice were measured and values wereexpressed as mean +/− standard deviation.

The number of podocytes was determined by counting their number in 10 randomly chosenEM sections of glomeruli with 3 different mice per genotype analyzed. The number ofpodocytes was expressed as mean +/− standard deviation.

ImmunostainingRat anti-mouse β1 integrin (MAB1997) was purchased from Chemicon. Rat anti-mouselaminin α1 mAb 8B3 was a gift from Dr. D. Abrahamson (St John et al., 2001) and rabbit anti-human podocin was a gift from Dr. Corinne Antignac (Roselli et al., 2002). Rabbit anti-lamininα5 (Miner et al., 1997), rabbit anti-sera specific for the mouse collagen type IV α4 chain (Minerand Sanes, 1994) and rabbit anti-nephrin (Holzman et al., 1999) have been described. Rabbitanti-chick integrin α3 was a gift from Mike Dipersio, Albany Medical College (DiPersio et al.,1995), CD2AP antibody was a gift from Andrey Shaw (St. Louis, Washington University);and ILK antibody (3862) was purchased from Cell Signaling. Anti-mouse CD31, rabbitpolyclonal antibodies to WT1 and anti-VEGF antibodies were purchased from Santa CruzBiotechnology. Monoclonal antibodies to entactin (clone ELM1 Rat monoclonal) werepurchased from Chemicon. Alexa 488- and Cy3- conjugated secondary antibodies werepurchased from Molecular Probes (Eugene, OR) and Chemicon (Temecula, CA), respectively.

Seven μm kidney frozen sections were incubated with the antibodies described above dilutedin PBS with 1% BSA followed by incubation with the appropriate secondary antibody. Thesections were subsequently mounted in 90% glycerol containing 0.1X PBS and 1 mg/ml p-phenylenediamine and viewed under epifluorescence with a Nikon Eclipse 800 compoundmicroscope. Images were captured with a Spot 2 cooled color digital camera (DiagnosticInstruments, Sterling Heights, MI).

Podocyte apoptosis was determined on kidney paraffin sections utilizing Apoptag ApoptosisDetection Kit, as described by the manufacturer (Chemicon). Positive apoptotic cells werecounted as podocytes when residing on the outer aspect of PAS-positive basement membrane.Apoptotic cells were determined in 25 random glomeruli per kidney with 4 different mice pergenotype analyzed. The number of podocytes was expressed as mean +/− standard deviation.

In situ hybridizationKidneys were dissected from mice on postnatal day 0, 10 and 21, washed briefly in RNase-free PBS and fixed overnight in DEPC-treated 4% paraformaldehyde. The tissues were thenplaced in 30% sucrose for 12–24 hours, embedded in Tissue-Tek OCT 4583 compound (SakuraFinetek USA Inc.) and snap frozen. Ten-micron tissue samples were cut on a Leica Jungcryostat (model CM3050; Leica Microsystems Inc.) and transferred to Superfrost microscopeslides (Fisher Scientific Co.). The slides were stored at −20°C until needed. Digoxigenin-labeled VEGF-A (kind gift of A. Nagy, Samuel Lunenfeld Research Institute, Toronto, Canada)

Pozzi et al.


NIH


NIH


NIH


probes were prepared according to the Roche Molecular Biochemicals protocol (RocheMolecular Biochemicals).

Slides were dried at room temperature for 2 hours, treated with 15ng/ml proteinase K in depc-PBS at 37°C for 5 minutes, washed in depc PBS, fixed in 4% PFA at room temperature for 7minutes and washed first in depc PBS and then in 2x SSC (PH 7) at room temperature. Slideswere prehybridized in mailers for 1 hour at 60°C using hybridization buffer (2.5 ml - 20x SSC,5 ml formamide, 250 μl – 20% CHAPS, 100 μl – 10% Triton X-100, 50 μl – 10 mg/ml yeastRNA, 25μl – 20 mg/ml Heparin, 100 μl –0.5 EDTA, 0.2 g blocking powder, 1.2 ml depc H2O)after which they were hybridized with the probe (1 ng/ml) diluted in hybridization buffer at60°’C overnight. The next day the slides were washed sequentially in 0.2XSSC and formamide/0.2x SSC. The slides were blocked with blocking solution from Roche as per instructions andthen incubated with anti-DIG antibody for 2 hours. Slides were washed and substrate was addedas per instructions from Roche. Slides were washed again and then immersed in Fast Red,dehydrated and mounted.

StatisticsThe Student’s t test was used for comparisons between two groups, and analysis of varianceusing Sigma Stat software was used for statistical differences between multiple groups. A pvalue less than 0.05 was considered statistically significant.

ResultsLoss of β1 integrin in the podocyte results in massive proteinuria and end stage renal failure

To determine the function of αsβ1 integrins expressed by podocytes in the glomerulus, micecarrying the floxed β1 integrin gene (β1flox/flox) (Raghavan et al., 2000) were crossed withmice expressing Cre recombinase under the control of the podocin promoter (pod-Cre)(Moeller et al., 2003), in which recombination occurs during the capillary loop stage inglomerular development. To verify that the β1 integrin subunit was deleted in the podocytes,we performed immunostaining for the β1 and α3 integrin subunits in P1, P10 and P21 mice.Expression of both of these subunits was significantly reduced in a segmental pattern in P1pod-Cre; β1flox/flox mice (Figure 1) and were further decreased in the glomeruli of P10 andP21 pod-Cre; β1flox/flox mice. β1 integrin subunit deletion was verified by crossing pod-cremice with another floxed β1 integrin mouse, in which a promoterless lacZ reporter gene wasintroduced after the downstream loxP site (Brakebusch et al., 2000) (data not shown).

Pod-Cre; β1flox/flox mice were born in the expected Mendelian ratio. There were noabnormalities observed in either the β1floxflox or the pod-Cre; β1flox/+ mice over time, howeverthe pod-Cre; β1flox/flox mice became less physically active than their littermate β1flox/flox

controls and developed severe edema 3 weeks after birth. Ninety percent (18/20) of the pod-Cre; β1flox/flox mice were euthanized between 4 and 5 weeks of age due to end stage renalfailure and nephrotic syndrome and only 10% (2/20) survived to 6 weeks of age (Figure 2A).The six week old pod-Cre; β1flox/flox mice had massive proteinuria, with the predominant bandrunning at the molecular weight of albumin (~ 66Kd) (Figure 2B). Autopsy of the mutant micerevealed smaller and paler kidneys than those isolated from age matched control animals, whichare characteristic features of end stage disease (Figure 2C). Numerous Bowmans capsules inpod-Cre; β1flox/flox mice were either empty or had partially disintegrated glomeruli as shownin Figures 2G. Interestingly the mesangium was only mildly hypercellular with little matrixexpansion (Figure 2I). In addition to the glomerular pathology, there was marked tubulardilatation and flattening of epithelial cells with extensive proteinaceous tubular casts (figure2E). Thus, all the mutant mice developed end stage kidney disease characterized bypathological changes in the glomeruli and tubulointerstitium.

Pozzi et al.


NIH


NIH


NIH


Glomerular capillary morphogenesis is normal in embryonic and newborn pod-Cre; β1flox/flox mice but becomes abnormal 10 days after birth

Due to the severity of the renal phenotype in the mutant mice we investigated when theabnormalities first became apparent. The podocin promoter is activated during the capillaryformation stage (Moeller et al., 2003), so we initially determined the histology of kidneysderived from E15.5 embryos from 10 wild type and mutant mice. Glomeruli wereindistinguishable from the wild type controls in all the pod-Cre; β1flox/flox mice when visualizedby light microscopy (Figures 3A and 3B). No overt abnormalities were apparent in mutant P1(n=10) mice (Figures 3C and 3D), however even at this early age the mice exhibited proteinuria(Figure 3M) but no hematuria.

In contrast to controls, kidneys from mutant P10 mice demonstrated tubular dilatation andmultiple cytoplasmic vacuoles within the tubular epithelial cells (Figures 3E-H), which areconsistent with heavy proteinuria. In addition, some glomeruli from mutant mice showedsegmentally “ballooned” capillary lumens (Figures 3H). By 3 weeks of age the tubules showedincreased dilatation and there were more “ballooned” capillary loops and mesangialhypercellularity (Figures 3J and 3L). When mesangial cell number was assessed, there weresignificantly more in the pod-Cre; β1flox/flox mice compared to the β1flox/flox mice (17.1 +/−3.6, vs 11.3 +/− 2.3. p < 0.05).

Deletion of β1 integrin in the podocyte results in foot process effacementTo evaluate the integrity of the glomerular filtration barrier, glomeruli of mice at various ageswere subjected to examination by electron microscopy. At E15 there was evidence of footprocess effacement in the pod-Cre; β1flox/flox mice but surprisingly the GBM was intact in bothgenotypes (Figure 4). Similar ultrastructural findings were present in the kidneys isolated fromE17.5 and E19.5 (data not shown) as well as P1 mice (Figure 4). There was no significantdifference in thickness of the GBM in β1flox/flox and pod-Cre; β1flox/flox P1 mice (130 +/− 25.9nm vs. 143 +/− 39.5: p=0.16). Extensive foot process effacement and early segmental splittingof the GBM was observed in P10 pod-Cre; β1flox/flox mice and these features became weremore evident in the P21 mutant mice (Figure 4).

To determine whether the abnormalities in the filtration barrier in newborn mice was due toaltered expression or localization of proteins involved in slit diaphragm formation or structuralproteins known to be associated with nephrotic syndrome, immunofluorescence was performedfor nephrin, podocin, CD2AP (Figure 5A) and synaptopdodin (not shown). Expression andlocalization of these proteins was similar in control and mutant mice by immunofluorescence.Similar results, with respect to mRNA expression, were found when in situ hybridization assayswere performed (data not shown). As integrins are thought to be critical for normal BMdevelopment, we examined expression of collagen IV and laminins in glomeruli.Unexpectedly, no differences in collagen IV α1, α3, α4, α5 and α6 chain expression wereobserved (Figure 5B). Similarly, expression of both laminin α1 and α5 chains as well as α2and β2 chains (data not shown) was similar in both genotypes (Figure 5C and data not shown).As morphogenesis of the glomerular capillaries appeared to be normal at this stage and noalterations in GBM components were present, we determined whether compensation byoverexpression of other receptors for GBM components had occurred. However, once again,no differences in expression of integrin αv, α6 and β4 subunits or dystroglycan were detectedin the mutant mice (data not shown).

Taken together these data suggest that during embryogenesis lack of integrin β1 results inpodocyte abnormalities characterized by dysmorphic foot processes, with no grossabnormalities in GBM composition or ultrastructure.

Pozzi et al.


NIH


NIH


NIH


Podocyte apoptosis occurs in pod-Cre; β1flox/flox mice within 10 days of birthBased on the phenotypes of the P21 and 6 week old pod-Cre; β1flox/flox mice, where there weredilated glomerular capillaries and subsequent glomerular disintegration, we determinedwhether the podocytes in the mutant mice were undergoing apoptosis. As seen in Figure 6A,a significant number of apoptotic podocytes were present in the P10 and P21 mutant mice.Furthermore, when expression of the specific podocyte markers WT1 (Figure 6B), podocin,CD2AP and synaptopodin (data not shown) was determined by immunofluoresence, they weresignificantly decreased in P10 and P21 mutant glomeruli suggesting the possibility of podocyteloss in the pod-Cre; β1flox/flox mice.

As deletion of the β1 integrin binding protein, integrin linked kinase (ILK), in podocytes resultsin focal segmental glomerulosclerosis and alteration in the distribution of integrin α3β1 startingat 4 weeks of age (Dai et al., 2006; El-Aouni et al., 2006), we investigated whether deletingthe β1 integrin subunit from podocytes altered glomerular ILK expression. As shown in Figure6B, no differences in glomerular ILK expression were observed in the glomeruli of P1 mutantand control mice. However, by P10 the mutant mice showed a marked decrease in glomerularILK expression and virtually no ILK was detected in glomeruli by P21. The pattern and timecourse of ILK expression in the mutant mice was very similar to that seen for β1 integrin (Figure1), suggesting that expression of the β1 integrin subunit might play a specific role in regulatingILK expression in the podocyte.

To specifically confirm podocyte loss in the P21 pod-Cre; β1flox/flox mice, the number ofpodocytes per glomerular section was determined utilizing electron microscopy. Whenformally quantified, P21 pod-Cre; β1flox/flox had about one third of the number podocytescompared to their β1flox/flox litter mates (Figure 6C). All together, these data suggest that thereare significantly less podocytes in the P21 pod-Cre; β1flox/flox mice.

Deletion of β1 integrin in the podocyte results in abnormalities of both capillary loops andthe mesangium

One of the most surprising findings of this study was that despite the obvious abnormalities ofthe podocytes, the principal lesions seen in P21 mutant mice was degeneration of the capillaryloops and mesangium with little glomerulosclerosis. To determine the possible mechanismunderlying these capillary loop abnormalities, we initially investigated the integrity of theendothelium by staining the kidneys with antibody to CD31, a specific endothelial cell marker.As shown in Figure 7A, the intensity and distribution of CD31 positive cells was similar in P1β1flox/flox and pod-Cre; β1flox/flox mice. By P10 there was significantly less CD31 staining inthe glomeruli of mutant mice and by P21 the CD31 staining was virtually undetectable in themutant mice, suggesting that endothelial cells had undergone cell death during this time period.To verify these findings, electron microscopy was performed on the glomerular capillary loops.The capillary loops in newborn, P10, and P21 β1flox/flox (Figure 7B) as well as newborn pod-Cre; β1flox/flox (data not shown) were normal. However at day 10 significant vacuolation, asign of cell damage, was present in the endothelial cells of mutant mice (Figure 7B), whichwas much worse by 3 weeks of age (Figure 7B).

As mesangial injury was evident in the mutant mice starting at 3 weeks of age (Figure 3), themesangium was studied in detail by EM. Electron microscopy of the mesangium from newborn,P10, and P21 β1flox/flox (Figure 7C) as well as newborn pod-Cre; β1flox/flox (data not shown)were normal; however there were multiple cytoplasmic vacuoles, indicative of cellular damage,in the mesangial cells of mutant mice starting at 10 days of age (Figure 7C). By 3 weeks therewas evidence of increased mesangial matrix with focal defects. These results suggest that themesangium, like the endothelium, was injured in the mutant mice.

Pozzi et al.


NIH


NIH


NIH


Vascular endothelial growth factor-A (VEGF), which is primarily produced by the podocytes,is a growth factor required for both normal capillary loop development as well as mesangialcell survival (Eremina et al., 2006; Eremina et al., 2003). In this context, mice lacking podocyte-produced VEGF, develop grossly abnormal glomerular capillary loops (Eremina et al., 2003).Furthermore, in mice hypomorphic for podocyte-produced VEGF, there is ballooning of theglomerular capillaries and mesangiolysis (Eremina et al., 2006). This finding suggests thatVEGF is required for the normal development and maintenance of glomerular integrity,especially with respect to the capillary loops and the mesangium. As the pod-Cre; β1flox/flox

mice reduce podocyte number with time, the levels of VEGF were analyzed in the glomeruliof β1flox/flox and pod-Cre; β1flox/flox mice. While the mRNA of this growth factor was similarin both genotypes at birth (Figure 7D), decreased mRNA expression was observed in 10 daysold mutant mice (Figure 7D) and by 3 weeks significantly less VEGF message was detectedin the glomeruli of the pod-Cre; β1flox/flox mice (Figure 7D). Thus, the timing of glomerulardegeneration correlates with the lack of VEGF expression by the podocytes.

Discussionβ1 integrins are highly expressed in the glomerulus of the kidney (Kreidberg and Symons,2000); however their role in glomerular morphogenesis and maintenance of the integrity of theglomerular filtration barrier is poorly understood. Recently the integrin α3 subunit wasselectively deleted in the podocyte, which resulted in mice that developed massive proteinuriain the first week of life and nephrotic syndrome by 5–6 weeks of age (Sachs et al., 2006).Newborn mice had complete podocyte foot process effacement and the glomeruli of the 6 weekold mice were severely sclerosed, demonstrated a disorganized GBM and prominent proteincasts in dilated proximal tubules. In this study we show that selectively deleting αsβ1 integrinsin the podocyte resulted in 1) normal morphogenesis of the glomerulus, despite podocyteabnormalities; 2) a defective glomerular filtration barrier present at birth; 3) podocytes lossover time; 4) capillary loop and mesangium degeneration with little evidence ofglomerulosclerosis and 5) the development of end stage kidneys characterized by bothtubulointerstitial and glomerular pathology by 3 to 6 weeks of age. Taken together these resultsdemonstrate that although the injury in the β1 integrin null mouse is more severe and has somedifferences in pattern, the overall phenotype is similar to that found in mice where the α3integrin subunit is selectively deleted in podocytes. This suggests that integrin α3β1 is theprincipal integrin required to maintain the structural integrity of the glomerulus and otherαsβ1 integrins play a relatively minor role.

The end stage kidneys of the pod-Cre; β1flox/flox mice were characterized by severetubulointerstitial disease in addition to the glomerular pathology. This was likely a consequenceof the heavy glomerular proteinuria that results in interstitial mononuclear cell accumulationand activation of interstitial fibroblasts to deposit collagens in the tubulointerstitialcompartment of the kidney (Eddy, 1994; Remuzzi, 1995; Remuzzi et al., 1997). As the defectsin the GBM also correlated with the decreased number of podocytes and proteinuria, it isprobable that the GBM changes result from podocyte loss and heavy proteinuria, rather thanlack of β1 integrin function by the podocytes per se during GBM development.

In contrast to the newborn total integrin α3-null mouse, where in addition to the podocyteabnormalities the GBM was disorganized and capillary loop numbers were reduced (Kreidberget al., 1996), the mutant mice in this study showed normal glomerular morphogenesis, GBMformation as well as expression of slit diaphragm and key cytoskeletal proteins. The differencesin glomerular morphogenesis between these two mice may be because: i) sufficient αsβ1integrins are expressed by the podocytes prior to Cre-mediated excision, as podocin isexpressed at the S-shape body stage of development (Moeller et al., 2003) ii) αsβ1 integrinsmight still be expressed by the podocytes following cre expression due to inefficient deletion

Pozzi et al.


NIH


NIH


NIH


of β1 integrin; iii) GBM might be predominantly derived from and/or organized by theendothelial cells in the pod-Cre; β1flox/flox mice, as both podocytes and the endothelium areknown to contribute to formation of the GBM (Abrahamson, 1985; St John and Abrahamson,2001); and iv) over-expression of αv- or β4-containig integrins or dystroglycan mightcompensate for the lack of αsβ1 integrins. This possibility, however, appears unlikely asexpression of these receptors was similar in both pod-Cre; β1flox/flox and β1flox/flox mice (notshown).

Although the overall phenotypes of the podocyte specific α3- and β1-null mice were similar,there were some discrepancies. The earlier onset and increased severity of injury in the pod-Cre; β1flox/flox could be due to the loss of interactions of podocytes with the GBM via thecollagen binding receptors, integrins α1β1and α2β1 and the laminin receptor integrin α6β1which are expressed by podocytes during development (Korhonen et al., 1990; Rahilly andFleming, 1992). Furthermore, we recently demonstrated that integrin α3β1 transdominantlyinhibits integrin αvβ3-dependent function (Borza et al., 2006) and podocytes express integrinαvβ3 in vivo (Borza et al in press J. Am. Soc. Neph.). Thus deleting integrin α3β1 integrinmight enhance podocyte adhesion to the GBM by increasing integrin αvβ3 affinity and avidity.

One of the striking and surprising differences in phenotype between our mice and the pod-Cre;α3flox/flox mice is that the GBM in our mice is relatively normal compared to the thickeningand severe lamellations seen in that mouse. Furthermore our results in the GBM contrast withthe improperly deposited epidermal BM in mice null for β1 integrin in keratinocytes(Brakebusch et al., 2000; Raghavan et al., 2000), however these changes might be organspecific as the BM is normal in mice where the β1 integrin was specifically deleted in themammary gland (Li et al., 2005). A possible explanation for the discrepancy in results betweenthe pod-Cre; α3flox/flox and the pod-Cre; β1flox/flox mice is that integrin α3β1 might negativelyregulate integrin α2β1-dependent glomerular collagen production. This hypothesis is based onthe observation that integrin α2β1 is a positive regulator of collagen synthesis (Ivaska et al.,1999); integrin α2β1 function can be negatively regulated by the expression of other integrinfamily members; and deleting integrin α2β1 results in diminished collagen IV production inmodels of glomerular injury (Pozzi and Zent unpublished data).

The abnormality of the glomerular filtration barrier was present at birth when only abnormalfoot process formation was present and this phenotype worsened with time and correlated withpodocyte apoptosis and loss. We were unable to determine whether the loss of podocytes wasdue to lack of β1 integrin-dependent cell adhesion to the GBM or due to cell injury andapoptosis as a consequence of lack of β1 integrin. However the temporal association ofpodocyte loss with increased hydrostatic pressure due to urinary flow following birth suggeststhe increase in force exerted on the podocytes may have induced podocyte detachment andsubsequent apoptosis.

One of the most interesting features of the podocyte β1-null mouse was the rapidity with whichthe glomeruli degenerated and the lack of glomerular fibrosis relative to the glomerular injury.This contrasts with mice deficient for specific collagen IV chains required for normal GBMformation, where glomerulosclerosis and proteinuria occurred (Cosgrove et al., 1996; Minerand Sanes, 1996) but the mice lived significantly longer than the podocyte integrinβ1-null mice.Similarly, mice deficient for ILK (R. Zent, unpublished data) (Dai et al., 2006; El-Aouni et al.,2006) or the integrin α3 subunit in the podocytes (Sachs et al., 2006) as well as lacking proteinsrequired for normal slit diaphragm formation (i.e. podocin and CD2-associated protein)(Roselli et al., 2004) (Shih et al., 1999) also had foot process effacement, proteinuria, anddeveloped severe glomerulosclerosis prior to developing end stage kidney. In addition to thesegenetic models, when 40% of podocytes were destroyed in rats by diphtheria toxin, the primarylesion observed was glomerulosclerosis (Wharram et al., 2005), which is consistent with the

Pozzi et al.


NIH


NIH


NIH


theory that if sufficient podocytes detach leaving a naked GBM, a circumscribed region offocal segmental sclerosis will initially form and eventually result in global glomerulosclerosis(Kriz, 2002). Thus foot process and slit diaphragm abnormalities or loss of up to 40% ofpodocytes in glomeruli does not explain why pod-Cre; β1flox/flox did not developglomerulosclerosis.

In an attempt to explain the lack of glomerulosclerosis and the severe mesangiolysis in ourmouse model, we noted that the glomerular phenotype of the pod-Cre; β1flox/flox mice at 3weeks of age had many features similar to those seen in mice expressing one hypomorphicVEGF-A allele (Eremina et al., 2006). Both these mice have dilated capillary loops and severeabnormalities within the mesangium. The phenotype in the VEGF hypomorphic mouse isproposed to be due to a requirement of podocyte-dependent VEGF production for bothendothelial cell proliferation and survival, and disruption of the endothelial compartment leadsto the mesangial defects (Eremina et al., 2006). Since the pod-Cre; β1flox/flox mice demonstratesignificant podocyte loss and the principal source of VEGF in the glomerulus is the podocytes,we postulate that when sufficient podocytes are lost in this mouse, the endothelial cells undergoapoptosis due to their dependence on this angiogenic factor, or other factors produced by thepodocyte, for survival. The lack of VEGF would not explain the mesangium phenotype asmesangial cells do not express receptors for VEGF (Eremina et al., 2006). However mesangialcells do require PDGF-β secretion by endothelial cells for their survival (Bjarnegard et al.,2004). Thus with progressive podocyte loss in the pod-Cre; β1flox/flox, we propose that theglomerulus loses VEGF and probably other podocyte-specific growth factors production,which results in endothelial cell death. The loss of growth factor production from both thesecells types subsequently results in the inability of the glomerulus to maintain integrity of themesangium.

In conclusion we provide evidence that podocyte expression of αsβ1 integrins is required forthe normal formation and integrity of the glomerular filtration barrier. Although the GBMappears to form relatively normally in newborn pod-Cre; β1flox/flox mice, podocyte foot processeffacement and proteinuria is seen. With the increase in glomerular hydrostatic pressure,podocytes are lost from the glomerulus, which promotes rapid destruction of the capillary loopsand mesangium with little glomerulosclerosis. The rapidly degenerating glomeruli promotetubulointerstitial disease which is likely due to the increased proteinuria. This phenotype is forthe most part similar but more severe than that seen in mice lacking the α3 integrin subunit inpodocytes, where proteinuria and glomerulosclerosis are the primary features (Sachs et al.,2006), suggesting that in addition to integrin α3β1, other αsβ1 integrins play a role inmaintaining the glomerular filtration barrier.

Acknowledgements

We thank Dr. Elaine Fuchs for providing the integrin β1flox/flox mice. This work was supported by R01-CA94849(AP); R01-DK74359 (AP); O’Brien Center Grant P50-DK39261-16 (AP, RZ, RCH); P01 DK65123 (DBB, BGH, RZ,AP, JHM), RO1-DK69921 (RZ); R01-DK064687 (JHM), RO1-DK51265 (RCH), Merit award from the Departmentof Veterans Affairs (RZ, RCH), Established Investigator Award from the American Heart Association (JHM), Grantin Aid American Heart Association (RZ), Research Grant from the National Kidney Foundation of Eastern Missouriand Metro East (GJ).

ReferencesAbrahamson DR. Origin of the glomerular basement membrane visualized after in vivo labeling of

laminin in newborn rat kidneys. J Cell Biol 1985;100:1988–2000. [PubMed: 3889015]Aumailley M, et al. A simplified laminin nomenclature. Matrix Biol 2005;24:326–332. [PubMed:

15979864]Bjarnegard M, et al. Endothelium-specific ablation of PDGFB leads to pericyte loss and glomerular,

cardiac and placental abnormalities. Development 2004;131:1847–57. [PubMed: 15084468]

Pozzi et al.


NIH


NIH


NIH


Borza CM, et al. Integrin alpha3beta1, a novel receptor for alpha3(IV) noncollagenous domain and atrans-dominant Inhibitor for integrin alphavbeta3. J Biol Chem 2006;281:20932–9. [PubMed:16731529]

Brakebusch C, et al. Skin and hair follicle integrity is crucially dependent on beta 1 integrin expressionon keratinocytes. Embo J 2000;19:3990–4003. [PubMed: 10921880]

Chen X, et al. Lack of integrin alpha1beta1 leads to severe glomerulosclerosis after glomerular injury.Am J Pathol 2004;165:617–30. [PubMed: 15277235]

Cosgrove D, et al. Collagen COL4A3 knockout: a mouse model for autosomal Alport syndrome. GenesDev 1996;10:2981–92. [PubMed: 8956999]

Dai C, et al. Essential role of integrin-linked kinase in podocyte biology: Bridging the integrin and slitdiaphragm signaling. J Am Soc Nephrol 2006;17:2164–75. [PubMed: 16837631]

De Arcangelis A, et al. Synergistic activities of alpha3 and alpha6 integrins are required during apicalectodermal ridge formation and organogenesis in the mouse. Development 1999;126:3957–3968.[PubMed: 10433923]

DiPersio CM, et al. alpha3beta1 Integrin is required for normal development of the epidermal basementmembrane. J Cell Biol 1997;137:729–42. [PubMed: 9151677]

DiPersio CM, et al. alpha 3A beta 1 integrin localizes to focal contacts in response to diverse extracellularmatrix proteins. J Cell Sci 1995;108(Pt 6):2321–36. [PubMed: 7673352]

DiPersio CM, et al. (α)3(β)1 and (α)6(β)4 integrin receptors for laminin-5 are not essential for epidermalmorphogenesis and homeostasis during skin development [In Process Citation]. J Cell Sci 2000;113(Pt 17):3051–3062. [PubMed: 10934043]

Eddy AA. Experimental insights into the tubulointerstitial disease accompanying primary glomerularlesions. J Am Soc Nephrol 1994;5:1273–87. [PubMed: 7893992]

El-Aouni C, et al. Podocyte-specific deletion of integrin-linked kinase results in severe glomerularbasement membrane alterations and progressive glomerulosclerosis. J Am Soc Nephrol2006;17:1334–44. [PubMed: 16611717]

Eremina V, et al. Vascular endothelial growth factor a signaling in the podocyte-endothelial compartmentis required for mesangial cell migration and survival. J Am Soc Nephrol 2006;17:724–35. [PubMed:16436493]

Eremina V, et al. Glomerular-specific alterations of VEGF-A expression lead to distinct congenital andacquired renal diseases. J Clin Invest 2003;111:707–16. [PubMed: 12618525]

Georges-Labouesse E, et al. Absence of integrin alpha 6 leads to epidermolysis bullosa and neonataldeath in mice. Nat Genet 1996;13:370–373. [PubMed: 8673141]

Haas CS, et al. Glomerular and renal vascular structural changes in alpha8 integrin-deficient mice. J AmSoc Nephrol 2003;14:2288–96. [PubMed: 12937305]

Hartner A, et al. The alpha8 integrin chain affords mechanical stability to the glomerular capillary tuftin hypertensive glomerular disease. Am J Pathol 2002;160:861–7. [PubMed: 11891185]

Holzman LB, et al. Nephrin localizes to the slit pore of the glomerular epithelial cell. Kidney Int1999;56:1481–91. [PubMed: 10504499]

Hynes R. Integrins: bidirectional, allosteric signaling machines. Cell 2002;110:673–87. [PubMed:12297042]

Ivaska J, et al. Integrin alpha2beta1 mediates isoform-specific activation of p38 and upregulation ofcollagen gene transcription by a mechanism involving the alpha2 cytoplasmic tail. J Cell Biol1999;147:401–16. [PubMed: 10525544]

Klinowska TC, et al. Epithelial development and differentiation in the mammary gland is not dependenton alpha 3 or alpha 6 integrin subunits. Dev Biol 2001;233:449–67. [PubMed: 11336507]

Korhonen M, et al. The alpha 1-alpha 6 subunits of integrins are characteristically expressed in distinctsegments of developing and adult human nephron. J Cell Biol 1990;111:1245–1254. [PubMed:2144000]

Kreidberg JA, et al. Alpha 3 beta 1 integrin has a crucial role in kidney and lung organogenesis.Development 1996;122:3537–3547. [PubMed: 8951069]

Kreidberg JA, Symons JM. Integrins in kidney development, function, and disease. Am J Physiol RenalPhysiol 2000;279:F233–42. [PubMed: 10919841]

Pozzi et al.


NIH


NIH


NIH


Kriz W. Podocyte is the major culprit accounting for the progression of chronic renal disease. MicroscRes Tech 2002;57:189–95. [PubMed: 12012382]

Li N, et al. Beta1 integrins regulate mammary gland proliferation and maintain the integrity of mammaryalveoli. Embo J 2005;24:1942–53. [PubMed: 15889143]

Miner JH. Renal basement membrane components. Kidney Int 1999;56:2016–24. [PubMed: 10594777]Miner JH. Building the glomerulus: a matricentric view. J Am Soc Nephrol 2005;16:857–61. [PubMed:

15728777]Miner JH, Li C. Defective glomerulogenesis in the absence of laminin alpha5 demonstrates a

developmental role for the kidney glomerular basement membrane. Dev Biol 2000;217:278–89.[PubMed: 10625553]

Miner JH, et al. The laminin alpha chains: expression, developmental transitions, and chromosomallocations of alpha1-5, identification of heterotrimeric laminins 8–11, and cloning of a novel alpha3isoform. J Cell Biol 1997;137:685–701. [PubMed: 9151674]

Miner JH, Sanes JR. Collagen IV alpha 3, alpha 4, and alpha 5 chains in rodent basal laminae: sequence,distribution, association with laminins, and developmental switches. J Cell Biol 1994;127:879–91.[PubMed: 7962065]

Miner JH, Sanes JR. Molecular and functional defects in kidneys of mice lacking collagen alpha 3(IV):implications for Alport syndrome. J Cell Biol 1996;135:1403–13. [PubMed: 8947561]

Moeller MJ, et al. Podocyte-specific expression of cre recombinase in transgenic mice. Genesis2003;35:39–42. [PubMed: 12481297]

Naylor MJ, et al. Ablation of beta1 integrin in mammary epithelium reveals a key role for integrin inglandular morphogenesis and differentiation. J Cell Biol 2005;171:717–28. [PubMed: 16301336]

Noakes PG, et al. The renal glomerulus of mice lacking s-laminin/laminin beta 2: nephrosis despitemolecular compensation by laminin beta 1. Nat Genet 1995;10:400–6. [PubMed: 7670489]

Raghavan S, et al. Conditional ablation of beta1 integrin in skin. Severe defects in epidermal proliferation,basement membrane formation, and hair follicle invagination. J Cell Biol 2000;150:1149–60.[PubMed: 10974002]

Rahilly MA, Fleming S. Differential expression of integrin alpha chains by renal epithelial cells. J Pathol1992;167:327–34. [PubMed: 1381436]

Remuzzi G. Abnormal protein traffic through the glomerular barrier induces proximal tubular celldysfunction and causes renal injury. Curr Opin Nephrol Hypertens 1995;4:339–42. [PubMed:7552100]

Remuzzi G, et al. Understanding the nature of renal disease progression. Kidney Int 1997;51:2–15.[PubMed: 8995712]

Roselli S, et al. Podocin localizes in the kidney to the slit diaphragm area. Am J Pathol 2002;160:131–9. [PubMed: 11786407]

Roselli S, et al. Early glomerular filtration defect and severe renal disease in podocin-deficient mice. MolCell Biol 2004;24:550–60. [PubMed: 14701729]

Sachs N, et al. Kidney failure in mice lacking the tetraspanin CD151. J Cell Biol 2006;175:33–9.[PubMed: 17015618]

Shih NY, et al. Congenital nephrotic syndrome in mice lacking CD2-associated protein. Science1999;286:312–5. [PubMed: 10514378]

St John PL, Abrahamson DR. Glomerular endothelial cells and podocytes jointly synthesize laminin-1and -11 chains. Kidney Int 2001;60:1037–46. [PubMed: 11532098]

St John PL, et al. Glomerular laminin isoform transitions: errors in metanephric culture are corrected bygrafting. Am J Physiol Renal Physiol 2001;280:F695–705. [PubMed: 11249861]

Wharram BL, et al. Podocyte depletion causes glomerulosclerosis: diphtheria toxin-induced podocytedepletion in rats expressing human diphtheria toxin receptor transgene. J Am Soc Nephrol2005;16:2941–52. [PubMed: 16107576]

Pozzi et al.


NIH


NIH


NIH


Figure 1. β1 integrin subunit is deleted in pod-Cre;β1flox/flox miceFrozen sections of kidneys from P1, P10 and P21 β1flox/flox and pod-Cre; β1flox/flox mice wereco-stained with anti-mouse integrinβ1 (red) and anti laminin-α5 chain (green) or anti-mouseintegrinα3 (green) and anti-entactin (red), respectively.

Pozzi et al.


NIH


NIH


NIH


Figure 2. Six week old pod-Cre;β1flox/flox mice develop severe proteinuria and end stage renaldisease(A–C) Six week old pod-Cre; β1flox/flox mice are smaller with evidence of severe edema (A),albuminuria (2 μl urine/lane) (B) and end stage kidneys (C) compared to aged matchedβ1flox/flox mice. (D–I) PAS staining of kidneys derived from the mice described above showingglomerular and tubular interstitial abnormalities in the mutant group. The arrows in panel Eshow dilated tubules filled with hyaline material and the asterisks in panel G show the remnantsof glomeruli in the pod-Cre; β1flox/flox mice. (D, E = 100 ×; F, G = 200 ×; H, I =630 ×).

Pozzi et al.


NIH


NIH


NIH


Figure 3. Kidneys from pod-Cre; β1flox/flox mice exhibit severe abnormalities in the glomerulusand tubulointerstitium(A, B) PAS staining of kidneys derived from E15.5 β1flox/flox and pod-Cre; β1flox/flox miceshowing no significant differences between the two genotypes (200 ×). (C, D) PAS stainingof kidneys derived from newborn β1flox/flox and pod-Cre; β1flox/flox mice showing no overallsignificant differences between the two genotypes (200 ×). (E, F) PAS staining of kidneysderived from P10 β1flox/flox and pod-Cre; β1flox/flox mice revealed tubulodilatation (arrow) inthe mutant mice (100 ×). (G, H) In P10 mutant mice there was evidence of “ballooned”glomerular capillary loops (asterisk) and protein containing vacuoles (arrow) in the tubules ofthe pod-Cre; β1flox/flox mice (400 ×). (I–L) PAS staining of kidneys derived from 3 week old

Pozzi et al.


NIH


NIH


NIH


β1flox/flox and pod-Cre; β1flox/flox mice revealed dilated tubules, evidence of “ballooned”capillary loops and mesangial hypercellularity (I, J = 200 ×; K, L= 400 ×). (M) Coomassiestaining of urine (2 μl/lane) from newborn β1flox/flox and pod-Cre; β1flox/flox mice showingalbuminuria in the latter group.

Pozzi et al.


NIH


NIH


NIH


Figure 4. Glomeruli from pod-Cre; β1flox/flox mice demonstrate podocyte foot process effacementEM analysis was performed on kidneys of mice at various ages. In the E15 mutant mice thereis evidence of foot process effacement of the podocytes but the GBM (arrow) was normal.Similar findings are present in P1 kidneys. In P10 mice, in addition to the foot processeffacement, there was evidence of very mild segmental splitting of the GBM (arrow) whichprogressed by day P21. Abbreviations: FP=Foot Process; Po=Podocyte

Pozzi et al.


NIH


NIH


NIH


Figure 5. Glomeruli from newborn pod-Cre; β1flox/flox mice demonstrate normal expression ofpodocyte-specific structural proteins glomerular basement membrane componentsFrozen sections of kidneys derived from newborn mice were stained with antibodies to (A)Nephrin, podocin or CD2AP, (B) the α4 (green) or α1 (red) chains of collagen IV and (C) theα5 (green) orα1 (red) laminin chains.

Pozzi et al.


NIH


NIH


NIH


Figure 6. Podocytes of pod-Cre; β1flox/flox mice undergo apoptosis(A) Examples of apoptosis, as determined by TUNEL assay, detected in the glomeruli ofβ1flox/flox or pod-Cre; β1flox/flox mice at the time points indicated. The number of apoptoticpodocytes expressed per 10 glomeruli is demonstrated graphically. The (*) indicates significantdifferences (p < 0.01) between the two genotypes. (B) Frozen sections of glomeruli were co-stained for WT-1 (green) and entactin (red) or ILK (green) and entactin (red) as described inthe Methods. (C) The number of podocytes in EM sections were evaluated as described in theMethods and expressed as mean +/− SD. (*) indicates significant differences (p < 0.05) betweenP21 β1flox/flox and P21 pod-Cre; β1flox/flox mice.

Pozzi et al.


NIH


NIH


NIH


Figure 7. Glomerular capillary and mesangium injury in pod-Cre; β1flox/flox mice(A) Frozen sections of kidneys derived from newborn, P10 and P21 mice were stained withCD31 antibodies to visualize the glomerular vasculature. (B) EM of kidneys from P21β1flox/flox (3,000x) as well as P10 (11,000x) and P21 (11,000x) pod-Cre; β1flox/flox micerevealed normal morphology of endothelial cells in the β1flox/flox mice. Vacuoles (arrow) inthe endothelial cells (arrow) were evident in the P10 and P21 mutant mice. Abbreviations: EC= endothelial cells; CL = capillary loops; (C) EM of kidneys emphasizing the mesangium.Note the presence of vesicles (arrow) in the mesangial cells of mutant mice by P10 (8,900 ×)and the increased mesangial matrix in the P21 pod-Cre; β1flox/flox mice (4th panel) (2,200 ×).

Pozzi et al.


NIH


NIH


NIH


Abbreviations: MC= mesangial cell. (D) In situ hybridizations on kidneys of newborn, P10and P21 day old mice showing VEGF mRNA expression.

Pozzi et al.


NIH


NIH


NIH


Β1 integrin expression by podocytes is required to maintain glomerular structural integrity

Documents