Cell Metabolism

Article

Insulin Signaling to the Glomerular PodocyteIs Critical for Normal Kidney FunctionGavin I. Welsh,1 Lorna J. Hale,1 Vera Eremina,5 Marie Jeansson,5 Yoshiro Maezawa,5 Rachel Lennon,1

Deborah A. Pons,1 Rachel J. Owen,3 Simon C. Satchell,1 Mervyn J. Miles,3 Christopher J. Caunt,4 Craig A. McArdle,4

Hermann Pavenstadt,6 Jeremy M. Tavare,2 Andrew M. Herzenberg,7 C. Ronald Kahn,8 Peter W. Mathieson,1

Susan E. Quaggin,5 Moin A. Saleem,1 and Richard J.M. Coward1,5,*1Academic Renal Unit2School of Biochemistry3School of Physics4Department of Molecular PharmacologySchool of Clinical Sciences, University of Bristol, Bristol BS8 1TH, UK5Samuel Lunenfeld Research Institute, Mount Sinai Hospital, University of Toronto, Toronto, Ontario M5G 1X5, Canada6Department of Internal Medicine D, Nephrology and Hypertension, University Clinics Muenster, Muenster 48149, Germany7Department of Pathology, University Health Network and University of Toronto, Ontario M5G 2C4, Canada8Joslin Diabetes Center and Harvard Medical School, Boston, MA 02215, USA*Correspondence: richard.coward@bristol.ac.ukDOI 10.1016/j.cmet.2010.08.015

SUMMARY

Diabetic nephropathy (DN) is the leading cause ofrenal failure in the world. It is characterized byalbuminuria and abnormal glomerular function andis considered a hyperglycemic ‘‘microvascular’’complication of diabetes, implying a primary defectin the endothelium. However, we have previouslyshown that human podocytes have robust responsesto insulin. To determine whether insulin signaling inpodocytes affects glomerular function in vivo, wegenerated mice with specific deletion of the insulinreceptor from their podocytes. These animalsdevelop significant albuminuria together with histo-logical features that recapitulate DN, but in a normo-glycemic environment. Examination of ‘‘normal’’insulin-responsive podocytes in vivo and in vitrodemonstrates that insulin signals through the MAPKand PI3K pathways via the insulin receptor anddirectly remodels the actin cytoskeleton of this cell.Collectively, this work reveals the critical importanceof podocyte insulin sensitivity for kidney function.

INTRODUCTION

In 2007 in the United States, 54% of patients requiring dialysis orrenal transplantation had diabetes mellitus and the annualeconomic cost of caring for these patients was in excess of $12billion (U.S. Renal Data System, USRDS 2007 Annual DataReport, http://www.usrds.org). This is similar in other parts ofthe world, and these numbers are predicted to increase greatlyin forthcoming years due to theglobal epidemic of type 2 diabetes(Zimmet et al., 2001).The natural history of diabetic nephropathy (DN) is dominated

by progressive albuminuria. Initially, small amounts of albumin

escape into the urine (microalbuminuria), and this developsinto overt albuminuria as nephropathy progresses. Not onlydoes albuminuria predict renal disease, it is also an independentcardiovascular risk factor (Perkovic et al., 2008). The presence ofalbuminuria suggests that the glomerular filtration barrier (GFB)is an early target in this disease. The GFB is composed of twocell types, podocytes and glomerular endothelial cells (GEnC)(Pavenstadt et al., 2003), which are located on either side ofthe glomerular basement membrane (GBM). The podocyte isparticularly important in maintaining the integrity of the GFB inhumans, as illustrated by a growing number of podocyte-specific mutations, all of which result in significant albumin lossinto the urine (Lavin et al., 2008; Brown et al., 2010). Albuminuriain DN has previously been considered to be predominantlya consequence of hyperglycemia directed against the microvas-culature (Nishikawa et al., 2000). However, microalbuminuriaalso occurs in the insulin-resistant metabolic syndrome (it isnow one of the diagnostic criteria as defined by theWorld HealthOrganization). This often occurs with a normal blood glucoselevel and supports a nonhyperglycemic pathway of GFB disrup-tion. Furthermore, there is growing evidence that podocytedysfunction plays an important role in the pathogenesis of DN(Wolf et al., 2005). We have previously shown that in respect toglucose uptake, podocytes and not GEnC are uniquely insulin-sensitive cells in the glomerulus of the kidney and are able toabsorb glucose via translocation of the glucose transportersGLUT1 and GLUT4 (Coward et al., 2005), dependent on thepodocyte protein nephrin (Coward et al., 2007). We now showthat insulin sensitivity is crucial in maintaining the integrity ofthe GFB and, when lost, recapitulates features of DN.

RESULTS

Specific Loss of Podocyte Insulin SignalingRecapitulates Features of Diabetic NephropathyWe generated two podocyte-specific insulin receptor knockoutmice (podIRKO) models by crossing floxed insulin receptor (IR)

Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc. 329

mice with podocyte-specific Cre recombinase mice driven byboth the nephrin and podocin promoters. As we were movingfacilities, this had the advantage that we could analyze mice inboth a conventional facility (Mount Sinai, Toronto, Ontario: podo-cin Cre IRKO) and a specific-pathogen-free facility (TorontoCenter for Phenogenomics: nephrin Cre IRKO) to allow us tocontrol for environment. Initially, we crossed our Cre-expressingmice with lacZ/EGFP (Z/EG) reporter mice and ascertained thatthere were good levels of excision exclusively in podocytes(Figure 1A). We confirmed that the IR was knocked down specif-ically in the podocyte in 3-week-old mice by studying genomicexcision of IR DNA in a variety of tissues in podIRKO and controlmice and found that genomic DNAwas excised in a pure glomer-ular extract (Figure 1B) only from podIRKO mice (Figure 1C). Wewent on to prove the IR was specifically knocked down in podo-

Figure 1. In Vivo Podocyte-Specific IRKnockdown in Transgenic Mice(A) Glomeruli of Z/EG reporter mice immuno-

stained with an anti-GFP antibody. Specific exci-

sion is demonstrated in podocytes (arrowed).

(B) Glomeruli were extracted and purified using

Dynabeads. This is light microscopic analysis of

the extract, showing a pure glomerular population

and the beads.

(C) PCR of genomic DNA from different tissues in

podIRKO compared with (cw) IRfl/fl Cre-negative

mice. Only podIRKO glomeruli have a smaller

excised genomic DNA band present (arrowed)

(G, glomeruli; S, spleen; E, ear).

(D) Podocyte isolation of CFP-positive podocytes

from 3-week-old podIRKO and control mice

demonstrates that the IR mRNA is knocked down

by over 90% in podocytes of the podIRKO mice

compared with other glomerular cells (OGC) that

were CFP-negative. In control mice, the IR is

increased by 60% in the podocyte fraction.

(E) There is increased production (100%) of IGF1

receptor mRNA in podocyte fraction of podIRKO

mice compared with a 50% reduction in the con-

trol mice.

cytes in the glomeruli of podIRKOmice byisolating a pure podocyte fraction andperforming real-time PCR analysis. Thisrevealed at least a 90% knockdown in IRmessage in podIRKO podocytes. Incontrast, there was increased IRmessagein the podocytes of control mice whencompared to the other cells of theirglomeruli (Figure 1D).

PodIRKO mice were born with normalMendelian frequency and initially ap-peared entirely normal. At 8 weeks, bothpodIRKO models were comparable insize to their littermates (Figures 2A and2B), were normoglycemic, and hadcomparable serum insulin levels (Figures2C and 2D). The kidney size of theseanimals was indistinguishable from theircontrols (Figure 2E).

Detailed renal investigation at 3 weeks of age was normalin podIRKO mice with no abnormality identifiable using eitherlight (data not shown) or electron microscopic analysis(Figure 3G). They were not albuminuric. However, at approxi-mately 5 weeks of age, both podIRKO models (nephrin andpodocin Cre) started to develop albuminuria. The only histolog-ical change at this time was loss of structure of the podocytefoot process detected using electron microscopy (Figure 3G).At 8 weeks of age, significant levels of albuminuria (Figures2F and 2G) and histological changes were present in podIRKOanimals. These histological changes included extensive loss offoot process structure of the podocyte (Figure 3G). Further-more, there was evidence of podocyte apoptosis, as shownby hyperchromatic condensed nuclei visualized by hematoxylinand eosin (H&E) stain at 8 weeks of age (Figure 3A) and by

Cell Metabolism

Insulin and the Podocyte

330 Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc.

TUNEL staining at 13 weeks of age (Figure 3D). Furthermore,podIRKO mice had increased amounts of glomerular matrix(Figure 3B), which included type IV collagen (Figure 3C), thick-ened GBM (Figures 3H and 3I) and increased levels of glomer-ulosclerosis (Figure 3B). These are classical histologicalfeatures of DN (Wolf and Ziyadeh, 1999). As they aged, thesepathological features became more prominent (Figure 3E),with some of the podIRKO mice developing macroscopicallyshrunken and sclerosed kidneys (Figure 3F). However, we didnot find that mesangial hypercellularity was present on objec-tive assessment (Figure 3B).We detected phenotypic variation between individual po-

dIRKO mice both within the podocin and nephrin Cre groupsand between these groups (Figures 2G and 3B). The podocinCre animals were more albuminuric and had greater histologicalabnormality and worse renal function (Figure 3B) in comparisonto the nephrin Cre mice. However, although there was variation,both groups as a whole were significantly different from theirlittermate controls. There are a number of possible explanationsfor the variability between and within groups; first, the mice weregenerated on mixed genetic backgrounds, so genetic modifierscould be important. This could also explain the variability withineach group. It is recognized that genetic background is impor-tant for severity of disease both in diabetic (Kulkarni et al.,2003) and albuminuric models of renal disease (Gurley et al.,2006). Second, the mice were housed in different facilities, andit is likely that the podocin Cre mice, which were cared for ina conventional facility, were exposed to more environmentalstressors, which may have resulted in a more severe phenotype.Third, there may be differences in the efficiency of the Cre re-combinase between the two lines, resulting in less IR excisionin the nephrin Cre mice.We were interested to assess if compensation of IR loss was

occurring by upregulation of the IGF1 receptor, as insulin isable to signal through both of these receptors (although signalsconsiderably less through the IGF1 receptor [Kurtzhals et al.,2000]). In vivo, we found that IGF1 receptor mRNA was upregu-lated in the podocytes of podIRKO mice (Figure 1E) comparedwith controls. This suggests there is a cellular attempt tocompensate for IR loss, presumably to allow insulin to continueto signal to the podocyte.

Insulin Signals to the Podocyte in the GlomerulusTo study insulin stimulated signaling in the glomerulus, weutilized a mousemodel in which wild-type (WT) mice were fastedovernight, thereby lowering insulin to basal levels, and then giveneither insulin or saline via an intraperitoneal route. After 15 min,they were sacrificed, and their kidneys were immediately snap-frozen in liquid nitrogen in an attempt to preserve insulin-inducedprotein phosphorylation. We elected to study the mitogen-acti-vated protein kinase (MAPK), phosphatidylinositol 3-kinase(PI3K), and cap/cbl pathways, as these are the signalingcascades described in other rapidly insulin-sensitive cells(adipocytes and muscle) (Chiang et al., 2001; Shepherd andKahn, 1999). Ex vivo, in the glomerulus, we could robustly iden-tify insulin-stimulated signaling. For MAPK, the signaling wasexclusively detected in the podocyte using phosphospecific(p)42 and 44 MAPK (Figures 4A and 4B). We assessed the PI3Kpathway by using pAKT (Ser473) and found again that signaling

was predominantly in the podocyte (Figure 4E). We confirmedthese observations by studying immortalized human podocytesand GEnC in isolation and found that when the cells wereexposed to low doses of insulin (1 nM), insulin signaled to podo-cytes via both theMAPK and PI3K pathways (Figures 4C, 4D, 4F,and 4G), and that there was a small amount of PI3K signaling toGEnC (Figure 4F) but not MAPK (Figure 4C). Finally, we ascer-tained that podocyte insulin signaling was through the IR andnot through the IGF1 receptor by repeating the experimentsin vivo using podIRKOmice and in vitro using short hairpin siRNAIR knockdown human podocytes (Figures 4B, 4C, 4E, and 4F).These both showed abrogation of bothMAPK and PI3K signalingpathways in response to insulin. We could not detect any effectof insulin on the cap/cbl pathway in podocytes or GEnC (data notshown).

Insulin Directly and Specifically Reorganizes the ActinCytoskeleton of PodocytesPodocyte function in maintaining the integrity of the GFB isdependent on its structure, particularly in the filamentous actin(F-actin)-rich foot processes of this cell (Pavenstadt et al.,2003). We therefore assessed if there was any role for insulin inpodocyte actin remodeling.In vitro, we found that insulin directly remodeled the actin cyto-

skeleton of human podocytes, but not GEnC, in a dose-depen-dent manner, causing cortical reorganization of F-actin, andthat this was IR dependent (Figures 5A–5C). Actin reorganizationresulted in changes in podocyte structure, particularly retractionof cellular processes (akin to foot processes), as shown by high-resolution atomic force microscopy (Figure 5D). Insulin alsoincreased podocyte cell motility and migration, as measured byscratch assays (Figure 5E), and reduced its monolayer resis-tance, as assessedby sensitive electrical cell impedance sensing(ECIS) methodology (Figure 5F). This is a method that can accu-rately detect changes in monolayer permeability and is particu-larly good for studying cells that do not form tight monolayers inculture, as is the case for podocytes in their differentiated form,which is how they exist in vivo (Pavenstadt et al., 2003), andwhen they are most insulin sensitive (Coward et al., 2005). Aswith F-actin reorganization, insulin did not alter the monolayerpermeability in GEnC treated in the same manner (Figure 5F).IGF1 has the ability to signal via the IR (Kurtzhals et al., 2000)

and shows considerable structural homology with insulin(Rinderknecht and Humbel, 1978). However, we found function-ally that IGF1 had no biological effect on actin remodeling, alter-ation in monolayer electrical resistance (Figure S1), or migrationand motility (Figure 5E) of podocytes within the same time frameas that observed for insulin.We went on to explore the mechanism of insulin’s action on

podocyte actin remodeling by studying its effect on smallGTPases. Using our in vitro human model, we found that insulinactivated the GTP form of RhoA within 5 min and switched offthe GTP form of cell division cycle 42 (CDC42) within 2 min(Figure 5G). This was confirmed using dominant-negativeconstructs that were virally transfected into the cells. C3 trans-ferase (dominant-negative Rho) prevented insulin-stimulatedcortical actin reorganization, and dominant-negative CDC42(N17CDC42) resulted in cortical F-actin patterning in basal cells(Figure 5H).

Cell Metabolism

Insulin and the Podocyte

Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc. 331

Cell Metabolism

Insulin and the Podocyte

332 Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc.

DISCUSSION

This study shows that insulin rapidly and directly signals to thepodocyte in the glomerulus, is able to dynamically remodel theactin cytoskeleton of this cell, and has a critically importantrole in maintaining the integrity of the GFB. Furthermore, whenpodocyte insulin signaling is lost, glomerular changes developthat recapitulate a number of features typical of DN.The central hypothesis underpinning this study was that podo-

cyte-specific insulin sensitivity was important for glomerularfunction. This had originated from our previous work on condi-tionally immortalized human podocytes and GEnC suggestinga specific response of differentiated podocytes to insulin inrespect to rapid glucose uptake. To explore this hypothesis,we generated transgenic podIRKO mice. These mice wereborn with normal Mendelian frequency, were indistinguishablefrom littermates, and had no evidence of renal disease, implyingthat insulin signaling to the podocyte is not critical for glomerulardevelopment. This is similar to the whole-body IR knockoutmouse that is normal at birth but dies of diabetic ketoacidosisat 72 hr (Accili et al., 1996), as well as an array of other tissue-specific IRKO models that are also initially normal but developphenotypes as they age (Biddinger andKahn, 2006). Froma renalperspective, we found no abnormality until 5 weeks of age, whenalbuminuria started to appear, and the only histological abnor-mality detectable was alterations in the structure of theF-actin-rich podocyte foot processes. This is in keeping withthe observation that early in human DN, the cytoskeletal archi-tecture of the podocyte is altered, as shown by foot processwidening (Bjørn et al., 1995; Pagtalunan et al., 1997). By 8 weeksof age, podIRKO mice had developed significant albuminuriatogether with histological changes typical of DN (Alsaad andHerzenberg, 2007). These included loss of podocytemorphology, production of large amounts of matrix (includingtype IV collagen), increased GBM thickness, and glomeruloscle-rosis. There was also evidence of podocyte apoptosis. This isimportant, as podocyte loss is a feature of DN that occurs rela-tively early in its natural history and is an excellent predictor ofdisease progression that is superior to other pathologicalfeatures, including mesangial expansion, matrix accumulation,and thickening of the GBM (Pagtalunan et al., 1997; Wolf et al.,2005). Interestingly, the kidneys of the podIRKO mice were notenlarged, and on objective analysis of the glomerular mesan-gium, we found no evidence of hypercellularity. These are alsocharacteristic features of human DN and imply that there maybe aspects of DN that are the result of podocyte insulin resis-tance (podocyte actin disorganization and loss of foot processstructure, increased matrix, thickening of GBM, podocyte

apoptosis, and glomerulosclerosis) and aspects that are not(kidney enlargement and mesangial expansion). It may be thatthe insulin-nonresistant podocyte features are consequencesof hyperglycemic or insulin-resistant effects on the other cellsof the kidney (GEnC and mesangium).Our results suggest that insulin resistance of the podocyte

may be an important initiator for many of the pathologicalprocesses observed in DN. In respect to glucose uptake andactin remodeling, we have found that GEnCs do not rapidlyrespond to insulin. However, there is evidence that insulin alsodirectly signals to the vascular endothelium (Potenza et al.,2009), controlling features such as vascular tone via nitric oxide.Currently, there are nomurinemodels that are able to specificallytarget the GEnC. If they become available, they will be helpful indefining the direct role of insulin signaling on GEnC and whetherthis is also important for glomerular function.Cellular insulin resistance is central to the development of the

metabolic syndrome and type 2 diabetes. A role for podocyteinsulin resistance in the progression of nephropathy in thesesituations is therefore conceptually easy to understand. En-hancement of podocyte insulin sensitivity may also explainhow thiazolidinediones delay or stop progression of nephropathyin these settings (Panchapakesan et al., 2005). Type 1 diabetes iscaused by insulin deficiency, but the associated nephropathy issimilar to type 2 diabetes in respect to both its natural history andhistological features. Conceptually, in type 1 DN, podocyteinsulin resistance would appear not to be important. However,there is compelling evidence that patients with type 1 diabeteswho develop nephropathy are also more insulin resistant ata cellular level compared to those patients who do not developkidney disease. Type 1 nephropathic patients often requirehigher doses of insulin to treat their diabetes (Andersen et al.,1983), have higher levels of cellular insulin resistance as deter-mined by objective euglycemic clamp studies (Orchard et al.,2002), and are more likely to have a positive family history ofcellular insulin resistance (Hadjadj et al., 2004). Furthermore,type 1 diabetes itself induces cellular insulin resistance whenprolonged (Yki-Jarvinen and Koivisto, 1986), and nephropathycommonly takes years to develop. Therefore, it is distinctlypossible that podocyte insulin resistance is also influential inthe development of nephropathy in type 1 diabetes. Therapeuti-cally, this is important, as it implies that cellular insulin sensitizerscould be beneficial in preventing or slowing progression ofnephropathy in this setting. This is supported by a number oftype 1 diabetic animal models (Isshiki et al., 2000; Nicholaset al., 2001; Zhang et al., 2008) in which insulin sensitizershave been shown to slow renal disease progression independentof glycemic control.

Figure 2. PodIRKO Mice Are Normoglycemic and Normally Grown at 8 Weeks of Age but Have Albuminuria(A) Macroscopically, mice were indistinguishable.

(B) No difference between podIRKO and littermate controls after formal weighing (mean ±SEM, n R 10 each group).

(C) They had comparable random blood glucose levels in both podocin and nephrin Cre groups (mean ±SEM, n R 10 each group).

(D) No difference in serum insulin levels in a subset of mice at 8 weeks.

(E) Macroscopically no difference in kidney size between podIRKO and control mice at 8 weeks.

(F) At 8 weeks, podIRKO mice become albuminuric. SDS-PAGE gel shows specific band at 66 kDa corresponding to albumin only in podIRKO mice (2 ml urine

loaded for each).

(G) Significant proteinuria in podIRKO mice generated on different Cre promoters, raised in different environments. However, there is variability both between

strains and within the groups (***p < 0.001, n R 10 each group, mean ±SEM shown).

Cell Metabolism

Insulin and the Podocyte

Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc. 333

Cell Metabolism

Insulin and the Podocyte

334 Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc.

It was not surprising that insulin signals to the podocytethrough both the MAPK and PI3K pathways, given the ex-tensive literature on other rapidly insulin-responsive cells,such as adipocytes and skeletal muscle (Shepherd and Kahn,1999), and work performed on podocytes by us (Coward et al.,2005) and other groups (Tejada et al., 2008). However, the impor-tance and specificity of this effect in the kidney has not beenpreviously reported. We found that the rapid effects of insulinin phosphorylating key molecules in these pathways (p42/p44MAPK and pAKT [Ser473]) occurred predominantly in the podo-cyte of the kidney. We went on to knock down the IR both in vivoand in vitro, revealing that in both models, these signaling path-ways were abrogated. This supports the notion that insulinsignals predominantly through the IR in podocytes and notthrough the IGF1 or hybrid IGF1/IR receptor.Insulin directly, rapidly, and specifically reorganized the actin

cytoskeleton of normal human podocytes, again via the IR, andloss of podocyte foot process cytoskeletal architecture was theearliest abnormality detected in theGFBof our podIRKOmodels.This raises the possibility that insulin is normally able to maintainthe structure of the foot process of the podocyte and henceprevent breakdown of the GFB. We speculate that in healthyconditions, insulin-induced dynamic remodeling of the podocyteand particularly its foot process occurs physiologically aftera meal to deal with the increased filtration load the GFB experi-encesat this time.Webelieve this is facilitatedby the insulin-stim-ulated uptake of glucose into this cell through translocation ofglucose transporters (Coward et al., 2005) as a source of energy.Podocyte actin reorganization was achieved through activa-

tion of RhoA and inhibition of CDC42. This is consistent withstudies that have examined the activation state of these smallGTPases in models of relaxation (Zhang et al., 2007) and tight-ening (Gao et al., 2007) of cell junctions between podocytes.Ourwork has shown that insulin-activated RhoA causes contrac-tion and cortical reorganization of F-actin. This is not the case forfibroblasts that have been most intensely studied (Nobes andHall, 1995), where RhoA induces stress fibers. However, cortical

reorganization has been shown to occur in other cell types (Mad-dox and Burridge, 2003), and RhoA-induced loss of processes isa commonfinding in neurons (Postmaet al., 1996), a cell typewitha number of similarities to podocytes (Kobayashi et al., 2004).Collectively, this work shows that the podocyte is a uniquely

insulin-sensitive cell in the kidney and is able to rapidly remodelits actin cytoskeleton and morphology in response to insulin.Furthermore, insulin signaling to this cell is critically importantfor the glomerulus to function normally and, when diminished,results in a number of pathological features that are similar toDN. Enhancing insulin sensitivity of the podocyte is an attractiveapproach in the treatment of the commonest cause of renalfailure in the world.

EXPERIMENTAL PROCEDURES

In Vivo StudiesTransgenic Mice Studies

PodIRKO mice were generated by crossing transgenic mice who had loxP

sites inserted on either side of exon 4 (Bruning et al., 1998) of their IR with

transgenic Cre recombinase-producing mice that targeted the podocyte by

virtue of their podocyte-specific podocin (Ding et al., 2006) and nephrin (Ere-

mina et al., 2003) promoters. PodIRKO mice were homozygous for the floxed

IR gene and heterozygous for Cre recombinase. These animals were on

a mixed genetic background, with contributions from C57BL/6, 129/SV, and

FVB strains. Mice were genotyped using PCR as previously described (Brun-

ing et al., 1998; Ding et al., 2006; Eremina et al., 2003). Control mice were litter-

mates of podIRKOs and included flox/flox IR mice without Cre together with

Cre-positive mice with no floxed IR present. They were all fed on standard

chow. Z/EG reporter mice (Novak et al., 2000) were initially crossed with the

Cre-producing mice as previously described to ensure specific excision in po-

docytes. All animal husbandry was carried out in accordance with the Cana-

dian Council on Animal Care protocols, and protocols were approved by the

animal care committee at the Samuel Lunenfeld Research Institute.

CFP Podocyte Isolation and Quantitative Real-Time PCR

To isolate podocytes from glomeruli, we utilized the nephrin cyan fluorescent

protein (CFP) mouse (Cui et al., 2005), which has CFP linked to the nephrin

promoter and is able to label podocytes. These mice were crossed with po-

dIRKO mice, and animals were generated that were heterozygous for nephrin

CFP and podocin Cre and homozygous for loxP IR. These were then

Figure 3. PodIRKO Mice Develop Histological Features of Diabetic Nephropathy(A) Light microscopy for podocin Cre podIRKO mice and controls aged 8 weeks. Top panels are low-power Periodic Acid Schiff (PAS) staining; middle, high-

power PAS; and lower, H&E staining at high power. PodIRKO (middle column) demonstrate dilated tubules containing proteinaceous casts. The glomeruli

show increased mesangial matrix production. Several glomeruli appear shrunken, associated with foci of segmental glomerulosclerosis. The affected glomerular

segments show adhesion to Bowman’s capsule, segmental scarring, and capillary insudative (hyalinotic) lesions. Occasional podocytes show small, hyperchro-

matic, condensed nuclei suggestive of apoptotic bodies (arrowed bottom middle panel). Focal glomeruli show podocyte and parietal epithelial cell swelling with

large protein absorption droplets.

(B) Objective histological glomerular scoring reveals significant levels of glomerulosclerosis and matrix accumulation in both the podocin and nephrin cre mice.

Histological abnormality is more marked in the podocin animals. No difference in mesangial hypercellularity was detected (*p < 0.05, **p < 0.005, ***p < 0.0005,

nR 10 each group, mean ±SEM shown). Plasma creatinine levels were measured in a subset of mice at 8 weeks. No difference was detected in the nephrin Cre

mice, but in the podocin Cre IRKO animals, there was a mildly raised creatinine level in mutants when compared to controls.

(C) Type IV collagen progressively accumulates in glomeruli of podIRKO mice (brown staining).

(D) Podocyte apoptosis is detected using TUNEL stain in 13-week-old podIRKO mice (nephrin promoter) (apoptotic nuclei arrowed).

(E) Light microscopy of 13-week-old podIRKO (nephrin promoter) showed massive amounts of matrix formation as well as increased glomerulosclerosis,

mesangial expansion, and protein in the tubules in comparison to the 8-week-oldmice. Right, high power (363magnification); left, low power (310magnification)

PAS staining.

(F) Some 13-week-old podIRKO mice had small sclerosed kidneys.

(G) Transmission electron microscopy of podIRKO and control mice. PodIRKO mice have normal GFB at 3 weeks of age. By age 5 weeks, the podocyte actin

cytoskeleton is beginning to become deranged, with broadening of foot processes (arrowed). Foot processes are completely lost by 8 weeks (arrowed).

No obvious changes were detected in the GEnC.

(H) Thickening of the GBM of podIRKO mice when measuring the lamina densa.

(I) Multiple sections from three mutant mice were compared with three control mice. Each section had at least ten measurements of lamina densa thickness per-

formed perpendicular with the edge of the GBM. **p = 0.007, mean ±SEM shown.

Cell Metabolism

Insulin and the Podocyte

Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc. 335

Figure 4. Insulin Signals to the Podocyte in the Glomerulus(A) MAPK signaling. Mice given insulin phosphorylated p42/44 MAPK only in the glomerulus (310 magnification). Green, p42/44MAPK; red, PECAM (endothelial

marker); blue, nephrin (podocyte marker).

(B) Using high-power confocal microscopy (363 magnification), insulin-stimulated p42/44 MAPK colocalizes with nephrin, demonstrating specific podocyte

signaling (middle row, representative of three independent experiments). PodIRKO mice were unable to phosphorylate p42/44 MAPK in response to insulin

stimulation (lower row).

Cell Metabolism

Insulin and the Podocyte

336 Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc.

compared to other control nephrin CFP-positive mice that were also hetero-

zygous for only podocin Cre or homozygous for the floxed IR gene. Animals

were sacrificed and their kidneys passed through a 105 mm sieve and caught

on a 71 mm sieve. A single-cell suspension was made by incubating the

glomeruli in digestion buffer containing 0.2% trypsin, 1 mg/ml collagenase

A, 2 mg/ml EDTA, and 100 U/ml DNase I at 37!C for 40 min. 7-aminoactino-

myocin was added to the cells just prior to FACS sorting to identify dead

cells, which were excluded in the sort. They were then FACS sorted, and

gating was set to isolate a pure CFP-positive population. These were

compared to CFP-negative cells from the glomerular sort. To ensure we

had isolated a pure podocyte population in the CFP-positive fraction, we

measured mRNA production of nephrin (podocyte marker) and platelet/endo-

thelial cell adhesion molecule 1 (PECAM) (endothelial marker) in both frac-

tions. We found substantially more nephrin and less PECAM in the CFP-posi-

tive fraction, indicating we had specifically isolated podocytes in our

experiments (Figure S2).

To obtain enough mRNA from these experiments, we needed to pool

three mice from both the CFP-positive podIRKO group and from the control

group. Messenger RNA was prepared using TRIzol and reverse transcribed

into cDNA using the reverse first-strand DNA synthesis kit (Fermentas; Bur-

lington, Ontario, Canada). Expression for the IR, IGF1 receptor, nephrin, PE-

CAM, and HPRT (housekeeping gene) was measured. The primers used are

shown in Table S1. The comparative DDCT method of relative quantification

was used to calculate for differences in gene expression using the software

for ABI Prism 7900 Sequence Detection system (Applied Biosystems; Carls-

bad, CA).

Glomerular Isolation and Detection of Genomic Excision of the IR

This was performed as previously described (Cui et al., 2005) with the use of

magnetic Dynabeads. Identification of IR gene excision was achieved by

RT-PCR with primers flanking the loxP sites situated in the IR gene. The primer

sequence used was forward, TCTATCATGTGATCAATGATTC and reverse,

ATACCAGAGCATAGGAG.

Glomerular Insulin Signaling in WT Mice

We used the same approach as previously described (Patel et al., 2008) to

prepare, dose with insulin, and isolate the kidneys of thesemice. Phosphospe-

cific antibodies against 42/44 MAPK and Ser473 AKT were purchased from

Cell Signaling (Danvers, MA).

Histological Analysis

Kidneys were processed as previously described (Eremina et al., 2003). We

used phospho-antibodies from Cell Signaling to detect insulin-stimulated

phosphorylated proteins in both the tissue immunofluorescence and in our

cell culture experiments. Nephrin (Acris; Herford, Germany), PECAM (Pharmin-

gen; San Diego, CA), and type IV collagen (Cedarlane; Burlington, NC) anti-

bodies were used in the mouse experiments. Immunohistochemistry was

performed on formalin-fixed tissue as previously described (Eremina et al.,

2003), as was the TUNEL stain (Wijsman et al., 1993). Electron microscopy

was achieved using a technique similar to one previously employed (Eremina

et al., 2003), with one variation. The fixation buffer consisted of 0.1 M sodium

cocodylate and 2% glutaraldehyde. Blinded analysis of glomerular histology

was performed by an experienced histopathologist (A.M.H.) (Supplemental

Experimental Procedures).

In Vitro StudiesImmortalized Cells

Conditionally immortalized human podocytes and GEnC were used as previ-

ously described (Saleem et al., 2002; Satchell et al., 2006). Podocytes were

studied 2 weeks after thermoswitching from 33!C to 37!C. Thermoswitching

silences a temperature-sensitive SV40 gene in the cells, which prevents

them from entering the cell cycle (as occurs at 33!C) and allows them to differ-

entiate. This is important, as in vivo podocytes are terminally differentiated

cells. Conditionally immortalized podocytes express cellular markers and

morphologically resemble differentiated podocytes. Two different WT podo-

cyte cell lines and one endothelial cell line were studied. Different clones of

each cell line were studied to ensure we were not observing a clonal effect.

Passage numbers between 10 and 15 were examined.

Insulin Stimulation

Cells were insulin and fetal calf serum (FCS) starved for at least 12 hr prior to

stimulation. They were treated with insulin doses ranging from 0.1 mM to

200 mM for 15 min to assess actin reorganization. To elucidate signaling path-

ways, a dose of 100 nM for 15 min was used, which is common practice in

insulin cell signaling studies.

Immunofluorescence

Cells were prepared in 96-well CellBIND plates (Nunc; Roskilde, Denmark) or

on glass coverslips (Menzel-Glaser; Braunschweig, Germany). Theywere fixed

in 4% paraformaldehyde for 15 min, permeabilized with 0.2% Triton, and then

washed with 0.1% sodium borohydride. The actin cytoskeleton was visualized

using phalloidin-conjugated Alexa Flour 488 or Texas Red (Invitrogen) and

nuclei using DAPI (Sigma). Cells were imaged using the IN Cell Analyzer

(GE Healthcare), and changes were assessed and processed using the multi-

target analysis algorithm software as recently described (Caunt et al., 2008).

The IN Cell Analyzer is able to detect cells by virtue of their nuclei (DAPI stain-

ing); it then objectively identifies actin patterning and assigns each cell as

having either cortical or cytoplasmic distribution.

Atomic Force Microscopy

Atomic force microscopy was carried out on a commercial instrument (Nano-

Scope IIIa Controller and Dimension head; Veeco; Cambridge, UK) using stan-

dard triangular silicon nitride cantilevers with a nominal spring constant of

0.32 N/m (Olympus ORC8; Veeco). Cells were fixed with 4% sucrose/2%

paraformaldehyde and then imaged using intermittent contact mode in liquid

(under phosphate-buffered saline) (Tamayo et al., 2001).

Activated GTPase Assays and Dominant-Negative Studies

Measurement of the GTP form of CDC42 was achieved using EZ-Detect Small

GTPase Kits (Pierce; Northumberland, UK). Activated GTP RhoA was per-

formed using the same techniques as recently reported (Lennon et al.,

2008). Rho and CDC42 dominant-negative adenoviral constructs were gener-

ated byMyc-tagged C3-transferase and N17CDC42mutants being subcloned

them into pacAd vectors. These were fully sequenced and transfected into

differentiated podocytes 24 hr prior to experimentation. We found that adeno-

viral transfection was excellent (greater than 95%) in podocytes using adeno-

viral GFP-tagged RhoA and CDC42 constructs (Figure S3).

Electrical Cell Substrate Impedance Sensing

Time course analysis and resistance measurement was performed using an

automated cell monitoring system, Electrical Cell-Substrate Impedance

Sensing (ECIS 1600R, Applied Biophysics; Rochester, NY), as recently

described (Satchell et al., 2007). This detects nanometer scale changes in

cell morphology in real time by virtue of the changes in electrical resistance

it detects. Briefly, immortalized podocytes and endothelial cells were seeded

onto gold microelectrodes in wells of ECIS arrays (eight wells, ten electrodes

per well). Confluent cell monolayers were treated with 100 nM insulin or vehicle

control in usual medium. Using the ECIS attachment mode resistance was

measured at regular time intervals of 30 s. The ratio of measured resistance

to baseline resistance was calculated for each well and plotted as a function

of time, normalized to control.

Scratch Assay

This was performed as previously described (Vollenbroker et al., 2009).

Short Hairpin Insulin Receptor Knockdown

Specific lentiviral transduction particles (Sigma) were used to produce stable

knockdown human podocyte cell clones. The lentiviral transduction particles

(C) p42/44 MAPK is switched on at 15 min with both low-dose (1 nM) and high-dose (100 nM) insulin in WT human podocytes. This does not occur in IR knock-

down podocyte cell lines and GEnC in response to low-dose insulin.

(D) Densitometry in WT podocytes reveals a 120% increase in insulin-stimulated p42/44MAPK at 15 min (p = 0.0093, n = 6, mean ±SEM shown).

(E) pAKT signaling is predominantly in podocytes after 15 min stimulation with insulin in WT mice (upper and middle panels, representative images from three

independent experiments). Insulin-stimulated AKT phosphorylation was abrogated in podIRKO mice (lower panels).

(F) AKT is phosphorylated inWT human podocytes by 15min at both low- and high-dose insulin. This does not occur in insulin receptor knockdown podocytes. In

contrast to MAPK, there is a small amount of pAKT detectable in human GEnC in response to low- and high-dose insulin.

(G) Densitometry shows a 200% increase in pAKT phosphorylation in response to insulin in human podocytes (**p = 0.007, n = 6, mean ±SEM shown).

Cell Metabolism

Insulin and the Podocyte

Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc. 337

Cell Metabolism

Insulin and the Podocyte

338 Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc.

consist of an average of 3–5 individual constructs targeting different regions of

the gene sequence. Therefore, a range of knockdown efficiency can be

expected. Proliferating podocytes at 33!C were coincubated with approxi-

mately 525 lentiviral transducing particles in standard culturemedia containing

Polybrene (Sigma) for 24 hr. Following viral incubation, the cells were washed

with normal media, then left to recover for a further 24 hr in fresh media before

the addition of puromycin dihydrochloride. Antibiotic selection pressure was

exerted on the cells until individual puromycin-resistant colonies of cells devel-

oped. These clones were then selected, expanded, and examined for their

level of receptor knockdown. We analyzed knockdown by western blot and

quantitative real-time PCR and studied a clone that had 70% knockdown of

the IR (Figure S4).

Statistics

Comparison between two groups was performed using Student’s t test. For

more than two groups, analysis of variance with a post hoc Bonferroni was

used. Statistical significance was defined as a p value less than 0.05. Standard

error of the mean (SEM) is shown for each population.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures,

Supplemental References, four figures, and one table and can be found with

this article online at doi:10.1016/j.cmet.2010.08.015.

ACKNOWLEDGMENTS

R.J.M.C. and L.J.H. are supported by the Medical Research Council (grant

number G0501901). The majority of the work presented here was funded by

this grant. R.L. and S.C.S. are supported by the Wellcome trust. A.M.H. was

supported by Canadian Diabetes Association Operating Grant #300349.

Some of this work was supported by a grant awarded by North Bristol NHS

Trust. We would like to thank Doug Holmyard, who processed the electron

microscopic pictures; Ken Harpal, who prepared the renal histology; and

Dionne White, who FACS sorted the glomerular cells. This work is dedicated

to the brilliant Andrew Herzenberg (1971–2010), who was a major driving force

within this project.

Received: January 10, 2010

Revised: May 21, 2010

Accepted: July 7, 2010

Published: October 5, 2010

REFERENCES

Accili, D., Drago, J., Lee, E.J., Johnson, M.D., Cool, M.H., Salvatore, P., Asico,

L.D., Jose, P.A., Taylor, S.I., and Westphal, H. (1996). Early neonatal death in

mice homozygous for a null allele of the insulin receptor gene. Nat. Genet. 12,

106–109.

Alsaad, K.O., and Herzenberg, A.M. (2007). Distinguishing diabetic nephrop-

athy from other causes of glomerulosclerosis: an update. J. Clin. Pathol. 60,

18–26.

Andersen, A.R., Christiansen, J.S., Andersen, J.K., Kreiner, S., and Deckert, T.

(1983). Diabetic nephropathy in Type 1 (insulin-dependent) diabetes: an epide-

miological study. Diabetologia 25, 496–501.

Biddinger, S.B., and Kahn, C.R. (2006). From mice to men: insights into the

insulin resistance syndromes. Annu. Rev. Physiol. 68, 123–158.

Bjørn, S.F., Bangstad, H.J., Hanssen, K.F., Nyberg, G., Walker, J.D., Viberti,

G.C., and Osterby, R. (1995). Glomerular epithelial foot processes and filtration

slits in IDDM patients. Diabetologia 38, 1197–1204.

Brown, E.J., Schlondorff, J.S., Becker, D.J., Tsukaguchi, H., Tonna, S.J.,

Uscinski, A.L., Higgs, H.N., Henderson, J.M., and Pollak, M.R. (2010). Muta-

tions in the formin gene INF2 cause focal segmental glomerulosclerosis.

Nat. Genet. 42, 72–76.

Bruning, J.C., Michael, M.D., Winnay, J.N., Hayashi, T., Horsch, D., Accili, D.,

Goodyear, L.J., and Kahn, C.R. (1998). A muscle-specific insulin receptor

knockout exhibits features of the metabolic syndrome of NIDDM without

altering glucose tolerance. Mol. Cell 2, 559–569.

Caunt, C.J., Rivers, C.A., Conway-Campbell, B.L., Norman, M.R., and

McArdle, C.A. (2008). Epidermal growth factor receptor and protein kinase C

signaling to ERK2: spatiotemporal regulation of ERK2 by dual specificity phos-

phatases. J. Biol. Chem. 283, 6241–6252.

Chiang, S.H., Baumann, C.A., Kanzaki, M., Thurmond, D.C., Watson, R.T.,

Neudauer, C.L., Macara, I.G., Pessin, J.E., and Saltiel, A.R. (2001). Insulin-

stimulated GLUT4 translocation requires the CAP-dependent activation of

TC10. Nature 410, 944–948.

Coward, R.J., Welsh, G.I., Yang, J., Tasman, C., Lennon, R., Koziell, A., Satch-

ell, S., Holman, G.D., Kerjaschki, D., Tavare, J.M., et al. (2005). The human

glomerular podocyte is a novel target for insulin action. Diabetes 54, 3095–

3102.

Coward, R.J., Welsh, G.I., Koziell, A., Hussain, S., Lennon, R., Ni, L., Tavare,

J.M., Mathieson, P.W., and Saleem, M.A. (2007). Nephrin is critical for the

action of insulin on human glomerular podocytes. Diabetes 56, 1127–1135.

Cui, S., Li, C., Ema, M., Weinstein, J., and Quaggin, S.E. (2005). Rapid isolation

of glomeruli coupled with gene expression profiling identifies downstream

targets in Pod1 knockout mice. J. Am. Soc. Nephrol. 16, 3247–3255.

Ding, M., Cui, S., Li, C., Jothy, S., Haase, V., Steer, B.M., Marsden, P.A.,

Pippin, J., Shankland, S., Rastaldi, M.P., et al. (2006). Loss of the tumor

suppressor Vhlh leads to upregulation of Cxcr4 and rapidly progressive

glomerulonephritis in mice. Nat. Med. 12, 1081–1087.

Eremina, V., Sood, M., Haigh, J., Nagy, A., Lajoie, G., Ferrara, N., Gerber, H.P.,

Kikkawa, Y., Miner, J.H., and Quaggin, S.E. (2003). Glomerular-specific alter-

ations of VEGF-A expression lead to distinct congenital and acquired renal

diseases. J. Clin. Invest. 111, 707–716.

Figure 5. Insulin Directly Remodels the Actin Cytoskeleton of Podocytes(A) Immortalized human podocytes and GEnC were insulin stimulated (100 nM) for 15 min, and their F-actin cytoskeletons were analyzed with phalloidin staining.

Podocytes cortically reorganized (arrowed) their cytoskeleton, and this did not occur in GEnC.

(B) Insulin receptor knockdown (IRKD) immortalized podocytes do not remodel their F-actin cytoskeleton in response to insulin (phalloidin staining).

(C) Objective computer-assisted analysis with the IN Cell Analyzer showed a dose response in insulin-stimulated actin reorganization in WT podocytes that was

statistically significant from 6.25 nM (ANOVA, p < 0.001; Bonferroni, *p < 0.05, **p < 0.001 compared with [cw] baseline; n = 8; mean ±SEM). This did not occur in

insulin-stimulated IRKD podocytes.

(D) Atomic force microscopy representative of four independent experiments. Insulin causes retraction of cell processes. Insulin was applied for 15 min.

(E) Cell motility is increased in podocytes given insulin but not with IGF1. Number of cells per unit area was significantly higher in the insulin-treated group 15 hr

after insulin stimulation. This did not occur in IGF1 treatment (n = 4, **p < 0.01, mean ±SEM shown).

(F) Electrical resistance is lost across podocyte monolayers within 5 min of insulin stimulation but returns to baseline by 30 min using ECIS analysis. ANOVA,

p< 0.01; Bonferroni, *p < 0.05, **p < 0.01. No effect on GEnC treated in the same manner (n = 12). Mean [symbol:plus/minus]SEM shown.

(G) Insulin activates RhoA by 5 min (*p = 0.018) and inhibits CDC42 by 2 min (*p = 0.02) of insulin stimulation. These are both back to baseline by 30 and 15 min,

respectively (n = 4 each, mean ±SEM shown).

(H) Differentiated human podocytes transfected with dominant-negative constructs for Rho (c3 transferase) and CDC42 (N17CDC42). There is diminished insulin

induced cortical actin reorganization in Rho dominant-negatively treated cells. CDC42 dominant-negative transfection resulted in a high proportion of cells being

cortically organized in their basal state.

Cell Metabolism

Insulin and the Podocyte

Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc. 339

Gao, S.Y., Li, C.Y., Shimokawa, T., Terashita, T., Matsuda, S., Yaoita, E., and

Kobayashi, N. (2007). Rho-family small GTPases are involved in forskolin-

induced cell-cell contact formation of renal glomerular podocytes in vitro.

Cell Tissue Res. 328, 391–400.

Gurley, S.B., Clare, S.E., Snow, K.P., Hu, A., Meyer, T.W., and Coffman, T.M.

(2006). Impact of genetic background on nephropathy in diabetic mice. Am. J.

Physiol. Renal Physiol. 290, F214–F222.

Hadjadj, S., Pean, F., Gallois, Y., Passa, P., Aubert, R., Weekers, L., Rigalleau,

V., Bauduceau, B., Bekherraz, A., Roussel, R., et al; Genesis France-Belgium

Study. (2004). Different patterns of insulin resistance in relatives of type 1 dia-

betic patients with retinopathy or nephropathy: the Genesis France-Belgium

Study. Diabetes Care 27, 2661–2668.

Isshiki, K., Haneda, M., Koya, D., Maeda, S., Sugimoto, T., and Kikkawa, R.

(2000). Thiazolidinedione compounds ameliorate glomerular dysfunction

independent of their insulin-sensitizing action in diabetic rats. Diabetes 49,

1022–1032.

Kobayashi, N., Gao, S.Y., Chen, J., Saito, K., Miyawaki, K., Li, C.Y., Pan, L.,

Saito, S., Terashita, T., and Matsuda, S. (2004). Process formation of the renal

glomerular podocyte: is there common molecular machinery for processes of

podocytes and neurons? Anat. Sci. Int. 79, 1–10.

Kulkarni, R.N., Almind, K., Goren, H.J., Winnay, J.N., Ueki, K., Okada, T., and

Kahn, C.R. (2003). Impact of genetic background on development of hyperin-

sulinemia and diabetes in insulin receptor/insulin receptor substrate-1 double

heterozygous mice. Diabetes 52, 1528–1534.

Kurtzhals, P., Schaffer, L., Sørensen, A., Kristensen, C., Jonassen, I., Schmid,

C., and Trub, T. (2000). Correlations of receptor binding and metabolic and

mitogenic potencies of insulin analogs designed for clinical use. Diabetes

49, 999–1005.

Lavin, P.J., Gbadegesin, R., Damodaran, T.V., and Winn, M.P. (2008). Thera-

peutic targets in focal and segmental glomerulosclerosis. Curr. Opin. Nephrol.

Hypertens. 17, 386–392.

Lennon, R., Singh, A., Welsh, G.I., Coward, R.J., Satchell, S., Ni, L., Mathieson,

P.W., Bakker, W.W., and Saleem, M.A. (2008). Hemopexin induces nephrin-

dependent reorganization of the actin cytoskeleton in podocytes. J. Am.

Soc. Nephrol. 19, 2140–2149.

Maddox, A.S., and Burridge, K. (2003). RhoA is required for cortical retraction

and rigidity during mitotic cell rounding. J. Cell Biol. 160, 255–265.

Nicholas, S.B., Kawano, Y., Wakino, S., Collins, A.R., and Hsueh, W.A. (2001).

Expression and function of peroxisome proliferator-activated receptor-

gamma in mesangial cells. Hypertension 37, 722–727.

Nishikawa, T., Edelstein, D., andBrownlee,M. (2000). Themissing link: a single

unifying mechanism for diabetic complications. Kidney Int. Suppl. 77,

S26–S30.

Nobes, C.D., and Hall, A. (1995). Rho, rac, and cdc42 GTPases regulate the

assembly of multimolecular focal complexes associated with actin stress

fibers, lamellipodia, and filopodia. Cell 81, 53–62.

Novak, A., Guo, C., Yang, W., Nagy, A., and Lobe, C.G. (2000). Z/EG, a double

reporter mouse line that expresses enhanced green fluorescent protein upon

Cre-mediated excision. Genesis 28, 147–155.

Orchard, T.J., Chang, Y.F., Ferrell, R.E., Petro, N., and Ellis, D.E. (2002).

Nephropathy in type 1 diabetes: a manifestation of insulin resistance and

multiple genetic susceptibilities? Further evidence from the Pittsburgh Epide-

miology of Diabetes Complication Study. Kidney Int. 62, 963–970.

Pagtalunan, M.E., Miller, P.L., Jumping-Eagle, S., Nelson, R.G., Myers, B.D.,

Rennke, H.G., Coplon, N.S., Sun, L., and Meyer, T.W. (1997). Podocyte loss

and progressive glomerular injury in type II diabetes. J. Clin. Invest. 99,

342–348.

Panchapakesan, U., Chen, X.M., and Pollock, C.A. (2005). Drug insight: thiazo-

lidinediones and diabetic nephropathy—relevance to renoprotection. Nat.

Clin. Pract. Nephrol. 1, 33–43.

Patel, S., Doble, B.W., MacAulay, K., Sinclair, E.M., Drucker, D.J., and Wood-

gett, J.R. (2008). Tissue-specific role of glycogen synthase kinase 3beta in

glucose homeostasis and insulin action. Mol. Cell. Biol. 28, 6314–6328.

Pavenstadt, H., Kriz, W., and Kretzler, M. (2003). Cell biology of the glomerular

podocyte. Physiol. Rev. 83, 253–307.

Perkovic, V., Verdon, C., Ninomiya, T., Barzi, F., Cass, A., Patel, A., Jardine,M.,

Gallagher, M., Turnbull, F., Chalmers, J., et al. (2008). The relationship between

proteinuria and coronary risk: a systematic review and meta-analysis. PLoS

Med. 5, e207.

Postma, F.R., Jalink, K., Hengeveld, T., and Moolenaar, W.H. (1996). Sphingo-

sine-1-phosphate rapidly induces Rho-dependent neurite retraction: action

through a specific cell surface receptor. EMBO J. 15, 2388–2392.

Potenza, M.A., Addabbo, F., and Montagnani, M. (2009). Vascular actions of

insulin with implications for endothelial dysfunction. Am. J. Physiol. Endocri-

nol. Metab. 297, E568–E577.

Rinderknecht, E., and Humbel, R.E. (1978). The amino acid sequence of

human insulin-like growth factor I and its structural homology with proinsulin.

J. Biol. Chem. 253, 2769–2776.

Saleem, M.A., O’Hare, M.J., Reiser, J., Coward, R.J., Inward, C.D., Farren, T.,

Xing, C.Y., Ni, L., Mathieson, P.W., and Mundel, P. (2002). A conditionally

immortalized human podocyte cell line demonstrating nephrin and podocin

expression. J. Am. Soc. Nephrol. 13, 630–638.

Satchell, S.C., Tasman, C.H., Singh, A., Ni, L., Geelen, J., von Ruhland, C.J.,

O’Hare, M.J., Saleem, M.A., van den Heuvel, L.P., and Mathieson, P.W.

(2006). Conditionally immortalized human glomerular endothelial cells ex-

pressing fenestrations in response to VEGF. Kidney Int. 69, 1633–1640.

Satchell, S.C., Buchatska, O., Khan, S.B., Bhangal, G., Tasman, C.H., Saleem,

M.A., Baker, D.P., Lobb, R.R., Smith, J., Cook, H.T., et al. (2007). Interferon-

beta reduces proteinuria in experimental glomerulonephritis. J. Am. Soc.

Nephrol. 18, 2875–2884.

Shepherd, P.R., and Kahn, B.B. (1999). Glucose transporters and insulin

action—implications for insulin resistance and diabetes mellitus. N. Engl.

J. Med. 341, 248–257.

Tamayo, J., Humphris, A.D., Owen, R.J., and Miles, M.J. (2001). High-Q

dynamic force microscopy in liquid and its application to living cells. Biophys.

J. 81, 526–537.

Tejada, T., Catanuto, P., Ijaz, A., Santos, J.V., Xia, X., Sanchez, P., Sanabria,

N., Lenz, O., Elliot, S.J., and Fornoni, A. (2008). Failure to phosphorylate AKT

in podocytes from mice with early diabetic nephropathy promotes cell death.

Kidney Int. 73, 1385–1393.

Vollenbroker, B., George, B., Wolfgart, M., Saleem, M.A., Pavenstadt, H., and

Weide, T. (2009). mTOR regulates expression of slit diaphragm proteins and

cytoskeleton structure in podocytes. Am. J. Physiol. Renal Physiol. 296,

F418–F426.

Wijsman, J.H., Jonker, R.R., Keijzer, R., van de Velde, C.J., Cornelisse, C.J.,

and van Dierendonck, J.H. (1993). A new method to detect apoptosis in

paraffin sections: in situ end-labeling of fragmented DNA. J. Histochem. Cyto-

chem. 41, 7–12.

Wolf, G., and Ziyadeh, F.N. (1999). Molecular mechanisms of diabetic renal

hypertrophy. Kidney Int. 56, 393–405.

Wolf, G., Chen, S., and Ziyadeh, F.N. (2005). From the periphery of the glomer-

ular capillary wall toward the center of disease: podocyte injury comes of age

in diabetic nephropathy. Diabetes 54, 1626–1634.

Yki-Jarvinen, H., and Koivisto, V.A. (1986). Natural course of insulin resistance

in type I diabetes. N. Engl. J. Med. 315, 224–230.

Zhang, H., Cybulsky, A.V., Aoudjit, L., Zhu, J., Li, H., Lamarche-Vane, N., and

Takano, T. (2007). Role of Rho-GTPases in complement-mediated glomerular

epithelial cell injury. Am. J. Physiol. Renal Physiol. 293, F148–F156.

Zhang, H., Saha, J., Byun, J., Schin, M., Lorenz, M., Kennedy, R.T., Kretzler,

M., Feldman, E.L., Pennathur, S., and Brosius, F.C., 3rd. (2008). Rosiglitazone

reduces renal and plasma markers of oxidative injury and reverses urinary

metabolite abnormalities in the amelioration of diabetic nephropathy.

Am. J. Physiol. Renal Physiol. 295, F1071–F1081.

Zimmet, P., Alberti, K.G., and Shaw, J. (2001). Global and societal implications

of the diabetes epidemic. Nature 414, 782–787.

Cell Metabolism

Insulin and the Podocyte

340 Cell Metabolism 12, 329–340, October 6, 2010 ª2010 Elsevier Inc.