Nck Adaptor Proteins Link Nephrin at the Podocyte Slit Diaphragm to the Hippo Regulator WTIP

Ava Keyvani Chahi

A Thesis presented to The University of Guelph

In partial fulfilment of requirements for the degree of Master of Science in Molecular and Cellular Biology

Guelph, Ontario, Canada

ABSTRACT NCK ADAPTOR PROTEINS LINK NEPHRIN AT THE PODOCYTE SLIT DIAPHRAGM TO THE HIPPO REGULATOR WTIP

Ava Keyvani Chahi Advisor: University of Guelph, 2014 Dr. Nina Jones

Podocytes are specialized epithelial cells that contribute to the kidney blood filtration barrier. Their unique cytoskeletal architecture and other complex biological signals are largely maintained by a modified adherens junction known as the slit diaphragm (SD). A major component of the SD is the transmembrane protein nephrin, which upon tyrosine phosphorylation of the intracellular domain regulates actin remodeling and cell survival. Nck adaptor proteins are a critical component of the filtration barrier and connect nephrin to actin-remodeling proteins by binding phosphotyrosine residues and proline-rich motifs. Herein we identify a novel Nck binding partner, Wilm’s tumor interacting protein (WTIP). WTIP is a transcription regulator in podocytes, though Nck is not nuclear localized under conditions known to induce WTIP nuclear accumulation. However, we demonstrate that WTIP is recruited to nephrin, upon nephrin tyrosine phosphorylation by Src Family Kinases, in an Nck-dependent manner. WTIP is an evolutionarily conserved negative regulator of the Hippo kinase pathway, which inhibits the transcription factor Yap. Yap activity promotes podocyte survival. We show in mice that cannot recruit Nck to nephrin, and by extension WTIP, that Yap protein levels are downregulated. Similarly, Yap downregulation is demonstrated in wildtype mice subjected the nephrotoxic serum model of podocyte injury, which is known to induce robust decrease in nephrin tyrosine phosphorylation. Overall, through identification of a novel Nck binding partner, WTIP, we have identified a novel mechanism through which Nephrin tyrosine phosphorylation is linked to the Hippo signaling cascade, which may in turn promote podocyte survival.

In dedication to

the memory of my friend,

Jason Kuchinsky

ACKNOWLEDGEMENTS

I am eternally grateful to my supervisor Dr. Nina Jones. Nina, thank you for giving me the freedom to disagree and the freedom to explore. Working with Nina and the members of the

Jones Laboratory has made me develop immensely as a person and as a researcher. I also owe

Dr. Terry Van Raay a debt of gratitude for really going above and beyond his role as a committee member. Dr. Van Raay is a wonderful teacher and no matter how busy he was he made his door open to offer me advice.

I would also like to thank past and present members of the Jones Lab and our surrounding supervisors and colleagues in the Van Raay, Bendall and Mosser Labs; especially Jahdiel (Jaddy)

Larraguibel, Rasmeet (Ras) Dhaliwal, Rachael (Katie) MacKenzie who were all awesome lab mates and housemates and without them I likely would not have found affordable housing during this degree. Others who have provided support and laughs include Agata Zienowicz, Kevin

Bosse, Colin Stringer, Cameron Harris, Christopher James (Jim) Cooper, Olivia Catherine

Anderson, Rebecca Rumney and so many more! The most important lab member that I need to thank is Dr. Una Adamicic-Bistrivoda, the former post-doctoral fellow in our lab, for mentorship, training and making sure I never became a “skirt-holder”. I also need to acknowledge those I collaborate with most frequently, Claire Elyse Martin and Dr. Laura A.

New, who were both instrumental in designing tools used in this thesis. Lastly, but far from least,

I need to thank my friend (Dr.–to-be) Melanie K.B. Wills who has been a role model for me and is always a source of scientific support, motivation and hysteric laughter.

I need to thank all of my wonderful friends and family for enduring the hardships of university with me- and hopefully for enjoying it with me. I can’t list everyone but I have to

iv thank Thomas Bracken Coyne for being a constant source of motivation and optimism; as well as for the countless number of times he listened while I rambled about my research.

I also need to thank all of the feline sensations on the Internet that have made me laugh until I cry and were my companions during many late-nights of western blotting. Also, my research’s financial supporters: the University of Guelph College of Biological Sciences, Ontario

Graduate Scholarship, and the Kidney Foundation of Canada all deserve acknowledgement.

Lastly, thank you to Maman, Baba and my spunky sister, Aram. Truly, my parents have made every sacrifice a person could make for their children and I accredit them for any and every one of my accomplishments.

v DECLARATION OF WORK PERFORMED

All of the work in this thesis was performed by Ava Keyvani Chahi, with the following exceptions. Isolation and generation of wildtype, Nck1 knockout and Nck1 and 2 knockout mouse podocyte cells was performed by Claire Elyse Martin. Subcloning of WTIPΔPRR from pEGFP-C1 to pCDNA3.1-HA was designed and organized by Ava Keyvani Chahi and performed by Olivia Catherine Anderson. All mouse organization and handling was done by Dr. Laura A. New and the University of Guelph Central Animal Facility. The β-actin forward and reverse primers were previously designed by Steve Hawley. Also, the Western Blot for Figure

16A was partially provided by Claire Elyse Martin.

vi TABLE OF CONTENTS

ACKNOWLEDGEMENTS…………………………………………………………………….iv

DECLARATION OF WORK PERFORMED………………………………………………...vi

TABLE OF CONTENTS……………………………………………………………………....vii

LIST OF TABLES…………………………………………………………………………...... x

LIST OF FIGURES……………………………………………………………………………..xi

LIST OF ABBREVIATIONS……………………………………………………………….....xii

CHAPTER 1: INTRODUCTION……………………………………………………………….1

SIGNIFICANCE OF KIDNEY RESEARCH………………………………………………………...... 2

KIDNEY ANATOMY………………………………………………………………………………..3

GLOMERULAR FILTRATION BARRIER……………………………………………………………5

PODOCYTES……………………………………………………………………………………….6

PODOCYTE CYTOARCHITECTURE………………………………………………………………...6

PODOCYTE EFFACEMENT…………………………………………………………………………9

PODOCYTE ADHESION STRUCTURES……………………………………………………………10

NEPHRIN: A MAJOR COMPONENT OF THE SLIT DIAPHRAGM……………………………………12

NCK ADAPTOR PROTEINS……………………………………………………………………….15

NCK AND NEPHRIN INTERACTION………………………………………………………………16

WILM’S TUMOR-1 INTERACTING PROTEIN (WTIP)……………………………………………19

WTIP IS A REGULATOR OF THE HIPPO KINASE PATHWAY…………………………………….21

SUMMARY AND RATIONALE……………………………………………………………………..24

CHAPTER 2: MATERIALS AND METHODS……………………………………………...26

TISSUE CULTURE………………………………………………………………………………...27

vii LPS TREATMENT………………………………………………………………………………..28

TRANSFECTION…………………………………………………………………………………..28

VECTORS……………………………………..…………………………………………………..28

CLONING OF GFP- AND HA- TAGGED WTIPΔPRR……………………………………………28

ANTIBODIES……………………………………………………………………………………...30

CD16 STIMULATION AND PP2 TREATMENT…………………………………………………….30

CELL LYSIS………………………………………………………………………………………31

CELL FRACTIONATION…………………………………………………………………………..31

IMMUNOBLOTTING………………………………………………………………………………32

IMMUNOPRECIPITATION………………………………………………………………………...33

GST PULLDOWN ASSAY…………………………………………………………………………33

RNA EXTRACTION AND CDNA PREPARATION…………………………………………………34

SEMI-QUANTITATIVE REAL TIME PCR………………………………………………………...34

Y3F NEPHRIN MICE AND NTS INJECTION………………………………………………………..35

GLOMERULI ENRICHMENT……………………………………………………………………...35

CHAPTER 3: RESULTS………………………………………………………………………37

NCK AND WTIP INTERACT……………………………………………………………………...38

MAPPING OF THE WTIP-NCK INTERACTION TO THE FIRST AND THIRD NCK SH3 DOMAINS...39

MAPPING OF THE WTIP-NCK INTERACTION TO THE PRR OF WTIP…………………………44

NCK SUBCELLULAR LOCALIZATION DOES NOT EXHIBIT SIMILAR TRENDS AS WTIP AND NCK

EXPRESSION DOES NOT INFLUENCE TRANSCRIPT LEVELS OF WT1 TARGET GENES……………46

WTIP IS RECRUITED TO CD16-NEPHIN IN AN NCK-DEPENDENT MANNER…………………….49

viii WTIP RECRUITMENT TO CD16-NEPHRIN IS DEPENDENT ON NEPHRIN PHOSPHORYLATION BY

SFK AND ON NCK’S PHOSPHO-TYROSINE- AND PROLINE- BINDING FUNCTIONS……………….50

TOTAL YAP LEVELS IN MOUSE GLOMERULI ARE INFLUENCED BY NEPHRIN

PHOSPHORYLATION……………………………………………………………………………...53

CHAPTER 4: DISCUSSION…………………………………………………………………..55

CHAPTER 5: CONCLUSIONS……………………………………………………………….67

REFERENCES………………………………………………………………………………….69

ix LIST OF TABLES

Table 1: Primer sequences (from 5’-3’)………………………………………………………....36 Table 2: Summary of Nck interaction partners and reinterpretation of mapping data……….…58

x LIST OF FIGURES

Figure 1: Schematic of a kidney and nephron……………………………………………………4

Figure 2: Schematic of glomerulus and glomerular filtration barrier…………………………….6

Figure 3: Podocyte Structure……………………………………………………………………..8

Figure 4: Podocyte Effacement………………………………………………………………….10

Figure 5: Schematic of adjacent foot processes connected by Slit Diaphragm………………....12

Figure 6: Schematic of Nck protein domains…………………………………………………...16

Figure 7: Schematic of WTIP protein domains and motifs……………………………………..20

Figure 8: Schematic of the Hippo kinase pathway in mammals………………………………...23

Figure 9: Summary of the molecular signals in healthy and injured podocytes relevant to the research in this thesis…………………………………………………………………………….25

Figure 10: Nck and WTIP interaction…………………………………………………………...39

Figure 11: Nck and WTIP interact via Nck SH3 domains……………………………………...41

Figure 12: Mapping of Nck and WTIP interaction to the first and third Nck SH3 domains……43

Figure 13: Mapping of the Nck and WTIP interaction to the WTIP proline rich region……….45

Figure 14: Nck nuclear localization in response to LPS and influence on gene expression in cultured podocytes……………………………………………………………………………….48

Figure 15: Nck dependent recruitment of WTIP to nephrin…………………………………….50

Figure 16: WTIP recruitment to CD16-nephrin is dependent on nephrin phosphorylation by SFK and on Nck’s phospho-tyrosine- and proline- binding functions………………………………...52

Figure 17: Nephrin tyrosine phosphorylation is a regulator of total Yap protein levels………..54

Figure 18: Schematic of proposed model of Yap regulation in mature podocytes through the nephrin-Nck-WTIP signaling complex…………………………………………………………..66

xi LIST OF ABBREVIATIONS AKI Acute Kidney Injury Arp Actin related protein BSA Bovine Serum Albumin CB Cell Body CD2AP Cluster of differentiation 2 associated protein ChIP Chromosome Immunoprecipitation CKD Chronic Kidney Disease CMV Cytomegalovirus DM Double Mutant DMEM Dulbecco's Modified Eagle's Medium DMSO Dimethyl sulfoxide EDTA Ethylenediaminetetraacetic acid EGTA Ethylene glycol tetraacetic acid ESRF End Stage Renal Disease ETEF Eukaryotic Translation Elongation Factor FAK Focal Adhesion Kinase FBS Fetal Bovine Serum FP Foot Process GBM Glomerular Basement Membrane GFP Green Fluorescent Protein Glepp Glomerular epithelial protein GST Glutathione S-Transferase HEK Human Embryonic Kidney IB Immunoblot IFN Interferon IgG Immunoglobin G ILK Integrin-Linked Kinase IP Immunoprecipitation LIM Lin11, Isl-1 and Mec-3 LPS Lipopolysaccharide

xii MPC Mouse Podocyte Cell N-WASp Neuronal- Wiskott–Aldrich Syndrome protein NAP Nck Associated Protein Nck Non-catalytic tyrosine kinase NES Nuclear Export Signal NIK Nck Interacting Kinase NLS Nuclear Localization Signal NTS Nephrotoxic Serum PAGE Polyacrylamide Gel Electrophoresis PAK p21-activated kinase PAN Puromycin aminonucleoside PBS Phosphate Buffered Saline PEI Polyethylenimine PI3K Phosphoinositide 3-kinase PINCH Particularly Interesting New Cysteine-Histidine rich protein PLC Phospholipase C PMSF Phenylmethanesulfonyl fluoride PP Primary Process PRR Proline Rich Region PVDF Polyvinylidene fluoride QRTPCR Quantitative Real Time Polymerase Chain Reaction Ras Rat Sarcoma SD Slit Diaphragm SDS Sodium dodecyl sulfate SFK Src Family Kinases SH Src Homology SOCS Suppressor of Cytokine Signaling TBST Tris Base Saline Tween WIP WASp interacting protein WT Wildtype WT1 Wilm's Tumor-1

xiii WTIP Wilm's Tumor-1 Interacting Protein ZO Zona Occludens

xiv CHAPTER 1: INTRODUCTION

1. Significance of Kidney Research

The worldwide prevalence of kidney disease is rapidly rising (Nahas and Bello, 2013).

Kidney disease may manifest due to genetic and/or congenital syndromes and these were once the most common causes of kidney disease (Reiser and Sever, 2013). However, today the leading cause of chronic kidney disease (CKD) is type II diabetes and hypertension (Pyram et al., 2012;

Jha et al., 2013). Kidney disease can be a diverse range of disorders that affect kidney structure and function and is distinguished based on duration as acute kidney injury (AKI) (less than 3 months) or CKD (more than 3 months) (Eckhardt et al., 2013). CKD is defined as a blood filtration rate lower than 60 mL/min per 1.73 m2 for minimum of 3 months and is associated with elevated protein excretion in urine (known as albuminuria or proteinuria) (Echkardt et al., 2013;

Jha et al., 2013). CKD is estimated to affect 8-16% of the population worldwide (Jha et al.,

2013), with the percentage of affected patients doubling over the age of 65 (Pyram et al., 2012), making CKD a significant public health and economic burden in both developed and developing nations (Pyram et al., 2012; Echkardt et al., 2013; Jha et al., 2013; Nahas and Bello, 2013).

Because the kidneys are highly involved in maintaining the body’s homeostasis, CKD is largely associated with cardiovasular complications as well as increased risk for AKI, decreased cognition, and mineral and bone disorders (Jha et al., 2013). Despite the enormous rise in kidney disease, the treatments available are limited and are not evolving as rapidly as their demand

(Reiser and Sever, 2013). Patients with CKD are treated with drugs that target the conditions causing their kidney damage, such as angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, corticosteroids, antibiotics, anti-inflammatory or anti-cancer drugs, with the primary intention of reducing or slowing the progression of CKD (Reiser and Sever, 2013),

2 rather than drugs that directly target the kidneys. When CKD progresses despite treatment, the patient can develop end-stage renal failure (ESRF) (Nahas and Bello, 2013; Reiser and Sever,

2013), which is when the blood filtration rate has declined below 15% (Eckardt et al., 2013). At this stage the only treatment options are dialysis or kidney transplantation (Reiser and Sever,

2013).

2. Kidney Anatomy

Kidneys are organs of the urinary system primarily responsible for blood filtration and excretion, but also for regulation of blood pressure through selective retention of water and solutes, homeostatic functions through maintenance of circulating glucose, amino acids, calcium and phosphate, and endocrine functions (Berndt, 1976). Within the kidney there are approximately one million functional units individually known as nephron (Quaggin and

Kreidberg, 2008). The first site of blood filtration through each nephron is the glomerulus, a tuft of capillaries surrounded by Bowman’s capsule. The ultrafiltrate in Bowman’s capsule travels through the continuous lumen of the proximal convoluted tubule, loop of Henle, distal convoluted tubule and collecting duct, with urine eventually entering the renal pelvis and ureter for excretion (Berndt, 1976) (Figure 1).

Figure 1: Schematic of a kidney and nephron. Cross section showing the relative locations of the renal cortex, medulla and pelvis and a magnified nephron indicating the directional flow of filtration from the glomerulus and Bowman’s capsule (1 and 2, respectively) to the proximal convoluted tubule (3), loop of Henle (4), and distal convoluted tubule (5). The final product, urine, exits to the renal pelvis via the collecting duct (6). (Adapted from Hall 12th edition)

Often the first site of kidney damage occurs at the glomerulus (Kriz, 2002; Kriz and

LeHir, 2005). A hallmark of glomerular damage is less selective and increased permeability of the glomerular filtration barrier resulting in protein passage into the nephron tubules and proteinuria. Consequently, the nephron tubules sustain damage because they are not molecularly equipped to mediate massive protein reabsorption (Kriz and LeHir, 2005). This can lead to nephron degeneration, glomerulosclerosis and eventually ESRF (Asanuma and Mundel, 2003).

Despite this disease burden, the molecular basis of renal physiology and disease is an area of research in its infancy. An increased understanding of the molecular basis of renal function and damage will thus eventually lead to the discovery and engineering of less invasive, more direct treatments for patients with kidney disease.

4 3. Glomerular Filtration Barrier

The glomerular filtration barrier is a capillary wall composed of three structurally distinct layers, which cooperatively enforce selectivity based on charge, shape and size. When blood enters the glomerular capillaries, the first layer of the glomerular filtration barrier encountered is the fenestrated endothelium. The second layer is an unusually thick basal lamina, called the glomerular basement membrane (GBM). The GBM is jointly in direct contact with the endothelial cells and the third layer of terminally differentiated polarized epithelial cells known as podocytes, which elaborate unique cellular structures known as foot processes (FP) that are fundamental for blood filtration at the glomerulus (Quaggin and Kreidberg, 2008; Miner and

Abrahamson, 2013) (Figure 2). Our understanding of the individual contributions of each of these layers to the selectivity of the glomerular filtration barrier is still limited. However, the unique structure of podocytes, which is simplified in disease states, is known to be critical for the permselectivity at the glomerulus (Asanuma and Mundel, 2003; Pavenstadt et al., 2003;

Mathieson, 2012).

Figure 2: Schematic of glomerulus and glomerular filtration barrier. A tuft of glomerular capillaries in the Bowman’s capsule, magnified to show a single capillary with the three layers of the glomerular filtration barrier, the first layer of fenestrated endothelial cells, the shared GBM and the last layer of podocytes that have unique cellular architecture including regularly spaced FP that are connected by slit diaphragm (Adapted from Tryggvason and Wartiovaara, 2005).

4. Podocytes

4.1. Podocyte Cytoarchitecture

Differentiated podocytes have three major cytosolic domains, the cell body (CB), primary processes (PP) and secondary FP (Pavenstadt et al., 2003) (Figure 2 and 3A and B). The cell body is large and protrudes into the urinary space of Bowman’s capsule. The PP are microtubule and intermediate filament rich extensions from the cell body that wrap around the capillary loop.

The secondary FP are basolateral cellular projections off the PP that are densely packed with filamentous (F-) actin, myosin and α-actinin based microfilaments. Within single podocytes, actin microfilaments of the FP are connected to each other in loops (Figure 3C), which are also

6 connected to the microtubules of the PP (Pavenstadt et al., 2003). FP of adjacent podocytes are organized in regularly spaced, parallel interdigitations and are connected by modified adherens junctions known as slit diaphragm (SD) (Pavenstadt et al., 2003; Miner and Abrahanson, 2013).

The SD concomitantly act as molecular sieves and hubs of cell signal transduction (Pavenstadt et al., 2003). A hallmark of glomerulopathies is continuous simplification of podocyte cytomorphology, a process known as FP effacement (Figure 4), which is associated with disorganization of FP actin cytoskeleton, loss of selectivity of glomerular filtration and proteinuria (Asanuma and Mundel, 2003; Mathieson, 2012).

Figure 3: Podocyte Structure. A) Scanning electron micrograph of the surface of a glomerular capillary showing the outer layer coated by the podocyte CB, PP and FP (Adapted from Pavenstadt et al., 2003). B) Transmission electron micrograph of a cross section of the glomerular capillary showing the three layers of the glomerular filtration barrier. The fenestrated endothelial cells are in closest proximity with the blood cells, above that is the GBM, on top of which sit the podocyte FP that are connected by a thin electron-dense junction, the SD (Adapted from Pavenstadt et al., 2003). C) Coronal section showing a podocyte PP and multiple projecting actin-rich FP. The arrows point to the loops of F-actin connecting the FP of a single podocyte to each other (Adapted from Faul et al., 2007).

8 4.2. Podocyte Effacement

Dynamics of the FP actin cytoskeleton are regulated in response to extracellular events communicated through transmembrane proteins at cell-to-cell and cell-to-matrix adhesions. In response to extracellular stressors such as reactive oxygen species (ROS), polyvalent cations

(protamine sulfate, puromycin), high glucose or endotoxins (lipopolysaccharide (LPS)), podocyte FP undergo broadening, retraction or effacement, resulting in compromised filtration function (proteinuria). Mechanisms of effacement include altered expression, modification and retrograde migration of SD proteins from the slit to apical membrane and migration of cell-to- matrix contacts to reorganize the actin cytoskeleton into a mat of cross-linked actin stress fibers which strengthen adhesion between the cell and matrix, and in turn compromise the integrity of the filtration slits (Pavenstadt et al., 2003). Effacement does not involve rearrangements of the microtubules, as vinblastin, a microtubule disruptor, to rat kidneys does not alter podocyte FP

(Tyson, 1977). Podocyte FP effacement is reversible, therefore representing a capacity for structural plasticity, which is hypothesized to be a mechanism to resist permanent damage

(Reiser and Sever 2013; Kriz, 2002; Kriz and LeHir, 2005; Faul et al., 2007). Permanent, irreversible damage to podocytes, such as apoptosis or detachment from the GBM (and anoikis) can lead to cellular hypertrophy for the remaining podocytes as they must extend further to cover the entire capillary surface, or present gaps in the GBM that can be invaded by the outer parietal epithelial cells of Bowman’s capsule (Kriz, 2002; Pavenstadt et al., 2003). Massive protein leak as a result of podocyte loss can damage nephron tubules leading to nephron degeneration. There is very little evidence of possible recovery after nephron degeneration (Kriz and LeHir, 2005).

Figure 4: Podocyte Effacement. Scanning electron micrographs of the surface of glomerular capillaries with A) healthy FP and B) effaced FP. Transmission electron micrographs indicating the location of the fenestrated endothelial cells (E) and GBM, and showing regularly spaced FP and filtration slits of a C) healthy glomerular filtration barrier compared to D) a glomerular filtration barrier with FP effacement and loss of SD (Adapted from Lefevre et al., 2010).

4.3. Podocyte Adhesion Structures

Differentiated podocytes adhere to the GBM through integrin-based focal adhesions at basal surfaces and tips of FP (Kretzler, 2002; Kagami and Kondo, 2004; Sachs and Sonnenberg,

2013). Due to space constraints and research focus, focal adhesions are not discussed at length in this thesis. The SD are the major cell-to-cell contacts formed between adjacent differentiated podocytes (Pavenstadt et al., 2003). Interdigitating FP form 30-45 nm (300-450Å) wide slits occupied by the SD. The SD is a zipper-like protein-based structure with 40Å pores, which are permeable to plasma solutes but not blood cells or proteins based on size (Rodewald and

Karnovsky, 1974). The SD is composed of many different transmembrane adhesion molecules, including typical adherens junction molecules (placental) P-cadherin (Aaltonen and Holthofer,

2007; Patrakka and Tryggvasson, 2010; Noris and Remuzzi, 2012), atypical protocadherin FAT1

10 (Inoue et al., 2001), unique adhesion molecules such as Nephrin (Ruotsalainen et al., 1999),

Neph1-3, podocalyxin (Nielsen and McNagny, 2009) and additional transmembrane proteins

TRPC6 and Glepp1 (Aaltonen and Holthofer, 2007; Noris and Remuzzi, 2012). The significance of the cadherin molecules at the SD is unclear, as P-cadherin knockout mice have no renal phenotype, however FAT1 (Ciani et al., 2003), Neph1 (Donoviel et al., 2001) and nephrin

(Putaala et al., 2001) knockout mice die soon after birth with massive proteinuria. The intracellular domains of SD proteins are associated with many adaptor proteins, including α-, β-,

γ- catenin, podocin (Schwarz et al., 2001; Saleem et al., 2002), CD2AP, ZO-1 and Nck, and effector molecules, including Mena, (Neuronal) Wiskott–Aldrich Syndrome Protein ((N-

)WASp), actin related protein (Arp) 2/3, α-actinin 4, synaptopodin and cortactin, to induce remodeling of the underlying FP actin cytoskeleton (Faul et al., 2007; Patrakka and Tryggvason;

2010; New et al., 2014) (Figure 5).

11 ,7M8$.1:&-676$?:.&966:&-"6$.NN.N&

,$F1& H>0:&*08#/"8QL& ,#0$8>&L.LP"81.&

;48$F101&G& -676$8>?@01& E>.##5& .D"01& H?18#:6#6701& 'ABC9&

9,!& !"#$%& '.#/& 2345& -676$01& '$(& )*+,-& '.#/"01& <& $8:.101& H>0:&L.LP"81.& &)87/."01& ;& =& O8N8>&L.LP"81.& 01:.Q"01& IJ4GK&1L& EOR&

9.1.N:"8:.7&B176:/.>0SL&

Figure 5: Schematic of adjacent foot processes connected by Slit Diaphragm. Adjacent foot processes are respresented above the fenestrated endothelial and glomerular basement membrane (GBM) layers of the filtration barrier. The apical, basal and slit components of the podocyte membrane are labeled. Foot processes are separated by a gap 30-45nm wide that is occupied by the slit diaphragm. The slit diaphragm is composed of transmembrane proteins (nephrin, Neph1- 3, FAT1, cadherin, podocalyxin, Glepp1, TRPC6) that are connected to intracellular adaptor proteins (CD2AP, podocin, Nck, ZO-1, catenins) and actin-associated proteins (α-actinin 4, synaptopodin).

5. Nephrin: a major component of the Slit Diaphragm

Nephrin is the central molecule of the SD (Aaltonen and Holthofer, 2007; Welsh and

Saleem, 2010). Nephrin is a transmembrane protein identified as the gene product of the NPHS1 locus, which is mutated in human cases of congenital nephrotic syndrome of the Finnish type

(Kestila et al., 1998; Ruotsalainen et al., 1999). Patients with congenital nephrotic syndrome of the Finnish type exhibit FP effacement and massive proteinuria in utero which becomes life threatening during infancy (Kestila et al., 1998). Testifying to the importance of nephrin at the

12 SD, nephrin deficient mice die by 24 hours after birth with massive proteinuria and edema

(Putaala et al., 2001) and podocyte-specific knock-in of rat nephrin in these mice rescues some lethality (Juhila et al., 2010).

Nephrin is a single-spanning transmembrane protein with eight N-terminal extracellular immunoglobulin (Ig)-like domains, a fibronectin type-III motif and a short cytoplasmic domain that has six tyrosine residues that are conserved between human, mouse and rat. In agreement with the proposed porous zipper-like ultrastructure of the SD, nephrin makes trans-homophilic interactions via its extracellular Ig-like domains (Tryggvason, 1999; Putaala et al., 2000;

Khoshnoodi et al., 2003). The nephrin extracellular domain is also able to heterophilically interact with the extracellular domain of the other essential SD proteins Neph1 (Barletta et al.,

2003; Gerke, 2003) and Neph3 (Heikkila et al., 2011).

Within the cytoplasmic tail of nephrin, six conserved tyrosine residues are targets of phosphorylation upon adhesion or clustering, highlighting nephrin’s role as a signaling receptor

(Lahdenpera et al., 2003). Nephrin is not a receptor tyrosine kinase and does not auto- phosphorylate its tyrosine residues; instead they are phosphorylated by Src family kinases (SFK)

Src, Yes and Fyn (Lahdenpera et al., 2003; Jones et al., 2006; Blasutig et al., 2008).

A mouse model has recently been designed and characterized by our research group in which nephrin is replaced by a variant with three tyrosine residues mutated to phenylalanine

(called nephrinY3F mice) (New, Jones, unpublished). NephrinY3F mice are phenotypically normal at birth but develop early and more severe proteinuria in adulthood compared to wildtype control littermates. Therefore, nephrin tyrosine phosphorylation (on these three sites) does not appear to be required during development, however is important for maintenance of mature podocytes

(New, Jones, unpublished). In mature podocytes, the majority of available evidence supports the

13 model that in unstimulated podocytes nephrin tyrosine phosphorylation and actin turnover is low

(Patrakka and Tryggvason, 2007) and nephrin phosphorylation increases when actin remodeling or (re-) formation of mature FP is required. In response to most injuries nephrin tyrosine phosphorylation is decreased and associated with FP effacement (Li et al., 2006; Jones et al.,

2009; New, Jones, unpublished). As effacement is reversible, nephrin clustering and tyrosine phosphorylation are upregulated to restore the mature structure of FP (Lahdenpera et al., 2003;

Faul et al., 2007; Moeller, 2007; Blasutig et al., 2008).

Nephrin interacts with intracellular proteins that have Src homology (SH)2 domains via phosphorylated tyrosine (Buday et al., 2002). These interactions initiate signaling cascades at the

SD that are mechanisms of not only actin rearrangement, but also regulation of more complex systems such as cell survival and differentiation (Tryggvason et al., 2006; Zenker et al., 2009).

These interactions and signaling events downstream of nephrin tyrosine phosphorylation are reviewed in Patrakka and Tryggvason (2007), Schlondorff (2008), Zenker et al. (2009), Welsh and Saleem (2010), and New et al. (2014) and will only be briefly summarized herein. Nephrin phosphorylation facilitates its interaction with podocin, resulting in appropriate localization of nephrin to lipid rafts at the slit membrane, competitively preventing association with and internalization by β-arrestin, thus allowing for downstream nephrin signaling (Huber et al.,

2003b; Quack et al., 2006). Nephrin phosphorylation regulates actin through interactions with the adaptor proteins Nck1 and Nck2 (Jones et al., 2006), which will be the focus of the research herein and discussed in detail below, and the adaptor protein CD2AP, which recruits synaptopodin and α-actinin-4 (Welsch et al., 2001; Asanuma et al., 2005). Furthermore, nephrin tyrosine phosphorylation allows for association with dendrin, effectively suppressing dendrin nuclear localization and pro-apoptotic signaling (Asanuma et al., 2007). As well, nephrin

14 tyrosine phosphorylation leads to recruitment of phosphatidylinositol 3-kinase (PI3K), which activates AKT-dependent survival signaling (Huber et al., 2003a). In accordance with the absence of developmental phenotype in the nephrinY3F mice, Done et al. (2008) report that in embryos that are nephrin deficient podocyte proliferation and survival during development are similar as wildtype control animals, therefore redundancy in survival signaling does exist. For example, AKT-dependent survival signaling is also regulated by integrin-linked kinase (Simons and Huber, 2008). Despite the lack of developmental importance, nephrin tyrosine phosphorylation is hypothesized to contribute significantly to mature podocyte function and survival (New, Jones, unpublished).

6. Nck adaptor proteins

Nck family of adaptor proteins, consisting of Nck1 (Nckα) and Nck2 (Nckβ or Grb4) and collectively called Nck, are ubiquitously expressed Src homology (SH)2/SH3 domain-containing cytoskeletal adaptor proteins. Nck has no intrinsic enzymatic activity but facilitates phosphotyrosine signal transduction pathways by connecting signaling and effector molecules via interactions mediated by SH2/SH3 domains. The domain composition of Nck includes three

N-terminal SH3 domains and one C-terminal SH2 domain. Nck binds phosphorylated tyrosines on signal-initiating proteins, such as nephrin (Jones et al., 2006) via its SH2 domain (Buday et al., 2002). Via its SH3 domains, Nck recruits downstream effector molecules by binding directly to proline rich motifs, preferentially the Pi-x-x-Pi+3 (P- proline, x- any amino acid) core consensus (Yu et al., 1994; Mayer, 2000) (Figure 6). Nck has a single conserved tryptophan in each SH3 domain (Nck1 W39, W143, W229 and Nck2 W39, W149, W235, in the first, second and third SH3 domains, respectively) and mutation to lysine abolishes the respective domain’s proline-binding capability (Tanaka et al., 1995).

Figure 6: Schematic of Nck protein domains. Nck has three N-terminal SH3 domains, each capable of binding P-x-x-P motifs (first, second and third, counted from N to C) and a C-terminal SH2 domain capable of binding phospho-tyrosine (pY) residues.

7. Nck and Nephrin Interaction

The interaction between the Nck SH2 domain and human nephrin was mapped using site- directed mutagenesis to three phosphotyrosine residues in the SFK target sequence I/L-Y-D-x-V

(I- isoleucine, L- leucine, Y- tyrosine, D- aspartate, V- valine) on human nephrin at positions

1176, 1193 or 1217 (mouse 1191, 1208, 1232) (Jones et al., 2006). These three tyrosines were mutated in the nephrinY3F mice (New et al., 2014). Nck2, but not Nck1 can be co-isolated in detergent-insoluble fractions of the cell membrane, representing lipid rafts, from cultured conditionally immortalized mouse podocytes (Verma et al., 2006). Antibody-induced clustering of nephrin or a CD16-nephrin chimera protein that has a CD16-extracellular domain and CD7 intracellular domain fused to nephrin’s intracellular tail in cultured MEF and NIH3T3 cells induces nephrin tyrosine phosphorylation (Lahdenpera et al., 2003), which recruits Nck (and N-

WASp) to the membrane and is associated with localized actin polymerization (Jones et al. 2006;

Verma et al., 2006; Blasutig et al., 2008). A multitude of biochemical evidence supports the

16 proposal that downstream of nephrin phosphorylation at the podocyte SD Nck regulates actin through recruitment and complex formation with N-WASp, WASp interacting protein (WIP) and

Actin-related protein (Arp)2/3 complex (Welsch et al., 2001; Rivera et al., 2004; Zhang et al.,

2005; Jones et al, 2006; Blasutig et al., 2008; Padrick et al., 2011; Ditlev et al., 2012; Banjade and Rosen, 2014).

Nck also links nephrin at the membrane of cultured podocytes to the actin cytoskeleton via p21-activated kinase (PAK) (Zhu et al., 2010). PAK, like Nck, is required for the normal morphology of differentiated podocytes (Zhu et al., 2010) and it was demonstrated elsewhere that Nck promotes PAK activation at the membrane (Lu et al., 1997; Howe et al., 2000; Vidal et al., 2002). Also, PAK dephosphorylation (inactivation) parallels nephrin dephosphorylation in the puromycin aminonucleoside (PAN) model of nephrosis in rats, suggesting that PAK is involved in podocyte cytoskeletal plasticity downstream of nephrin and Nck (Zhu et al., 2010).

Nck also has many more actin-regulating interaction partners such as Nck interacting kinase

(NIK) (Buday et al., 2002); Nck associated protein (NAP)4 (also known as Suppressor Of

Cytokine Signaling (SOCS)7); NAP5 (Matuoka et al., 1997); NAP1 (Kitamura et al., 1996); focal adhesion kinase (FAK) (Goicoechea et al., 2002); and Particularly Interesting New

Cysteine and Histidine rich protein (PINCH)1 (Tu et al., 1998). However, the physiological relevance of these interactions in podocytes or downstream of nephrin signaling are undefined.

Overall, there appears to be vast potential for characterization of novel signaling pathways downstream of nephrin and Nck at the SD.

In addition to Nck binding phosphorylated nephrin, Nck is associated with the enhancement or maintenance of nephrin phosphorylation (in vitro and in vivo) (New et al.,

2013). One proposed mechanism for this is that upon Nck binding to phosphotyrosine residues of

17 nephrin, there is consequent recruitment of Fyn kinase by the Nck SH3 domains. Thus, if Nck cannot bind nephrin, Fyn recruitment and nephrin phosphorylation is low (New et al., 2013). The

Nck-dependent positive feedback of nephrin tyrosine phosphorylation is correlated with activation of downstream nephrin signaling to AKT and cofilin (New et al., 2013). Furthermore, this phenomenon is significantly higher when mediated by Nck2 compared to Nck1 (New et al.,

2013), perhaps consistent with successful co-isolation of Nck2 from podocyte lipid rafts (Verma et al., 2006) where nephrin tends to localize (Huber et al., 2003b). This is suggestive of independent or non-overlapping functions for Nck1 and Nck2 in podocytes. However, independent roles for Nck proteins in podocytes remain unclear, as mice with podocyte specific deletion of one of the Nck proteins are phenotypically normal (Jones et al., 2006).

In vivo evidence supports that Nck is a critical factor for podocyte development and maintenance (Jones et al., 2006; Li et al., 2006; Verma et al., 2006; Jones et al., 2009). Jones et al. (2006) developed a mouse model in which Nck1 and 2 were deleted specifically in podocytes in utero and found that mutant mice died from massive proteinuria by 3.5 weeks of age and electron microscopy showed that mutant embryos had reduced FP formation (Jones et al., 2006).

Further studies by Jones et al. (2009) in which Nck1 and 2 were deleted from adult mouse podocytes resulted in proteinuria within 2-4 weeks of knockout induction. Therefore, Nck is not only responsible for formation of FP during development but also maintenance of FP in adult mice. Based on the available biochemical evidence it was hypothesized that Nck regulates podocyte cytomorphology through its interaction with nephrin and connection to the actin cytoskeleton. However, more recent evidence from nephrinY3F mice indicates that loss of nephrin tyrosine phosphorylation on sites that recruit Nck has a less severe phenotype than deletion of Nck in podocytes (New, Jones, unpublished). Therefore, Nck must have nephrin-

18 independent roles that are required for podocyte development and maintenance. In light of these results, an ongoing research initiative has been the investigation of Nck as a linker between cell- to-cell and cell-to-matrix adhesions by associating with focal adhesion complexes through SH3 domain-mediated interactions with FAK (Goicoechea et al., 2002) and PINCH1 (Tu et al.,

1998). A parallel research initiative has been the identification of novel Nck binding partners in podocytes.

We have identified Wilm’s tumor-1 interacting protein (WTIP) as a putative Nck SH3 domain binding partner, and the major objective of my research described in this thesis was to investigate this novel interaction and its relevance in podocyte physiology.

8. Wilm’s tumor-1 Interacting Protein (WTIP)

WTIP was first identified in a yeast-two-hybrid screen for Wilm’s tumor-1 (WT1) binding partners. In the mature glomerulus WT1 is exclusively expressed in podocytes where it acts as a transcription activator and repressor for many genes, including some involved in maintaining FP structure, such as podocalyxin (Drossopoulou et al. 2009; Palmer et al. 2001), nephrin (Guo et al., 2002; Wagner et al. 2004; Takashi et al., 2010) and Neph3 (Ristola et al.,

2012). WT1 is also an activator of the amphiregulin promoter (Lee et al., 1999) and binds tightly to the Fyn promoter (Hartwig et al., 2010). In further examination, Srichai et al. (2004) validate that WTIP is expressed in podocytes. Also Srichai et al. (2004) show that WTIP and WT1 are binding partners and that WTIP suppresses WT1-dependent transcriptional activation in the amphiregulin-promoter luciferase reporter assay.

WTIP is a member of the ajuba family of LIM domain-containing proteins that includes three mammalian homologues; ajuba, lim-domain containing protein 1 (Limd1) and WTIP. The members of the ajuba family, including WTIP, contain three highly similar C-terminal LIM

19 (Lin11, Isl-1 and Mec-3) domains. WTIP also has a proline-rich pre-LIM region, N-terminal classical nuclear export signal (NES) (AALLLAGLGLRES), which is recognized by Crm1- export receptor, and a VTEL PDZ-binding motif (Srichai et al., 2004). WTIP was identified as a putative Nck binding partner by Dr. Nina Jones (unpublished) because of the presence of five conserved P-x-x-P motifs in the pre-LIM proline rich region (PRR) (Srichai et al., 2004) (Figure

7). Likewise, ajuba has been shown to interact via its pre-LIM PRR with SH3 domains of a similar adaptor protein as Nck, Grb2 (Goyal et al., 1999).

Figure 7: Schematic of WTIP protein domains and motifs. WTIP has an N-terminal nuclear export signal (NES), a proline rich linker region (PRR) that contains five conserved P-x-x-P motifs, three LIM domains and a C-terminal PDZ-binding motif coded as VTEL.

In standard cultured podocytes tagged WTIP localizes with markers of focal adhesions; paxillin and vinculin, until cell-to-cell adhesions begin to form, at which point WTIP localizes with markers of developing and mature cadherin-based adherens junctions; including cadherin,

β-catenin (Srichai et al., 2004; Kim et al., 2012) and ZO-1 (Rico et al., 2005). WTIP is believed to be responsible for formation of adherens junctions (Kim et al., 2012), however the mechanisms of adherens junction regulation by WTIP are still poorly understood. Knockdown studies in cultured podocytes show that reduced WTIP leads to failure of stress fiber elaboration at adherens junctions and the F-actin of adjacent cells is not organized in a parallel manner as it is in control cells (Kim et al., 2010). Kim et al. (2010) postulate that WTIP regulates F-actin at cell-to-cell contacts by regulating RhoA activation, however the evidence for this is inconclusive.

20 WTIP subcellular localization is not exclusive to the cytoplasm, and like many LIM domain-containing proteins (Marie et al., 2003), a population of WTIP does exist in the nucleus

(Srichai et al., 2004). In podocytes, WTIP reversibly accumulates in the nucleus in response to multiple models of induced podocyte injury (Kim et al., 2010). In cultured podocytes, nuclear accumulation of WTIP in response to the PAN injury model is associated with decreased expression of retinoblastoma binding protein (Rbbp7), a target of WT1, supporting the notion that WTIP is a repressor of WT1-dependent transcriptional activity (Rico et al., 2005).

Interestingly, Nck is also reported to exist in the nucleus of many cell types (Lawe et al., 1997), despite the lack of a nuclear localization signal (NLS) or NES. However there is currently very little known about how Nck is shuttled into or out of the nucleus or Nck’s functions in the nucleus.

10. WTIP is a Regulator of the Hippo Kinase Pathway

In addition to their role as cytoskeletal regulators, the ajuba family of proteins has also been reported as negative regulators of the Hippo pathway (Thakur et al., 2010). The Hippo pathway is conserved from S. Cerevisiae and D. Melanogaster to mammals and in mammals is a signaling pathway that regulates proliferation, apoptosis, differentiation and survival (Varelas,

2014). In vertebrates the Hippo pathway is regulated by mechanical cues from cell-to-cell contacts (and constitutes a mechanism of contact inhibition of growth) (Zhao et al., 2007; Kim et al., 2011), also by focal adhesions, F-actin, and chemical cues communicated through G-protein coupled receptors (Varelas, 2014). The core Hippo pathway is composed of mammalian kinases

MST1/2 and Lats1/2, and adaptors WW45 and Mob1/2 (Figure 8). Activation of the Hippo pathway, for example by formation of cell-to-cell contacts, results in phosphorylation and nuclear exclusion (thus inhibition) of the transcription regulators Yap and Taz by Lats1/2 (Figure

21 8). Phosphorylation of S127 on Yap or S298 on Taz results in relocalization of Yap/Taz from the nucleus to the cytoplasm where they are inhibited mainly through binding 14-3-3. Furthermore, phosphorylation of S127/ S298 of Yap/Taz primes them for an additional phosphorylation by

Lats1/2 on S297/ S111 respectively, which targets Yap/Taz for ubiquitination and degradation.

In the nucleus Yap/Taz activate the TEAD 1-4 transcription factors, leading to cell proliferation or survival and generally inhibition of apoptosis (Varelas, 2014) (Figure 8). Currently there has been very little published data on Yap/Taz activity specifically in podocytes. Yap knockdown from cultured podocytes (Campbell et al., 2013) and podocyte specific knockout of Yap

(Schwartzman et al., 2014) increases podocyte apoptosis; therefore Yap is considered an antiapoptotic factor in healthy podocytes.

WTIP binds Lats1/2 or WW45 via its LIM domains to inhibit Lats1/2 mediated phosphorylation and inhibition of Yap/Taz (Thakur et al., 2010, Codelia et al., 2012; Sun and

Irvine, 2013; Reddy and Irvine, 2013; Rauskolb et al., 2014). Upstream signals that promote

WTIP-mediated inhibition of Lats1/2 in non-podocyte cells are also signals initiated in response to podocyte injury models (Harris, 2011; Bollee et al., 2011; Advani et al., 2011; Codelia et al.,

2012; Kim et al., 2012; Sun and Irvine, 2013; Reddy and Irvine, 2013). Therefore, I hypothesize that the Hippo pathway is regulated in response to podocyte injury, possibly by WTIP. Also, I hypothesize that the interaction between Nck and WTIP may link Nck and nephrin to regulation of the Hippo pathway in podocytes.

22 WW45

Proliferation/ Apoptosis Apoptosis Survival

Figure 8: Schematic of the Hippo kinase pathway in mammals (Adapted from Varelas, 2014).

23 11. Summary and Rationale

Nck is an adaptor protein that is critical for podocyte FP development and maintenance, and it acts downstream of phosphorylated nephrin at the SD to initiate actin remodeling to promote mature FP. In a research initiative to identify novel Nck interaction partners, our laboratory has identified the adaptor protein WTIP.

WTIP is reported to regulate the podocyte cytoskeleton downstream of adherens junctions, and in response to podocyte injury, WTIP accumulates in the nucleus where it is able to repress the transcriptional activator WT1. WT1 normally maintains the structural integrity of

FP. WTIP is a negative regulator of the Hippo kinase pathway, which is involved in the regulation of podocyte survival (Figure 9).

The overall objective of the research presented in this thesis is to characterize the interaction between Nck and WTIP and its signaling role in podocytes. To this end, we validated that Nck and WTIP are binding partners and mapped the interaction using mutated variants.

Based on our knowledge of the putative roles of WTIP, we first investigated nuclear functions for Nck with WTIP. Next, we explored whether Nck could recruit WTIP to nephrin. Lastly, we pursued the connection between the nephrin-Nck-WTIP signaling complex and the Hippo kinase pathway. Overall, we have identified a novel signaling interaction that may be important in podocyte function.

Figure 9: Summary of the molecular signals in healthy and injured podocytes relevant to the research in this thesis. A) In mature podocytes, nephrin is tyrosine phosphorylated which recruits Nck to facilitate actin remodeling. In parallel, WTIP is localized at cadherin-catenin to regulate actin. In the nucleus, Yap and WT1 are active to promote survival and expression of genes required for maintenance of foot processes, respectively. Based on Yap activity, it is hypothesized that Lats1/2 in the cytoplasm has low activity. B) In response to injury, nephrin phosphorylation is decreased, the actin cytoskeleton is disorganized, and WTIP translocates to the nucleus where it inhibits WT1. It is hypothesized that since WTIP (a negative regulator of Lats1/2) is now nuclear, Lats1/2 may be active to phosphorylate and downregulate Yap.

25 CHAPTER 2: METHODS

26 Tissue Culture

Human embryonic kidney (HEK)293T cells were obtained from the American Type

Culture Collection (Manassas, VA, USA). HEK293T cells were grown in Dulbecco’s Modified

Eagle Medium (DMEM) (Sigma-Aldrich, Oakville, ON, Canada) supplemented with 10% Fetal

Bovine Serum (FBS), 200 Units/mL penicillin and 200 µg/mL streptomycin (Invitrogen,

Burlington, ON, Canada) and maintained at 37°C and 5% CO2. Cells used for a particular experiment were always the same passage number and plated at equal density.

Mouse podocyte cells (MPC) were isolated from wildtype and Nck1 null Nck2 flox mice

(Jones et al., 2006) and exposed to Cre recombinase to generate Nck1 and Nck2 null cells

(Martin, Jones, unpublished). As a result there are three MPC lines, wildtype (55C3GFPC1; also referred to as WT), Nck1 null Nck2 flx/flx (42C4noAdeno; also referred to as Single Mutant

(SM)), Nck1 null and Nck2 null (42C4cre1C3; also referred to as double mutant (DM)). MPC were cultured on sterile collagen-coated dishes. To collagen coat the dishes, Rat collagen Type

IV (Cultrex®) was diluted 1/100 in 0.02 M acetic acid (final concentration 50 µg/mL) and incubated on the surface of the dish for one hour. The dishes were then washed with phosphate buffered saline solution (PBS) to remove excess acid and stored at 4°C for up to 3 weeks.

Undifferentiated MPC were grown in DMEM/F12 (Sigma-Aldrich, Oakville, ON, Canada) supplemented with 10% FBS, 200 Units/mL penicillin and 200 µg/mL streptomycin (Invitrogen,

Burlington, ON, Canada) and 0.1 µg/mL interferon γ and maintained at 33°C and 5% CO2. To differentiate MPC they were plated at a density of 30-40,000 cells/mL and maintained in

DMEM/F12 with 10% FBS, 200 Units/mL penicillin and 200 µg/mL streptomycin at 37°C and

5% CO2 for 8 or 9 days then re-plated into collagen-coated 10cm-dishes at a density of 70-80,000 cells/mL to ensure equal density of differentiated cells prior to experimental treatments.

27 Experiments were performed between day 12-14 of differentiation. MPC were not used for experiments beyond passage 16 and passage numbers were matched between cell lines for each experiment.

LPS Treatment

Cells treated with LPS were incubated for indicated hours in standard culture conditions with final concentration of 1 µg/mL LPS.

Transfections

All transfections were performed on HEK293T cells using polyethylenimine (PEI) (final concentration 4 µg/mL) when cells were 60-80% confluent. Transient expression was allowed for 24 to 42 hours at 37°C.

Vectors

Human Nck1 (BC006403) WT, Nck1SH3x3* (W39K, W143K, W229K), Nck1SH2* (R308M),

Nck1SH3(1)*, Nck1SH3(2)*, Nck1SH3(3)*, Nck1SH3(1,2)*, Nck1SH3(1,3)*, Nck1SH3(2,3)*,

Nck2 (BC000103) WT, Nck2 SH3x3*(W39K, W149K, W235K) and Nck2SH2* (R311M) were previously cloned into the Creator mammalian expression vector V180 (CMV) with an N- terminal triple-Flag tag and described in Blasutig et al. (2008). Human Nck1 SH3x3, SH3x3*,

SH3(1), SH3(2) and SH3(3) were previously cloned into pGEX-4T-1 as glutathione S- transferase (GST) fusion constructs. Mouse WTIP (BC054125) was previously cloned into pEGFP-C1 and pCDNA3.1-HA by Dr. Nina Jones. CD16-nephrin-GFP and CD16-nephrinY3F-

GFP was previously cloned into pEBB and described in Blasutig et al. (2008).

Cloning of GFP- and HA- tagged WTIPΔPRR

WTIPΔPRR-GFP was constructed by PCR-based cloning using pEGFP-C1-WTIP as template.

351bp (bases128-479, inclusive) were deleted from the WTIP cDNA; to effectively delete 117

28 amino acids of the proline rich linker region of WTIP without introducing any new amino acids into the sequence. The 5’ end of WTIP was cloned into pEGFP-C1 by destroying a BglII and

BamHI site, therefore forward (Fwd) and reverse (Rev) primers for the N-terminal of WTIP

(herein called P1) were designed to reconstruct a 5’ BglII restriction site and to introduce a 3’

HindIII site. The primers used to make P1 are called WTIP-P1-Fwd and WTIP-P1-Rev (Table

1). Forward and reverse primers for C-terminal of WTIP (herein called P2) were designed to include 5’ HindIII and 3’ EcoRI restriction sites and were called WTIP-P2-Fwd and WTIP-P2-

Rev (Table1). P1 and P2 were PCR amplified using Phusion HF Polymerase (New England

Biolabs) with the GC Phusion buffer and 5% DMSO for P1 and the HF Phusion buffer and 0%

DMSO for P2. The following thermocycling conditions were used: 95°C for 5 minutes; 30 rounds of 95°C for 30 seconds, 55°C for 30 seconds and 72°C for 100 seconds; followed by one round of 72°C for 20 minutes. PCR products from four reactions of P1 and two reactions of P2 were pooled and purified using PureLink™ PCR Purification kit (Invitrogen). P1 was digested using BglII (Invitrogen) and Fast Digest® HindIII (Fermentas), P2 was digested using Fast

Digest® HindIII and EcoRI (New England Biolabs) and pEGFP-C1 was digested using BglII and

EcoRI. Digestions were performed at 37°C for 1 hour in 1X Buffer H (Invitrogen). All three pieces were separated from restriction enzymes by gel extraction by running them through 2% agarose and using Pure Link™ Quick Gel extraction kit (Invitrogen). All three pieces were simultaneously ligated using T4 DNA ligase (Invitrogen) in the approximate ratio of 1:4:2

(vector:P1:P2) overnight at room temperature. The entire ligation reaction was transformed into competent Escherichia coli (E. coli) DH5α which grew on LB-Agar with 30 µg/mL kanamycin selection. To subclone WTIPΔPRR from pEGFP-C1 into pCDNA3.1-HA the sequence for

WTIPΔPRR was PCR amplified from the pEGFP-C1 vector using the WTIPΔPRR-Fwd primer

29 (Table 1), which mutates the BglII site into a BamH1 site and the WTIP-P2-Rev with the same

PCR conditions that produced P1 except with 120 second elongation time. BamHI (Invitrogen) and EcoRI (Invitrogen) were used in 1x REact Buffer 3 (Invitrogen) at 37°C for 1 hour. The transformed ligation product was grown on LB-Agar with 100 µg/mL ampicillin. The vectors were purified using Pure Link™ Quick Plasmid Miniprep Kit (Invitrogen) and sequenced by the

Advanced Analysis Centre (University of Guelph, Science Complex) using the EGFPseq and T7 primers (Table 1).

Antibodies

The following antibodies were obtained commercially: rabbit anti-GFP (ab290) (Abcam), mouse anti-CD16 (sc-19620) (Santa Cruz Biotechnology), mouse anti-β-actin (AC-15) (A5441) (Sigma-

Aldrich), mouse anti-FLAG clone M2 (F3165) (Sigma-Aldrich), mouse anti-GAPDH (clone

1D4) (G041) (Applied Biological Materials Inc., Richmond, BC, Canada), mouse anti-Lamin

A/C (4C11) (#4777)(Cell Signaling Technology), rabbit anti-pan-cadherin (ab16505) (Abcam), rabbit anti-Yap (D8H1XP(R)) (Cell Signaling) and rabbit anti-Phospho-Ser127 Yap (Yap- pS127) (#4911) (Cell Signaling). Mouse anti-HA (clone 12CA5) were provided by Sunnybrook

Hybridoma Bank, Toronto, Ontario, Canada. Rabbit anti-pan-NckLL was generously provided by Dr. Louise Larose of McGill University Department of Medicine and mouse anti-β-catenin from Dr. Terry Van Raay of the University of Guelph, Department of Molecular and Cellular

Biology. Generation and testing of the phospho-specific rabbit anti-nephrin pTyr1217 #2423-1

(EPITOMICS) was previously reported in Jones et al. (2009).

Secondary horseradish peroxidase-conjugated polyclonal goat anti-mouse #170-6516 or goat anti-rabbit #170-6515 antibody were purchased from BioRad (Mississauga, ON, Canada).

CD16 Stimulation and PP2 Treatment

30 This protocol was described previously (Blasutig et al., 2008; New et al., 2013). Twenty-four hours after transfection media was removed cells and replaced with serum-free DMEM for at least 18 hours. In some experiments, PP2 kinase inhibitor was added to serum-free media on cells to final concentration of 10 µM PP2 in DMSO (or DMSO alone) for 3 hours and CD16 stimulation occurred in the continued presence of PP2. Clustering was induced using 1 µg anti-

CD16 per mL serum-free media for 10 min at 37 °C.

Cell Lysis

Media was removed from dishes and cells were washed with chilled PBS solution. Cells were detached from 10-cm plates by scraping in 1 mL PBS supplemented with protease inhibitors 0.1 mg/mL aprotinin, 0.1 mg/mL leupeptin, 1 mM sodium orthovanadate and 1mM phenylmethanesulfonyl fluoride (PMSF) (PBS+). Cell pellets were lysed in 800-900 µL

Phospholipase C (PLC) Lysis buffer (10% Glycerol, 50 mM Hepes pH 7.5, 150 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA, 10 mM NaPPi, 100 mM NaF, 1%Triton X-100) supplemented with the same protease inhibitors as PBS+ (PLC+). Cells expressing CD16-nephirn (WT/Y3F)-GFP were lysed by vortexing for 30 seconds, sonicating using Vibra-Cell (Sonics & Materials Inc.) for 30 seconds and incubating on ice for 10 minutes. Experiments that did not include CD16- nephrin (WT/Y3F)-GFP were lysed by vortexing 30 seconds and incubating on ice for 10 minutes. Lysates were prepared in 5XSDS Loading Buffer (50% Glycerol, 0.3125 M Tris Base,

20% SDS, 25% β-mercaptoethanol) for analysis by western blot.

Cell Fractionation

This protocol was adapted from a protocol provided on the website of Dr. David Ron at the

University of Cambridge Metabolic Research Laboratories (URL inactivated). This protocol was used to segregate the cytoplasmic components from the nuclear components. All procedures

31 were done on ice or at 4°C and all solutions were chilled to 4°C. Cells were washed with PBS and harvested by scraping in 1 mL 1 mM EDTA in PBS+ and centrifuging at 3,000 rpm for 5 minutes. Cell pellets were resuspended by pipetting in at least 6 volumes of cytoplasmic extraction buffer (10 mM HEPES pH 7.9, 50 mM NaCl, 0.5 M Sucrose, 0.1 mM EDTA, 0.5%

Triton X-100, with freshly added 1 mM DTT, 100 mM NaF, and protease inhibitors as indicated in Cell Lysis) and incubated on ice for 5 minutes then centrifuged at 1000 rpm for 5 minutes.

The supernatant (cytoplasmic fraction) was separated and centrifuged for an additional 15 minutes at 14,000 rpm to clear away any impurities. The pellet (nuclear faction) was washed in

500 µL of wash buffer (10 mM HEPES pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA and freshly added 1 mM DTT and protease inhibitors as in Cell Lysis). The nuclear pellet was resuspended in nuclear extraction buffer (10 mM HEPES pH 7.9, 500 mM NaCl, 0.1 mM EDTA,

0.1 mM EGTA, 0.1% NP40 and freshly added 1 mM DTT and protease inhibitors as in Cell

Lysis) and vortexed for 15 minutes. Next the mixture was and centrifuging at 14,000 rpm for 10 minutes and the supernatant was retained as the nuclear fraction and the pellet was discarded.

Fractions were validated by western blot of cytoplasmic, nuclear and membrane control proteins.

Immunoblotting

Protein samples were separated by 12% SDS-PAGE and semi-dry transfer of separated proteins to Polyvinylidene fluoride (PVDF) membranes (Roche, Mississauga, ON, Canada) was conducted at 0.25 A (per gel) for 1.5 hours. Membranes were blocked for minimum of 30 minutes in 5% skim milk/Tris Base Salt Tween buffer (TBST) (0.02 M Tris Base, 0.15 M NaCl,

0.005% Tween20) or 5% bovine serum albumin fraction V (BSA) (Roche Applied Science,

Werk Penzberg 82372 Penzberg, Germany)/TBST for phospho-antibodies. Membranes were probed overnight at 4°C with primary antibodies diluted 1:1000 in TBST; with the exception of

32 rabbit anti-GFP ab290, which was diluted in 5% skim milk/TBST, rabbit anti-Yap and rabbit anti-YAPpS127, which were diluted in 1% BSA/TBST and mouse anti-βactin, which was diluted

1:5000. Membranes were then washed three times for 10 minutes with TBST and incubated with secondary antibody diluted 1:10,000 in blocking solution for 1-1.5 hours. Secondary antibody was then washed from the membranes for a minimum of three times for 10 minutes with TBST.

Chemiluminescent staining was done using 1:1 ECL Western Blotting Substrates (#32106)

(Pierce). Membranes were exposed on X-ray blue film (VWR) and developed automatically using Konica Minolta Automatic Film Developer SRX-101A.

Immunoprecipitation

Equal volumes (approximately 600 µL) of cell lyate for each sample were incubated with

0.5 µL of antibody per sample overnight at 4°C on a nutator. Simultaneously, Sepharose protein

A (CL4B) beads (Sigma #P3391) were blocked in 5% BSA/PLC+ overnight at 4°C on a nutator.

The following day the beads were washed three times with PLC+ and combined with the lysate- antibody mixure and incubated for 3 hours at 4°C. The beads were then washed three times with

PLC+ and the immunocomplex was eluted from the beads in 30 µL 2x SDS Loading Buffer

(20% Glycerol, 0.125 M Tris base, 4% SDS, 10% β-mercaptoethanol) by boiling at 100°C for 2 minutes.

GST Pulldown Assay

GST-fusion proteins had to first be purified. The vectors were transformed into competent E. coli DH5α and plated on LB-Agar dishes with 100 µg/mL ampicillin selection. A colony was selected and grown in a 50 mL LB-ampicillin culture until O.D.600 was between

0.5-1 then expression of the GST-fusion proteins was induced with 0.5 mM isopropyl β-D-1- thiogalactopyranoside (IPTG) for 3 hours. Cells were harvested by centrifuging 5000 g for 10

33 minutes at 4°C and lysed in PBS+ on ice by vortexing, sonicating and centrifuging at 12,000 g for 15 minutes at 4°C. Supernatant (i.e. lysate) was incubated with 100 µL glutathione sepharose™ 4B (GE Healthcare) bead bed for 30 minutes at 4°C. Beads were washed three times with PBS+ and stored in PBS+ and 1mM DTT as a total 250 µL 40% slurry. Purification of the

GST-fusion proteins was assessed by running 10 µL of the slurry eluted in 10 µL 2xSDS

Loading Buffer on an SDS-PAGE and staining with coomassie brilliant blue.

Beads were incubated for 3 hours at 4°C with lysates from HEK293T that expressed

WTIP-GFP or WTIPΔPRR-GFP, then washed with PLC+ three times. The pull-down complex was eluted in 30 µL 2xSDS loading buffer by boiling at 100°C for 2 minutes. Pull down was assessed by western blot for GFP.

RNA Extraction and cDNA Preparation

RNA was extracted from differentiated MPC. Cells were scraped from 10-cm dishes in

PBS+. Cells from 5 plates were pooled to collect a cell pellet of approximately 3.5-4 million cells. Pellets were homogenized in 1 mL of TRIzol® (Ambion Life Technologies) and RNA extraction was done following manufacturer’s protocol. RNA pellet was resuspened in 20 µL nuclease-free water. RNA samples were treated with DNase1 (ThermoScientific EN0521) for 30 minutes at 37°C to remove any contaminating genomic DNA and DNase1 activity was stopped by addition of 5 mM EDTA for 10 minutes at 65°C. RNA was reverse transcribed into cDNA using Quanta qScriptTM (Quanta Biosciences Inc P/N84034) following manufacturer’s protocol.

Semi-Quantitative Real Time PCR

Semi-QRTPCR was done for the following mouse genes: β-actin, Eukaryotic Translational

Elongation Factor II (ETEFII), Podocalyxin and Fyn using forward and reverse primers in Table

1. Cycling conditions were optimized for each primer. All reactions included one initial round of

34 95°C for 5 minutes and one final round of 72°C for 10 minutes. The chain reaction was done using DreamTaq DNA Polymerase (Thermo Scientific) for a specified number of rounds of 95°C for 30 seconds, 53°C for 40 seconds, and 72°C for 1 minute. For β-actin, ETEFII, and Fyn 2 µg of cDNA was used and samples were thermocycled for 30 rounds. For podocalyxin 4 µg of cDNA was used and samples were thermocycled for 50 rounds. PCR products were analyzed by running samples through a 3% agarose gel and visualized using ChemiDoc™ XRS (BioRad).

NephrinY3F Mice and NTS Injection

NephrinY3F knockin mice were generated by homologous recombination targeting mouse

Y 1191, 1208 and 1232 for mutation to F (New, Jones, unpublishe). NTS was provided by David

Salant (Boston University). NTS and control IgG injection was performed by the Central Animal

Facility at the University of (New, Jones, unpublished).

Glomeruli Enrichment

Two kidneys were dissected per animal and the kidney medulla was separated and discarded.

The kidney cortex was minced using a 10 gauge scalpel and incubated with 1 µg/mL collagenase in PBS at 37°C with 140 rpm shaking for 30 minutes. Cortex was passed through 100 µM sterile cell strainer with 45 mL chilled PBS and pelleted for 1 min at 1000 g. Red blood cells were lysed in sterile Ack lysis buffer (0.15 M NH4Cl, 7 mM KCO3, 0.1 mM EDTA) at room temperature by inverting and immediately pelleting for 1 min at 1000 g. Cortices from two kidneys were lysed in

2 mL PLC+ as described in “Cell Lysis” with sonicating.

35 Table 1: Primer sequences (from 5’-3’). Primer Name Sequence WTIP-P1-Fwd CGGACTCAGATCTACGGCCATG WTIP-P1-Rev GCCGCAGCCAAGCTTGCCTCG. WTIP-P2-Fwd GCCGAGCGGAAGCTTGAGGCG WTIP-P2-Rev GCAGAATTCTGCTTCTCAGAGCTC WTIPΔPRR-Fwd CGGACTCGGATCCACGGCCATG EGFPseq AACCACTACCTGAGCACCCAGTCC T7 TAATACGACTCACTATAGGG β-actin-Fwd CCTGAACCCTAAGGCCAACC β-actin-Rev TGCTGATCCACATCTGCTGG ETEFII-Fwd CAGGCCATGTGGACTTCTCT ETEFII-Rev ACACACGCCAGACACACAGT Podocalyxin-Fwd ACTGTTGCAGTGTGGAGACG Podocalyxin-Rev ACTGTTGCAGTGTGGAGAC Fyn-Fwd AACAGCTCGGAAGGAGACTG Fyn-Rev CGGCCAAGTTTTCCAAAGTA

36 CHAPTER 3: RESULTS

Nck and WTIP Interact.

Based on the presence of conserved P-x-x-P motifs in the primary sequence of WTIP, and its expression in podocytes, WTIP was hypothesized to be a candidate interaction partner for the SH3 adaptor protein Nck. To validate this hypothesis first HEK293T cells were transfected to coexpress GFP-tagged WTIP or GFP and Flag-tagged Nck1 or empty vector control (Figure

10A; INPUT). WTIP-GFP and GFP were immunoprecipitated from cell lysates using an antibody against GFP. Specific coimmunoprecipitation of Nck1-Flag upon Flag immunoblotting was detected with WTIP-GFP but not with GFP (Figure 10A; IP: GFP lanes 1 and 2). Therefore,

Nck1-Flag complexes with WTIP-GFP and this is not through non-specific association with

GFP, the beads or the GFP antibody. Lysates expressing WTIP-GFP or GFP alone with empty flag vector serve as negative controls and verify that the immunoprecipitation conditions do not result in non-specific detection of a 48 kDa protein by the Flag antibody in the immunoblot

(Figure 10A; lane 3 and 4). Likewise, Nck2-Flag specifically coimmunoprecipitates with WTIP-

GFP from HEK293T lysates (Figure 10B). Therefore, WTIP interacts with both of the Nck proteins and the coimmunoprecipiation is not artificially mediated by non-specific binding.

(These immnoprecipitation assays were performed 2 times for Nck1 and 4 times for Nck2.)

38 !" !!!"#$!%&#! !!!!!!!!"'#()! *)"#+%&#!!!!!!!,!!!!!!+!!!!!!,!!!!!!+!!!!!!!!!!!,!!!!!!!+!!!!!,!!!!!!+!!!!!! %&#!!!!!!!!!!!!!!!!!!+!!!!!!,!!!!!!+!!!!!!,!!!!!!!!!!!+!!!!!!!,!!!!!+!!!!!!,!! '-.<+&012!!!!!!!!,!!!!!!,!!!!!!+!!!!!!+!!!!!!!!!!!,!!!!!!,!!!!!!+!!!!!!+!!!!!

34!!!!! "9$!%&#!:*)"#;! 78! 56!

84!!!!!

/4! 78! 56! "9$!&012!:'-.<;!

#" ! !!"#$!%&#!! !!!!!!!!"'#()! *)"#+%&#!!!!!!!,!!!!!!+!!!!!!,!!!!!!+!!!!!!!!!!!,!!!!!!!+!!!!!,!!!!!!+!!!!!! %&#!!!!!!!!!!!!!!!!!!+!!!!!!,!!!!!!+!!!!!!,!!!!!!!!!!!+!!!!!!!,!!!!!+!!!!!!,!! '-./+&012!!!!!!!!,!!!!!!,!!!!!!+!!!!!!+!!!!!!!!!!!,!!!!!!,!!!!!!+!!!!!!+!!!!! 34!!!!! "9$!%&#!:*)"#;! 78! 56!

84!!!!!

/4!

78! "9$!&012!:'-./;! 56!

Figure 10: Nck and WTIP interaction. Transient co-expression of WTIP-GFP or GFP with A) Nck1-Flag or B) Nck2-Flag in HEK293T cells. Cell lysates (INPUT) and immunoprecipitation of GFP were assessed by immunoblotting for GFP or Flag. Both Nck1 and Nck2 coimmunoprecipitate with WTIP-GFP but not GFP alone.

Mapping of the WTIP-Nck interaction to the first and third Nck SH3 domains.

It was hypothesized the interaction between Nck and WTIP is mediated by the Nck SH3 domains binding proline rich motifs of WTIP. To test this, previously designed Nck1 and Nck2 mutated variants were used for coimmunoprecipitation with WTIP-GFP. Introduction of a point mutation from R to M in the SH2 domain of Nck1 or Nck2 obstructs phospho-tyrosine binding

39 (Figure 11A, C). Also, W to K point mutations in the Nck1 or Nck2 SH3 domains prevents binding to proline rich motifs (Figure 11A, C). HEK293T cells were transfected to coexpress

Nck1-Flag or Nck2-Flag WT, SH2* or SH3x3* (Figure 11A, C) and lysates were assessed by westernblot for GFP (WTIP) and Flag (Nck or variants) (Figure 11B, D; INPUT). Cell lysates were used for immunoprecipitation of GFP (WTIP), which was assessed by immunoblotting for

GFP (WTIP) and Flag (Nck and variants). Mutations in the SH3 domains, but not SH2 domain, prevented coimmunoprecipitation of Nck1-Flag or Nck2-Flag with WTIP (Figure 11B, D).

Therefore, as hypothesized, WTIP and Nck interact through the proline-binding function of

Nck’s SH3 domains. (These immnoprecipitation assays were performed 3 times for Nck1 and 2 times for Nck2.)

The presence of multiple SH3 domains within Nck prompted me to next examine which of these domains might mediate binding to the PRR of WTIP. To determine which of the three

Nck SH3 domains are required to bind to WTIP, both GST pull down assay using GST-fused

Nck1 SH3 domains and coimmunoprecipitation of full length Nck variants was performed.

40 .% 0% ()!":%&"$$$$$$$$$;$$$$$$$;$$$$$$$$$;$ ,12@:&./0$$$$$$$()$$<=5>5?$<=3?$

<=5$ <=5$ <=5$ <=3$ 67$$$$$ !"#$%&"$'()!"*$ '@*$ '3*$ '5*$ 45$ !+#$%&"$ !"#$%&'% 45$ !"#$%&"$'()!"*$ !"#$%()-,% 89$ !+#$&./0$',12@*$ D5E9F$ 67$$$$$ !,"-)$ !"#$%()*+*,% 45$ !+#$%&"$'()!"*$ (5ABC$(@85BC$(33AB$ 45$$$$$ !,"-)$ 89$ !+#$&./0$',12@*$

/% 1% ()!":%&"$$$$$$$$;$$$$$$$$;$$$$$$$;$ ,123:&./0$$$$$()$$$<=5>5?$$$<=3?$

<=5$ <=5$ <=5$ <=3$ 67$$$$$ !"#$%&"$'()!"*$ '@*$ '3*$ '5*$ !+#$%&"$ !"#-%&'% 45$ 45$ !"#$%&"$'()!"*$ !"#-%()-,% 89$ !+#$&./0$',123*$ D5@@F$ 67$$$$$ !,"-)$ !"#-%()*+*,% 45$ !+#$%&"$'()!"*$ (5ABC$(@8ABC$(357B$ 45$$$$$ !,"-)$ 89$ !+#$&./0$',123*$

Figure 11: Nck and WTIP interact via Nck SH3 domains. A) Schematic of Nck1 variants used for coimmunoprecipitation with WTIP-GFP. All the domains represented are present in the protein and X indicates that domain has an inactivating point mutation. Nck1 SH2* has the point mutation R308M, which prevents the SH2 domain from binding phospho-tyrosine. NckSH3x3* has three point mutations: W39K, W143K, W229K, one in each SH3 domain, which renders all of the domains incapable of binding proline rich motifs. B) Immunoblot for GFP and Flag of HEK293T cell lysates (INPUT) co-expressing WTIP-GFP with Nck1-Flag WT, SH3x3* or SH2*. Lysates were used for immunoprecipitation of GFP (WTIP) and coimmunoprecipitation was assessed by Immunoblot for GFP and Flag. C) Schematic of Nck1 variants used for coimmunoprecipitation with WTIP-GFP. Nck2 SH2* has the point mutation R311M and Nck2SH3x3* has three point mutations: W39K, W149K, W235K, one in each SH3 domain. Obstruction of the domain’s binding function is analogous to Nck1 variants. D) Immunoblot for GFP and Flag of HEK293T cell lysates (INPUT) co-expressing WTIP-GFP with Nck2-Flag WT, SH3x3* or SH2*. Lysates were used for immunoprecipitation of GFP (WTIP) and coimmunoprecipitation was assessed by Immunoblot for GFP and Flag. Mutation of the Nck1 or Nck2 SH3 domains prevents immunocomplex formation between Nck1 or Nck2 and WTIP.

For GST pulldown assay, first GST fusion proteins (Figure 12A) were purified from

E.Coli cultures and immobilized on glutathione-conjugated sepharose. Successful purification of

GST fused to all three Nck1 SH3 domains (SH3x3), all three domains with W to K mutations

41 (SH3x3*), the first SH3 domain (SH3(1)), the second SH3 domain (SH3(2)) or the third SH3 domain (SH3(3)) (Figure 12A) was assessed by SDSPAGE and Coomassie staining (Figure

12C). Next, lysate from HEK293T cells expressing WTIP-GFP was incubated with these immobilized GST fusion proteins and binding was assessed by immunoblotting for GFP. GST-

Nck1 SH3x3 and GST-Nck1 SH3(3) are the only fusion proteins capable of precipitating WTIP-

GFP from the lysate (Figure 12C). Next, we validated the mapping results from the GST pulldown assay in the context of the full length Nck1 protein. To do this, all the possible Flag- tagged Nck1 variants with one, two or three SH3 domains containing a W to K mutation were used for coimmunoprecipitation with WTIP-GFP (Figure 12B). HEK293T cells were transfected to coexpress WTIP-GFP and Nck1-Flag WT or variants and lysates were assessed by immunoblotting for GFP or Flag (Figure 12D; INPUT). Lysates were then used for immunoprecipitation of GFP (WTIP) and coimmunoprecipitation was assessed by immunoblotting for GFP and Flag (Figure 12D). As demonstrated previously (Figure 10), WTIP coimmunoprecipitates with Nck1-Flag WT but not SH3x3* (Figure 12D; lane 1 and lane 2).

Mutation of the second SH3 domain ((2)*) of Nck1 has less of an effect on the coimmunoprecipitation of Nck with WTIP than mutation of either the first ((1)*) or third ((3)*)

SH3 domains (Figure 12D; lane 3-5), indicating that the second SH3 domain is likely less involved in mediating the interaction between Nck1 and WTIP than the first or third SH3 domains of Nck1. Nck1 with mutations in the first and second ((1,2)*), first and third ((1,3)*), or second and third ((2,3)*) SH3 domains are all very ineffective at interacting with WTIP (Figure

12D; lane 6-8). (The GST pulldown assay was performed 2 times and the immunoprecipitation of WTIP-GFP with all the Nck1-Flag variants was performed 2 times.)

42 3% 6% LM:$ LM:$ LM:$ LM<$ LM:$ LM:$ LM:$ !"#$%&'% '2*$ '<*$ ':*$ 1('2!"#$()*+*% %L)$ '2*$ '<*$ ':*$ !"#$%()*+*,% (:PQ=$(2A:Q=$(<

()!"7%&"$$$$$$$$8$$$$$$$8$$$$$$$$8$$$$$$$8$$$$$$$8$$$$$$8$$$$$$$$8$$$$$$$8$ /0127&,-.$$$$$$()$$$$$9:;$$$'2*;$$'<*;$$':*;$'2=<*;$'2=:*;$'<=:*;$$ %L)7$ >?$$$$$ !"#$%&"$'()!"*$ !+#$%&"$ /012LM:#$$$$9:;$$9:$$$'2*$$$'<*$$':*$$$$$$$!/"C)$ @:$ >?$$$$$ !+#$%&"$ !"#$%&"$'()!"*$ @:$ AB$$$$$ !+#$&,-.$'/01*$$ :D$EF04GHE$ !"#$%&"$'()!"*$ >?$$$$$ !+#$&,-.$'/01*$$ @:$ )6JO,FE$ AB$$$$$ ?$IJG5KFE$ AB$$$$ !"#$%&"$'()!"*$ :?$$$$$ LJG.,FE$ AB$$$$$ !+#$&,-.$'/01*$$ ?$$$$$ !/"C)$ @:$ !+#$%&"$'()!"*$

!/"C)$ AB$$$$$ !+#$&,-.$'/01*$

Figure 12: Mapping of Nck and WTIP interaction to the first and third Nck SH3 domains. A) Schematic of GST fusion proteins used for GST pull down assay of WTIP-GFP. B) Schematic of Flag-tagged Nck1 full length variants used for coimmnoprecipitation with WTIP-GFP. C) Coomassie stained SDS PAGE with arrows indicating if is fused to three Nck domains (Triples) or one Nck domain (Singles). Immunoblot of GFP for HEK293T lysate expressing WTIP-GFP (INPUT) and the results of GST pull down assay. D) Cell lysates (INPUT) co-expressing WTIP- GFP and Flag-tagged Nck1 WT, SH3x3*, SH3(1)*, SH3(2)*, SH3(3)*, SH3(1,2)*, SH3(1,3)*, or SH3(2,3)* and immunoprecipitation of GFP (WTIP) were immunoblotted for GFP and Flag. Nck1 WT and Nck1 SH3(2)* coimmunoprecipitate with WTIP-GFP most effectively.

43 Mapping of the WTIP-Nck interaction to the PRR of WTIP.

We next wanted to demonstrate that the PRR of WTIP is mediating the interaction with the Nck SH3 domains. To this end, we created a WTIP variant lacking 117 amino acids starting from G128 until R479, in which there are 5 P-x-x-P motifs as well as other proline rich sequences, and validated the mutated cDNA by sequencing (Figure 13A). This variant, denoted

WTIPΔPRR, was successfully cloned into pEGFP-C1 and pCDNA3.1-HA (data not shown).

Lysates from HEK293T cells expressing WTIP-GFP or WTIPΔPRR-GFP were incubated with

GST fused to all three Nck1 SH3 domains each with a point mutation obstructing proline- binding function (GST-Nck1SH3x3*) or all three wildtype Nck1 SH3 domains (GST-

Nck1SH3x3) and pull down was assessed by western blot for GFP. As expected, neither WTIP nor WTIPΔPRR bind to the mutated Nck1 SH3 domains. The interaction between Nck SH3 domains and WTIP is reduced when the PRR of WTIP is deleted (Figure 13B). Furthermore, we wanted to assess whether WTIPΔPRR could still bind other WTIP interaction partners. A known interaction partner of WTIP is β-catenin. This interaction is likely mediated by the WTIP LIM domains (Marie et al., 2003), therefore is expected to be unaffected by deletion of the PRR. To test if WTIPΔPRR can bind β-catenin, vectors encoding WTIP-GFP or WTIPΔPRR-GFP were transected into HEK293T cells and cells were harvested for immunoprecipitation (Figure 13C;

INPUT). Immunoprecipitation of the GFP tag and coimmunoprecipitation of endogenous β- catenin was assessed by immunoblot (Figure 13C). As expected, although WTIPΔPRR is lacking a large portion of its primary sequence, it is still capable of binding β-catenin. (The GST pull down of WTIP-GFP and WTIPΔPRR-GFP was done once and the coimmunoprecipitation of β- catenin with WTIP-GFP and WTIPΔPRR-GFP was done once.)

44 $"

!" )>&($2?<>@;A;B! )>&($2?<>@;A;! "$#%&!

'&"#()*#!!!!!!!!!!!!!!!!!'&!!!!!+#,,!!!!!!!!!!!'&!!!!!!!!+#,,!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!'&!!!!!!!!!+#,,!

89!!!!!

:;!

#" !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"#!!!!!!!!!!!!!!!!!!!!!"$#%&!!!!!!!!!!!!!! '&"#()*#!!!!!!!!!!'&!!!!!+#,,!!!!!!!'&!!!!!!!+#,,! 89!!!!! "-.!)*#!/'&"#0! :;!

<==! "-.!1(2345676!

Figure 13: Mapping of Nck and WTIP interaction to WTIP proline rich region. A) Primary sequence of WTIP aligned with WTIPΔPRR showing the deleted P-x-x-P motifs in blue, NES in purple, LIM domains in orange and the PDZ-binding motif in red. B) GST pull down of WTIP- GFP and WTIPΔPRR-GFP from HEK293T lysates (INPUT) using GST-Nck1SH3x3* and GST- NckSH3x3. Neither WTIP or WTIPΔPRR is pulled down with Nck1SH3x3* and pull down with wildtype Nck1SH3x3 is reduced when the PRR is deleted from the WTIP sequence. C) Transient expression of WTIP-GFP or WTIPΔPRR-GFP in HEK293T cells and immnuprecipitation of GFP. Lysate (INPUT) and coimmunoprecipitation are assessed by immunoblotting for GFP and β-catenin and show that WTIPΔPRR retains its ability to bind β-catenin.

45 Nck subcellular localization does not exhibit similar trends as WTIP and Nck expression does not influence transcript levels of WT1 target genes.

While the role of WTIP within podocytes remains to be firmly established, it has been shown in cultured cells to translocate to the podocyte nucleus at 6 hours following LPS-mediated podocyte injury and relocalize back to the cytoplasm after 24 hours (Kim et al., 2010). We hypothesized that if Nck is involved in the injury response through its interaction with WTIP, that they would exhibit similar subcellular translocation patterns. Cultured differentiated podocytes were treated with 1 µg/mL LPS for 0, 1, 6 and 24 hours and subjected to cellular fractionation (Figure 14A). The nuclear and cytoplasmic controls (Lamin A/C and GAPDH, respectively) show the fractionation assay successfully segregated cytoplasmic and nuclear components. While Nck appears to be distributed both in the cytoplasm and the nucleus, as expected from previous reports in other cell types (Lawe et al., 1997), its localization does not appear to be dynamic in response to the LPS model of injury. Of note however, the fractionation assay does not effectively separate plasma membrane components from either the cytoplasmic or nuclear fractions, thus membrane-associated molecules, such as Nck, may be artificially appearing in the nuclear fraction. Due to antibody limitations, namely an effective antibody against WTIP does not exist (Kim et al., 2010; Thakur et al., 2010; Kim et al., 2012; Reddy and

Irvine, 2013; Sun and Irvine 2014), we were unable to confirm the relocalization of WTIP in this assay that was reported previously (Kim et al., 2010).

Additionally, we sought to determine whether Nck influenced gene expression in cultured differentiated podocytes. We began by testing expression of WT1 target genes in cultured differentiated podocytes that are Nck wildtype (WT), have stable Nck1 knockout (SM) or stable

Nck1 and Nck2 knockout (DM). Analysis of WT1-target genes, podocalyxin and Fyn, by semi-

46 QRTPCR using ETEFII as a housekeeping control showed no striking differences in transcript levels between WT, DM and SM MPC (Figure 14B). (The LPS and fractionation assay was done once and the semi-QRTPCR of all three podocyte cell lines was performed two times.

Additionally semi-QRTPCR for the same targets was done once more for WT and DM MPC

(data not shown).)

!" ''''''''''''''''''''''':01-23*45'''''''''''''''''''H$+3&$4'' @9A)'''''''''B'''''='''''D'''''E('''''''''''''B'''''='''''D'''''E(' YZ)'2*.[H+I' (X''''' WB'4&+-.64'

YZ)'2*.[H+I' (X''''' V'5".$1&4'

UV''''' YZ)'@*5".'J?:'' DW''''' H$+3&*%'+-.1%-3'

YZ)'MJ9NO' WV''''' :01-23*45"+'' +-.1%-3'

=WV'''' YZ)'2*.[+*6<&%".'' =BB''''' 8&5F%*.&'' UV''''' +-.1%-3'

#" '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''\Q[''''''''''''' ]Q''''''N8'''''A8'''''9:\['''']Q'''''N8''''A8'''''''' 9-6-+*30^".'

!0.'

_Q_!YY'

`[*+,.'

Figure 14: Nck nuclear localization in response to LPS and influence on gene expression in cultured podocytes. A) Differentiated WT MPC were treated with 1 µg/mL LPS for 0, 1, 6 or 24 hours and cells were fractionated and fractions were assessed by IB. GAPDH, Lamin A/C and pan-cadherin are used as cytoplasmic, nuclear and membrane controls, respectively. Nck is present in the cytoplasmic and nuclear fractions at all time points. B) Semi-QRTPCR analysis of WT1 target genes in standard cultured WT, DM and SM MPC. Relative to control ETEFII, podocalyxin and Fyn expression in the three cell lines are constant. β-Actin was used as a control to test RT-PCR negative samples, which are RNA samples that were not reverse transcribed, and showed no nuclear DNA contamination. PCR- control included all components of PCR except cDNA.

!"#$%&'()'!%*+,-.*,-.'-/'+01-23*45'*.6'.$+3&$4'/%-5'6"7&%&.,*1&6'89:'1%&*1&6';"1<'=>#?5@'@9A' 48 /-%'BC'=C'D'*.6'E('<-$%4'*.6'"55$.-F3-G&6'/-%'H+IC'3*5".'J?:'K.$+3&*%'+-.1%-3LC'MJ9NO' K+01-23*45"+'+-.1%-3LC'2*.'+*6<&%".'K5&5F%*.&'+-.1%-3LP'Q<&%&'*22&*%'1-'F&'.-'4"#."R+*.1'+<*.#&4' ".'1<&'4$F+&33$3*%'3-+*3"S*,-.'-/'H+I'".'%&42-.4&'1-'1<&'@9A'".T$%0'5-6&3P'' WTIP is recruited to CD16-nephrin in an Nck-dependent manner.

The prominent role of Nck in nephrin signaling prompted us to next explore whether

WTIP might be recruited into this complex. In order to manipulate nephrin phosphorylation and complex formation a chimeric protein was used containing the intracellular domain of nephrin fused to an extracellular CD16 domain and intracellular CD7 domain, denoted as CD16-nephrin.

Stimulating the cells with an antibody against CD16 results in clustering of the chimera leading to phosphorylation of the intracellular nephrin domain by SFK and recruitment of molecules such as Nck (Jones et al., 2006; Blasutig et al., 2008). HEK293T cells were transfected to coexpress CD16-nephrin-GFP, Nck2-Flag and WTIP-HA. CD16-nephrin was clustered using anti-

CD16 antibody and cells were harvested for co-immunoprecipitations (Figure 15, INPUT).

Immunoprecipitation of nephrin using anti-GFP antibody and coimmunoprecipitation was assessed by western blot of GFP (CD16-nephrin), Flag (Nck2) and HA (WTIP). As shown previously (Jones et al., 2006; Blasutig et al., 2008), Nck2-flag coimmunoprecipitates with

CD16-nephrin-GFP independent of WTIP expression (Figure 15, lane 1 and 2). However, WTIP coimmunoprecipitation with nephrin is enhanced when Nck2 is overexpressed (Figure 15, lane

2) compared to in the absence of exogenous Nck2, WTIP still shows some interaction with nephrin, which may be due to endogenous Nck (Figure 15, lane 3). This suggests that Nck mediates recruitment of WTIP to nephrin. (The panel in Figure 15 was reproduced 3 times.)

CD>E$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$F$$$$$$$$F$$$$$$$F$ CD>EG()*+,-.G%&"$$$$$$$$$$$F$$$$$$$$F$$$$$$$F$ (56HGI34$$$$$$$$$$$$$$$$$$$$$$$$$$$$F$$$$$$$$F$$$$$$$$G$ 9:!"G78$$$$$$$$$$$$$$$$$$$$$$$$$$$$$G$$$$$$$$F$$$$$$$$F$ >??$$$$$$ !"#$%&"$'()*+,-./0$!1#$%&"$ <=$$$$$$ @A$$$$$$ !"#$%&"$'()*+,-./0$!1#$&234$'(56/$ @A$$$$$$ !"#$%&"$'()*+,-./0$!1#$78$'9:!"/$ B=$$$$$$ >??$$$$$$ !(";:0$!1#$%&"$'()*+,-./$ <=$$$$$$ @A$$$$$$ !(";:0$!1#$&234$'(56/$ @A$$$$$$ !(";:0$!1#$78$'9:!"/$ B=$$$$$$

Figure 15: Nck dependent recruitment of WTIP to nephrin. HEK293T cells were stimulated with &-4J,)$=#$9:!"$-K$,)5,J-L)M$LN$.)*+,-.$-.$3.$(56GM)*).M).L$O3..),P$'anti-CD16 antibody and co-expressed CD16-nephrin-GFP, Nck2-flag!/$:,3.K-).L$ and WTIPL,3.KQ)5RN.-HA. Lysates$NQ$ 7STHUB:$5)22K$V-L+$CD>EG.)*+,-.G%&"0$(56HGI34$3.M$9:!"G78$3.M$3,RW5-32$52JKL),-.4$NQ$.)*+,-.$(INPUT) and Immunoprecipitation of GFP (nephrin) were assessed by IB for GFP, Flag and HA. JK-.4$3.RGCD>E$3.RXNMY$,)KJ2LK$-.$Nck2-flag coimmunoprecipitated with5N!" CD16$NQ$.)*+,-.$V-L+$(56H$,)43,M2)KK$NQ$9:!"$)Z*,)KK-N.$3.M$-nephrin-GFP independent of WTIP-HA expression 5N!"but coimmunoprecipitation$NQ$9:!"$V-L+$.)*+,-.$-K$-.5,)3K)M$V+).$(56H$-K$ of WTIP-HA with CD16N[),)Z*,)KK)M-nephrin-GFP P$$is enhanced in the presence of overexpressed Nck2-flag compared to in the presence of only endogenous Nck.

WTIP recruitment to CD16-nephrin is dependent on nephrin phosphorylation by SFK and on

Nck’s phospho-tyrosine- and proline- binding functions.

In order to validate that WTIP recruitment to nephrin is dependent on nephrin

phosphorylation, as is Nck’s recruitment, we compared WTIP recruitment to CD16-nephrin with

or without CD16 clustering and in the presence or absence of a SFK inhibitor, PP2. Cells

coexpressing CD16-nephrin-GFP, Nck2-Flag and WTIP-HA, then treated with PP2 or vehicle

for 3 hours and CD16 stimulation in the continued presence of PP2 were harvested or

coimmunoprecipitation (Figure 16A; INPUT). Immunoprecipitation of nephrin using anti-GFP

antibody demonstrates, as previously, that phosphorylation of nephrin (on Y1217) is enhanced in

the presence of Nck overexpression (New et al., 2013); highest after CD16 clustering; and lowest

in the presence of PP2 (Lahdenpera et al., 2003; Jones et al., 2006; Verma et al., 2006; Blasutig

50 et al., 2008) (Figure 16A). Likewise, coimmunoprecipitation of Nck with CD16-nephrin is greatest upon CD16 clustering and lowest in the presence of PP2 (Figure 16A).

Coimmunoprecipitation of WTIP-HA with CD16-nephrin is greatest upon CD16 stimulation and in the presence of Nck2-Flag overexpression (Figure 16A). Thus, WTIP is recruited to nephrin upon phosphorylation by SFK in an Nck-dependent manner. In order to validate that the Nck- dependent mechanism of WTIP recruitment to nephrin is through Nck’s phospho-tyrosine- and proline- binding functions the assay was repeated using Nck and nephrin mutated variants.

WTIP-HA was co-expressed with CD16-nephrin-GFP and Nck2-flag WT; Nck2-flag SH3x3* that can bind nephrin but not WTIP; Nck2-flag SH2* that can bind WTIP but not nephrin; or with CD16-nephrinY3F-GFP that cannot recruit Nck. Cells were CD16-stimulated and harvested for coimmunoprecipitation (Figure 16B; INPUT). Nephrin phosphorylation on Y1217 in these lysates exhibit the same patterns as reported in New et al. (2013); that mutations in the SH3 domains of Nck reduces Nck’s ability to enhance nephrin phosphorylation and mutation in the

SH2 domain of Nck nearly abolishes Nck’s ability to enhance nephrin phosphorylation.

Immunoprecipitation of these lysates for GFP (nephrin) results in coimmunoprecipitation of

WTIP-HA with CD16-nephrin-GFP only when both the intracellular domain of nephrin and

Nck2-flag are wildtype (Figure 16B). Thus, WTIP recruitment to nephrin is mediated through binding of Nck SH2 domain to nephrin phospho-tyrosine residues after phosphorylation by SFK and binding of Nck SH3 domains to the WTIP PRR. (The panel in Figure 16A was reproduced 2 times, as well an additional time without the PP2 treatment; and the panel in Figure 16B was reproduced 3 times.)

51 #" ##6''''''''''''''''''''''''''''''''''''''C''''''''''C''''''''C''''''''C''''''''C''''''''C'''''''''D''''''''D''''''''D'' @A5B'''''''''''''''''''''''''''''''''''C''''''''''C''''''''C''''''''D'''''''D''''''''D''''''''D''''''''D''''''''D''' @A5BC"-./012C*+#''''''''''D'''''''''D'''''''D''''''''D'''''''D''''''''D''''''''D''''''''D''''''''D'''''''' ";<6CF9:'''''''''''''''''''''''''''D'''''''''D''''''''C''''''''D'''''''D''''''''C'''''''''D''''''''D''''''''C' ?%!#C=>''''''''''''''''''''''''''''C'''''''''D''''''''D''''''''C''''''''D'''''''D'''''''''C'''''''''D''''''''D' 5KK'''''' !#)'*+#',"-./0123&'!()'*+#' 7J''''''

5KK'''''' !#)'*+#',"-./0123&'!()'.45657' 7J'''''' LM'''''' !#)'*+#',"-./0123&'!()'+89:',";<3'

!#)'*+#',"-./0123&'!()'=>',?%!#3' LM''''''

5KK'''''' !"#$%&'!()'*+#',"-./0123' 7J'''''' 5KK'''''' !"#$%&'!()'.45657' 7J''''''

LM'''''' !"#$%&'!()'+89:',";<3'

!"#$%&'!()'=>',?%!#3' LM'''''' !" @A5BC"-./012C*+#'''''''''''''''''D''''''''D'''''''D'''''''4E+' ";<6CF9:'''''''''''''''''''''''''''''''''?%''G=EHEI'G=6I'?%' ?%!#C=>'''''''''''''''''''''''''''''''''''''D''''''''D'''''''D''''''D'

5KK'''''' !#)'*+#',"-./0123&'!()'*+#' 7J'''''' 5KK'''''' !#)'*+#',"-./0123&'!()'.45657' 7J'''''' !#)'*+#',"-./0123&'!()'+89:',";<3' LM''''''

LM'''''' !#)'*+#',"-./0123&'!()'=>',?%!#3'

5KK'''''' !"#$%&'!()'*+#',"-./0123' 7J'''''' 5KK'''''' !"#$%&'!()'.45657' 7J'''''' !"#$%&'!()'+89:',";<3' LM'''''' !"#$%&'!()'=>',?%!#3' LM''''''

Figure 16: WTIP recruitment to CD16-nephrin is dependent on nephrin phosphorylation by SFK and on Nck’s phospho-tyrosine- and proline- binding functions. A) Immunoblot of HEK293T lysates that are coexpressing CD16-nephrin-GFP, Nck2-Flag and WTIP-HA and were either treated with PP2 and/or CD16 stimulated and used for immunoprecipitation of GFP (INPUT). Coimmunoprecipitation was assessed by immunoblot for GFP, Flag and HA and shows Nck binding nephrin most after CD16 stimulation and least in the presence of PP2 and WTIP binding nephrin most in the presence of exogenous Nck and after CD16 stimulation. B) Immunoblot of HEK293T lysates that are coexpressing CD16-nephrin-GFP WT or Y3F, Nck2-Flag WT, SH3x3* or SH2* and WTIP-HA and were CD16 stimulated and used for immunoprecipitation of GFP (INPUT). Coimmunoprecipitation of Nck2-Flag WT and SH3x3* with CD16-nephrin-GFP WT but not Y3F and WTIP-HA coimmunoprecipitation with CD16-nephrin-GFP WT only in the presence of Nck2-Flag WT.

Total Yap levels in mouse glomeruli are influenced by nephrin phosphorylation.

We hypothesized that complex formation of nephrin-Nck-WTIP may be involved in regulation of the Hippo pathway, as WTIP is a known negative regulator of the Hippo pathway kinases Lats1/2. To test this, glomeruli were isolated from control wildtype mice and mice expressing nephrinY3F that cannot recruit Nck, and by extension does not bind WTIP, and immunoblotted for Yap and phospho-Yap (pS127; inactive) (Figure 17A). NephrinY3F kidney cortices were previously validated to have equal total nephrin transcript and protein expression as wildtype animals (New, Jones, unpublished). Immunoblot for Yap and Yap phospho-S127 shows that nephrinY3F glomeruli have strikingly lower total Yap levels, but similar phospho-Yap levels, compared to WT control littermates (Figure 17A). Therefore, in glomerular enriched fractions of nephrinY3F mice the ratio of phosphorylated Yap to total Yap is higher than WT control littermates (not quantified). To further assess the role of nephrin phosphorylation in regulation of Yap phosphorylation and total levels wildtype animals were challenged with the nephrotoxic serum (NTS) model of podocyte injury because it is known to robustly reduce nephrin phosphorylation (New, Jones, unpublished). Mice were injected with NTS or IgG control serum and immunoblot of glomerular enriched fractions for nephrin phospho-Y1208

(detected by anti-pY1217) revealed, as expected, reduced phosphorylation (Figure 17B).

Immunoblot of Yap and phospho-Yap (pS127; inactive) showed that total Yap levels are decreased in glomeruli of challenged animals, similar to the unchallenged nephrinY3F animals

(Figure 17B). However, unlike the glomerular enriched fractions from untreated neprhinY3F mice, phospho-Yap is also decreased. Therefore, decreased nephrin phosphorylation is associated with decreased levels of total Yap protein. (Both of these experiments were done in duplicate and shown in Figure 17.)

53 !" ##########&'###########################(")#

!"# ./0#(12324,5$#

$%###### ./0#(12## !"# 5-#6789:;6# $%###### ./0#(12## !"# 5#<=:>?76#

"%# ./0#HIJKL#

#" ####.GH######B'4#########.GH#######B'4##

,+-# ./0#B72CD=:#

./0#2(,5,$# EB72CD=:F# ,+-#

$%# ./0#(12324,5$# !"# ./0#(12## "-#6789:;6# !"# ./0#(12## !"# 5#<=:>?76# *+# ./0#@318A:#

Figure 17: Nephrin tyrosine phosphorylation is a regulator of total Yap protein levels. A) Immunoblot of glomerular enriched fractions from WT and nephrinY3F kidneys showing decreased total Yap in nephrinY3F glomeruli compared to WT littermate controls (n=2). B) Immunoblot of glomerular enriched fractions from WT mice injected with NTS or IgG after 24 hours (n=2). Glomerlui subjected to NTS challenge have reduced nephrin phosphorylation on Y1208 (IB: pY1217) and reduced levels of total Yap, similar to unchallenged nephrinY3F mice.

54 CHAPTER 4: DISCUSSION

The results presented in this thesis describe the novel interaction between Nck and WTIP.

The interaction is mediated by Nck’s SH3 domains and WTIP’s PRR and may be involved in nephrin signaling to the transcription regulator Yap in podocytes.

Nck and WTIP interact and the interaction is mediated by Nck’s first and third SH3 domains and

WTIP’s PRR.

Using recombinant proteins we demonstrate that WTIP interacts with both Nck1 and

Nck2 (Figure 10A,B). It is commonly reported that Nck interaction partners can bind both Nck1 and Nck2 (Buday et al., 2002). However, binding assays are usually performed with recombinant or purified proteins (Buday et al., 2002) and are not indicative of binding preference for a particular Nck paralogue in vivo. A major limitation to validating the interaction between Nck and WTIP is that there are no effective antibodies against WTIP, therefore recombinant, exogenous species were employed and the interaction between endogenous Nck and WTIP is yet to be shown. To somewhat overcome this limitation we sought to map the interaction between

Nck and WTIP, to show despite the use of recombinant proteins that interaction is specific and can be ablated under certain conditions. First, using Nck1 and Nck2 variants with mutations that disturb the phospho-tyrosine binding function of the SH2 domain or the proline-binding function of all three SH3 domains we demonstrate that the interaction requires functional SH3 domains but not the SH2 domains (Figure 11B,D). To further refine this mapping, single Nck1 SH3 domains were tested as GST-fusions for their ability to bind WTIP-GFP and demonstrated that only the third SH3 domain of Nck1 binds WTIP (Figure 12C). However, using full-length Nck1 variants with mutated SH3 domains we show that the interaction between Nck and WTIP is mediated through both the first and third SH3 domains of Nck1 (Figure 12D). The favoured

56 model of binding is that Nck and WTIP interact via both the first and third SH3 domains as protein functions are known to be regulated differentially in the context of an isolated domain (as used in GST pull down) compared to the full length sequence (as used in coimmunoprecipitation) (Pawson and Nash 2003; Pawson, 2007).

It is not uncommon or surprising that multiple Nck SH3 domains can mediate the interaction with a single binding partner (Wunderlich et al., 1999b; Buday et al., 2002). Only a few Nck binding partners are shown to associate more to a single SH3 domain than to a construct containing all three SH3 domains (Kitamura et al., 1996; Matuoka et al., 1997) (Table 2). The interactions between Nck and Cbl (Wunderlich et al., 1999b), Abl (Wunderlich et al., 1999b), N-

WASp (Rohatgi et al., 2001), IRS-1 (Tu, Liang et al., 2001) and Dock 180 (Tu, Kucik et al.,

2001) have been mapped to multiple Nck SH3 domains (Table 2). Additionally, upon reinterpretation of the results presented for some Nck interaction partners that are claimed to bind a single Nck SH3 domain (Rivero-Lezcano et al., 1995; Wunderlich et al., 1999a;

Wunderlich et al., 1999b) it becomes clear that tandem SH3 domains can increase binding (Table

2). Furthermore, although the third Nck SH3 domain was the only isolated domain capable of precipitating WTIP in the GST pull down assay it was not as effective as the GST-Nck1SH3x3.

It is hypothesized therefore that a GST-Nck1SH3(1,3) construct would bind to WTIP more than

SH3(3) and more like the GST-Nck1SH3x3.

57 Table 2: Summary of Nck interaction partners and reinterpretation of mapping data. Where data could not be reinterpreted N/A is noted and where reassessment of data yielded same conclusions there is a check (✓). Interaction Partner Nck SH3 Reinterpretation of data Reference domain NAP1 1 Single domain binds more Kitamura et al., 1996 than all three SH3 domains NAP4 2 Single domain binds more Wunderlich et al., 1999b than all three SH3 domains NAP5 2 Single domain binds more Wunderlich et al., 1999b than all three SH3 domains SAM68 1 N/A Lawe et al., 1997 RRas 2 N/A Wang et al., 2000 Pak1 3 SH3-2,3 binds more than Wunderlich et al., 1999b SH3-3 alone Sos 3 SH3-1,2,3 binds more than Wunderlich et al., 1999a SH3-3 WASp 3 SH3-1,2,3 binds more than Rivero-Lezcano et al., SH3-3 1995 WIP 3 Single domain binds similar Anton et al., 1998 to all three SH3 domains Cbl 1,2 ✓ Wunderlich et al., 1999b Abl 1,2,3 ✓ Wunderlich et al., 1999b N-WASp 1,2,3 ✓ Rohatgi et al., 2001 IRS-1 2,3 ✓ Tu, Liang et al., 2001 Dock180 2,3 ✓ Tu, Kucik et al., 2001

Wunderlich et al. (1999b) propose two reasons why Nck would have multiple SH3 domains, first that it can bind three effectors at the same time, the second and their favoured explanation, to allow for higher avidity for interaction partners and tighter complex formation.

This coincides with the theory of cooperative binding reviewed in Pawson and Nash (2003) and

Pawson (2007), which simply stated is the theory that cooperation between domains of an adaptor protein can alter binding functions of the whole protein both for and after the initiation of binding. A benefit of cooperative binding is that multiple domains with low affinity for their ligand can have increased affinity in the context of all the protein domains, to both increase affinity and specificity during recruitment of signaling effectors. Conversely, this can also result

58 in significant reduction of affinity for one interaction partner compared to another that would not have been apparent if analyzing the interaction in the context of single domains (Pawson and

Nash, 2003; Pawson, 2007). Evolutionarily this could explain the reoccurrence of similar domains in multiple proteins, yet distinct functions of these proteins, contributing to complexity of cell biology in higher order species.

GST-pull down of WTIP and WTIPΔPRR shows reduced binding of the Nck SH3 domains with WTIP when the PRR is deleted compared to wildtype WTIP (Figure 13B).

Inclusion of GST-Nck1SH3x3* as a negative control provides an effective reference for background binding that can occur during the pull down assay or western blotting protocols. This assay allows for the conclusion that the interaction between Nck and WTIP is primarily mediated via the PRR region of WTIP. However, the GST pull down assay, which employs GST-

Nck1SH3x3* as a reference for negative control, is ineffective for detection of low levels of residual binding between Nck and WTIPΔPRR that can be mediated via the WTIP LIM domains or a 6th P-x-x-P motif present in the third WTIP LIM domain. Further experiments to validate presence or absence of residual binding of Nck1 or Nck2 to WTIPΔPRR would be very interesting as there is one documented case in which Nck2 but not Nck1, binds to PINCH1 via its third SH3 domain to the fourth LIM domain of PINCH1 in a proline-independent manner (Tu et al., 1998). The interaction between Nck2 and PINCH1 is specific (Tu et al., 1998) but considered transient because of low binding affinity (Vaynberg et al., 2005). In the case of the

Nck2-PINCH1 interaction the residues primarily involved at the interaction interface are D257 of

Nck and R-R-P (197-199) of PINCH1 (Vaynberg et al., 2005). Primary sequence alignment of the forth LIM domain of human PINCH1 (188-251) with the primary sequence of human WTIP using NCBI BLASTp resulted in 32% identities, 50% positives and conservation of 198/199 R-P

59 at positions 355/356 in the third LIM domain of WTIP (similar results are obtained for human

PINCH1 LIM4 and mouse WTIP, including RP conservation). Although primary sequence alignment is not as strong a prediction technique as structural alignment, this is suggestive of the potential for transient LIM-SH3 interaction between Nck and WTIP.

Nck does not accumulate in the nucleus of cultured differentiated podocytes after LPS treatment and does not appear to influence transcript levels WT1-target genes.

Nck exists in the nucleus of many cell types (Lawe et al., 1997; Kremer et al., 2007), though its role in the podocyte nucleus is unknown. Since WTIP accumulates in podocyte nuclei in response to multiple models of injury, including LPS, ROS, UV, and PAN (Kim et al., 2010), it was hypothesized that Nck subcellular localization during injury might parallel that of WTIP in these cells. To test this, cultured differentiated podocytes were treated with LPS and Nck localization was assessed by a fractionation assay that separates cytoplasmic from nuclear components. WTIP is reported to accumulate in the nucleus after 6 hours of LPS treatment and return to the cytoplasm after 24 hours; however, Nck subcellular localization does not appear to be affected by LPS treatment at these time points (Figure 14A). These experiments were confounded by contamination of the nuclear fraction (as seen by pan-cadherin/membrane marker), inability to detect endogenous WTIP (owing to lack of suitable reagents) and differences in timepoints for validation of injury response (phospho-p38 and phospho-JNK, as reported by Kim et al. (2010)). An alternate approach to assess shuttling of proteins between the cytoplasm and nucleus would be via immunofluorescence microscopy; however, cross-reactivity of Nck antibodies for endogenous Nck using this technique, coupled with the difficulty in transfection of tagged proteins into podocytes, produce a distinct series of challenges.

60 We also investigated whether Nck influences gene expression in podocytes, starting with genes regulated by WT1 and by extension WTIP. Differentiated podocytes with stable Nck1 knockout (SM), stable Nck1 and Nck2 knockout (DM) or wildtype controls were cultured under standard conditions and transcript levels of WT1 target genes, podocalyxin and Fyn, were compared by semi-QRTPCR using ETEFII as the housekeeping control. Preliminary results showed no differences (Figure 14B), suggesting that Nck is not involved in WT1’s transcriptional regulation role under standard culture conditions.

Overall, we were unable to present any evidence suggesting that Nck functions in the podocyte nucleus with WTIP and further experiments are required to explore the significance of

Nck in the podocyte nucleus.

WTIP is recruited to nephrin upon nephrin phosphorylation by SFK in an Nck-dependent manner and may be involved in connecting nephrin signaling to Yap.

Next we investigated whether the interaction between Nck and WTIP can be connected to nephrin at the SD. To investigate this we used the recombinant chimera protein with a CD16 extracellular domain and CD7 intracellular domain fused to the cytoplasmic tail of nephrin expressed in HEK293T cells as this system can be easily manipulated to induce nephrin tyrosine phosphorylation (Blasutig et al., 2008). First the coimmunoprecipitation between CD16-nephrin-

GFP and WTIP-HA is demonstrated to be enhanced in the presence of overexpressed Nck2-Flag

(Figure 15), which suggests that nephrin and WTIP can complex in an Nck-dependent manner.

To follow up, we validated that complex formation between CD16-nephrin-GFP and WTIP-HA, like recruitment of Nck (Jones et al., 2006), is dependent on nephrin phosphorylation by comparing coimmunoprecipitation of CD16-nephrin-GFP and WTIP-HA with and without CD16

61 stimulation as well as in the presence or absence of the SFK inhibitor, PP2 (Figure 16A). Next, we demonstrated that the nephrin-Nck-WTIP complex formation is blocked when Nck’s SH3 domains are mutated or binding between Nck and nephrin is perturbed by mutation of the Nck

SH2 domain or the nephrin tyrosine residues (Figure 16B). Therefore, we propose that the nephrin-Nck-WTIP complex is mediated through binding of the Nck SH2 domain to phospho- tyrosine residues on nephrin and the Nck SH3 domains to WTIP. This suggests that WTIP is recruited to nephrin at the plasma membrane under conditions known to result in actin reorganization (Jones et al., 2006).

It is possible that WTIP recruitment to nephrin is involved in actin remodeling at the SD, similar to its function downstream of cadherin in cultured podocytes (Srichai et al., 2004; Kim et al., 2012). Although WTIP is necessary for normal stress fiber organization and is recruited to cadherin in cultured podocytes, its function as an actin regulator downstream of cadherin has not been confirmed in vivo. It would be interesting, therefore, to investigate whether WTIP- dependent actin remodeling at the SD occurs downstream of nephrin, cadherin or both. This is a topic of interest since the tools used in this thesis (transfected HEK293T cells) and by Dr. John

Sedor’s research group (cultured differentiated podocytes) (Srichai et al., 2004; Kim et al., 2012) are not entirely accurate representations of mature podocytes in vivo. Specifically, cultured differentiated podocytes are known to lose expression of nephrin (Takano et al., 2007), which is the primary adhesion molecule of the SD in vivo (Kestila et al., 1998; Khoshnoodi et al., 2003;

Juhila et al., 2010), and they have high cadherin expression, which normally declines in mature podocytes (Yaoita et al., 2002). Furthermore, there exists conflicting data regarding expression of P-cadherin in mature SD (Yaoita et al., 2002; Usui et al., 2003; Aaltonen and Holthofer, 2007;

Patrakka and Tryggvason, 2010; Norris and Remuzzi, 2012). The results of this thesis therefore

62 offer a novel, possibly more physiologically relevant, receptor upstream of WTIP’s actin- regulating functions.

It was previously proposed by Srichai et al. (2004) that WTIP could facilitate cross talk of signaling events from adherens junctions to nephrin. This hypothesis was made because of extensive characterization of WTIP functions at cadherin structures (Srichai et al., 2004; Kim et al., 2010; Kim et al., 2012) and colocalization of WTIP with CD2AP, an adaptor known to associate with nephrin, in cultured podocytes (Srichai et al., 2004). However, colocalization of

WTIP and CD2AP is not direct evidence of WTIP localization with nephrin. Also, Srichai et al.

(2004) propose that WTIP can facilitate cross talk by converging on F-actin, which is entirely valid but can also be stated for most of the signaling and adaptor proteins at the SD, as F-actin dynamics are widely known to initiate inside-out signals and stabilize adhesion structures

(Pavenstadt et al., 2003; Faul et al., 2007; Weber et al., 2011; Hong et al., 2013; Engl et al.,

2014). The results presented in this thesis provide an alternate and more intimate mechanism for cross talk between nephrin and cadherin, which is through WTIP’s interaction with Nck.

However, expanding our understanding of the mechanism and physiological relevance of bridging nephrin and cadherin signals at the SD through WTIP is beyond the scope of the research presented herein.

Alternatively, we sought to investigate if WTIP could link nephrin signaling in podocytes to the Hippo pathway, as WTIP is a known negative regulator of the Hippo pathway kinases

Lats1/2. We compared the Hippo pathway target Yap and Yap phospho-S127 (inactive) in glomerular enriched factions from wildtype and nephrinY3F mice, because nephrinY3F animals are unable to recruit Nck (New, Jones, unpublished), and by extension WTIP. Total Yap levels in glomeruli of nephrinY3F animals are significantly lower and the ratio of phospho-Yap (pS127) to

63 total Yap is higher compared to glomeruli of wildtype control littermates (Figure 17A). We hypothesize that the decreased Yap expression observed in the nephrinY3F glomeruli is a result of higher basal activation of Hippo signaling leading to phosphorylation of S297 of Yap and subsequent ubiquitin-mediated degradation (Varelas, 2014). However, transcript analysis is required to validate that the gene expression has remained constant in nephrinY3F glomeruli compared to wildtype.

Next we investigated Yap regulation in the NTS model of nephropathy, as it is a system known to induce nephrin dephosphorylation (New, Jones, unpublished). As shown previously

(New, Jones, unpublished), nephrin phosphorylation (on Y1208) is reduced 24 hours after injection of NTS compared to control animals injected with IgG (Figure 17B). Also, similar to the nephrinY3F animals, mice injected with NTS have reduced total Yap levels (Figure 17B). The evidence from nephrinY3F mice and the NTS injury model support a model in which nephrin phosphorylation stabilizes Yap, possibly through recruitment of Nck and WTIP, while nephrin dephosphorylation promotes Yap destabilization, likely through activation of the Hippo kinase pathway. In corroboration of this data, we should next characterize the expression of Yap target genes, such as connective tissue growth factor and amphiregulin (Zhang et al., 2009; Gumbiner and Kim, 2014) in glomeruli of nephrinY3F mice and control littermates and in response to podocyte injury models.

In general, regulation of the Hippo pathway is largely through mechanotransduction, the biochemical transduction of mechanical forces that can derive from the cell-cell, cell-matrix or cell-lumen interfaces. Adhesion molecules, such as cadherin, are the initial receptors of mechanical forces (Zhang and Labouesse, 2012). Nephrin has not yet been identified as a mechanotransducer, however it is widely hypothesized that it must be involved in responding to

64 mechanical forces from the glomerular capillaries such as hypertension (Simons and Huber,

2008). Although the core Hippo pathway and some of its regulators are highly evolutionarily conserved from D. Melanogaster to mammals it is noted to have highly divergent and cell type- specific upstream regulators in higher order species (Pan, 2010; Bossuyt et al., 2014). We propose that through mechanosensing nephrin acts as a novel receptor upstream of the Hippo pathway to promote cell survival in podocytes. Nephrin pro-survival signaling is already well documented, as nephrin phosphorylation recruits PI3K and AKT, leading to AKT-dependent activation of antiapoptotic Bad (Huber et al., 2003a; Chuang and He, 2010). Also, nephrin phosphorylation recruits dendrin, which similarly to WTIP, translocates to the nucleus in response to podocyte injury and promotes pro-apoptotic gene expression (Asunuma et al., 2007).

We propose that nephrin mechanosensing and signaling to Yap may be mediated through recruitment of Nck and WTIP. One hypothesis is that recruitment of WTIP to nephrin by Nck may yield similar results as Rauskolb et al. (2014), where they report that in Drosophilia jub

(Ajuba family homologue) recruitment to cadherin results in sequestration of Warts (Lats1/2 homologue) away from activators, thus inhibiting the Hippo pathway. Thus, we would expect that nephrin phosphorylation and recruitment of WTIP would result in sequestration of Lats1/2 away from activators, thereby inhibiting the Hippo pathway to render Yap nuclear and active

(Figure 18). This is in line with the results that the nephrinY3F mice, which cannot recruit Nck and by extension WTIP to nephrin, exhibit increased Yap phosphorylation (downregulation) and overall lower basal protein levels of Yap. Also in agreement with this model is the expected

WTIP nuclear localization in response to podocyte injury (Kim et al., 2010). If WTIP is in fact nuclear localized after 24 hours of NTS, then WTIP would not be able to inhibit Lats1/2 and Yap would become phosphorylated and possibly targeted for degradation. A critical experiment to

65 validate this hypothesis is examination of binding between WTIP and Lats1/2 downstream of nephrin stimulation. If this hypothesis is representative of a true mechanism regulating the Hippo pathway then binding between WTIP and Lats1/2 would increase upon induction of nephrin phosphorylation.

!"#$%&'( )*+,-".( /-(012(

)*+,-./%

[email protected]%A*AB-1/*%

!""#$% 7% !""#$% !""&'% 7% !""&'% )56% 0123"4(% !"("#% 7% 89:7% 0123"4(% !"("#% 7% !1+% !1+%

;*<-1;1=>/%

89:7% !1+%

Figure 18: Schematic of proposed model of Yap regulation in mature podocytes through the nephrin-Nck-WTIP signaling complex. Nephrin phosphorylation in healthy podocytes and recruitment of Nck and WTIP is proposed to facilitate Lats1/2 inhibition, thus Yap is active in the nucleus promoting expression of pro-survival genes. When nephrin phosphorylation is reduced, such as in the case of nephrinY3F mice or in response to podocyte injury, Nck and WTIP are uncoupled from nephrin, WTIP can be nuclear localized and unable to inhibit Lats1/2 and Yap becomes phosphorylated and targeted for degradation.

66 CHAPTER 5: CONCLUSION

Within this thesis it has been demonstrated that Nck and WTIP are binding partners and the interaction is mediated by Nck’s first and third SH3 domain binding to the PRR of WTIP.

While we were not able to show parallel nuclear localization for Nck and WTIP, we were however able to demonstrate interaction between WTIP and nephrin via Nck in the cytoplasm.

Furthermore, we show using transgenic mice and the NTS podocyte challenge model that decreased nephrin phosphorylation, and likely reduced complex formation between nephrin, Nck and WTIP, is associated with decreased total Yap levels in glomeruli. We therefore propose that

WTIP via Nck can link nephrin signaling to the Hippo pathway in podocytes. This signaling pathway may be important in podocyte mechanotransduction, which may in turn affect actin regulation and/or survival.

68 REFERENCES

Aaltonen, P. & Holthöfer, H. The nephrin-based slit diaphragm: new insight into the signalling platform identifies targets for therapy. Nephrology, Dialysis, Transplantation 22, 3408–10 (2007). Advani, A. et al. Inhibition of the epidermal growth factor receptor preserves podocytes and attenuates albuminuria in experimental diabetic nephropathy. Nephrology 16, 573–81 (2011). Antón, I. M., Lu, W., Mayer, B. J., Ramesh, N. & Geha, R. S. The Wiskott-Aldrich syndrome protein-interacting protein (WIP) binds to the adaptor protein Nck. The Journal of Biological Chemistry 273, 20992–5 (1998). Asanuma, K., Campbell, K. N., Kim, K., Faul, C. & Mundel, P. Nuclear relocation of the nephrin and CD2AP-binding protein dendrin promotes apoptosis of podocytes. Proceedings of the National Academy of Sciences of the United States of America 104, 10134–9 (2007). Asanuma, K. et al. Synaptopodin regulates the actin-bundling activity of α-actinin in an isoform- specific manner. Journal of Clinical Investigation 115, (2005). Asanuma, K. & Mundel, P. The role of podocytes in glomerular pathobiology. Clinical and Experimental Nephrology 7, 255–9 (2003). Banjade, S. and Rosen, K. M. Phase transitions of multivalent proteins can promote clustering of membrane receptors. Elife doi: 10.7554/eLife.04123. (2014). Barletta, G.M., Kovari, I.A., Verma, R. K., Kerjaschki, D. & Holzman, L. B. Nephrin and Neph1 co-localize at the podocyte foot process intercellular junction and form cis hetero- oligomers. The Journal of Biological Chemistry 278, 19266–71 (2003). Berndt, W. O. Renal function tests: what do they mean? A review of renal anatomy, biochemistry, and physiology. Environmental Health Perspectives 15, 55–71 (1976). Blasutig, I. M. et al. Phosphorylated YDXV motifs and Nck SH2/SH3 adaptors act cooperatively to induce actin reorganization. Molecular and Cellular Biology 28, 2035–46 (2008). Bollée, G. et al. Epidermal growth factor receptor promotes glomerular injury and renal failure in rapidly progressive crescentic glomerulonephritis. Nature Medicine 17, 1242–50 (2011). Bossuyt, W. et al. An evolutionary shift in the regulation of the Hippo pathway between mice and flies. Oncogene 33, 1218–28 (2014). Buday, L., Wunderlich, L. & Tamás, P. The Nck family of adapter proteins: regulators of actin cytoskeleton. Cellular Signalling 14, 723–31 (2002). Chaung & He signaling in regulation of podocyte phenotype. Nephron Physiology 111, 1–10 (2010). Ciani, L., Patel, A. & Allen, N. D. Mice Lacking the Giant Protocadherin mFAT1 Exhibit Renal Slit Junction Abnormalities and a Partially Penetrant Cyclopia and Anophthalmia Phenotype. Molecular and Cellular Biology 23, 3575–3582 (2003). Codelia, V. A, Sun, G. & Irvine, K. D. Regulation of YAP by Mechanical Strain through Jnk and Hippo Signaling. Current biology 24, 2012–7 (2014). Ditlev A. J. Stoichiometry of Nck-dependent actin polymerization in living cells. The Journal of Cellular Biology. 197, 643-658 (2012). Doné, S. C. et al. Nephrin is involved in podocyte maturation but not survival during glomerular development. Kidney International 73, 697–704 (2008).

69 Donoviel, D. B. et al. Proteinuria and Perinatal Lethality in Mice Lacking NEPH1, a Novel Protein with Homology to NEPHRIN. (2001). Molecular and Cellular Biology. 21, 4829- 4836. Drossopoulou, G. I., Tsotakos, N. E. & Tsilibary, E. C. Impaired transcription factor interplay in addition to advanced glycation end products suppress podocalyxin expression in high glucose-treated human podocytes. American Journal of Physiology. Renal Physiology 297, F594–603 (2009). Eckardt, K.U. et al. Evolving importance of kidney disease: from subspecialty to global health burden. Lancet 382, 158–69 (2013). Engl, W., Arasi, B., Yap, L. L., Thiery, J. P. & Viasnoff, V. Actin dynamics modulate mechanosensitive immobilization of E-cadherin at adherens junctions. Nature Cell Biology 16, 587–94 (2014). Faul, C., Asanuma, K., Yanagida-Asanuma, E., Kim, K. & Mundel, P. Actin up: regulation of podocyte structure and function by components of the actin cytoskeleton. Trends in Cell Biology 17, 428–37 (2007). Gerke, P. Homodimerization and Heterodimerization of the Glomerular Podocyte Proteins Nephrin and NEPH1. Journal of the American Society of Nephrology 14, 918–926 (2003). Goicoechea, S. M. et al. Nck-2 interacts with focal adhesion kinase and modulates cell motility. The International Journal of Biochemistry & Cell Biology 34, 791–805 (2002). Goyal, R. K. et al. Ajuba, a novel LIM protein, interacts with Grb2, augments mitogen-activated protein kinase activity in fibroblasts, and promotes meiotic maturation of Xenopus oocytes in a Grb2- and Ras-dependent manner. Molecular and Cellular Biology 19, 4379–89 (1999). Gumbiner, B. M. & Kim, N.G. The Hippo-YAP signaling pathway and contact inhibition of growth. Journal of Cell Science 127, 709–17 (2014). Guo, J.-K. et al. WT1 is a key regulator of podocyte function: reduced expression levels cause crescentic glomerulonephritis and mesangial sclerosis. Human Molecular Genetics 11, 651–9 (2002). Hall, J.E. Ed. Physiologic Anatomy of the Kidneys. Guyton and Hall Textbook of Medical Physiology 12th edition. Harris, R. EGFR signaling in podocytes at the root of glomerular disease. Nature medicine 17, 1188–9 (2011). Hartwig, S. et al. Genomic characterization of Wilms’ tumor suppressor 1 targets in nephron progenitor cells during kidney development. Development 137, 1189–203 (2010). Heikkilä, E. et al. Trans-interaction of nephrin and Neph1/Neph3 induces cell adhesion that associates with decreased tyrosine phosphorylation of nephrin. The Biochemical Journal 435, 619–28 (2011). Hong, S., Troyanovsky, R. B. & Troyanovsky, S. M. Binding to F-actin guides cadherin cluster assembly, stability, and movement. The Journal of Cell Biology 201, 131–43 (2013). Howe, A. K. Cell adhesion regulates the interaction between Nck and p21-activated kinase. The Journal of Biological Chemistry 276, 14541–4 (2001). Huber, T. B. et al. Nephrin and CD2AP Associate with Phosphoinositide 3-OH Kinase and Stimulate AKT-Dependent Signaling. Molecular and Cellular Biology 23, 4917–4928 (2003)a.

70 Huber, T. B. Molecular basis of the functional podocin-nephrin complex: mutations in the NPHS2 gene disrupt nephrin targeting to lipid raft microdomains. Human Molecular Genetics 12, 3397–405 (2003)b. Inoue, T. et al. FAT is a component of glomerular slit diaphragms. Kidney International 59, 1003–12 (2001). Jha, V. et al. Chronic kidney disease: global dimension and perspectives. Lancet 382, 260–72 (2013). Jones, N. et al. Nck adaptor proteins link nephrin to the actin cytoskeleton of kidney podocytes. Nature 440, 818–23 (2006). Jones, N. et al. Nck proteins maintain the adult glomerular filtration barrier. Journal of the American Society of Nephrology 20, 1533–43 (2009). Juhila, J. et al. Inducible nephrin transgene expression in podocytes rescues nephrin-deficient mice from perinatal death. The American Journal of Pathology 176, 51–63 (2010). Kagami, S. & Kondo, S. Beta1-integrins and glomerular injury. The Journal of Medical Investigation 51, 1-13 (2004). Kestilä, M. et al. Positionally cloned gene for a novel glomerular protein nephrin is mutated in congenital nephrotic syndrome. Molecular Cell 1, 575–82 (1998). Khoshnoodi, J. et al. Nephrin promotes cell-cell adhesion through homophilic interactions. The American Journal of Pathology 163, 2337–46 (2003). Kim, J. H. et al. Podocyte injury induces nuclear translocation of WTIP via microtubule- dependent transport. The Journal of Biological Chemistry 285, 9995–10004 (2010). Kim, J. H., Mukherjee, A., Madhavan, S. M., Konieczkowski, M. & Sedor, J. R. WT1- interacting protein (Wtip) regulates podocyte phenotype by cell-cell and cell-matrix contact reorganization. American Journal of Physiology. Renal Physiology 302, F103–15 (2012). Kim, N., Koh, E., Chen, X. & Gumbiner, B. M. E-cadherin mediates contact inhibition of proliferation through Hippo signaling-pathway components. Proceedings of the National Academy of Science 108, 11930-11935 (2011). Kitamura, T. et al. Molecular cloning of p125Nap1, a protein that associates with an SH3 domain of Nck. Biochemical and Biophysical Research Communications 219, 509–14 (1996). Kremer, B. E., Adang, L. A. & Macara, I. G. Septins Regulate Actin Organization and Cell- Cycle Arrest through Nuclear Accumulation of NCK Mediated by SOCS7. Cell 7, 837– 850 (2007). Kretzler, M. Regulation of adhesive interaction between podocytes and glomerular basement membrane. Microscopy Research and Technique 57, 247–53 (2002). Kriz, W. Podocyte is the major culprit accounting for the progression of chronic renal disease. Microscopy Research and Technique 57, 189–95 (2002). Kriz, W. & LeHir, M. Pathways to nephron loss starting from glomerular diseases — Insights from animal models. Kidney International 67, 404–419 (2005). Lahdenpera, J. et al. Clustering-induced tyrosine phosphorylation of nephrin by Src. Kidney International 64, 404–413 (2003). Lawe, D. C., Hahn, C. & Wong, A. J. The Nck SH2/SH3 adaptor protein is present in the nucleus and associates with the nuclear protein SAM68. Oncogene 14, 223–231 (1997). Lee, S. B. et al. The Wilms Tumor Suppressor WT1 Encodes a Transcriptional Activator of amphiregulin. Cell 98, 663–673 (1999).

71 Lefevre, G. M., Patel, S. R., Kim, D., Tessarollo, L. & Dressler, G. R. Altering a histone H3K4 methylation pathway in glomerular podocytes promotes a chronic disease phenotype. PLoS genetics 6, e1001142 (2010). Li, H. et al. Rat nephrin modulates cell morphology via the adaptor protein Nck. Biochemical and Biophysical Research Communications 349, 310–6 (2006). Lu, W., Katz, S., Gupta, R. & Mayer, B. J. Activation of Pak by membrane localization mediated by an SH3 domain from the adaptor protein Nck. Current Biology 7, 85–94 (1997). Reddy, B. V. V. G. & Irvine, K. D. Regulation of Hippo Signaling by EGFR-MAPK Signaling through Ajuba Family Proteins. Developmental Cell 24, 459–471 (2013). Marie, H. et al. The LIM protein Ajuba is recruited to cadherin-dependent cell junctions through an association with alpha-catenin. The Journal of Biological Chemistry 278, 1220–8 (2003). Mathieson, P. W. The podocyte cytoskeleton in health and in disease. Clinical Kidney Journal 5, 498–501 (2012). Matuoka, K., Miki, H., Takahashi, K. & Takenawa, T. A novel ligand for an SH3 domain of the adaptor protein Nck bears an SH2 domain and nuclear signaling motifs. Biochemical and Biophysical Research Communications 239, 488–92 (1997). Nahas, A. M. & Bello, A. K. Chronic kidney disease: the global challenge. Lancet 365, 331–40 (2010). Mayer, B. J. SH3 domains: complexity in moderation. Signal Transduction and Cellular Organization. 114, 1253-1263 (2000). Miner, J. H. & Abrahamson, D. R. Molecular and Cellular Mechanisms of Glomerular Capillary Development. Seldin and Giebisch’s The Kidney 891–910 (2013). Moeller, M. J. Dynamics at the slit diaphragm--is nephrin actin’? Nephrology, dialysis, transplantation 22, 37–9 (2007). New, L. A., Keyvani Chahi, A. & Jones, N. Direct regulation of nephrin tyrosine phosphorylation by Nck adaptor proteins. The Journal of Biological Chemistry 288, 1500– 10 (2013). New, L. A., Martin, C. E. & Jones, N. Advances in slit diaphragm signaling. Current Opinion in Nephrology and Hypertension 23, 420–30 (2014). Nielsen, J. S. & McNagny, K. M. The role of podocalyxin in health and disease. Journal of the American Society of Nephrology 20, 1669–76 (2009). Noris, M. & Remuzzi, G. Non-muscle myosins and the podocyte. Clinical Kidney Journal 5, 94– 101 (2012). Padrick, S. B., Doolittle, L. K., Brautigam, C. A., King, D. S. & Rosen, M. K. Arp2/3 complex is bound and activated by two WASP proteins. Proceedings of the National Academy of Sciences of the United States of America 108, E472–9 (2011). Palmer, R. E. et al. WT1 regulates the expression of the major glomerular podocyte membrane protein Podocalyxin. Current biology 11, 1805–9 (2001). Pan, D. The hippo signaling pathway in development and cancer. Developmental Cell 19, 491– 505 (2010). Patrakka, J. & Tryggvason, K. Molecular make-up of the glomerular filtration barrier. Biochemical and Biophysical Research Communications 396, 164–9 (2010). Pavenstädt, H., Kriz, W. & Kretzler, M. Cell biology of the glomerular podocyte. Physiological Reviews 83, 253–307 (2003).

72 Pawson, T. Dynamic control of signaling by modular adaptor proteins. Current Opinion in Cell Biology 19, 112–6 (2007). Pawson, T. & Nash, P. Assembly of cell regulatory systems through protein interaction domains. Science 300, 445–52 (2003). Putaala, H., Sainio, K., Sariola, H. & Tryggvason, K. Primary Structure of Mouse and Rat Nephrin cDNA and Structure and Expression of the Mouse Gene. Journal of the American Society of Nephrology 11, 911-1001 (2000). Putaala, H., Soininen, R., Kilpeläinen, P., Wartiovaara, J. & Tryggvason, K. The murine nephrin gene is specifically expressed in kidney, brain and pancreas: inactivation of the gene leads to massive proteinuria and neonatal death. Human Molecular Genetics 10, 1–8 (2001). Pyram, R., Kansara, A., Banerji, M. A. & Loney-Hutchinson, L. Chronic kidney disease and diabetes. Maturitas 71, 94–103 (2012). Quack, I. et al. beta-Arrestin2 mediates nephrin endocytosis and impairs slit diaphragm integrity. Proceedings of the National Academy of Sciences of the United States of America 103, 14110–5 (2006). Quaggin, S. E. & Kreidberg, J. a Development of the renal glomerulus: good neighbors and good fences. Development 135, 609–20 (2008). Rauskolb, C., Sun, S., Sun, G., Pan, Y. & Irvine, K. D. Cytoskeletal Tension Inhibits Hippo Signaling through an Ajuba-Warts Complex. Cell 158, 143–56 (2014). Raiser, J. & Sever, S. podocyte biology and pathogenesis of kidney disease. Annual Review of Medicine 357–366 (2013). Rico, M. et al. WT1-interacting protein and ZO-1 translocate into podocyte nuclei after puromycin aminonucleoside treatment. American Journal of Physiology. Renal Physiology 289, F431–41 (2005). Ristola, M., Arpiainen, S., Saleem, M. a, Holthöfer, H. & Lehtonen, S. Transcription of nephrin- Neph3 gene pair is synergistically activated by WT1 and NF-κB and silenced by DNA methylation. Nephrology, Dialysis, Transplantation 27, 1737–45 (2012). Rivera, G. M., Briceño, C. a., Takeshima, F., Snapper, S. B. & Mayer, B. J. Inducible Clustering of Membrane-Targeted SH3 Domains of the Adaptor Protein Nck Triggers Localized Actin Polymerization. Current Biology 14, 11–22 (2004). Rivero-Lezcano, O. M., Marcilla, a, Sameshima, J. H. & Robbins, K. C. Wiskott-Aldrich syndrome protein physically associates with Nck through Src homology 3 domains. Molecular and Cellular Biology 15, 5725–31 (1995). Rodewald, R. & Karnovsky, M. J. Porous substructure of the glomerular slit diaphragm in the rat and mouse. The Journal of Cell Biology 60, 423–33 (1974). Rohatgi, R., Nollau, P., Ho, H. Y., Kirschner, M. W. & Mayer, B. J. Nck and phosphatidylinositol 4,5-bisphosphate synergistically activate actin polymerization through the N-WASP-Arp2/3 pathway. The Journal of Biological Chemistry 276, 26448– 52 (2001). Ruotsalainen, V. et al. Nephrin is specifically located at the slit diaphragm of glomerular podocytes. Proceedings of the National Academy of Sciences of the United States of America 96, 7962–7 (1999). Saleem, M. a et al. Co-localization of nephrin, podocin, and the actin cytoskeleton: evidence for a role in podocyte foot process formation. The American Journal of Pathology 161, 1459– 66 (2002).

73 Sashs, N. & Sonnenberg, A. Cell-matrix adhesion of podocytes in physiology and disease. Nature Reviews Nephrology 9, 200–210 (2013). Schlondorff, J. Nephrin AKTs on actin: The slit diaphragm-actin cytoskeleton signaling network expands. Kidney International 73, 524–6 (2008). Schwarz, K. et al. Podocin, a raft-associated component of the glomerular slit diaphragm, interacts with CD2AP and nephrin. The Journal of Clincal Investigation 108, 1583–1587 (2001). Schwartzman M.H. et al. Podocyte-specific Deletion of Yes-associated protein leads to progressive glomerulosclerosis. Journal of American Society of Nephrology: Abstract Supplement 25, 742 (2014). Simons, M. & Huber, T. B. It’s not all about nephrin. Kidney International 73, 671–3 (2008). Srichai, M. B. et al. A WT1 co-regulator controls podocyte phenotype by shuttling between adhesion structures and nucleus. The Journal of Biological Chemistry 279, 14398–408 (2004). Sun, G. & Irvine, K. D. Ajuba Family Proteins Link JNK to Hippo Signaling. Science Signaling 6, ra81 (2013). Takashi, K., Shinya, M. & Miyuki, K. The decreases of nephrin and nuclear WT1 in podocytes may cause albuminuria during the experimental sepsis in mice. Biomedical Research 31, 363–369 (2010). Tanaka, M., Gupta, R. & Mayer, B. J. Differential inhibition of signaling pathways by dominant- negative SH2/SH3 adapter proteins. Molecular and Cellular Biology 15, 6829–37 (1995). Thakur, M.D. et al. Ajuba LIM proteins are negative regulators of the Hippo signaling pathway. Current Biology 20, 657–62 (2010). Tryggvason, K. Unraveling the Mechanisms of Glomerular Ultrafiltration: Nephrin, a Key Component of the Slit Diaphragm. Journal of the American Society of Nephrology 10, 2440-2445 (1999). Tryggvason, K., Patrakka, J. & Wartiovaara, J. Hereditary proteinuria syndromes and mechanisms of proteinuria. The New England Journal of Medicine 354, 1387–401 (2006). Tryggvason, K. & Wartiovaara, J. How does the kidney filter plasma? Physiology 20, 96–101 (2005). Tu, Y., Li, F. & Wu, C. Nck-2, a novel Src homology2/3-containing adaptor protein that interacts with the LIM-only protein PINCH and components of growth factor receptor kinase-signaling pathways. Molecular Biology of the Cell 9, 3367–82 (1998). Tu, Y., Kucik, D. F. & Wu, C. Identification and kinetic analysis of the interaction between Nck- 2 and DOCK180. 491, 193–199 (2001). Tu, Y., Liang, L., Frank, S. J. & Wu, C. Src homology 3 domain-dependent interaction of Nck-2 with insulin receptor substrate-1. Journal of Biochemistry 322, 315–322 (2001). Tyson, G. E. Scanning electron microscopic study of the effect of vinblastine on podocytes of rat kidney. Virchows Archiv. B: Cell pathology 25, 105–15 (1977). Usui, J. et al. Localization of intercellular adherens junction protein p120 catenin during podocyte differentiation. Anatomy and Embryology 206, 175–84 (2003). Varelas, X. The Hippo pathway effectors TAZ and YAP in development, homeostasis and disease. Development 141, 1614–26 (2014). Vaynberg, J. et al. Structure of an ultraweak protein-protein complex and its crucial role in regulation of cell morphology and motility. Molecular Cell 17, 513-523 (2005).

74 Verma, R. et al. Nephrin ectodomain engagement results in Src kinase activation, nephrin phosphorylation, Nck recruitment, and actin polymerization. The Journal of Clincal Investigation 116, (2006). Vidal, C., Geny, B., Melle, J., Jandrot-Perrus, M. & Fontenay-Roupie, M. Cdc42/Rac1- dependent activation of the p21-activated kinase (PAK) regulates human platelet lamellipodia spreading: implication of the cortical-actin binding protein cortactin. Blood 100, 4462–9 (2002). Wagner, N., Wagner, K.D., Xing, Y., Scholz, H. & Schedl, A. The major podocyte protein nephrin is transcriptionally activated by the Wilms’ tumor suppressor WT1. Journal of the American Society of Nephrology 15, 3044–51 (2004). Wang, B. R-Ras Contains a Proline-rich Site That Binds to SH3 Domains and Is Required for Integrin Activation by R-Ras. Journal of Biological Chemistry 275, 5222–5227 (2000). Weber, G. F., Bjerke, M. a & DeSimone, D. W. Integrins and cadherins join forces to form adhesive networks. Journal of Cell Science 124, 1183–93 (2011). Welsch, T., Endlich, N., Kriz, W. & Endlich, K. CD2AP and p130Cas localize to different F- actin structures in podocytes. American Journal of Physiology. Renal Physiology 281, F769–77 (2001). Welsh, G. I. & Saleem, M. A. Nephrin — signature molecule of the glomerular podocyte? Journal of Pathology 220, 328–337 (2010). Wunderlich, L., Faragó, A. & Buday, L. Characterization of interactions of Nck with Sos and dynamin. Cellular Signalling 11, 25–9 (1999)a. Wunderlich, L., Gohér, A., Faragó, A., Downward, J. & Buday, L. Requirement of multiple SH3 domains of Nck for ligand binding. Cellular Signalling 11, 253–62 (1999)b. Yaoita, E., Sato, N., Yoshida, Y., Nameta, M. & Yamamoto, T. Cadherin and catenin staining in podocytes in development and puromycin aminonucleoside nephrosis. Nephrology, Dialysis, Transplantation 17, 16–19 (2002). Yu, H. et al. Structural basis for the binding of proline-rich peptides to SH3 domains. Cell 76, 933–45 (1994). Zenker, M., Machuca, E. & Antignac, C. Genetics of nephrotic syndrome: new insights into molecules acting at the glomerular filtration barrier. Journal of Molecular Medicine 87, 849–57 (2009). Zhang, H. & Labouesse, M. Signalling through mechanical inputs - a coordinated process. Journal of Cell Science 125, 4172–4172 (2012). Zhang, J. et al. YAP-dependent induction of amphiregulin identifies a non-cell-autonomous component of the Hippo pathway. Nature Cell Biology 11, 1444–50 (2009). Zhang, W., Wu, Y., Du, L., Tang, D. D. & Gunst, S. J. Activation of the Arp2/3 complex by N- WASp is required for actin polymerization and contraction in smooth muscle. American Journal of Physiology. Cell physiology 288, C1145–60 (2005). Zhao, B. et al. Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control. Genes & Development 21, 2747–61 (2007). Zhao, Tumaneng & Guan Hippo in organ size control, tissue regeneration and stem cell self- renewal. Nature Cell Biology 13, 877-883 (2011). Zhu, J. et al. P21-Activated Kinases Regulate Actin Remodeling in Glomerular Podocytes. American Journal of Physiology: Renal Physiology 298, F951–61 (2010).