Views for Entrez

Home , ABI gene family member 3, ACTR3, BCKDHA, CARKD, CYTH2, DPYSL2

BASIC RESEARCH www.jasn.org

Phosphoproteomic Analysis Reveals Regulatory Mechanisms at the Kidney Filtration Barrier

†‡ †| Markus M. Rinschen,* Xiongwu Wu,§ Tim König, Trairak Pisitkun,¶ Henning Hagmann,* † † † Caroline Pahmeyer,* Tobias Lamkemeyer, Priyanka Kohli,* Nicole Schnell, †‡ †† ‡‡ Bernhard Schermer,* Stuart Dryer,** Bernard R. Brooks,§ Pedro Beltrao, †‡ Marcus Krueger,§§ Paul T. Brinkkoetter,* and Thomas Benzing*

*Department of Internal Medicine II, Center for Molecular Medicine, †Cologne Excellence Cluster on Cellular Stress | Responses in Aging-Associated Diseases, ‡Systems Biology of Ageing Cologne, Institute for Genetics, University of Cologne, Cologne, Germany; §Laboratory of Computational Biology, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland; ¶Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; **Department of Biology and Biochemistry, University of Houston, Houston, Texas; ††Division of Nephrology, Baylor College of Medicine, Houston, Texas; ‡‡European Molecular Biology Laboratory–European Bioinformatics Institute, Hinxton, Cambridge, United Kingdom; and §§Max Planck Institute for Heart and Lung Research, Bad Nauheim, Germany

ABSTRACT Diseases of the kidney filtration barrier are a leading cause of ESRD. Most disorders affect the podocytes, polarized cells with a limited capacity for self-renewal that require tightly controlled signaling to maintain their integrity, viability, and function. Here, we provide an atlas of in vivo phosphorylated, glomerulus- expressed proteins, including podocyte-specific gene products, identified in an unbiased tandem mass spectrometry–based approach. We discovered 2449 phosphorylated proteins corresponding to 4079 identified high-confidence phosphorylated residues and performed a systematic bioinformatics analysis of this dataset. We discovered 146 phosphorylation sites on proteins abundantly expressed in podocytes. The prohibitin homology domain of the slit diaphragm protein podocin contained one such site, threonine 234 (T234), located within a phosphorylation motif that is mutated in human genetic forms of proteinuria. The T234 site resides at the interface of podocin dimers. Free energy calculation through molecular dynamic simulations revealed a role for T234 in regulating podocin dimerization. We show that phosphorylation critically regulates formation of high molecular weight complexes and that this may represent a general principle for the assembly of proteins containing prohibitin homology domains.

J Am Soc Nephrol 25: 1509–1522, 2014. doi: 10.1681/ASN.2013070760

The kidney ﬁlter consists of three layers: fenestrated proteinuria. Alteration of these proteins results in de- endothelial cells, the glomerular basement mem- fective signaling causing podocyte dysfunction, pro- brane, and podocytes.1 Damage to any of these gressive glomerulosclerosis, and kidney failure. The compartments becomes clinically evident as proteinuria and the development of kidney disease.2 Of particular importance for the regulation of Received July 22, 2013. Accepted November 15, 2013. podocyte biology through signaling is the slit P.T.B. and T.B. are senior authors. diaphragm, a specialized intercellular junction that Published online ahead of print. Publication date available at bridges the 40-nm gap in between foot processes of www.jasn.org. neighboring podocytes. It also serves as a signaling Correspondence: Dr. Paul T. Brinkkoetter or Dr. Thomas Benzing, platform regulating podocyte function. Mutations Department of Internal Medicine II, University of Cologne, Kerpener in genes encoding for components of the slit dia- Str. 62, 50937 Cologne, Germany. Email: paul.brinkkoetter@uk-koeln. phragm, such as nephrin,3 podocin,4 CD2AP,5 and de or [email protected] TRPC6,6,7 are important causes of genetic forms of Copyright © 2014 by the American Society of Nephrology

J Am Soc Nephrol 25: 1509–1522, 2014 ISSN : 1046-6673/2507-1509 1509 BASIC RESEARCH www.jasn.org slit diaphragm protein complex is a lipid-multiprotein super- confidence (localization score.0.75). These phosphorylation complex.8 Of central importance to the integrity and function of sites were used for further analysis and are accessible online the protein complex is the prohibitin homology (PHB) domain in the GloPhos database (https://helixweb.nih.gov/ESBL/ protein podocin,9 which forms multimeric complexes and is Database/GloPhos/GloPhos.htm). The phosphorylation sites required to control signal transduction through associated trans- showed a typical distribution of serine, threonine, and tyro- membrane proteins.10,11 sine phosphorylated proteins, with most phosphorylation Signaling processes governing podocyte function, integrity, sites residing on serine residues (Supplemental Figure 2A). and survival largely depend on signaling processes involving The majority of proteins were found to contain one or two phosphorylation.12,13 Comprehensive analyses of the signal- phosphorylated residues (Supplemental Figure 2B). Phospho- ing events in podocytes in vivo have been hampered by the fact serines and phosphothreonines localized to substantially dif- that interference with these signaling cascades by genetic de- ferent protein classes than phosphotyrosines (Figure 1D, letion often results in massively disrupted and dysfunctional protein classes based on Panther classification). Phosphoser- podocytes. One of the primary aims of this study was to use ines and phosphothreonines were predominantly localized on phosphoproteomics to analyze thousands of phosphorylation structural proteins, transcription factors, and cytoskeleton- sites in native murine glomeruli within single samples. Within associated proteins. Tyrosines, however, mainly localized to this study, we show that this approach allows the introduction receptors, kinases, and cytoskeletal proteins. The list of tyro- of new concepts into signaling processes at the kidney filtra- sine phosphorylated proteins is depicted in Supplemental Ta- tion barrier. ble 1. Gene ontology (GO) term analysis using a previously published mouse phosphorylation atlas as background revealed overrepresentation of phosphorylation sites on pro- RESULTS teins involved in key functions in the glomerulus such as organization of polarity, cell–cell contacts, and cytoskeleton Phosphoproteomic and Proteomic Analyses of Murine (Supplemental Tables 2 and 3). Glomeruli We performed a categorization of phosphorylation motifs We freshly isolated murine glomeruli to obtain a comprehen- using a binary decision tree algorithm.18 Most serine and thre- sive dataset of in vivo phosphorylated glomerular proteins. onine phosphorylation sites were either part of basophilic The isolated glomeruli showed preservation of podocyte mi- ([RK]-x-x-[ST]) or of proline-directed motifs ([ST]-P) (Fig- crostructure as demonstrated by light and electron micros- ure 2A). Position-weighted matrices for the three major kinase copy, even after maintaining them for several minutes ex motifs (proline-directed, acidic, and basophilic) are depicted vivo.14 We performed immunoblotting of protein lysates to in Figure 2B. This finding is representative of a variety of validate the purity of the preparation. In the glomerular frac- phosphorylation datasets obtained with similar ap- tion, there was an enrichment of podocin and nephrin and a proaches.18–21 Using the PhosphoSitePlus repository,17 we ex- de-enrichment for two tubular markers, Na-K-ATPase and tracted respective kinases of basophilic and proline-directed Tamm–Horsfall protein (Supplemental Figure 1A). We also phosphorylation motifs, indicating that substrates of protein verified the solubilization of detergent-resistant glomerular kinase A, protein kinase C (PKC), extracellular signal-regulated proteins such as nephrin and podocin by a lysis buffer con- kinase, and cyclin-dependent kinase 1 (CDK1), CDK2, and taining 8 M urea (Supplemental Figure 1B).15 CDK5 are most frequent (Figure 2, C and D). We also analyzed We performed both proteomic and phosphoproteomic the phosphorylation sites for conservation across multiple spe- profiling of the murine glomerular lysates and identified 4671 cies using the CPhos algorithm (Supplemental Figure 3).22 expressed proteins based on previously published peptide and protein identification criteria.16 We confirmed the majority of Phosphorylation Regulates Podocyte-Specific and Slit previously mass spectrometry (MS)–based identified proteins Diaphragm–Associated Proteins in murine glomeruli (Figure 1A). In addition, this study com- We next generated a list of 48 phosphoproteins known to be prises .90% of all proteins, which are significantly more specifically expressed in podocytes.16 The 146 confident phos- highly expressed in podocytes than in nonpodocyte glomer- phorylation sites corresponding to these proteins are outlined ular cells, based on a previous study16 (termed podocyte- in Supplemental Table 4. In the overall dataset, we also found specific proteins) (Figure 1B). phosphorylation sites on additional several bona fide podocyte- The phosphoproteomic analysis identified 2449 phospho- specific proteins such as phospholipase A2 receptor,23 FAT1,24 proteins. These phosphoproteins covered a substantial per- nephrin, Pdlim2,25 and Lat326 (Slc43a1) (Supplemental Table 5). centage of podocyte-specific proteins as well as proteins shown Representation of different classes of phosphorylation motifs to be expressed in glomeruli previously (Figure 1C).16 We within these candidates was similar compared with the total found 6868 phosphorylated residues. Of these phosphoryla- dataset (Figure 3, A and B; data not shown). tion sites, 1717 were not found in the version of the Phospho- The podocyte phosphoprotein synaptopodin (Synpo)was SitePlus database released on March 2013.17 We found that one of the proteins with the highest number of phosphoryla- 4079 of these sites were unambiguously localized with high tion sites (n=18) in the entire glomerular dataset and in the

1510 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 www.jasn.org BASIC RESEARCH

Figure 1. Analysis of the glomerular proteome and phosphoproteome obtained by LC-MS/MS. (A) Venn diagram indicating the overlap of protein identifications between this study and the previous murine glomerular proteome by Boerries et al.16 (B) Venn diagram indicating the overlap of protein identifications between proteins obtained in this study and proteins determined as podocyte specificin this study. (C) Bar graph indicating the number of glomerular and podocyte-specific proteins found as nonphosphoprotein or phosphoprotein. (D) Classification of glomerular phosphoproteins identified by LC-MS/MS. Protein classification is performed using the Panther database. The horizontal axis shows the relative abundance of serines, threonines, and tyrosines. podocyte-specific protein population (Figure 3, A and B). phosphorylation sites identified on synaptopodin were proline Phosphorylation affected all parts of the protein (Figure directed (Figure 3D). Figure 3E depicts the most frequent res- 3C). Interestingly, most of the phosphorylated residues were idues surrounding all 121 serines on synaptopodin and localized on the C-terminal tail of synaptopodin comprising shows a particular frequent localization of proline after these amino acids 684–909 (UniProt ID: Q9355-2). Of note, all residues.16

J Am Soc Nephrol 25: 1509–1522, 2014 Phosphoproteomic Analysis of Murine Glomeruli 1511 BASIC RESEARCH www.jasn.org

distinct phosphorylation sites within a 11 amino acid sequence window as obtained from the Human Gene Mutation Database (HGMD) (March 2013)29 (Figure 4B). Re- gions with .5 point mutations within an 11 amino acid window are significantly enriched for the occurrence of mutations (P,0.05). It appeared that human T232 (the site corresponding to murine T234) resides in a region, which is frequently affected by mutations. In fact, the residue itself is affected in a patient with sporadic kidney disease (heterozygous T232I mutation).30 Residue T232 of human podocin is part of a highly conserved basophilic phosphorylation motif (H-R-X-X-T). This phosphorylation motif is affected by two further nonconservative amino acid mutations in this patient group (R229L and H228D)31 (Figure 4C). In addition, the polymorphism R229Q (risk allele) also af- fects this motif.28,30 By contrast, only one patient mutation resided close to the human residue number S360 and none close to S380 (Figure 4C). It has been postulated that phosphorylation sites with high biologic significance fi Figure 2. Analysis of phosphorylation motifs based on con dently localized phos- often represent acommonphosphorylation fi phorylation sites. (A) Classi cation of phosphorylation motifs in four classes based on site in different members of a domain a binary decision tree. (B) Position-weighted matrices of the three main phosphory- protein family.32 To further substantiate lation motifs (proline-directed, acidic, and basophilic). (C and D) Abundance of known substrates for basophilic (C) and proline-directed (D) phosphorylation sites as indexed this hypothesis, we analyzed the occur- within the PhosphoSitePlus database. CamK2, calmodulin-dependent kinase 2; SGK, rence of PHB domain phosphorylation at serum glucocorticoid-regulated kinase 1; mTOR, mammalian target of rapamycin; JNK, the domain region corresponding to mu- c-Jun N-terminal kinase. rine podocin T234. The algorithm priori- tizes post-translational modifications within protein domains by searching for We were able to identify 16 phosphorylation sites on major statistical enrichment of phosphorylation sites with actual slit diaphragm proteins (nephrin, Neph1, podocin, TrpC6, and proteomic evidence compared with random sampling of Cd2ap) (Table 1). Kinase preferences or binding motifs attrib- phosphorylation sites across the entire domain.32 The analysis uted to these phosphorylation sites were analyzed using the revealed that phosphorylation sites on PHB domains were Scansite software package (Table 1).27 significantly overrepresented within the domain area aligning with the murine podocin residue T234 (Figure 4D). Thus, it Biologic Relevance of PHB Domain Phosphorylation of represents a “phosphorylation hotspot,”32 which supports its Podocin likeliness to play a pivotal role for PHB domain function.32 To show the relevance of these novel phosphorylation sites, we focused on the phosphorylation sites found on the slit di- In Silico Characterization of Podocin T234 aphragm protein podocin. We identified phosphorylation sites Phosphorylation localized quite centrally in the PHB domain of podocin (T234) The quaternary structure of the homologous PHB domain of and at the distal C terminus (S362, S382) (Figure 4A). stomatin has recently been solved: two PHB domains of Mutations of the podocin gene NPHS2 are a major cause of stomatin form a banana-shaped dimer via main chain inter- steroid-resistant nephrotic syndrome and ESRD requiring di- action of residues 196–199.33 The residues responsible for alysis in children and are therefore of major importance in interaction of dimers are well conserved in human and medicine.28 To prioritize the phosphorylation sites for further mouse podocin as well as in MEC-2, the podocin ortholog studies, we analyzed the frequency of known patient point in Caenorhabditis elegans (Figure 5A). Moreover, the basophilic mutations associated with these diseases surrounding the phosphorylation motif within the PHB domain is also conserved

1512 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 www.jasn.org BASIC RESEARCH

published crystal structure of stomatin PHB domain dimers (Figure 5, B and C; Protein Data Bank ID 4FVF) to generate a homology model of podocin PHB domains (Supplemental Figure 4, A and B). The dimerization and phosphorylation of mono- mersanddimersarerepresentedbya thermodynamic cycle (Supplemental Figure 4C). According to Hess’s law, the effect of phosphorylation on dimerization was cal- culatedbythedifferenceinfreeenergy changes during phosphorylation of monomers and dimers, which can be directly simulated (equation in Supplemental Fig- ure 4C). Both banana-shaped dimers are predicted to bend more upon phosphorylation (Figure 6A) (10° bending for stomatin, 4° bending for podocin in the phosphorylated state). The calculated free energy changes due to phosphorylation for monomer and dimers, as well as for the dimerization, are listed in Table 2. These results predict that phosphorylation promotes stomatin dimerization (DDG=26.161.4 kcal/mol) but prevents podocin from forming dimers (DDG=8.162.2 kcal/ Figure 3. Phosphorylation sites on synaptopodin, a podocyte-specificprotein.(A)A mol). Consistent with this prediction, bar graph depicting the number of high-confident phosphorylation sites normalized the phosphorylation-dependent change over protein residue numbers in the glomerulus. Synaptopodin (Synpo) is one of the in dipole moment was negative for podo- major phosphoproteins in the glomerulus. For the analysis, only phosphoproteins with cin (216 debyes) and positive for stoma- .10 phosphorylation sites are considered. The values in parentheses denote the tindimerformation(+92debyes)(Table absolute count of phosphorylation sites. (B) A bar graph depicting the number of high- 3). The reason for the differential effect fi con dent phosphorylation sites normalized over protein residue number within the of phosphorylation on interaction is the fi . podocyte-speci c protein population. For the analysis, only phosphoproteins with 2 result of a negatively charged (E235) res- phosphorylation sites are considered. The values in parentheses denote the absolute iduepresentinpodocin,butnotinsto- count of phosphorylation sites. (C) Scheme of murine synaptopodin and localization of detected phosphorylation sites. The residue numbers refer to the kidney-specific matin (Figure 6B). This residue is predicted isoform (UniProt ID Q8CC35-2). (D) Sequence logo of all phosphorylation sites dis- to repulse phosphate groups localized at the covered on synaptopodin. The distribution of amino acids in this dataset is used as other protein domain. The force at the in- a background. (E) Sequence logo of all serines and threonines on synaptopodin interaction surface is thereby strong enough dicating the high proline content of the protein, especially at the +1 position. to overcome all opposite forces and prevents dimer formation. with some differences (Figure 5A). We used this information to Effect of Podocin T234 Phosphorylation on visualize the structural context corresponding to murine po- Multimerization docin T234 in murine stomatin S161.34 Interestingly, the res- To further corroborate our in silico findings, we performed idue S161 localized on the PHB domain of the first dimer biochemical analysis of native stomatin and podocin protein resided very close to the residue S161 on the PHB domain of complexes using the blue native gel electrophoresis (BN-PAGE) the second protein (Figure 5B). The distance between both technique.37 We introduced phosphoablating (stomatin serine hydroxyl groups is 9.8 Å (Figure 5C). Thus, we hypoth- S161A, podocin T234A) or phosphomimicking (stomatin esized that this phosphorylation site is involved in regulating S161D, podocin T234D) mutations into the full-length pro- dimerization of the protein. tein. Both phosphoablating and phosphomimicking mutants To further substantiate the effect of phosphorylation on of stomatin were equally expressed (Figure 7A). As previously PHB domain interaction in silico, we performed molecular published, stomatin forms multimeric megadalton complexes dynamics simulations to calculate free energies using the (Figure 7A).38 Consistent with the results obtained in silico, CHARMM algorithms.35,36 We could rely on the previously stomatin showed an increased multimerization when the

J Am Soc Nephrol 25: 1509–1522, 2014 Phosphoproteomic Analysis of Murine Glomeruli 1513 BASIC RESEARCH www.jasn.org

Table 1. Phosphorylation sites on slit diaphragm proteins Gene Symbol Name (UniProt ID) Phosphopeptide Sequence Site Scansite Prediction Nphs1 Nephrin (Q9QZS7) LAEEIS*EKT*EAGS*EEDR S1112, T1115, S1119 Casein kinase sites LAEEISEKT*EAGSEEDRIR T1115a Casein kinase site TEAGS*EEDRIR S1119 Casein kinase site Nphs2 Podocin (Q9QZS) SLT*EILLER T234 AGC family kinase site AQGS*INYPSSSKPVEPLNPK S362 KKDS*PML S382 Trpc6 Short transient receptor potential FGISGS*HEDLSK S814 Casein kinase site channel 6 (Q61143) Cd2ap CD2-associated protein (Q9JLQ0) S*VDLDAFVAR S458 Casein kinase site FNGGHS*PTQS*PEK S510, S514 AEADDGKRNS*VDELR S580 ERK D-domain binding Kirrel Neph1 (Q80W68) VETVNREPLT*MHSDREDDTASISTATR T581 VETVNREPLTMHS*DREDDT*ASISTATR S584, T590 Casein kinase site VETVNREPLTMHS*DREDDTAS*ISTATR S584, S592 VETVNREPLTMHSDREDDTAS*IST*ATR S592, T595 LSHS*SGYAQLNTYSR S676b FSYTSQHS*DYGQR S775 PDK1 binding motif Asterisks (*) indicate the localization of phosphate. ERK, extracellular signal-regulated kinase; PDK1, phosphoinositide-dependent protein kinase 1. aAmbiguous phosphorylation site due to localization score,0.75, appropriate spectrum on manual examination. bAmbiguous phosphorylation site (S676 versus S675) upon manual examination of spectrum.

phosphomimicking mutation S161D was present (lack of low podocin showed a lower multimeric complex when the T234D molecular mass complexes, with a molecular mass,1000 kDa) mutation was introduced (Figure 7B). We conﬁrmed these (Figure 7A). In contrast and consistent with the simulations, ﬁndings in coimmunoprecipitation assays. Compared with

Figure 4. Phosphorylation sites discovered on murine podocin and its corresponding human sites. (A) Murine and corresponding human residue numbers are mapped to the structure of podocin (residue numbers based on Pfam entry [http://pfam.sanger.ac.uk/]). (B) Frequency of described point mutations within the podocin protein. All currently known mutations associated with either FSGS or nephrotic are extracted from the HGMD. The plot shows the frequency of found mutations within an 11-amino-acid sequence window surrounding the respective residue number. (C) An 11-amino-acid sequence window of murine podocin phosphosites. Mutations or risk alleles previously described in patients with either FSGS or nephrotic syndrome are mapped to the sequence. (D) Phosphorylation enrichment analysis of PHB/Band7-like protein domains. The position of aligned murine Podocin T234 residue corresponds to a phosphorylation hotspot within the PHB domain. Asterisk (*) indicates signiﬁcant enrichment.

1514 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 www.jasn.org BASIC RESEARCH

Figure 5. Structural context of podocin phosphorylation site T234. (A) Alignment of podocin and stomatin demonstrating conservation of interaction residues as well as the phosphorylation motif. (B) Visualization of the residue corresponding to podocin T234, stomatin S161 on the crystal structure of stomatin PHB domain dimer. (C) Magnification of the structure showing a distance between both phosphorylable serine hydroxyl groups of 9.8 Å. the phosphoablating mutant (F.T234A), phosphomimicking enhanced the understanding of epithelial function in the mutant (F.T234D) interacted to a lesser degree with V5.T234A tubular system41–44 and are cornerstones for hypothesis- podocin (Figure 7, C and D). When both T234 residues (on V5 driven studies in many laboratories.45–47 Here, we present and FLAG-tagged mutants) were mutated to D, there was also a the first available atlas of phosphorylated residues of proteins marked decrease in homophilic interactions (Figure 7, C and D). present in native murine glomeruli revealed by an advanced Fyn was used as a similar-sized protein not interacting with MS-based method. podocin.39 Converse results were obtained for the respective The phosphopeptides analyzed in this study may derive phosphomimicking stomatin mutants (Figure 7, E and F). from any glomerular cell type, including mesangial cells. Atypical PKC (aPKC) is a master regulator of podocyte However,webelieve that thisapproach isnot only morefeasible biology and is critically involved in regulating cellular polar- but is also superior to single-cell separations and long flow- ization and differentiation.11,40 We tested whether aPKC may sorting procedures because these may dramatically alter regulate podocin phosphorylation at T234. To this end, we phosphorylation patterns in any cell type. We performed cotransfected wild-type murine podocin with active (T410E) bioinformatics analysis to retrieve phosphorylation sites in or inactive (T410A) aPKC and subjected whole cell lysates to podocyte-specific proteins and added several bona fide podocyte- phosphoproteomic analysis. Label-free quantification re- specific proteins to this list. Phosphorylation motifs at least for vealed an increased phosphorylation of T234 with the active the 146 phosphorylation sites on known podocyte-specificpro- version of the kinase (Supplemental Figure 5). The phosphor- teins were not strikingly different compared with nonpodocyte ylation at S382, the second discovered site in this setting, was proteins (data not shown). This suggests that global baseline not increased by cotransfection of the active kinase. kinase activity in the different cell types is comparable. Hence, differences in morphology and cell cycle most likely result from signaling within microdomains such as the foot pro- DISCUSSION cesses. Our study is also consistent with the concept of different responsibilities for tyrosine and serine/threonine phosphoryla- Phosphoproteomic approaches have been applied to the tion sites in maintaining cell morphology in general (Figure 1D, kidney with a focus on the tubular system. These studies Supplemental Table 2).

J Am Soc Nephrol 25: 1509–1522, 2014 Phosphoproteomic Analysis of Murine Glomeruli 1515 BASIC RESEARCH www.jasn.org

conceivable that dynamic phosphorylation at this region (by proline-directed kinases, such as extracellular signal-regulated kinase and CDK1, CDK2, or CDK5) may be involved in regulating this interaction and their modulation of the podocyte cytoskeleton. The podocyte slit diaphragm is a structure with a direct pathophysiologic effect. This dataset yielded 16 previously unde- scribed phosphorylation sites on five major slit diaphragm proteins: Neph1, nephrin, podocin, CD2AP, and Trpc6. Interestingly, Neph1, nephrin, Trpc6, and CD2AP all had phosphorylation sites within acidic regions and may thus be targets of acidic kinases such ascaseinkinases,akinaseclassnotimplicated in glomerular biology. S814 at TrpC6 appears nottobephosphorylatedbycaseinkinase2.50 The previously described tyrosin phosphorylation site on nephrin was not detected, which may either be the result of the not optimal access of trypsin to this site or due to a low baseline phosphorylation because Figure 6. In silico analysis of podocin phosphorylation on T234 through molecular dynamics. (A) Result of molecular dynamics simulation predicting bending of the several studies have shown that nephrin ty- banana-shaped dimer by 10° (stomatin, conformations before and after phosphory- rosine phosphorylation is inducible upon li- – lation are colored green and purple, respectively) as well as 4° (podocin, conformations gation of the extracellular domain.51 53 before and after phosphorylation are colored yellow and red, respectively). (B) Details at We focused on studying phosphoryla- the interface explaining different effect of phosphorylation on podocin and stomatin tion of the PHB domain protein podocin. dimerization. Protein backbones are shown as helical ribbons for helices and as arrowed Mutations in the podocin gene (NPHS2) ribbons for b-strands with arrows toward the C terminus. Positively and negatively are a major cause of steroid-resistant ne- charged residues are colored blue and red, respectively. The side chains of S161 in phrotic syndrome in children.4 Three co- stomatin and T234 in podocin, as well as the phosphate groups attached to them, are herent lines of evidence suggest a major shown as sticks. The differential effect of phosphorylation on free energy can be ex- importance of the phosphorylation site plained by the existence of a negatively charged residue, E235 in podocin, but not in T234 on murine podocin function. First, stomatin. The negative charge of E235 repulses the phosphate group so that the phosphorylated podocin has less tendency to be dimerized (indicated by gray arrows). this site is highly conserved and localizes within a conserved functional protein do- Our analysis revealed that synaptopodin is among the main. Within the PHB domain, this phosphorylation site re- phosphoproteins with highest number of phosphorylation sides within a phosphorylation hotspot, a protein domain sites within the podocyte and the glomerulus (Figure 3, A and region with significantly enriched evidence for phosphoryla- B). The majority of sites localized within the C terminus of the tion32 (Figure 4D). Such regions have been shown to be im- kidney-specific isoform and were mostly proline directed, a portant for protein functions in general.32 Second, mutation finding that may be attributed to the large number of proline of the phosphorylation site T232 in humans was recently residues within this protein (Figure 3, C–E). The C terminus is found in a patient with FSGS.30 This mutation (T232I) is one of the main interaction sites for a-actinin-4,48 a regulator nonphosphoryable. In addition, the functional basophilic of podocyte cytoskeletal dynamics. Mutations in the ACTN4 phosphorylation motif (H-R-x-x-T) is affected by two other gene encoding for a-actinin-4 are a known cause of genetic patient mutations (R229L, H228D) and the frequent risk poly- forms of kidney disease and hypertension.49 It is highly morphism R229Q.4,28,31 Genetic data obtained from the model organism C. elegans point to a similar direction. There is a strong similarity of the mechanosensory machinery in C. Table 2. Free energy changes (kcal/mol) of phosphorylation elegans and the complex organization of the slit diaphragm.8 for monomer and dimer proteins The PHB domain protein MEC-2 is a crucial member of this D D DD System G1 G2 G complex responsible for touch sensitivity.54 In fact, a dominant Podocin 2272.660.8 2537.162.1 8.162.2 mutation (T246I) in the PHB domain of MEC-2, affecting the 2 6 2 6 2 6 Stomatin 275.3 0.4 556.7 1.6 6.1 1.4 same area of the domain as T232I in podocin (Figure 5A),

1516 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 www.jasn.org BASIC RESEARCH

Table 3. Dipole moments (debyes) of stomatin and podocin monomers and dimers before and after phosphorylation Podocin Stomatin Native T234Ph Phosphorylation Change Native S161Ph Phosphorylation Change Monomer 422 620 198 185 324 139 Dimer 411 791 380 150 520 370 Dimerization change 2433 2549 216 2220 2128 92 Ph, phosphorylation of the respective residue. enhances touch insensitivity.55 The phosphorylation site itself, Sonication on ice (1.5 minutes, 0.5-second pulses) was performed however, is not conserved in C. elegans (Figure 5A). Third, based and the supernatant was saved for further analysis. This cycle was on the crystal structure of the homologous PHB domain of repeated twice in order to ensure the maximal amount of solubiliza- stomatin, this site shows an extremely prominent localization tion of membrane and nuclear proteins. All three lysates were pooled close to the interface responsible for the interaction with other and protein amount was determined using a commercial bicincho- PHB domains (Figure 5, B and C). Using in silico and in vitro ninic acid assay. We saved 40 ml for immunoblot analysis. The exper- methods, we show that this phosphorylation site is involved in iment was repeated twice. All animal procedures were performed organizing podocin multimeric complexes. Interestingly, the ho- according to protocols approved by federal agencies. mologous phosphorylation sites on stomatin (within protein domains) have an opposite effect on protein multimerization. Phosphopeptide and Peptide Purification Phosphorylation induced dimerization for stomatin, whereas Phosphopeptides were purified using a protocol previously used for phosphorylation reduced dimerization for podocin. This was phosphoproteomic studies with some modifications.20 One and one predicted by molecular dynamics simulations (Figure 6, Tables half milligrams of protein was reduced using 5 mM dithiothreitol and 2 and 3) and was confirmed for the respective full-length protein alkylated using 50 mM of iodoacetamide in the dark. Urea was diluted using native gels (Figure 7, A and B). to a concentration of 1.5 M and trypsin was added at a 1:50 w/w ratio. The PHB domain phosphorylation site discovered in this Digestion was performed at 37° for 16 hours. Aqueous peptide solu- study may prove to be a modulator of the oligomerization tion was desalted using Oasis HLB columns (Waters Corporation, status of podocin at the slit diaphragm in vivo, thereby altering Milford, MA). Briefly, peptides were fractionated using a polysul- resistance and mechanics in the podocyte in response to var- foethyl A column (4.6 mm internal diameter320 cm length, 5-mm ious intracellular and extracellular stimuli affecting AGC particle size, 300-Å pore size) on a Finnigan HPLC machine. Strong kinases.56,57 In fact, aPKC, a kinase with relevance for podocyte cation exchange chromatography was conducted at a 1 ml/min flow biology,11,40 can induce phosphorylation of this site when rate using the following flow gradient: 100% solvent A and 0% solvent coexpressed with podocin in human embryonic kidney Bfor2minutes;0%–20% solvent B for 40 minutes; 20%–100% (HEK293T) cells (Supplemental Figure 5). solvent B for 5 minutes; and 100% solvent B held for 5 minutes

Taken together, we use a combination of complex tissue (solvent A: 5 mM KH2PO4, 25% acetonitrile (ACN) pH 2.7; solvent isolation, liquid chromatography coupled to tandem mass B: 5 mM KH2PO4, 25% ACN, 350 mM KCl, pH 2.7). Seven to nine spectrometry (LC-MS/MS)–based phosphoproteomics, and fractions were collected. Phosphopeptides were enriched using computational analysis with subsequent in vitro and in silico FeNTA Immobilized metal ion affinity chromatography (IMAC) col- experiments to understand molecular signaling mechanisms umns (Thermo Fisher Scientific). The IMAC elutes containing phos- in kidney health and disease. phopeptides were subjected to further MS analysis. After cleanup using ZipTips (Millipore, Bonn, Germany), samples were submitted to LC-MS/MS analysis. One half of the sample was analyzed using an CONCISE METHODS LTQ Orbitrap XL, and the other half was analyzed using a Q Exactive mass spectrometer. IMAC flowthroughs containing mainly nonphos- Isolation of Murine Glomeruli phorylated peptides were analyzed on an LTQ Orbitrap XL mass Murine glomeruli were obtained as previously described.14 Briefly, spectrometer only to enhance proteomic coverage. 6- to 8-week-old female Bl6N wild-type mice were euthanized by cervical dislocation and the kidneys were excised and perfused with MS HBSS containing magnetic beads. Subsequently, kidneys were digested One half of the respective phosphopeptide samples were analyzed with collagenase for 1 minute at 37° and purified by sieving and a using an LTQ Orbitrap XL mass spectrometer and collision induced strong magnet. Isolated glomeruli were washed two times and were dissociation (CID) fragmentation. Analyses using reversed-phase pelleted down using spin of 1000 rpm for 3 minutes at 4°. Kidneys, liquid chromatography coupled to nanoflow electrospray tandem MS glomeruli, and lysates were kept on ice during the whole procedure. were carried out using an EASY nLC-II system (Proxeon/Thermo Glomeruli were suspended in 200 ml lysis buffer (8 M urea and 50 Fisher Scientific) with a 150 mm C18 column (75 mminternaldi- mM ammonium bicarbonate) containing protease and phosphatase ameter; Dr. Maisch GmbH, Ammerbuch, Germany) coupled to a inhibitor cocktails (Thermo Fisher Scientific, Bremen, Germany). LTQ Orbitrap Discovery mass spectrometer (Thermo Fisher

J Am Soc Nephrol 25: 1509–1522, 2014 Phosphoproteomic Analysis of Murine Glomeruli 1517 BASIC RESEARCH www.jasn.org

Figure 7. Characterization of PHB domain phosphorylation in vitro. (A) BN-PAGE analysis of Stomatin S161A and stomatin S161D multimeric complexes. Lower molecular weight (LMW) complexes are marked by LMW. The same lysates are subjected to SDS-PAGE followed by immunoblotting. (B) BN-PAGE analysis of podocin T234A and T234D multimeric complexes. The same lysates are subjected to SDS-PAGE followed by immunoblotting. All BN-PAGE analyses are performed in triplicate. (C and D) Coimmunoprecipitation

1518 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 www.jasn.org BASIC RESEARCH

Scientific). Peptide separation was performed at a flow rate of PXD000266; http://www.ebi.ac.uk/pride). All phosphorylation data 300 nl/min over 90 minutes (5%–7% ACN in 5 minutes, 7%–45% are made accessible at the online GloPhos database (https://helixweb. in 60 minutes, 45%–50% in 5 minutes, 50%–97% in 5 minutes, wash nih.gov/ESBL/Database/GloPhos/GloPhos). Protein categories were at 100%) (buffer A: 0.1% formic acid in H2O; buffer B: 0.1% formic obtained from the Panther database based on gene symbols of the acid in acetonitrile). Survey full-scan MS spectra (m/z 300–2000) of respective proteins (http://www.panther.org). intact peptides were acquired in the Orbitrap at a resolution of 30,000 Further analyses of bioinformatics data were performed using using m/z 445.12003 as a lock mass. The mass spectrometer acquired Perseus and Microsoft Excel, as well as previously published National spectra in data-dependent mode and automatically switched between Heart, Lung, and Blood Institute (NHLBI) in-house software.60,61 GO MS and MS/MS acquisition. Signals with unknown charge state were terms were annotated via the Perseus package using UniProt acces- excluded from fragmentation. The dynamic exclusion option was sions as the reference. For enrichment analysis, a Fisher’sexacttest enabled (1 minute). The five most intense ions (charge state z$2) with correction for multiple testing (Benjamini–Hochberg FDR) was were isolated and fragmented in the linear ion trap using CID frag- performed. A cutoffofBenjamini–Hochberg FDR of 0.02 (for mentation. The other half of the samples was analyzed using a Q serines/threonines) or 0.05 (for tyrosins) was chosen. The five GO terms Exactive mass spectrometer using higher-energy collisional dissoci- with highest enrichment were selected (Supplemental Tables 2 and 3). ation (HCD) fragmentation as described.58,59 Briefly, enriched As a background dataset, we selected the following: a dataset contain- phosphopeptides were loaded on a reversed-phase column packed ing all phosphoproteins containing serine and threonine phosphory- in-house with C18-AQ 13 mm resin material (Dr. Maisch GmbH) via lation in a comprehensive mouse tissue phosphoproteomic study an Easy nLC nanoflow ultra-HPLC system (Thermo Fisher Scien- (Supplemental Table 2)21; and a dataset containing all phosphoproteins tific). Separation was performed by a linear gradient from 5% to containing tyrosin phosphorylation in the same study (Supplemental 60% buffer B (80% ACN and 0.1% formic acid) at a flow rate of Table 3). The latter two datasets were extracted from PhosphoSite 250 ml/min over 150 minutes. Survey scan MS spectra were measured (http://www.phosphosite.org).17 All protein entries with multiple in the Orbitrap (m/z 300–1750) with a resolution of 70,000 (at m/z phosphorylation sites were counted as one phosphoprotein in the 400) after accumulation of 106 ions. The 10 most intense peptides respective analysis. The CPhos program was used for analysis of site (charge state z$2) were isolated and fragmented in the octapole col- conservation with default settings for mouse datasets.22 lision cell by HCD. Phosphorylation enrichment analysis of PHB/Band7-like protein domains was performed as previously described.32 For this analysis, Bioinformatic Analyses we used 198,773 phosphorylation sites derived from previously pub- Raw files were searched using MaxQuant (version 1.3.0.5). Raw file lished MS experiments across 15 species obtained from the PTMfunc spectra were searched against the mouse UniProt reference database database (http://www.ptmfunc.com). All proteins containing a using the target-decoy strategy. Mass accuracy was 20 ppm in the first Band_7 PFAM domain were aligned to a representative Band_7 con- search and 4.5 ppm in the second search with the deisotoping option taining protein (the mouse Flot2, UniProt: FLOT2_MOUSE) and 30 enabled. For fragment ions, mass tolerance was 0.5 Da for CID data phosphosites from these proteins were mapped to the Flot2 sequence. and 20 ppm for HCD data. Fixed modifications were carbamidome- We then used a sliding window of 10 amino acids and random sam- thylation of cysteines (+57 Da). Variable modifications were phos- pling to identify regions within the Band_7 containing Flot2 protein phorylation at S,T,Y (+80 Da) and oxidation of methionine (+16 Da) that are enriched for phosphosites. Any 10 amino acid peptide with a with a number of four allowed modifications. Protein, peptide, and significant (P,0.01) enrichment of phosphosites compared with site false discovery rate (FDR) were adjusted to ,0.01. All peptides random distribution of sites was defined a potential regulatory region that passed these criteria were taken into consideration. A protein was or regulatory hot spot. Enrichment for mutations within the podocin defined as expressed with one or more unique peptides matched to protein was performed using the same method. Mutations were re- the sequence.16 Class I phosphorylation sites with a localization score trieved from the HGMD database (March 2013, professional ver- .0.75 were termed as confident and were used for further analysis. sion). Eleven amino acid windows surrounding residues were tested. All raw files and corresponding search results were uploaded to the Randomizations were performed 10.000 times. Enrichment of mu- ProteomeXchange/PRIDE repository (ProteomeXchange accession: tations was considered significant when P,0.05.

experiments of different podocin phosphoablating and phosphomimicking mutants. All coimmunoprecipitation experiments are performed with five replicates. Quantification of densitometric intensities is performed. (C) V5-tagged phosphoablating mutant (V5.T234A) is coimmunoprecipitated with FLAG-tagged phosphoablating and phosphomimicking mutants (F.T234A, F.T234D). Total lysates (Lys.) are also immunoblotted. (D) The V5-tagged phosphomimicking mutant (V5.T234D) is coimmunoprecipitated with FLAG-tagged phosphoablating and phosphomimicking mutants (F.T234A, F.T234D). (E and F) Coimmunoprecipitation experiments of different stomatin phosphoablating and phosphomimicking mutants. All coimmunoprecipitation experiments are performed with six replicates. Quantifi- cation of densitometric intensities is performed. (E) The V5-tagged phosphoablating stomatin mutant (V5.S161A) is coimmunoprecipitated with FLAG-tagged phosphoablating and phosphomimicking mutants (F.S161A, F.S161D). (F) The V5-tagged phosphomimicking stomatin mutant (V5.S161D) is coimmunoprecipitated with FLAG-tagged phosphoablating and phosphomimicking stomatin mutants (F.S161A, F.S161D). *Statistical differences in densitometric values (two-tailed t test, P,0.05).

J Am Soc Nephrol 25: 1509–1522, 2014 Phosphoproteomic Analysis of Murine Glomeruli 1519 BASIC RESEARCH www.jasn.org

Structure Visualization and Free Energy Calculation Quantitative Analysis of Podocin Phosphorylation in Sequences were aligned using the ClustalWX algorithm. Structures HEK293T Cells (Protein Data Bank ID 4FVF) were visualized using the Cn3D For quantitative phosphoproteomics analysis of podocin phosphor- program obtained from the National Center for Biotechnology Infor- ylation in HEK293T cells, 5 mg F. murine podocin cloned into a mation website (http://www.ncbi.nlm.nih.gov). Distance between phos- pcDNA6.2 vector was cotransfected with either active PKCz phorylable serine hydroxyl groups was measured using JMol (http:// (T410E) or inactive PKCz (T410A). These plasmids were obtained www.jmol.org). Molecular dynamics were simulated using the from Addgene (Addgene 10801 and 10804).65 Cells were lysed in 8 M CHARMM program.35,36 We used the perturbation method to calcu- urea buffer and subjected to phosphopeptide enrichment with IMAC late free energy differences of a series of 154 windows along the path columns as described above in the phosphopeptide puriﬁcation sec- linking the state before and after phosphorylation in both directions. tion and analyzed with an LTQ Orbitrap XL machine as described In each free energy simulation, 1540-ps simulations were performed. above in the MS section. Raw data were searched using the Sequest Both molecular dynamics simulation and force momentum–based search algorithm implemented in the Proteome Discoverer platform. self-guided Langevin dynamics (SGLDfp) simulation methods were Data were further analyzed using the NHLBI in-house QUOIL pro- utilized.62 For SGLDfp simulations, we used a local average time of gram for visualization, normalization, and label-free quantiﬁcation 0.2 ps, a guiding factor of 1.0, and a friction constant of 1/ps. of MS1-precursor ion intensities as previously described (n=3).66

Cell Culture and Coimmunoprecipitation, and ACKNOWLEDGMENTS Immunoblotting HEK293T cells were maintained in DMEM supplemented with 10% FBS. HEK293T cells were transiently transfected separately with The authors thank Ruth Herzog, Stefanie Keller, and Astrid Wilbrand- F.T234A, F.T234D podocin constructs, and F.Fyn constructs cloned Hennes for excellent technical help; Martin Höhne, Malte Bartram, into a pcDNA6.2 vector using the calcium phosphate method. For and Bodo Beck for helpful discussions; and Mark Knepper (NHLBI) each of these dishes, one additional dish was transfected either with for hosting the GloPhos database. V5.T234A podocin or V5.T234D podocin (cloned into a pcDNA6.2 This work was supported by the German Research Foundation vector). Two days after transfection, cells were harvested with ice-cold PBS. (Sonderforschungsbereich 635 to T.B. and BR2955 to P.T.B.]) and the Cells were lysed in a 1% Triton X-100 buffer (1% Triton X-100, 20 mm German Federal Ministry of Education and Research (GerontoSys2 program Sybacol to T.B.). M.M.R. was supported by a Köln-Fortune Tris–HCl, pH 7.5, 50 mm NaCl, 50 mm NaF, 15 mm Na4P2O7,2mm grant from the University of Cologne as well as a Fritz-Scheler sti- Na3VO4,andRocheCompleteProteaseInhibitorMix)for15minuteson ice. After centrifugation at 20,000g (15 minutes, 4°C), a small aliquot of pendium by the German Kidney Foundation and KfH Foundation for each supernatant was preserved and diluted with 23 SDS–PAGE sample Preventative Medicine. buffer for later Western blot analysis (i.e., lysate). Lysates containing V5-tagged protein (either V5.T234A or V5.T234D, respectively) were DISCLOSURES subsequently pooled and same volumes were added to the F9-tagged None. protein lysate containing FLAG-tagged protein to reduce expression var- iability in V5-tagged bait protein. Pooled lysates were incubated for 1 hour at 4°C with anti-FLAG (M2) antibody covalently coupled to agarose beads REFERENCES (Sigma-Aldrich). The beads were washed fivetimeswithlysisbufferwith 0.25% sodium deoxycholate added, and bound proteins were resolved by fl 1. Brinkkoetter PT, Ising C, Benzing T: The role of the podocyte in albumin SDS-PAGE, blotted onto polyvinylidene uoride membranes, and probed filtration. Nat Rev Nephrol 9: 328–336, 2013 with polyclonal rabbit anti-V5 antibody or anti-FLAG antibody (1:2000 2. Shankland SJ: The podocyte’s response to injury: Role in proteinuria dilution). Protein bands were visualized with enhanced chemilumines- and glomerulosclerosis. Kidney Int 69: 2131–2147, 2006 cence. Densitometry of raw band intensities was performed using the 3. Kestilä M, Lenkkeri U, Männikkö M, Lamerdin J, McCready P, Putaala H, ImageJ/FIJI suite. Statistics were performed as two-tailed t tests. Ruotsalainen V, Morita T, Nissinen M, Herva R, Kashtan CE, Peltonen L, Holmberg C, Olsen A, Tryggvason K: Positionally cloned gene for a novel glomerular protein—nephrin—is mutated in congenital ne- Blue Native Gels phrotic syndrome. Mol Cell 1: 575–582, 1998 A3%–13% BN-PAGE was performed as described by Wittig et al.37 4. Boute N, Gribouval O, Roselli S, Benessy F, Lee H, Fuchshuber A, with the following modifications. We solubilized 100 mg whole cells Dahan K, Gubler MC, Niaudet P, Antignac C: NPHS2, encoding the glomerular protein podocin, is mutated in autosomal recessive steroid- in solubilization buffer (50 mM NaCl, 5 mM 6-aminohexanoic acid, resistant nephrotic syndrome. Nat Genet 24: 349–354, 2000 50 mM imidazole/HCl pH 7.0, 50 mM KPi buffer pH 7.4, 10% 5. Kim JM, Wu H, Green G, Winkler CA, Kopp JB, Miner JH, Unanue ER, glycerol) with 4 g/g protein digitonin (1.0% w/v) or 2.5 g/g protein Shaw AS: CD2-associated protein haploinsufficiency is linked to glo- dodecylmaltoside (0.625% w/v), respectively. Digitonin solubilized merular disease susceptibility. Science 300: 1298–1300, 2003 crude yeast mitochondria were used as a molecular weight marker.63 Gels 6. Winn MP, Conlon PJ, Lynn KL, Farrington MK, Creazzo T, Hawkins AF, fl Daskalakis N, Kwan SY, Ebersviller S, Burchette JL, Pericak-Vance MA, were blotted on polyvinylidene uoride membrane and decorated with Howell DN, Vance JM, Rosenberg PB: A mutation in the TRPC6 cation anti-FLAG M2 antibody (F1804; Sigma-Aldrich). Expression of channel causes familial focal segmental glomerulosclerosis. Science proteins was confirmed by 10% Tris-tricine SDS-PAGE.64 308: 1801–1804, 2005

1520 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 www.jasn.org BASIC RESEARCH

7. Reiser J, Polu KR, Möller CC, Kenlan P, Altintas MM, Wei C, Faul C, aminonucleoside nephrosis and neonatal kidney. Kidney Int 68: 542– Herbert S, Villegas I, Avila-Casado C, McGee M, Sugimoto H, Brown D, 551, 2005 Kalluri R, Mundel P, Smith PL, Clapham DE, Pollak MR: TRPC6 is a 25. Sistani L, Dunér F, Udumala S, Hultenby K, Uhlen M, Betsholtz C, glomerular slit diaphragm-associated channel required for normal renal Tryggvason K, Wernerson A, Patrakka J: Pdlim2 is a novel actin-regulating function. Nat Genet 37: 739–744, 2005 protein of podocyte foot processes. Kidney Int 80: 1045–1054, 2011 8. Schermer B, Benzing T: Lipid-protein interactions along the slit di- 26. Sekine Y, Nishibori Y, Akimoto Y, Kudo A, Ito N, Fukuhara D, Kurayama R, aphragm of podocytes. JAmSocNephrol20: 473–478, 2009 Higashihara E, Babu E, Kanai Y, Asanuma K, Nagata M, Majumdar A, 9. Huber TB, Schermer B, Müller RU, Höhne M, Bartram M, Calixto A, Tryggvason K, Yan K: Amino acid transporter LAT3 is required for Hagmann H, Reinhardt C, Koos F, Kunzelmann K, Shirokova E, podocyte development and function. J Am Soc Nephrol 20: 1586– Krautwurst D, Harteneck C, Simons M, Pavenstädt H, Kerjaschki D, 1596, 2009 Thiele C, Walz G, Chalfie M, Benzing T: Podocin and MEC-2 bind 27. Obenauer JC, Cantley LC, Yaffe MB: Scansite 2.0: Proteome-wide cholesterol to regulate the activity of associated ion channels. Proc Natl prediction of cell signaling interactions using short sequence motifs. Acad Sci U S A 103: 17079–17086, 2006 Nucleic Acids Res 31: 3635–3641, 2003 10. Huber TB, Hartleben B, Kim J, Schmidts M, Schermer B, Keil A, Egger L, 28. Machuca E, Hummel A, Nevo F, Dantal J, Martinez F, Al-Sabban E, Lecha RL, Borner C, Pavenstädt H, Shaw AS, Walz G, Benzing T: Baudouin V, Abel L, Grünfeld JP, Antignac C: Clinical and epidemio- Nephrin and CD2AP associate with phosphoinositide 3-OH kinase and logical assessment of steroid-resistant nephrotic syndrome associated stimulate AKT-dependent signaling. Mol Cell Biol 23: 4917–4928, 2003 with the NPHS2 R229Q variant. Kidney Int 75: 727–735, 2009 11. Hartleben B, Schweizer H, Lübben P, Bartram MP, Möller CC, Herr R, 29. Stenson PD, Mort M, Ball EV, Howells K, Phillips AD, Thomas NS, Wei C, Neumann-Haefelin E, Schermer B, Zentgraf H, Kerjaschki D, Cooper DN: The Human Gene Mutation Database: 2008 update. Ge- Reiser J, Walz G, Benzing T, Huber TB: Neph-Nephrin proteins bind the nome Med 1: 13, 2009 Par3-Par6-atypical protein kinase C (aPKC) complex to regulate podocyte 30. Tonna SJ, Needham A, Polu K, Uscinski A, Appel GB, Falk RJ, Katz A, Al- cell polarity. J Biol Chem 283: 23033–23038, 2008 Waheeb S, Kaplan BS, Jerums G, Savige J, Harmon J, Zhang K, Curhan 12. Benzing T: Signaling at the slit diaphragm. J Am Soc Nephrol 15: 1382– GC, Pollak MR: NPHS2 variation in focal and segmental glomerulo- 1391, 2004 sclerosis. BMC Nephrol 9: 13, 2008 13. George B, Holzman LB: Signaling from the podocyte intercellular 31. Abid A, Khaliq S, Shahid S, Lanewala A, Mubarak M, Hashmi S, Kazi J, junction to the actin cytoskeleton. Semin Nephrol 32: 307–318, 2012 Masood T, Hafeez F, Naqvi SAA, Rizvi SAH, Mehdi SQ: A spectrum of 14. Höhne M, Ising C, Hagmann H, Völker LA, Brähler S, Schermer B, novel NPHS1 and NPHS2 gene mutations in pediatric nephrotic syn- Brinkkoetter PT, Benzing T: Light microscopic visualization of podocyte drome patients from Pakistan. Gene 502: 133–137, 2012 ultrastructure demonstrates oscillating glomerular contractions. Am J 32. Beltrao P, Albanèse V, Kenner LR, Swaney DL, Burlingame A, Villén J, Pathol 182: 332–338, 2013 Lim WA, Fraser JS, Frydman J, Krogan NJ: Systematic functional pri- 15. Yuan H, Takeuchi E, Salant DJ: Podocyte slit-diaphragm protein oritization of protein posttranslational modifications. Cell 150: 413– nephrin is linked to the actin cytoskeleton. Am J Physiol Renal Physiol 425, 2012 282: F585–F591, 2002 33. Brand J, Smith ESJ, Schwefel D, Lapatsina L, Poole K, OmerbašicD, 16. Boerries M, Grahammer F, Eiselein S, Buck M, Meyer C, Goedel M, Kozlenkov A, Behlke J, Lewin GR, Daumke O: A stomatin dimer mod- Bechtel W, Zschiedrich S, Pfeifer D, Laloë D, Arrondel C, Gonçalves S, ulates the activity of acid-sensing ion channels. EMBO J 31: 3635– Krüger M, Harvey SJ, Busch H, Dengjel J, Huber TB: Molecular fin- 3646, 2012 gerprinting of the podocyte reveals novel gene and protein regulatory 34. Wang Y, Geer LY, Chappey C, Kans JA, Bryant SH: Cn3D: Sequence networks. Kidney Int 83: 1052–1064, 2013 and structure views for Entrez. Trends Biochem Sci 25: 300–302, 2000 17. Hornbeck PV, Kornhauser JM, Tkachev S, Zhang B, Skrzypek E, Murray 35. Brooks BR, Brooks CL 3rd, Mackerell AD Jr, Nilsson L, Petrella RJ, Roux B, Latham V, Sullivan M: PhosphoSitePlus: A comprehensive resource B, Won Y, Archontis G, Bartels C, Boresch S, Caflisch A, Caves L, Cui Q, for investigating the structure and function of experimentally de- Dinner AR, Feig M, Fischer S, Gao J, Hodoscek M, Im W, Kuczera K, termined post-translational modifications in man and mouse. Nucleic Lazaridis T, Ma J, Ovchinnikov V, Paci E, Pastor RW, Post CB, Pu JZ, Acids Res 40[Database issue]: D261–D270, 2012 Schaefer M, Tidor B, Venable RM, Woodcock HL, Wu X, Yang W, York 18. Villén J, Beausoleil SA, Gerber SA, Gygi SP: Large-scale phosphorylation DM, Karplus M: CHARMM: The biomolecular simulation program. analysis of mouse liver. Proc Natl Acad Sci U S A 104: 1488–1493, 2007 J Comput Chem 30: 1545–1614, 2009 19. Macek B, Mann M, Olsen JV: Global and site-specific quantitative 36. Brooks BR, Bruccoleri RE, Olafson BD, States DJ, Swaminathan S, phosphoproteomics: Principles and applications. Annu Rev Pharmacol Karplus M: CHARMM: A program for macromolecular energy, mini- Toxicol 49: 199–221, 2009 mization, and dynamics calculations. J Comput Chem 4: 187–217, 1983 20. Rinschen MM, Yu M-J, Wang G, Boja ES, Hoffert JD, Pisitkun T, 37. Wittig I, Braun H-P, Schägger H: Blue native PAGE. Nat Protoc 1: 418– Knepper MA: Quantitative phosphoproteomic analysis reveals vaso- 428, 2006 pressin V2-receptor-dependent signaling pathways in renal collecting 38. Bauer M, Pelkmans L: A new paradigm for membrane-organizing and duct cells. Proc Natl Acad Sci U S A 107: 3882–3887, 2010 -shaping scaffolds. FEBS Lett 580: 5559–5564, 2006 21. Huttlin EL, Jedrychowski MP, Elias JE, Goswami T, Rad R, Beausoleil SA, 39. Völker LA, Schurek EM, Rinschen MM, Tax J, Schutte BA, Lamkemeyer Villén J, Haas W, Sowa ME, Gygi SP: A tissue-specificatlasofmouse T, Ungrue D, Schermer B, Benzing T, Höhne M: Characterization of a protein phosphorylation and expression. Cell 143: 1174–1189, 2010 short isoform of the kidney protein podocin in human kidney. BMC 22. Zhao B, Pisitkun T, Hoffert JD, Knepper MA, Saeed F: CPhos: A pro- Nephrol 14: 102, 2013 gram to calculate and visualize evolutionarily conserved functional 40. Hartleben B, Widmeier E, Suhm M, Worthmann K, Schell C, phosphorylation sites. Proteomics 12: 3299–3303, 2012 Helmstädter M, Wiech T, Walz G, Leitges M, Schiffer M, Huber TB: 23. Beck LH Jr, Bonegio RGB, Lambeau G, Beck DM, Powell DW, Cummins aPKCl/i and aPKCz contribute to podocyte differentiation and glo- TD, Klein JB, Salant DJ: M-type phospholipase A2 receptor as target merular maturation. J Am Soc Nephrol 24: 253–267, 2013 antigen in idiopathic membranous nephropathy. NEnglJMed361: 41. Hoffert JD, Pisitkun T, Wang G, Shen RF, Knepper MA: Quantitative 11–21, 2009 phosphoproteomics of vasopressin-sensitive renal cells: Regulation of 24. Yaoita E, Kurihara H, Yoshida Y, Inoue T, Matsuki A, Sakai T, Yamamoto aquaporin-2 phosphorylation at two sites. Proc Natl Acad Sci U S A 103: T: Role of Fat1 in cell-cell contact formation of podocytes in puromycin 7159–7164, 2006

J Am Soc Nephrol 25: 1509–1522, 2014 Phosphoproteomic Analysis of Murine Glomeruli 1521 BASIC RESEARCH www.jasn.org

42. Bansal AD, Hoffert JD, Pisitkun T, Hwang S, Chou CL, Boja ES, Wang processes are highly tyrosine phosphorylated in the normal rat kidney. G, Knepper MA: Phosphoproteomic profiling reveals vasopressin- Nephrol Dial Transplant 25: 1785–1795, 2010 regulated phosphorylation sites in collecting duct. JAmSoc 55. Gu G, Caldwell GA, Chalfie M: Genetic interactions affecting touch Nephrol 21: 303–315, 2010 sensitivity in Caenorhabditis elegans. Proc Natl Acad Sci U S A 93: 43. Gunaratne R, Braucht DWW, Rinschen MM, Chou CL, Hoffert JD, 6577–6582, 1996 Pisitkun T, Knepper MA: Quantitative phosphoproteomic analysis re- 56. Anderson M, Kim EY, Hagmann H, Benzing T, Dryer SE: Opposing ef- veals cAMP/vasopressin-dependent signaling pathways in native renal fects of podocin on the gating of podocyte TRPC6 channels evoked by thick ascending limb cells. Proc Natl Acad Sci U S A 107: 15653–15658, membrane stretch or diacylglycerol. Am J Physiol Cell Physiol 305: 2010 C276–C289, 2013 44. Feric M, Zhao B, Hoffert JD, Pisitkun T, Knepper MA: Large-scale 57. Kim EY, Anderson M, Wilson C, Hagmann H, Benzing T, Dryer SE: phosphoproteomic analysis of membrane proteins in renal proximal NOX2 interacts with podocyte TRPC6 channels and contributes to their and distal tubule. Am J Physiol Cell Physiol 300: C755–C770, 2011 activation by diacylglycerol: Essential role of podocin in formation of 45. Nedvetsky PI, Tabor V, Tamma G, Beulshausen S, Skroblin P, Kirschner this complex. Am J Physiol Cell Physiol 305: C960–C971, 2013 A, Mutig K, Boltzen M, Petrucci O, Vossenkämper A, Wiesner B, 58. Kelstrup CD, Young C, Lavallee R, Nielsen ML, Olsen JV: Optimized fast Bachmann S, Rosenthal W, Klussmann E: Reciprocal regulation of and sensitive acquisition methods for shotgun proteomics on a quad- aquaporin-2 abundance and degradation by protein kinase A and p38- rupole orbitrap mass spectrometer. JProteomeRes11: 3487–3497, MAP kinase. J Am Soc Nephrol 21: 1645–1656, 2010 2012 46. Saritas T, Borschewski A, McCormick JA, Paliege A, Dathe C, Uchida S, 59. Michalski A, Damoc E, Hauschild J-P, Lange O, Wieghaus A, Makarov Terker A, Himmerkus N, Bleich M, Demaretz S, Laghmani K, Delpire E, A, Nagaraj N, Cox J, Mann M, Horning S: Mass spectrometry-based Ellison DH, Bachmann S, Mutig K: SPAK differentially mediates vaso- proteomics using Q Exactive, a high-performance benchtop quadrupole pressin effects on sodium cotransporters. J Am Soc Nephrol 24: 407– Orbitrap mass spectrometer. Mol Cell Proteomics 10: M111.011015, 2011 418, 2013 60. Cox J, Mann M: 1D and 2D annotation enrichment: A statistical method 47. Moeller HB, Praetorius J, Rützler MR, Fenton RA: Phosphorylation of integrating quantitative proteomics with complementary high-throughput aquaporin-2 regulates its endocytosis and protein-protein interactions. data. BMC Bioinformatics 13[Suppl 16]: S12, 2012 Proc Natl Acad Sci U S A 107: 424–429, 2010 61. Tchapyjnikov D, Li Y, Pisitkun T, Hoffert JD, Yu M-J, Knepper MA: 48. Asanuma K, Kim K, Oh J, Giardino L, Chabanis S, Faul C, Reiser J, Proteomic profiling of nuclei from native renal inner medullary col- Mundel P: Synaptopodin regulates the actin-bundling activity of alpha- lecting duct cells using LC-MS/MS. Physiol Genomics 40: 167–183, actinin in an isoform-specific manner. JClinInvest115: 1188–1198, 2010 2005 62. Wu X, Brooks BR: Force-momentum-based self-guided Langevin dy- 49. Kaplan JM, Kim SH, North KN, Rennke H, Correia LA, Tong HQ, Mathis namics: A rapid sampling method that approaches the canonical en- BJ, Rodríguez-Pérez JC, Allen PG, Beggs AH, Pollak MR: Mutations in semble. J Chem Phys 135: 204101, 2011 ACTN4, encoding alpha-actinin-4, cause familial focal segmental glo- 63. Tatsuta T, Model K, Langer T: Formation of membrane-bound ring merulosclerosis. Nat Genet 24: 251–256, 2000 complexes by prohibitins in mitochondria. MolBiolCell16: 248–259, 50. Bousquet SM, Monet M, Boulay G: The serine 814 of TRPC6 is phos- 2005 phorylated under unstimulated conditions. PLoS ONE 6: e18121, 2011 64. Schägger H, von Jagow G: Tricine-sodium dodecyl sulfate-polyacrylamide 51. Verma R, Kovari I, Soofi A, Nihalani D, Patrie K, Holzman LB: Nephrin gel electrophoresis for the separation of proteins in the range from 1 to ectodomain engagement results in Src kinase activation, nephrin 100 kDa. Anal Biochem 166: 368–379, 1987 phosphorylation, Nck recruitment, and actin polymerization. JClinIn- 65. Chou MM, Hou W, Johnson J, Graham LK, Lee MH, Chen CS, Newton vest 116: 1346–1359, 2006 AC, Schaffhausen BS, Toker A: Regulation of protein kinase C zeta by PI 52. Verma R, Wharram B, Kovari I, Kunkel R, Nihalani D, Wary KK, Wiggins 3-kinase and PDK-1. Curr Biol 8: 1069–1077, 1998 RC, Killen P, Holzman LB: Fyn binds to and phosphorylates the kidney 66. Wang G, Wu WW, Pisitkun T, Hoffert JD, Knepper MA, Shen R-F: slit diaphragm component Nephrin. JBiolChem278: 20716–20723, Automated quantification tool for high-throughput proteomics us- 2003 ing stable isotope labeling and LC-MSn. Anal Chem 78: 5752–5761, 53. Venkatareddy M, Cook L, Abuarquob K, Verma R, Garg P: Nephrin 2006 regulates lamellipodia formation by assembling a protein complex that includes Ship2, filamin and lamellipodin. PLoS ONE 6: e28710, 2011 54. Zhang Y, Yoshida Y, Nameta M, Xu B, Taguchi I, Ikeda T, Fujinaka H, Magdeldin S, Tsukaguchi H, Harita Y, Yaoita E, Yamamoto T: Glomer- This article contains supplemental material online at http://jasn.asnjournals. ular proteins related to slit diaphragm and matrix adhesion in the foot org/lookup/suppl/doi:10.1681/ASN.2013070760/-/DCSupplemental.

1522 Journal of the American Society of Nephrology J Am Soc Nephrol 25: 1509–1522, 2014 Supplemental material

Supplemental figures

Supplemental Fig. S1: Characterization of murine glomerular preparation. (A) Glomerular lysates (Glom) and cortex kidney lysates (remnants of the glomerular preparation) were subjected to immunoblotting and membranes were probed with antibodies against Nephrin and Podocin as glomerular markers as well as NaKATPase (alpha subunit) and the Tamm-Horsefall-glycoprotein (THP). (B) Glomeruli isolated by magnetic beads were lysed in 8M Urea and subjected to three consecutive sonication steps. The urea unsoluble pellet was sonicated and dissolved in 2% SDS buffer. Same volumes of lysates were analyzed for content of Nephrin and Podocin.

Supplemental Fig. S2: Characterization of phosphorylation datasets. (A) Comparison of distribution of serines, threonines and tyrosin phosphorylation sites. (B) Bar graph depiciting the number of phosphorylation sites per protein.

Supplemental Fig. S3: Results of conservational analysis of detected phosphorylation sites A. Automated batch-alignment was performed using the cPhos software and phosphosites were evaluated in terms of conservation of phosphorylation sites and phosphorylation residues. The results show the conservation score of phosphorylation sites and surrounding motifs. (B, C) Examples for high or low conserved phosphorylation sites and motifs in the phosphoproteomic dataset. 9 vertebrate species of the respective homologene group are listed. The central residue (marked with yellow) shows the phosphorylation site. (B) Src pY424 is a highly conserved site (site conservation score = 1.0, motif conservation score = 0.92) (C) Akap1 pS55 is a low conserved phosphorylation sites (site conservation score = 0.65, motif conservation score = 0.66). Residues are colour-coded according to their electrostatic and structural properties (red= negatively charged residues [D;E]; blue: positively charge residues [R;K]; green: prolines)

Supplemental Fig. S4: Modeling and free energy simulation of PHB domain interaction. (A) Modeling of the effect of phosphorylation on Stomatin PHB domain conformation. The crystal structure PDB ID 4FVF was chosen as a reference for this simulation. (B) Construction of homologous model of Podocin based on the stomatin PHB domain dimers and analysis of the effect of phosphorylation on PHB domain conformation. (C) Thermodynamic cycle of Podocin PHB domain phosphorylation and dimerization. The free energy G denotes the effect of phosphorylation on the dimer and can be simulated based on the given equation. (D) Free energy simulation of dimers was performed in a 93.31×62.21×46.66 Å3 sized periodic box of water to analyze the effect of phosphorylation on the dimerization. (E) Monomer was extracted from dimer structure and was dissolved in a 77.758×46.655×46.655 Å3 sized periodic box of water for the free energy simulation of phosphorylation.

Fig. S5: Further characterization of Podocin phosphorylation at T234. Phosphoproteomic analysis of Podocin PHB-domain phosphorylation at T234 in the presence of an inactive (T410A) or constitutively active (T410E) protein kinase C zeta (rPKCz). A representative reconstructed ion chromatogram of the MS1 precursor (“ion current”) is shown.

Supplemental tables

Supplemental Table S1: Tyrosine phosphorylation sites on glomerular proteins

This table depicts high-confidence tyrosin phosphorylation sites on glomerular proteins. In the modified sequence column the original sequence of the phosphopeptide is depicted. “*” denotes serine, threonine and tyrosin phosphorylation whereas “#” denotes (methionin-) oxidation.

Gene UniProt symbol Protein Name ID Residue Phosphopeptide sequence Abhd14a Abhydrolase domain-containing Q922Q6 Y59 IFY*REVLPIQQACR protein 14A Abi3 ABI gene family member 3 Q8BYZ1 Y318 VVT*LY*PY*T*RQK Abi3 ABI gene family member 3 Q8BYZ1 Y320 VVT*LY*PY*T*RQK Ank3 ankyrin-3 G5E8K5 Y1461 KRY*SYLTEPSMSPQS*PCER Ankrd61 Ankyrin repeat domain-containing Q9CQM6 Y315 NGES*PIY*M#Y*LQR protein 61 Ankrd61 Ankyrin repeat domain-containing Q9CQM6 Y317 NGES*PIY*M#Y*LQR protein 61 Arhgap35 Rho GTPase-activating protein 35 Q91YM2 Y1105 NEEENIY*SVPHDSTQGK Arntl2 Aryl hydrocarbon receptor nuclear Q2VPD4 Y258 KFHT*VHCT*GY*LR translocator-like protein 2 Bcar1 Breast cancer anti-estrogen Q61140 Y310 GLLSSSHHSVY*DVPPSVSK resistance protein 1 Bckdha 2-oxoisovalerate dehydrogenase Q3U3J1 Y346 IGHHS*TSDDSSAY*RSVDEVNYWDKQDHPIS subunit alpha, mitochondrial R Brca2 Breast cancer type 2 P97929 Y581 FIY*S*VS*DDAS*LQGK susceptibility protein homolog Bzrap1 Peripheral-type benzodiazepine Q7TNF8 Y1766 T*S*RPM#VAAFDY*NPR receptor-associated protein 1 Carkd ATP-dependent (S)-NAD(P)H- K3W4M4 Y103 IGIVGGCQEY*TGAPYFAGISALK hydrate dehydratase Ccdc105 Coiled-coil domain-containing Q9D4K7 Y372 FNQEM#Y*VTRGIIK protein 105 Ccdc147 coiled-coil domain containing 147 B2RW38 Y245 QAEIKAM#QQY*MHKSK Ccdc158 Coiled-coil domain-containing Q8CDI6 Y106 FY*LRQSVIDLQTK protein 158 Ccdc39 Coiled-coil domain-containing Q9D5Y1 Y642 Y*EILT*VVM#LPPEGEEEK protein 39 Cdk2 Cyclin-dependent kinase 2 P97377 Y15 IGEGTY*GVVYK Cntnap1 Contactin-associated protein 1 O54991 Y105 YMLLY*GDR Ctnnd1 Catenin delta-1 E9Q8Z8 Y96 LNGPQDHNHLLY*STIPR Ctnnd1 Catenin delta-1 E9Q8Z8 Y280 FHPEPY*GLEDDQRSMGYDDLDYGMM#SDY GTAR Ctnnd1 Catenin delta-1 E9Q8Z8 Y865 SQSSHSY*DDSTLPLIDR Cyth2 Cytohesin-2 D3YU95 Y391 DEWIKS*IQAAVS*VDPFY*EM#LAAR Dock3 Dedicator of cytokinesis protein 3 F8VPQ1 Y1203 LMERLLDY*RDCM#K Dpy19l3 Protein dpy-19 homolog 3 Q71B07 Y612 TQAVY*QIY*AKR Dpy19l3 Protein dpy-19 homolog 3 Q71B07 Y615 TQAVY*QIY*AKR Drg2 Developmentally-regulated GTP- Q9QXB9 Y332 Y*ALVWGT*S*TKYS*PQR binding protein 2 Dyrk1a Dual specificity tyrosine- Q61214 Y321 IYQY*IQSR phosphorylation-regulated kinase 1A Dyrk4 Dual specificity tyrosine- Q8BI55 Y379 VYTY*IQSR phosphorylation-regulated kinase 4 Eef1a1 Elongation factor 1-alpha 1 P10126 Y418 SGDAAIVDMVPGKPMCVESFSDY*PPLGR Epb4.1 Protein 4.1 A2A841 Y665 ERLDGENIY*IR Epb41l3 Band 4.1-like protein 3 Q9WV92 Y479 DSVSAAEVGTGQY*ATTK Erbb2ip Protein LAP2 B7ZNX6 Y1097 RTEGDY*LSYR

Fam180a Protein FAM180A Q8BR21 Y135 EDFERIT*LTLAYT*AY*R Fat1 FAT tumor suppressor homolog 1 F2Z4A3 Y3062 SNAEIT*Y*T*LFGSGAEK precursor Fgd1 FYVE, RhoGEF and PH domain- P52734 Y93 ALRFS*Y*HLEGSQPR containing protein 1 Gab1 GRB2-associated-binding protein Q9QYY0 Y660 SSGSGSSM#ADERVDY*VVVDQQK 1 Gm4955 G3UZV2 Y20 MVNY*YKQIVLLSGLEYM#NDY*NFR Protein Gm4955 Gm6434 F7CXS7 Y67 PY*VNY*RAPAGIDT*PLT*VK Protein Gm6434 Gm6434 F7CXS7 Y70 PY*VNY*RAPAGIDT*PLT*VK Protein Gm6434 Gm8298 uncharacterized protein J3QPI0 Y235 EY*KHGPFLT*R LOC666803 precursor Gpx1 Glutathione peroxidase 1 P11352 Y147 Y*IIWSPVCR Grm7 Metabotropic glutamate receptor G5E8D5 Y492 NGDAPGRYDIFQYQTT*NT*TNPGY*R 7 Gstt4 Glutathione S-transferase theta-4 Q9D4P7 Y6 GLELY*M#DLLS*APCR Hes1 Transcription factor HES-1 P35428 Y110 AQM#TAALSTDPSVLGKY*R Hipk3 Homeodomain-interacting protein A2AQH3 Y359 TVCSTY*LQSR kinase 3 Idh3b isocitrate dehydrogenase 3, beta Q91VA7 Y120 VAIIGKIYT*PM#EY*K subunit Ifitm3 Interferon-induced Q9CQW9 Y27 IKEEY*EVAEMGAPHGSASVR transmembrane protein 3 Ipo13 Importin-13 Q8K0C1 Y735 M#Y*ST*VPQAS*ALDLTR Kdr Vascular endothelial grow th factor P35918 Y1212 FHY*DNTAGISHYLQNSK receptor 2 Kiaa1731 Centrosomal protein KIAA1731 Q8BQ48 Y586 LLEY*QT*VLK homolog Kif23 kinesin family member 23 E9Q5G3 Y380 TCM#EVLRENQT*Y*GT*NK Lyn Tyrosine-protein kinase Lyn P25911 Y194 SLDNGGYY*ISPR Mapk1 Mitogen-activated protein kinase P63085 Y185 VADPDHDHTGFLTEY*VATR 1 Mapk12 Mitogen-activated protein kinase O08911 Y185 QADSEM#TGY*VVTR 12 Mapk3 Mitogen-activated protein kinase D3Z3G6 Y205 IADPEHDHTGFLTEY*VATR 3 Mbp myelin basic protein F6RT34 Y99 PGLCHMY*KDS*HTR Mmrn1 Multimerin-1 B2RPV6 Y1205 YPPVTTFS*GY*LLY*R Mmrn1 Multimerin-1 B2RPV6 Y1208 YPPVTTFS*GY*LLY*R Mrps22 28S ribosomal protein S22, Q9CXW2 Y102 T*FRPAIQPLKPPT*Y*K mitochondrial Mrps35 28S ribosomal protein S35, Q8BJZ4 Y289 EELLGTKEVEDY*QK mitochondrial Myh10 Myosin-10 Q3UH59 Y680 IVGLDQVT*GMT*ET*AFGSAY*K Myo10 Unconventionnal myosin-X F8VQB6 Y1482 LY*TKLLNEAT*RWSSAIQNVT*DT*K Myo1a Unconventional myosin-Ia O88329 Y585 NLYS*KNPNY*IRCIK Ncf2 Neutrophil cytosol factor 2 O70145 Y351 ELKLS*VPM#PY*M#LK Nedd9 Enhancer of filamentation 1 O35177 Y344 ANPEERDGVY*DVPLHNPADAK Olfr318 olfactory receptor 318 Q5NCD7 Y123 Y*IAICQPLHY*PVLM#S*RK Olfr318 olfactory receptor 318 Q5NCD7 Y132 Y*IAICQPLHY*PVLM#S*RK Papss2 Bifunctional 3'-phosphoadenosine O88428 Y21 STNVVY*QAHHVSR 5'-phosphosulfate synthase 2 Pdk1 [Pyruvate dehydrogenase Q8BFP9 Y163 NRHNDVIPT*M#AQGVT*EY*K [lipoamide]] kinase isozyme 1, mitochondrial Polr2a DNA -directed RNA polymerase II P08775 Y413 RGNS*QY*PGAKY*IIR subunit RPB1 Polr2a DNA -directed RNA polymerase II P08775 Y418 RGNS*QY*PGAKY*IIR subunit RPB1 Prpf4b Serine/threonine-protein kinase Q61136 Y849 LCDFGSASHVADNDITPY*LVSR PRP4 homolog Ptk2 Focal adhesion kinase 1 P34152 Y614 YMEDSTY*YK Ptp4a2 Protein tyrosine phosphatase O70274 Y11 PAPVEISY*ENM#RFLITHNPT*NAT*LNK type IVA 2

Pxn Paxillin F8VQ28 Y118 AGEEEHVY*SFPNK Sat1 Diamine acetyltransferase 1 A2BES2 Y27 LIKELAKY*EY*M#EDQVILT*EK Sat1 Diamine acetyltransferase 1 A2BES2 Y29 LIKELAKY*EY*M#EDQVILT*EK Shh Sonic hedgehog protein Q62226 Y45 LT*PLAY*KQFIPNVAEKT*LGAS*GR Slc22a1 Solute carrier family 22 member 1 O08966 Y545 AKENTIY*LQVQTGK Slc6a20a Sodium- and chloride-dependent Q8VDB9 Y204 VVY*FTAS*M#PY*CVLIIYLVR transporter XTRP3A Sorbs1 Sorbin and SH3 domain- Q62417 Y325 SIY*EYQPGK containing protein 1 Sorcs1 VPS10 domain-containing E9QQ02 Y495 T*FIT*Y*NK receptor SorCS1 Spsb1 SPRY domain-containing SOCS Q9D5L7 Y19 T*VDM#RDPT*Y*R box protein 1 Src Neuronal proto-oncogene P05480 Y424 LIEDNEY*TAR tyrosine-protein kinase Src Stat3 Signal transducer and activator of P42227 Y705 YCRPESQEHPEADPGSAAPY*LK transcription 3 Stk38 Serine/threonine-protein kinase Q91VJ4 Y115 DT*GHVY*AM#KILR 38 Susd1 sushi domain containing 1 E9Q3H4 Y685 LY*YGEY*YNAPLKTGNEY*CILLR precursor Susd1 sushi domain containing 1 E9Q3H4 Y696 LY*YGEY*YNAPLKTGNEY*CILLR precursor Tex2 Testis-expressed sequence 2 Q6ZPJ0 Y234 QESDTVS*Y*KPPDSK protein Tjp2 Tight junction protein ZO-2 Q9Z0U1 Y1095 IEIAQKHPDIY*AVPIK Tln1 Talin-1 P26039 Y26 TMQFEPSTMVY*DACR Tln1 Talin-1 P26039 Y70 ALDY*YM#LR Tmc8 Transmembrane channel-like B0QZP7 Y45 M#LPY*AM#ADKRFIR protein 8 Tnk2 Activated CDC42 kinase 1 G3X9X7 Y874 VS*ST*HY*Y*LLPER Tnk2 Activated CDC42 kinase 1 G3X9X7 Y875 VS*ST*HY*Y*LLPER Tns1 tensin 1 E9Q0S6 Y1480 HAAYGGY*STPEDR Tns1 tensin 1 E9Q0S6 Y1558 AGSLPNY*ATINGK Trrap transformation/transcription E9PWT1 Y892 AELM#QALWRT*LRNPADSISHVAY*R domain-associated protein Ttll1 Probable tubulin polyglutamylase Q91V51 Y268 WTVNNLRLY*LES*TRGR TTLL1 Ttyh2 Protein tw eety homolog 2 Q3TH73 Y494 Y*ENVPLIGR Vcl Vinculin Q64727 Y822 SFLDSGY*R Vmn1r202 vomeronasal 1 receptor 202 Q8R259 Y218 RVLY*LHS*S*R Vmn2r74 vomeronasal receptor Vmn2r74 E9PW21 Y79 HLIFS*VY*LALEEINK precursor Yes1 Tyrosine-protein kinase Yes Q04736 Y192 GAY*SLSIR Yes1 Tyrosine-protein kinase Yes Q04736; Y424 LIEDNEY*TAR P05480 Zdhhc5 Palmitoyltransferase ZDHHC5 Q8VDZ4 Y533 LLPTGPPHREPS*PVRY*DNLSR 2310030G0 Uncharacterized protein C11orf52 Q9D8L0 Y79 TSNSTSEESDLHY*ADIHVLR 6Rik homolog

Supplemental Table S2.

Gene ontology terms enriched and deenriched in phosphoproteins with serine/threonine phosphorylation sites. The five most enriched molecular function and cellular component GO-Term categories are depicted. A phosphoproteomic mouse tissue atlas dataset of serine/threonine phosphorylated proteins by Huttlin et al. was used as a background dataset (Huttlin et al. Cell 2010).

Enrichment Benj. Hoch. Category GO Term factor -lg(P value) FDR Count Molecular guanyl-nucleotide exchange 1.62 5.23 0.00125 22 Function factor activity phospholipid binding 1.47 6.92 0.00011 36 nucleoside-triphosphatase 1.4 5.63 0.00056 59 regulator activity GTPase regulator activity 1.4 5.63 0.00064 32 enzyme regulator activity 1.39 6.8 0.00008 62 Cellular cell-cell adherens junction 1.86 3.5 0.01667 22 Component lamellipodium 1.8 4.83 0.00207 36 adherens junction 1.78 7.22 0.00005 59 focal adhesion 1.77 4.21 0.00468 32 anchoring junction 1.75 7.16 0.00003 62

Deenriched Enrichment Benj. Hoch. Category GO Term factor -lg(P value) FDR Count Molecular gated channel activity 1.86 4.49 0.01667 15 Function Cellular ion channel complex 1.8 3.75 0.00207 7 Component

Supplemental Table S3.

Gene ontology terms enriched in phosphoproteins with tyrosin phosphorylation sites. A phosphoproteomic mouse tissue atlas dataset of tyrosin phosphorylated proteins by Huttlin et al. was used as a background dataset (Huttlin et al. Cell 2010). No phosphorylated proteins were found to be deenriched.

Enrichment Benj. Hoch. Category GO Term Factor -lg(p-value) FDR count Molecular signal transducer activity 2.87 4.2 0.021 14 function molecular transducer 2.87 4.2 0.041 14 activity Cellular focal adhesion 3.8 3.95 0.017 9 compartment cell-substrate adherens 3.8 3.95 0.025 9 junction cell-substrate junction 3.78 4.32 0.021 10

Supplemental Table S4: phosphorylation sites on podocyte specific proteins. This table outlines phosphorylation sites on proteins shown to be significantly higher expressed in podocytes. These candidate proteins are based on a proteomic study by Boerries et al., Kidney International 2013. In the modified sequence column the original sequence of the phosphopeptide is

depicted. “*” denotes serine, threonine and tyrosin phosphorylation whereas “#” denotes (methionin- )oxidation.

Residue Gene symbol Protein name UniProtID number Modified sequence 0610010K14Rik Chromatin complexes subunit BAP18 D3Z687 S96 VYEDSGIPLPAES*PKKGPK Medium-chain specific acyl-CoA dehydrogenase, Acadm mitochondrial P45952 T351 RNT*YYASIAK Actr3 Actin-related protein 3 Q99JY9 S418 HNPVFGVM#S* Canx Calnexin P35564 S582 AEEDEILNRS*PR QKSDAEEDGVTGS*QDEEDS Canx Calnexin P35564 S563 KPK QKS*DAEEDGVTGSQDEEDS Canx Calnexin P35564 S553 KPK Cd2ap CD2-associated protein Q9JLQ0 S510 FNGGHS*PTQS*PEK Cd2ap CD2-associated protein Q9JLQ0 S580 AEADDGKRNS*VDELR Cd2ap CD2-associated protein Q9JLQ0 S514 FNGGHS*PTQS*PEK Cd2ap CD2-associated protein Q9JLQ0 S458 S*VDLDAFVAR Bifunctional ATP-dependent dihydroxyacetone AS*YISSAQLDQPDPGAVAAA Dak kinase/FAD-AMP lyase (cyclizing) Q8VC30 S545 AIFR Dlg1 Disks large homolog 1 Q811D0 S619 FGDILHVINAS*DDEWWQAR Dlg1 Disks large homolog 1 Q811D0 S573 EQM#M#NSSVSSGS*GSLR Dlg1 Disks large homolog 1 Q811D0 S575 EQM#M#NSSVSSGSGS*LR Dpysl2 Dihydropyrimidinase-related protein 2 O08553 T509 GLYDGPVCEVSVT*PK TVT*PAS*SAKTSPAKQQAPP Dpysl2 Dihydropyrimidinase-related protein 2 O08553 T514 VR Dync1h1 Cytoplasmic dynein 1 heavy chain 1 Q9JHU4 S4366 TDS*TSDGRPAWMR Dync1h1 Cytoplasmic dynein 1 heavy chain 1 Q9JHU4 S1228 RKDS*AIQQQVANLQMK Epb41l5 Band 4.1-like protein 5 Q8BGS1 S581 EDS*LLTHK Epb41l5 Band 4.1-like protein 5 Q8BGS1 S348 FRYS*GKTEYQTTK SWKGDDS*PVANGAEPAGQ Epn3 Epsin-3 Q91W69 S257 R Heterogeneous nuclear ribonucleoprotein U-like AVEEQGDDQDS*EKSKPAGS Hnrnpul2 protein 2 Q00PI9 S183 DGER Heterogeneous nuclear ribonucleoprotein U-like Hnrnpul2 protein 2 Q00PI9 S159 S*GDETPGSEAPGDK Ilk Integrin-linked protein kinase O55222 S186 NGTLNKHS*GIDFK Ilk Integrin-linked protein kinase O55222 S232 S*RDFNEECPR Iqgap2 Ras GTPase-activating-like protein IQGAP2 Q3UQ44 S764 IQSLLRAS*K Iqgap2 Ras GTPase-activating-like protein IQGAP2 Q3UQ44 S1461 S*IKLDGKAEPK Iqgap2 Ras GTPase-activating-like protein IQGAP2 Q3UQ44 S16 YGS*IVDDER VETVNREPLTMHS*DREDDT Kirrel Kin of IRRE-like protein 1 Q80W68 S584 ASISTATR Kirrel Kin of IRRE-like protein 1 Q80W68 S775 FSYTSQHS*DYGQR VETVNREPLTMHSDREDDTA Kirrel Kin of IRRE-like protein 1 Q80W68 T595 S*IST*ATR VETVNREPLTMHS*DREDDT Kirrel Kin of IRRE-like protein 1 Q80W68 S592 AS*ISTATR VETVNREPLTM#HS*DREDD Kirrel Kin of IRRE-like protein 1 Q80W68 T590 T*ASISTATR Cation-dependent mannose-6-phosphate GVGDDQLGEES*EERDDHLL M6pr receptor P24668 S268 PM Membrane-associated guanylate kinase, WW and Magi2 PDZ domain-containing protein 2 Q9WVQ1 S1013 IIPQEELNS*PTSAPSSEK Mapt Microtubule-associated protein A2A5Y6 S510 SGYSSPGS*PGTPGSR Mapt Microtubule-associated protein A2A5Y6 S507 SGYSS*PGS*PGTPGSR AKTDHGAEIVYKS*PVVS*GD Mapt Microtubule-associated protein A2A5Y6 S708 T*SPR TDHGAEIVYKS*PVVS*GDTS* Mapt Microtubule-associated protein A2A5Y6 S704 PR Mapt Microtubule-associated protein A2A5Y6 S506 SGYS*SPGS*PGTPGSR

Mapt Microtubule-associated protein A2A5Y6 T539 VAVVRT*PPKS*PSASK Mapt Microtubule-associated protein A2A5Y6 S570 IGS*TENLKHQPGGGK Mapt Microtubule-associated protein A2A5Y6 S664 IGS*LDNITHVPGGGNKK TTPSPKT*PPGSGEPPKSGE Mapt Microtubule-associated protein A2A5Y6 T489 R Mapt Microtubule-associated protein A2A5Y6 S522 S*RTPS*LPTPPTREPK Nes Nestin Q6P5H2 S728 ERQES*LKSPEEEDQQAFR Nes Nestin Q6P5H2 S816 LVEKES*QES*LKSPEEEDQR Nes Nestin Q6P5H2 S813 LVEKES*QES*LKSPEEEDQR Nes Nestin Q6P5H2 S688 S*PEEDQQAFRPLEK Nes Nestin Q6P5H2 S731 ERQESLKS*PEEEDQQAFR Nes Nestin Q6P5H2 S819 LVEKESQESLKS*PEEEDQR Nphs2 Podocin Q91X05 S362 AQGS*INYPSSSKPVEPLNPK Nphs2 Podocin Q91X05 T234 SLT*EILLER Nphs2 Podocin Q91X05 S382 KKDS*PML Palld Palladin Q9ET54 S901 IAS*DEEIQGTK Palld Palladin Q9ET54 S1146 SRDS*GDENEPIQER Pard3 Partitioning defective 3 homolog B7ZNY3 S906 GRGCNES*FRAAIDK Pard3b Partitioning defective 3 homolog B Q9CSB4 S801 SYDGPEEADADGLS*DKSSR Pard3b Partitioning defective 3 homolog B Q9CSB4 S1088 GGS*ADPVDYLTASPR Pard3b Partitioning defective 3 homolog B Q9CSB4 S575 SMS*MEGNIR Pard3b Partitioning defective 3 homolog B Q9CSB4 S780 GRGCNES*FR Pard3b Partitioning defective 3 homolog B Q9CSB4 S990 DGRPLS*PDHLEGLYAK Pard3b Partitioning defective 3 homolog B Q9CSB4 S730 SDS*PGKDFGPTLGLKK Pard3b Partitioning defective 3 homolog B Q9CSB4 S635 NDSSILYPFGTYS*PQDKRK Pard3b Partitioning defective 3 homolog B Q9CSB4 S346 ASSPEGEEPAS*PQQSK Pard3b Partitioning defective 3 homolog B Q9CSB4 S1182 GSS*PDQYPYR Parva Alpha-parvin Q9EPC1 S28 KKDDS*FLGK Pdlim5 PDZ and LIM domain protein 5 Q8CI51 S228 RGS*QGDIKQQNGPPR AIGGIILT*ASHNPGGPNGDF Pgm2 Phosphoglucomutase-1 A2CEK3 T133 GIK Membrane-associated progesterone receptor Pgrmc2 component 2 Q80UU9 S98 KRDFS*LEQLR Membrane-associated progesterone receptor LLKPGEEPSEYT*DEEDTKDH Pgrmc2 component 2 Q80UU9 T205 SK cAMP-dependent protein kinase type II-alpha RVS*VCAETFNPDEEEEDND Prkar2a regulatory subunit Q8K1M3 S97 PR Ptpro Receptor-type tyrosine-protein phosphatase O E9Q612 T869 ECGAGT*FVNFASLER Rhpn1 Rhophilin-1 E9Q7Q7 S31 KGYGS*FVQNQPGQLQSHR Sec22b Vesicle-trafficking protein SEC22b O08547 S137 NLGS*INTELQDVQR Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r1 RF1 P70441 S285 SAS*SDTSEELNSQDSPKR Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r1 RF1 P70441 S289 SAS*S*DTS*EELNSQDSPK Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r1 RF1 P70441 S337 SS*KRAPQMDWSK Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r1 RF1 P70441 S275 EALVEPASES*PRPALAR Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r1 RF1 P70441 S297 SASSDTSEELNSQDS*PKR Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r1 RF1 P70441 T288 SASSDT*SEELNSQDSPK Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r2 RF2 Q9JHL1 S298 RDPFQES*GLHLSPTAAEAK Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r2 RF2 Q9JHL1 S280 SDLPGS*EKDNEDGSTWK Na(+)/H(+) exchange regulatory cofactor NHE- Slc9a3r2 RF2 Q9JHL1 S303 RDPFQESGLHLS*PTAAEAK

KPSLVSDLPWEGASPQSPSF Sntb2 Beta-2-syntrophin Q61235 S213 SGSEDSGS*PK Sntb2 Beta-2-syntrophin Q61235 S75 GLGPPS*PPAPPR Sntb2 Beta-2-syntrophin Q61235 S238 NLS*MPDLENR Snx1 Sorting nexin-1 Q9WV80 S280 AVGTQALS*GAGLLK Snx1 Sorting nexin-1 Q9WV80 S188 RFS*DFLGLYEK Sord Sorbitol dehydrogenase Q64442 S171 RGSVS*LGNK Stx4 Syntaxin-4 P70452 S15 QGDNIS*DDEDEVR Stx4 Syntaxin-4 P70452 S117 AIEPQKEEADENYNS*VNTR DRAS*PAAAEEAVPEWASCL Synpo Synaptopodin E9Q3E2 S689 KS*PR DRAS*PAAAEEAVPEWASCL Synpo Synaptopodin E9Q3E2 S672 K Synpo Synaptopodin E9Q3E2 S1002 MRS*PQPAS*PAR Synpo Synaptopodin E9Q3E2 S512 AWAPPASSMADRS*PQPQR Synpo Synaptopodin E9Q3E2 S1007 MRS*PQPAS*PAR Synpo Synaptopodin E9Q3E2 S1091 RGS*LPTEASCTT Synpo Synaptopodin E9Q3E2 S535 LLGQRS*PVLER Synpo Synaptopodin E9Q3E2 S1023 S*PLPIGPSSCASPR Synpo Synaptopodin E9Q3E2 S1084 GWNGS*LR Synpo Synaptopodin E9Q3E2 S946 SVS*PLRS*ETEARPPSR Synpo Synaptopodin E9Q3E2 S942 SVS*PLRSETEARPPSR Synpo Synaptopodin E9Q3E2 S1018 GAAFS*PIPR Synpo Synaptopodin E9Q3E2 S134 STS*FTENDLK Synpo Synaptopodin E9Q3E2 S1037 S*PQAAPSRPFPYR Synpo Synaptopodin E9Q3E2 S1051 S*PTDSDVSLDSEDSGLK Synpo Synaptopodin E9Q3E2 S833 AAS*PAKPSSLDLVPNLPR VRS*PPSYSTLYPSSDPKPSH Synpo Synaptopodin E9Q3E2 S567 LK Synpo Synaptopodin E9Q3E2 S258 VAS*EEEEVPLVVYLK Synpo2 Synaptopodin-2 E9Q1U2 S895 AQS*PTPSLPASWK VQIPVSHPDPEPVS*DNEDD Tjp1 Tight junction protein ZO-1 B9EHJ3 S125 S*YDEEVHDPR FNHNLLPSETVHKPELSSKT* Tjp1 Tight junction protein ZO-1 B9EHJ3 T1484 PT*SPK Tjp1 Tight junction protein ZO-1 B9EHJ3 S622 SREDLS*AQPVQTKFPAYER Tjp1 Tight junction protein ZO-1 P39447 S912 IDS*PGLKPASQQK Tjp1 Tight junction protein ZO-2 B9EHJ3 S912 IDS*PGLKPASQQVYR Tjp1 Tight junction protein ZO-1 P39447 S175 S*VASSQPAKPTK Tjp1 Tight junction protein ZO-1 B9EHJ3 S1058 FEEPAPLS*YDSR Tjp1 Tight junction protein ZO-1 P39447 S617 S*REDLSAQPVQTK AVPVS*PSAVEEDEDEDGHT Tjp1 Tight junction protein ZO-1 B9EHJ3 S1534 VVATAR Tjp2 Tight junction protein ZO-2 Q9Z0U1 S968 DAS*PPPAFKPEPPK Tjp2 Tight junction protein ZO-2 Q9Z0U1 Y1095 IEIAQKHPDIY*AVPIK Tjp2 Tight junction protein ZO-2 Q9Z0U1 S895 MSYLTAMGADYLS*CDSR Tjp2 Tight junction protein ZO-2 Q9Z0U1 S404 RQQYS*DQDYHSSTEK Tjp2 Tight junction protein ZO-2 Q9Z0U1 S213 S*IDRDYDRDYER Tjp2 Tight junction protein ZO-2 Q9Z0U1 S1136 GSYGS*DPEEEEYR SYHEAYEPDYGGGYS*PSYD Tjp2 Tight junction protein ZO-2 Q9Z0U1 S239 RR Tjp2 Tight junction protein ZO-2 Q9Z0U1 S395 SFS*PEER Tjp2 Tight junction protein ZO-2 Q9Z0U1 S441 S*TGDITAAGVTEASREPR

Tjp2 Tight junction protein ZO-2 Q9Z0U1 S209 GLDRDFVSRDHS*R Tjp2 Tight junction protein ZO-2 Q9Z0U1 S107 KVQVAPLQGS*PPLSHDDR Tmod3 Tropomodulin-3 Q9JHJ0 S71 LLS*YLEK Tpd52 Tumor protein D52 F8WHQ1 S198 NSPTFKS*FEEKVENLK Tpd52 Tumor protein D52 Q62393 S170 KLEDVKNS*PTFK Tpd52l2 Tumor protein D54 A2AUD5 S189 NSATFKS*FEDR Tpd52l2 Tumor protein D54 A2AUD5 S96 S*WHDVQVSTAYVK GRM#S*MKEVDEQMLNVQN Tubb4b Tubulin beta-4B chain P68372 S322 K Cytochrome b-c1 complex subunit 1, Uqcrc1 mitochondrial Q9CZ13 S212 RLS*RTDLTDYLNR AAQAS*LNALNDPIAVEQALQ Utrn Utrophin E9Q6R7 S933 EK Voltage-dependent anion-selective channel Vdac1 protein 1 Q60932 S117 LTFDSSFS*PNTGKK Vim Vimentin P20152 S73 LRSS*VPGVR Vim Vimentin P20152 S56 SLYSSS*PGGAYVTR Vim Vimentin P20152 S55 SLYSS*SPGGAYVTR Vim Vimentin P20152 S72 LRS*SVPGVR QVQS*LTCEVDALKGTNESL Vim Vimentin P20152 S325 ER Vim Vimentin P20152 S51 S*LYSSSPGGAYVTR

Supplemental Table S5: phosphorylation sites on bona fide podocyte specific proteins. This table outlines phosphorylation sites on proteins known to be almost exclusively expressed in podocytes. The references for the respective hypothesis-driven studies can be found in the main text.

Gene symbol Protein name UniProtID Residue number Modified sequence

Pla2r1 Secretory phospholipase A2 receptor Q62028 S1486 GRPICIS*P

Pdlim2 PDZ and LIM domain protein 2 Q8R1G6 S124 SACFS*PVSLSPR

Pdlim2 PDZ and LIM domain protein 2 Q8R1G6 S199 VLLHS*PGRPSS*PR

Pdlim2 PDZ and LIM domain protein 2 Q8R1G6 S210 FSS*LDLEEDSEVFK

Pdlim2 PDZ and LIM domain protein 2 Q8R1G6 T276 ALAT*PPKLHTCEK

Slc43a1 Large neutral amino acids transporter small subunit 3 G3X8X3 S450 DGAS*TKFTRPR

Nphs1 Nephrin Q9QZS7 S1112 LAEEIS*EKT*EAGS*EEDR

Nphs1 Nephrin Q9QZS7 S1119 TEAGS*EEDRIR

Fat1 FAT tumor suppressor homolog 1 precursor F2Z4A3 Y3062 SNAEIT*Y*T*LFGSGAEK

Fat1 FAT tumor suppressor homolog 1 precursor F2Z4A3 T3063 SNAEIT*Y*T*LFGSGAEK

A Glom Kidney kDa Nephrin

130

Podocin 70

Na K ATPase 100

THP 100

B 8 M Urea Pellet Sonication 2% SDS 1 2 3 kDa Nephrin

130

Podocin 70

Fig. S1 A es 3000

2000

1000 Number of phosphosi t

S T Y B

1200

eins 1000 o t

800

600

400

200 Number of phosphop r 0 1 2 3 4 5 6-10 >10 Number of phosphosites

Fig. S2 A Motif conservation Site conservation high (score: 0.95-1) medium (score: 0.8-0.95) low (score: <0.8)

B Src, pY424 HID: 21120 RefSeq Species Amino acid window NP_033297.2 M. musculus 418 LIEDNEYTARQGA 430 NP_001003837.2 D. rerio 411 LIEDNEYTARQGA 423 NP_990788.2 G. gallus 410 LIEDNEYTARQGA 422 NP_001104274.1 B. taurus 419 LIEDNEYTARQGA 431 XP_865870.1 C. lupus 413 LIEDNEYTARQGA 425 NP_114183.1 R. norvegicus 419 LIEDNEYTARQGA 431 XP_001092180.1 M. mulatta 413 LIEDNEYTARQGA 425 XP_001141228.2 P. troglodytes 413 LIEDNEYTARQGA 425 NP_938033.1 H. sapiens 413 LIEDNEYTARQGA 425

C Akap1, pS55

HID: 31165 RefSeq Species Amino acid window NP_001036006.1 M. musculus 49 IKDRRLSEEACPG 61 NP_001091649.2 D rerio 52 EGSNGMVDKSCPA 64 XP_415913.1 G. gallus 54 NDSPPKTEACVPQ 66 XP_002695633.1 B. taurus 53 AQEVLPVEDSRPG 65 XP_003435322.1 C. lupus 52 IKEPLSMEDPCPA 64 NP_446117.1 R. norvegicus 49 IEDRLPTEEACPG 61 XP_001105339.1 M. mulatta 53 IKEPLPMEDVCPE 65 XP_001172289.1 P. troglodytes 54 IKEPLPMEDVCPK 66 NP_003479.1 H. sapiens 54 IKEPLPVEDVCPK 66

Fig. S3 A B

Stomatin dimer (PDB: 4FVF) Podocin dimer Podocin dimmerer

Phosphoryla on at T234

C G(Dimerization) Podocin (Podocin)2

Gmono (Phosphorylation) Gdimer (Phosphorylation)

Podocin-T234Ph (Podocin-T234Ph)2 Gphos (Dimerization)

G Gphos (Dimerization) - G(Dimerization)

Gdimer (Phosphorylation) - 2 Gmono (Phosphorylation) D E

Fig. S4 SLT*EILLERK (+2): NPHS2, T234

)

6 Podocin +PKCz T410A 0 5.80

(

4.35

s 2.90

e t 1.45

)

6 Podocin +PKCz T410E 0 5.80

( 4.35

s 2.90

e t 1.45

31.8 32.0 32.2 32.4 32.6 32.8 Retention Time Fig. S5