Disruption of the exocyst induces podocyte loss and dysfunction

Although the slit diaphragm proteins in podocytes are uniquely organized to maintain glomerular filtration assembly and function, little is known about the underlying mechanisms that participate in trafficking these proteins to the correct location for development and homeostasis. Identifying these mechanisms will likely provide novel targets for therapeutic intervention to preserve podocyte function following glomerular injury. Analysis of structural variation in cases of human nephrotic syndrome identified rare heterozygous deletions of EXOC4 in two patients. This suggested that disruption of the highly-conserved eight-protein exocyst trafficking complex could have a role in podocyte dysfunction. Indeed, mRNA profiling of injured podocytes identified significant exocyst down-regulation. To test the hypothesis that the exocyst is centrally involved in podocyte development/function, we generated homozygous podocyte-specific Exoc5 (a central exocyst component that interacts with Exoc4) knockout mice that showed massive proteinuria and died within 4 weeks of birth. Histological and ultrastructural analysis of these mice showed severe glomerular defects with increased fibrosis, proteinaceous casts, effaced podocytes, and loss of the slit diaphragm. Immunofluorescence analysis revealed that Neph1 and Nephrin, major slit diaphragm constituents, were mislocalized and/or lost. mRNA profiling of Exoc5 knockdown podocytes showed that vesicular trafficking was the most affected cellular event. Mapping of signaling pathways and Western blot analysis revealed significant up-regulation of the mitogen-activated protein kinase and transforming growth factor-β pathways in Exoc5 knockdown podocytes and in the glomeruli of podocyte-specific Exoc5 KO mice. Based on these data, we propose that exocyst-based mechanisms regulate Neph1 and Nephrin signaling and trafficking, and thus podocyte development and function.

Diseases that cause podocyte dysfunction and loss are the leading causes of end-stage kidney disease worldwide. Targeting the early molecular events in podocyte development and/or homeostasis may lead to an effective therapy to manage podocyte injury and recovery, which is critical given the paucity, and toxicity, of current therapies. The slit diaphragm proteins Nephrin and Neph1 are known to regulate podocyte actin dynamics through their ability to assemble a signaling cascade that leads to structural and functional changes in podocytes (1)(2)(3). The membrane organization of Nephrin and Neph1 is critical for cell survival, and several recent reports highlight changes in their localization, where they are displaced from the slit diaphragm, in response to glomerular injury (2)(3)(4)(5). Reports have also emerged that demonstrate endocytic trafficking of these proteins as a critical event in podocyte function (6).
The exocyst is a highly-conserved octameric protein complex composed of Exoc1-8 that mediates the targeting and docking of vesicles carrying membrane proteins (7). Various cellular processes have been shown by us and others that involve the exocyst, including cell polarity (8,9), ciliogenesis (10 -13), and protein translation (14,15). Although the exocyst complex is a hetero-octameric protein complex, Exoc5 is a central component of the exocyst (9,13), loss of which leads to disintegration of this complex (13,16). Exocyst complex members were also shown to be mutated in families with diseases that can affect the kidney, such as Joubert syndrome (17); however, the role of the exocyst complex in podocyte biology has not been investigated.
Our recent results identified two patients with exocyst 4 (EXOC4) deletions from a cohort of 256 patients with nephrotic syndrome, which prompted us to further investigate the role of the exocyst complex in podocytes by generating podocyte-specific Exoc5 knockout (KO) mice. In this study, we demonstrate the phenotypic and mechanistic analyses of these mice, which show that loss of Exoc5 induced podocyte cell death and proteinuria. We further demonstrate that mislocalization of Nephrin and Neph1 is a major contributor to this phenotype.

Identification of structural variation revealed deletions in the exons of EXOC4, a member of the exocyst complex, in two patients with nephrotic syndrome
Structural variation (SV) 3 represents a form of genomic variation that can cause or be associated with human disease (18), and it is known to impact between 15 and 20% of all children born with abnormally developed kidneys (19). Therefore, SV analysis was performed on 256 patients with nephrotic syndrome enrolled in the Nephrotic Syndrome Study Network (NEPTUNE) (20), who had undergone whole-genome sequencing and Illumina ExomeArray Genotyping. By doing this, we identified two patients with heterozygous deletions on chromosome 17 impacting EXOC4. EXOC4 is an exocyst component that directly interacts with EXOC5 (21,22). One patient had a 522-kb deletion (Chr. 17: 132,970,005-133,492,495, deleting exons 3-10), and the other had a 9-kb deletion (Chr. 17: 132,987,872-132,997,017 deleting exon 3) in the EXOC4 gene (Fig. 1, A and B, black dots represent SNPs within the copy-number variation (CNV) region and have a lower log R ratio). These heterozygous deletions were independently validated by mapping the intensity of the signals for these SNPs in the entire cohort, identifying the two cases with deletions as having the lowest intensity (Fig. 1C). In the 2,504 member "1000 Genomes Project cohort," there was only one individual with an SV overlapping the EXOC4-coding sequence, and this affected only exon 15. Although both patients had nephrotic syndrome, the patient with 522-kb deletion had collapsing focal segmental glomerular sclerosis (FSGS) phenotype, and the patient with 9-kb deletion displayed a milder minimal-change phenotype. These genetic data are supportive of a hypothesis for the involvement of the exocyst in human glomerular disease.

Injury to podocytes significantly down-regulated members of exocyst complex and its regulator Cdc42
To further determine whether the exocyst complex participates in podocyte injury, we performed mRNA profiling (RNA-Seq) of cultured podocytes treated with the glomerular injury-inducing agent puromycin aminonucleoside (PAN). The analysis showed significant down-regulation of multiple proteins (EXOC2-5 and -7) in response to injury (Fig. 1D). Interestingly, CDC42, a known regulator for the exocyst complex (12,23,24), was also down-regulated, further implicating involvement of the exocyst complex in podocyte injury.

Podocyte-specific KO of the central exocyst component, Exoc5, disrupted podocyte function and induced proteinuria
To investigate the role of exocyst in podocytes, the exocyst complex was disrupted by genetically deleting Exoc5 specifi-

Exocyst disruption induces podocyte loss
cally in podocytes. Thus, the Exoc5fl/fl mouse line that we previously generated (25) was crossed with the Podocin-Cre driver mouse line (26). Ai14 Cre reporter mice that harbor a loxPflanked STOP cassette preventing transcription of a CAG promoter-driven red fluorescent protein variant (tdTomato), congenic on the C57BL/6J genetic background, were mated with Exoc5fl/fl mice to induce tdTomato expression in a cell where Exoc5 has been inactivated by Cre. The target tdTom-Exoc5fl/ fl;Pod-Cre/ϩ mice selectively labeled podocytes in which Exoc5 was deleted with red fluorescence (Fig. 2A). The kidney sections from these mice were stained with Exoc5 antibody, and as shown in Fig. 2B, the staining of Exoc5 did not overlap with the red fluorescence indicating Exoc5 deletion in podocytes. Additionally, the red podocytes were isolated by FACS sorting and subjected to quantitative PCR analysis, which showed more than 90% loss of Exoc5 (Fig. 2C). Collectively, these results confirmed genetic deletion of Exoc5 specifically in podocytes.

Podocyte-specific Exoc5 KO showed foot process effacement and loss of slit diaphragm
Podocyte foot process effacement is the hallmark of many glomerular diseases that affect podocyte structure and function, including FSGS (27)(28)(29). To further evaluate the effect of

Exocyst disruption induces podocyte loss
Exoc5 deletion on podocyte structure, ultrastructural analyses of tdTom-Exoc5fl/fl;Pod-Cre/ϩ and control mice were performed using scanning (SEM) and transmission electron (TEM) microscopic procedures. Compared with the control littermates, the SEM images of Exoc5fl/fl;Pod-Cre/ϩ mice showed significant alterations to glomerular morphology, with loss of secondary and tertiary foot processes (Fig. 3B). In addition, the TEM analysis showed foot process fusion and effacement with loss of the slit diaphragm (Fig. 3C). These profound changes to podocyte structure suggest that loss of Exoc5 induces structural and functional damage to podocytes, and it may affect the development and/or maintenance of these cells.

Ciliogenesis does not contribute to the proteinuric phenotype in Exoc5 KO mice
Because multiple studies have shown the requirement of the exocyst complex for ciliogenesis in kidney tubule cells (10 -13), we investigated whether ciliogenesis-based mechanisms contributed to this phenotype. Thus, we generated mice with podocyte-specific KO of a ciliary protein, intraflagellar transport protein 88 (Ift88) (Fig. 4A, both panels). Ift88 is required for ciliogenesis, and loss of this protein leads to the absence of cilia (30 -32). Interestingly, the Ift88fl/fl:Pod-Cre/ϩ mice developed normally and did not show any signs of proteinuria or any renal abnormality even after 6 months of monitoring (Fig. 4B). These results indicate that exocyst mechanisms other than ciliogenesis contribute to the pathogenic phenotype observed in podocyte-specific Exoc5 KO mice. Additionally, the deletion of Ift88 in podocytes had no effect on sensitivity of mice toward glomerular injury induced by NTS and/or adriamycin (Fig. 4, C and D), further supporting the idea that ciliogenesis does not contribute to podocyte structure and/or function.

Exoc5 knockdown in cultured podocytes down-regulated intracellular movement and localization events
To further investigate the mechanism through which loss of Exoc5 results in podocyte damage, we generated Exoc5 knockdown cultured human podocytes using Exoc5-specific shRNA (Fig. 5A). The mRNA isolated from scramble control and Exoc5 knockdown podocytes were subjected to profiling analysis. RNA-Seq data were submitted to the Gene Expression Omnibus (GEO) database (accession number GSE127736). The volcano plot shows 3,921 differentially expressed genes in Exoc5 knockdown human podocytes, of which 1,661 were up-regulated and 2,260 were down-regulated (Fig. 5B). The heat map of hierarchical clustering from RNA-Seq data showed differentially expressed genes in control and Exoc5 knockdown podocytes cells (Fig. 5C). Gene ontology enrichment analysis was performed, which identified several major biological processes (p Ͻ 0.05), including cellular movement, apoptosis, locomotion, vesicle transport, and protein localization ( Fig. 5D and Table 1). These data highlight the role of Exoc5 in cellular trafficking events in podocytes.

Loss of Exoc5 induced mislocalization of slit diaphragm proteins Nephrin and Neph1
Because the mRNA-Seq analysis suggested involvement of Exoc5 in cellular transportation and localization events and the podocyte-specific Exoc5 knockout resulted in loss of the slit diaphragm leading to podocyte dysfunction, we hypothesized that Exoc5 is involved in trafficking of the slit diaphragm proteins Nephrin and Neph1. Because the slit diaphragm is a modified cell junction, which is constructed through homo-and heterophilic interactions between Nephrin and Neph1 (33, 34), we further hypothesized that the loss of the slit diaphragm in Exoc5fl/fl;Pod-Cre/ϩ mice was due to mislocalization of Nephrin and Neph1. Indeed, immunofluorescence analysis of kidney sections from these mice showed significant changes to the distribution of Neph1 and Nephrin. Both proteins failed to colocalize with the podocyte marker synaptopodin in Exoc5fl/ fl;Pod-Cre/ϩ mice (Fig. 6A). To further validate these observations, Exoc5 knockdown podocyte cells were generated, which in addition to the loss of Exoc5 showed down-regulation of the other exocyst components, except for Exoc6 (Fig. 6B). When analyzed, these cells showed an increase in Neph1, SMAD3, and ERK phosphorylation, which indicates progression of injury (Fig. 6, C and D) and is consistent with previous reports in which injury induced Neph1 phosphorylation (35). Induction of glomerular injury is known to activate the MAPK (ERK) pathway (36,37), and we previously showed that knockout of Exoc5 in kidney tubules increased MAPK expression (10,13,25). Here, we note a similar increase for ERK phosphorylation in Exoc5 KD podocytes (Fig. 6C).
The induction of fibrosis in the kidney sections from Exoc5 knockout mice (Fig. 3A) prompted us to investigate whether the loss of Exoc5 affects TGF-␤ signaling. Hence, immunoblot analysis of cell lysate from the Exoc5 KD cultured podocytes was performed, which showed increased expression of ␣-SMA and phosphorylated SMAD3 (Fig. 6C). A similar increase in the expression of phosphorylated SMAD3 was also noted in the Exoc5 knockout zebrafish line (Fig. S1). Collectively, these results highlight a mechanism that involves activation of . Podocyte-specific Ift88 knockout mice are not proteinuric and are not more susceptible to glomerular injury. A, podocyte-specific Ift88 knockout mice were generated (Ift88fl/fl;Pod-Cre/ϩ). Immunoblot and RT-PCR analysis of glomeruli isolated from Ift88 KO and control mice. ***, p Ͻ 0.001. B, SDS-PAGE of the urine from Ift88 podocyte-specific KO mice showed no evidence of proteinuria. C and D, podocyte-specific Ift88 knockout and control mice (Ift88fl/fl;Pod-Cre/ϩ) were subjected to acute (NTS) (C) and chronic (adriamycin) (D) glomerular injury, and the extent of albuminuria was assessed. No changes in albuminuria, compared with controls, were noted in either of these models.

Exocyst disruption induces podocyte loss
canonical and noncanonical TGF-␤ signaling pathways that may contribute to the disease phenotype.
To further confirm that the loss of Exoc5 affects the expression of Exoc4 (also known as Sec8) in these podocytes, immunoblotting was performed, which showed significant downregulation of Exoc4 in these cells (Fig. 6, C and D). Thus, these results are in agreement with our previously published results showing that loss of Exoc5 destabilizes the exocyst complex, most likely leading to its degradation (13). These results further validate Exoc5 as the central exocyst component.
Additional analysis of Exoc5 KD podocytes showed significant reduction in the localization of Neph1 at the podocyte cell membrane providing further support for the role of Exoc5 in trafficking (Fig. 7, A and B). Because Nephrin protein remains undetectable in cultured podocytes, similar analyses could not be performed for endogenous Nephrin. To further confirm these observations, we constructed chimeric Nephrin and Neph1 constructs, where a FLAG-tag was inserted in the extra-cellular domain of these proteins, immediately following the signal peptide sequence, but before the functional IgG domains (Fig. 7C). These constructs were transfected into Exoc5 KD podocytes, and their surface/membrane localization in live cells was assessed using FLAG antibody (Fig. 7, D and G). Similar to our in vivo results, a significant reduction in membrane localization of Nephrin and Neph1 was noted in Exoc5 KD cultured podocytes (Fig. 7, E, F, H, and I). These data support the hypothesis that Exoc5 regulates trafficking of slit diaphragm proteins Neph1 and Nephrin.
Importantly, the loss of Neph1 and Nephrin membrane localization in human Exoc5 KD podocytes was rescued by transducing cells with packaged viruses expressing WT mouse Exoc5 (Fig. 8), which was not susceptible to the shRNA given a difference in primary base-pair structure (Fig. S2).
To further demonstrate the effect of Exoc5 on the trafficking of slit diaphragm proteins, we examined the movement of Neph1 vesicles in cultured podocytes. To do this, we generated

Exocyst disruption induces podocyte loss
cultured podocytes stably expressing mCherry-Neph1 (Fig. S3). Exoc5 was knocked down in these podocytes, and the movement (micrometers) and speed (micrometers/s) of mCherry-Neph1 vesicles were evaluated. The results showed that loss of Exoc5 significantly decreased the movement and speed (ϳ30 and ϳ20% respectively) of mCherry-Neph1 vesicles as shown by time-lapse live-cell imaging analysis (Fig. 9, A and B).

Podocyte-specific Exoc5 knockout induced apoptosis in podocytes
Because podocyte loss leads to proteinuria (38) and our mRNA profiling analysis indicated apoptosis as a major cellular event in Exoc5 KD podocytes, we investigated whether Exoc5 deletion induced apoptosis in podocytes. Indeed, TUNEL staining of kidney sections showed increased apoptotic nuclei in Exoc5fl/fl;Pod-Cre/ϩ mice, when compared with the control mice (Fig. 10, A and B). Interestingly, apoptotic nuclei were also observed in cells surrounding the glomeruli. This is not totally surprising given the fact that there is significant cross-talk between podocytes and other glomerular cell types, including endothelial cells and the glomerular basement membrane (39).

Discussion
We report several novel findings in this study. Importantly, this is the first study to examine the role of the exocyst complex in podocyte development and disease. The hypothesis for involvement of the exocyst was prompted by the discovery of two rare heterozygous deletions impacting EXOC4 in patients with nephrotic syndrome that were not seen in a reference population of 2,504 individuals. Independent support for the genetic dependence on the exocyst for podocyte development and/or health was demonstrated in mice using podocyte-specific KO of Exoc5. Loss of Exoc5 in podocytes resulted in significant glomerular damage leading to death in all mice within 30 days of birth. Whether exocyst inhibition prevents normal development of the podocytes, inhibits podocyte homeostasis, or both, is something that needs to be studied further using timed matings.
Signaling of Nephrin and Neph1 is a critical event that governs subsequent trafficking of these proteins in response to injury. Our results indicate that Exoc5 and the exocyst may participate in regulating the signaling of slit diaphragm proteins, particularly Nephrin and Neph1. To determine whether ciliogenesis contributed to the pathogenic phenotype in Exoc5 KO mice, we generated mice with podocyte-specific KO of the ciliary protein Ift88. Many reports have shown that Ift88 is required for ciliogenesis, and loss of this protein leads to the absence or severe stunting of primary cilia (30,40). Interestingly, these mice remained healthy and showed no signs of proteinuria or increased susceptibility to glomerular injury, indicating that exocyst-mediated mechanisms other than ciliogenesis contributed to the pathogenic phenotype that we observed in Exoc5 KO mice.
Previous studies using various injury models have shown that glomerular injury induces redistribution of slit diaphragm proteins Nephrin and Neph1, which is associated with podocyte dysfunction and proteinuria (5,41,42). Our podocyte-specific Exoc5 KO mice showed similar results suggesting that loss of Exoc5 replicated events that cause podocyte injury. Because Nephrin and Neph1 have been shown to be critical for podocyte survival, and loss of Exoc5 significantly affected Nephrin and Neph1 expression and localization, we hypothesized that the exocyst plays a major role in regulating their trafficking and signaling, thus affecting podocyte development and/or maintenance. The role of Exoc5 in the trafficking of Nephrin and Neph1 was further substantiated using cultured human Exoc5 knockdown podocytes, where loss of Exoc5 significantly attenuated the membrane localization of chimeric Nephrin and Neph1 proteins. Moreover, the profiling analyses of Exoc5 knockdown podocytes further confirmed trafficking as the major cellular event affected due to loss of Exoc5. Thus, it is likely that trafficking associated events are the major contributor to the proteinuric phenotype observed in Exoc5 knockout mice.
Glomerular injury-induced Neph1 phosphorylation and induction of the MAPK pathway have been reported previously (43), and similar observations were noted in our Exoc5 knockdown cells. Although little is known about the signaling events that are initiated in response to Nephrin and Neph1 activation, the phosphorylation and therefore the induction of both Neph1 and Nephrin, and the MAPK pathway in Exoc5 KO mice, provide evidence for the participation of the exocyst in generating the injury phenotype. Importantly, our results also showed that TGF-␤ signaling is induced in Exoc5 knockdown podocytes. Although TGF-␤ is known to participate in podocyte pathogenesis (44 -46), the results from this study suggest that Exoc5 may negatively regulate TGF-␤ signaling. However, these conclusions need to be investigated further through a detailed analysis of TGF-␤ signaling in Exoc5-depleted and -overexpressing systems, which will be the subject of future investigations.
We previously demonstrated that Cdc42 regulates the exocyst complex (12,23,24), and it is interesting to note that our podocyte-specific Exoc5 KO mice phenocopied podocyte-specific Cdc42 KO mice (47). Based on these results, and the over-

Exocyst disruption induces podocyte loss
all results presented in this study, we hypothesize that fundamental mechanisms that regulate trafficking of slit diaphragm proteins also regulate podocyte structure and function. These data, for the first time, identify the exocyst as a novel component of podocytes that actively participates in their development and/or homeostasis.

Structural variation analysis
We used low-depth whole-genome sequencing (WGS) and Illumina Infinium Exome genotypes that were collected from the NEPTUNE patients. SV calls from WGS were made by the GenomeSTRiP pipeline, and we looked for those that affected an exon(s) and were not present in participants from the 1,000 Genome Project Phase 3, which we used as a reference (48). For independent confirmation of this finding, we examined intensities of the EXOC4 SV region using the Exome Chip genotypes.

Generation of podocyte-specific Exoc5 knockout mice
To investigate the role of the exocyst in podocytes, we generated podocyte-specific Exoc5 KO mice by crossing Exoc5fl/fl with Podocin-Cre (Pod-Cre) mice on a C57BL6/J background. Target mice, designated Exoc5fl/fl;Pod-Cre/ϩ, were confirmed by PCR-based genotyping. Genotyping primers used were GCCTGTAACTCACAGAGATC (forward) and GCTG-GCATTCTAAGTCATGG (reverse) for Exoc5 flox and AGG-TTCGTGCACTCATGGA (forward) and TCGACCAGT-TTAGTTACCC (reverse) for Cre recombinase. To further demonstrate the fidelity of Cre recombinase, Exoc5fl/fl mice were mated with Ai14 Cre reporter mice that harbor a loxPflanked STOP cassette preventing transcription of a CAG promoter-driven red fluorescent protein variant (tdTomato), congenic on a C57BL/6J genetic background. This allowed expression of tdTomato in podocytes where Exoc5 was inactivated by Cre. The tdTomato expression was seen in tdTom-Exoc5fl/fl;Pod-Cre/ϩ and tdTom-Exoc5fl/ϩ;Pod-Cre/ϩ, but not in tdTomato Exoc5fl/fl, mice (Fig. 2). Urinalysis was performed by SDS-PAGE, and histological and morphological analysis of kidney sections using immunofluorescence and histological procedures. A similar strategy was followed to generate podocytespecific Ift88 KO mice by crossing Ift88fl/fl with Pod-Cre mice on a C57BL6/J background to determine whether ciliogenesis is required for podocyte growth and development.  (n ϭ 10). C, schematic of the chimeric Neph1 and Nephrin constructs, where FLAG-tag was inserted following the signal peptide sequence. D-I, chimeric Nephrin and Neph1 were expressed in cultured podocytes (D and G), and stable cell lines were created. Nephrin and Neph1 in these cells were extracellularly labeled with FLAG antibody and live-cell staining was performed. The surface localization of Neph1 and Nephrin was analyzed by immunofluorescence using confocal microscopy (E and H), and the images were quantified using ImageJ, which showed significant loss of surface Neph1 and Nephrin in Exoc5 knockdown podocytes (n ϭ 10) (F and I). Bar, 25 m. In each experiment, 2 million cells were plated on each coverslip, and 10 or more cells were analyzed for quantification and statistical analysis. **, p Ͻ 0.001; *, p Ͻ 0.01, paired Student's t test.

Exocyst disruption induces podocyte loss Histological and immunological characterization of Exoc5fl/fl; Pod-Cre/؉ mice
Murine kidneys were perfused with Hanks' buffered salt solution (HBSS) for 5 min followed by 4% paraformaldehyde in HBSS for 15 min at 80 -100 mm Hg retrograde to the intrarenal aorta of anesthetized mice using a 24-gauge angiocatheter (BD Biosciences). Mice were anesthetized with Avertin (2,2,2-tribromoethanol) before the procedure. Perfused kidneys were decapsulated, washed with 1ϫ PBS, transected, fixed for 4 -12 h in 4% paraformaldehyde, and rinsed, and sequential alcohol treatments were performed. The samples were submitted to the Abramson Histology Core at MUSC and the University of South Carolina's Histology Core facility. Paraffin-embedded sections were obtained and deparaffinized with xylene-ethanol (EtOH). Paraffin-embedded sections were stained with Masson's trichome for histological analysis. Staining of various proteins, including Neph1, Nephrin, and synaptopodin, was performed using specific antibodies as described previously (49,50). Briefly, antigen retrieval was performed using 10 mM Tris, 1 mM EDTA (pH 9.0) at 65°C for Ͼ18 h. Kidney sections were encircled by use of a Pap pen, blocked in TBS, 0.025% Triton X-100 (TBST) plus 3% bovine serum albumin (BSA), incubated with primary antibody (Ab), washed five times, re-blocked with TBST plus 10% goat serum (Invitrogen) plus 1% BSA, incubated with Alexa Fluor secondary Ab (Invitrogen) at 1:1,000), washed five times, and mounted with Prolong Gold antifade reagent (Invitrogen). IgG control primary antibody were used as negative control (Fig. S4). A Zeiss inverted microscope (Zeiss Axio Observer D1m) fitted with ϫ63 and ϫ25 oil immersion objectives was used for fluorescence microscopy, whereas confocal microscopy was performed at MUSC on a Leica microscope. All parameters were maintained at a constant level throughout acquisition of the images, including the exposure time, and the images were collected in 16-bits format. ImageJ software was used for quantitative analysis. Multiple images from three or more mice were used for the quantitative analysis. Representative images from three or more independent experiments were used in each figure. For SEM and TEM analyses, the kidney samples were kept in a mixture of 2% glutaraldehyde and 2% paraformaldehyde solution overnight for fixation. Samples

Exocyst disruption induces podocyte loss
were processed, sectioned, and imaged at the Department of Pathology, MUSC. Five to 10 glomeruli per sample from at least three mice were evaluated, and slit diaphragm frequency and thickness of glomerular basement membranes were assessed quantitatively by counting the number of junctions/micron of basement membrane using ImageJ software.

Immunoblotting, antibodies, and reagents
Primary kidney cells and human podocytes were lysed in RIPA buffer, and protein estimation was performed using the BCA method. 20-g protein samples from lysate were used for immunoblotting experiments. To ensure equivalent protein loading, the membranes were simultaneously incubated with

shRNA-based Exoc5 knockdown in cultured podocytes
Exoc5 knockdown in cultured human podocytes was achieved using Exoc5-specific shRNA. To target Exoc5, we screened Mission Lentiviral Transduction particle shRNAs in pLKO.1-puro (Exoc5 MISSION Lentiviral Transduction shRNA particle, commercially purchased from Sigma, catalog no. SHCLNV-NM_006544 and TRC no. TRCN0000061963, TRCN0000061964, TRCN0000061965, TRCN00000286419, and TRCN0000061966). Transfection of the shRNA plasmids was performed with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Transfected cells were grown in 2.5 g/ml puromycin-containing medium for the selection of stable transfectants, and the Exoc5 knockdown was confirmed by Western blotting.

Generation of viral expression constructs for FLAG-Nephrin, FLAG-Neph1, mcherry Neph1, and mouse Exoc5
A standard PCR-cloning technique was used to generate the mammalian expression plasmid encoding FLAG-tagged fulllength Nephrin in the pBABE-puro vector as described previously (55). The FLAG tag was inserted between the signal peptide and the start of the extracellular domains of Nephrin. The PCR-amplified FLAG-Nephrin was cloned at the EcoRI site in the retroviral vector pBABE-puro. The following two forward and one reverse primer sets were used for cloning: forward primer (FP) 1: 5Ј-CCCTGCTGCTCGCGGGAATGCTGACC-ACAGGCCTGGCCCAGTCGGACTACAAGGATGACGA-TGACAAACCAGTCCCCACCTCAGCACC-3Ј, FP 2: 5Ј-ATGGGAGCTAAGGAAGCCACAGTCAGAGGCCCCGGG-GCAAGCCCAGTGCACAGAACTTGCCACCTGATTCCCC-TGCTGCTCGCGGGAATGC-3Ј, and reverse primer (RP): 5Ј-AGGAATTCCATCACACCAGATGTCCCCTCAG-3Ј, respectively. The fidelity of the constructs was determined by restriction digestion and DNA sequencing. Neph1 was also cloned into the pBABE vector at the EcoRI and SalI site employing a similar strategy using following sets of primers: FP 5Ј-GAATTCATGACTC-TGGAGAGCCCTAGCACA-3Ј and RP as CGAGTCGACCTA-CACGTGAGTCTGCATGCG. Subsequently, the FLAG tag was inserted between the signal peptide and the extracellular domain of Neph1 using the following sets of primers: forward primer GAC-TACAAGGATGACGATGACAAAACTCAGACTCGCTTC and reverse primer TTTGTCATCGTCATCCTTGTAGTCC-CCGGGCAAAGCCAA. Finally, retroviruses overexpressing FLAG-Neph1 and FLAG-Nephrin were generated by the transfection of the respective plasmids (cloned into the pBABE vector) into Phoenix cells according to the manufacturer's instructions. The cultured podocytes were transfected with these viruses, and the expression of FLAG-Nephrin and FLAG-Neph1 was evaluated by Western blotting. mCherry Neph1 was constructed in a pBABE Puro retroviral vector using standard PCR-cloning strategies, as described previously (55). Similarly, the lentiviral construct, where Exoc5 was cloned into the EcoRV site of the pLenti-GIII-UbC vector, was commercially obtained from Applied Biological Materials (ABM, Canada). Viruses expressing mouse Exoc5 were generated as per the manufacturer's instructions and used in the rescue experiment to express mouse Exoc5 in Exoc5 KD human cultured podocytes. The differences in the CDS sequences of mouse and human Exoc5 targeted by the shRNA lentiviral constructs were analyzed and are presented in Fig. S2.

Isolation of murine glomeruli
Mice were anesthetized using isoflurane and perfused with 8 ϫ 10 7 Dynabeads diluted in 40 ml of PBS through the heart. Before the perfusion, the right atrium was cut to achieve the open circulation during perfusion. The kidneys were removed, minced into 1-mm 3 pieces, and digested in collagenase (1 mg/ml collagenase A, 100 units/ml DNase I in PBS) at 37°C for 15 min with gentle agitation. The collagenase-digested tissue was gently pressed through a 70-m cell strainer using a flattened pestle, and the cell strainer was then washed with 50 ml of PBS. The filtered cells were passed through a new 40-m cell strainer without pressing, and the cell strainer was washed with 50 ml of PBS. The cell suspension was then centrifuged at 200 ϫ g for 5 min. The supernatant was discarded, and the cell pellet was resuspended in 2 ml of PBS. Finally, glomeruli containing Dynabeads were gathered by a Magnetic Particle Concentrator and washed at least three times with PBS. During the procedure, kidney tissues were kept at 4°C, except for the collagenase digestion step at 37°C. The glomeruli were used for isolation of RNA by the Qiagen RNA prep kit.

RNA-Seq
RNA samples isolated from cultured human podocytes were quantitatively and qualitatively analyzed using Nanodrop (for RNA purity (OD 260 /OD 280 ), agarose gel electrophoresis (for RNA integrity and potential contamination), and Agilent 2100 (check RNA integrity again) by the Novogene Genome

Exocyst disruption induces podocyte loss
Sequencing Co. Briefly, library construction was done using mRNA purified from human podocytes by poly(T) oligo-attached magnetic beads. The mRNA was fragmented randomly by addition of fragmentation buffer, and the library was prepared. After selecting the appropriate library, quality assessment was performed, and we quantified the effective concentration (Ͼ2 nM). Libraries were fed into HiSeq/MiSeq Illumina machines for RNA-Seq using a standard method by the Novogene Genome Sequencing Co. Bioinformatics and the downstream analysis was also performed by Novogene Genome Sequencing using a combination of programs, including Bow-tie2, Tophat2, HTseq, Cufflink, and the wrapped scripts at Novagen. Briefly, the reference genome and gene model annotation files were obtained from the NCBI/UCSC/Ensemble genome browser. Indexes of the reference genome were built using Bowtie version 2.0.6 and paired-end clean reads were aligned to the reference genome using TopHat version 2.0.9. Quantification of the gene expression level was analyzed using HTSeq version 0.6.1, and read numbers counts were mapped for each gene. Reads per kilobase of exon model (RPKM) of each gene were calculated based on the length of the gene, and gene expression levels were analyzed. Differential expression analysis was performed using the DESeq2 R package (2_1.6.3). DESeq2 provided statistical routines for determining differential expression in digital gene expression data using a model based on the negative binomial distribution. The resulting p values were adjusted using the Benjamini and Hochberg's approach, and adjusted p values Ͻ0.05 found by DESeq2 were assigned as differentially expressed. Differential expression analysis of two conditions was performed using the DEGSeq R package (1.12.0). The p values were also adjusted using Benjamini and Hochberg method. Corrected p values of 0.005 and log2(fold change) of 1 were set as the threshold for determining significant differential expression. Correlations were determined using the corr.test function in R with options set alternative ϭ "greater" and method ϭ "Spearman." To identify the correlation between the differences, we clustered different samples using expression level RPKM with hierarchical clustering and generation of a heat map. Self-organization mapping and k means using the silhouette coefficient to adapt the optimal classification with default parameter in R. Gene ontology (GO) and KEGG enrichment analyses of the differentially expressed genes were analyzed using the GOseq R package. GO terms with a corrected p value less than 0.05 were considered significant. KOBAS software was used to test the statistical enrichment of differential expression genes in KEGG pathways.

Quantitative real-time RT-PCR
To evaluate the Exoc5 and Ift88 expression levels in Exoc5 and Ift88 KO mice, quantitative real-time RT-PCR (qRT-PCR) was done. Additionally, the expression of exocyst components (Exoc1-8) in Exoc5 knockdown cultured podocytes was evaluated by qRT-PCR. The specific primers were designed using Gene runner. The sequences of primers used is listed in Table 2. Quantitative RT-PCR experiments were performed using an CFX96 real-time thermal cycler (Bio-Rad) equipment and a SYBR Green PCR master mix (Bio-Rad). The thermocycling conditions utilized were s follows: an initial step of 95°C for 3 min, 40 cycles of 95°C for 10 s, and 60°C for 30 s. The specificity of the amplified products was evaluated by analysis of dissociation curves generated by the equipment. Three independent RNA samples were used for analysis. The gene encoding actin protein was used as the calibrator gene. Determination of the expression ratios was performed using the threshold cycle (⌬⌬C T ) method.

Immunofluorescence microscopy and immunohistochemistry
Cells were transiently transfected with vectors using the Lipofectamine 2000 transfection agent following the manufacturer's instructions. The cells were selected for stable knockdown using puromycin. Twenty four hours post-plating and staining, the images were taken using confocal microscopy. Images of the cells were taken using a confocal microscope (Leica). Immunofluorescence staining was performed on 4-m frozen sections from mice, and cultured podocytes were grown on coverslips as described previously (42). Each experiment was carried out in three independent sets, and the images were collected on a Leica confocal microscope (Leica TCS SP5).

Live-cell microscopy for evaluating the movement of Neph1containing vesicles in Exoc5 knockdown cultured podocytes
mCherry-Neph1-expressing cultured podocytes, with and without Exoc5 KD, were plated on 35-mm glass-bottom cell culture plates, and the movement of Neph1-containing vesicles was analyzed using live-imaging microscopy. Images were captured for 2 min with 10-s intervals using Leica confocal microscopy (Leica TCS SP5), and images were analyzed using ImageJ software. The displacement of Neph1-containing vesicles was plotted against time. Each experiment was repeated at least three times. In each experiment, 2 million cells were plated on each coverslip, and 10 or more cells were analyzed for quantification.

Exocyst disruption induces podocyte loss Apoptosis/TUNEL assay
Frozen sections from kidneys of WT and Exoc5 knockout mutants were stained using an in situ cell death detection kit, fluorescein (Sigma), according to the manufacturer's instructions. The in situ Cell Death Detection protocol is based on the detection of single-strand DNA and dsDNA breaks (TUNEL technology: TdT-mediated dUTP-X nick end-labeling) that occur at the early stages of apoptosis.
Kidney sections were fixed and permeabilized. Subsequently, the kidney sections were incubated with the TUNEL reaction mixture that contains TdT and fluorescein dUTP. Images of the apoptotic cells in kidney sections were photographed using a confocal microscopy (Leica TCS SP5).

Animal study approval
All animal studies were conducted as per the protocols approved by the Medical University of South Carolina and/or the Ralph H. Johnson VAMC, Institutional Animal Care and Use Committee, and National Institutes of Health guidelines for the Care and Use of Laboratory Animals. Treatment of mice, including housing, injections, and surgery was in accordance with the institutional guidelines. Isoflurane anesthesia (5% induction, 2% maintenance) was used to perform all nonsurvival surgeries.

Human study approval
The CNV analysis was approved by the IRB of the University of Michigan and NEPTUNE. The human samples were deidentified.

Statistics
Each data set was presented as mean Ϯ S.E. Statistical analysis, paired one-tailed Student's t test, and one-way analysis of variance was performed using the GraphPad Prism 7 software. A p value of Յ0.05 was considered as statistically significant.