The carboxyl terminus of Neph family members binds to the PDZ domain protein zonula occludens-1.

The PSD95/Dlg/ZO-1 (PDZ) domain-containing protein zonula occludens-1 (ZO-1) selectively localizes to the cytoplasmic basis of the slit diaphragm, a specialized cell-cell contact in between glomerular podocytes necessary to prevent the loss of protein in the urine. However, the function of ZO-1 at the slit diaphragm has remained elusive. Deletion of Neph1, a slit diaphragm protein of the immunoglobulin superfamily with a cytoplasmic PDZ binding site, causes proteinuria in mice. We demonstrate now that Neph1 binds ZO-1. This interaction was mediated by the first PDZ domain of ZO-1 and involved the conserved PDZ domain binding motif present in the carboxyl terminus of the three known Neph family members. Furthermore, Neph1 co-immunoprecipitates with ZO-1 from lysates of mouse kidneys, demonstrating that this interaction occurs in vivo. Both deletion of the PDZ binding motif of Neph1 as well as threonine-to-glutamate mutation of the threonine within the binding motif abrogated binding of ZO-1, suggesting that phosphorylation may regulate this interaction. ZO-1 binding was associated with a strong increase in tyrosine phosphorylation of the cytoplasmic tail of Neph1 and dramatically accelerated the ability of Neph1 to induce signal transduction. Thus, our data suggest that ZO-1 may organize Neph proteins and recruit signal transduction components to the slit diaphragm of podocytes.

Hereditary nephrotic syndrome is a heterogeneous disease characterized by heavy proteinuria and renal failure. The recent description of gene defects of the podocyte resulting in hereditary nephrotic syndrome has provided a completely new understanding of the glomerular filter and unraveled important aspects of the pathogenesis of proteinuric kidney diseases (1). The most severe hereditary disorder is the congenital nephrotic syndrome of the Finnish type, caused by mutations in NPHS1, 1 the gene encoding for nephrin. Nephrin is an integral membrane protein of the immunoglobulin superfamily located at opposing sites of the secondary foot processes formed by podocytes, specialized epithelial cells that ensure size-and charge-selective ultrafiltration (for reviews, see Refs. 2 and 3). The precise function of nephrin is unknown; however, nephrin is a critical structural component of the slit diaphragm, an ultrathin zipper-like structure that bridges the ϳ40-nm-wide slit between interdigitating podocyte foot processes. We have recently shown that nephrin is a signaling protein and that signal transduction can be augmented by another podocyte protein called podocin (4). Podocin is a hairpin-like protein at the slit diaphragm encoded by NPHS2, the gene disrupted in a steroid-resistant hereditary nephrotic syndrome (5,6). In addition to nephrin and podocin, other podocyte proteins including CD2AP (7,8), ␣-actinin 4 (10), and Neph1 (11) have recently been associated with the development of proteinuria. Neph1 contains five extracellular immunoglobulin-like domains and is structurally related to nephrin. It is abundantly expressed in the kidney, and disruption of the Neph1 gene in mice results in effacement of glomerular podocytes, heavy proteinuria, and early postnatal death (11). Neph1 belongs to a family of three closely related proteins that bind to podocin (12). The interaction of nephrin with the Src homology 3 (SH3)containing adaptor protein CD2AP (8), likely involved in protein trafficking and endocytosis (13), suggests that the protein complex at the slit diaphragm is highly dynamic and regulated by protein-protein interactions that organize the filtration barrier and initiate intracellular signaling. Although many of the key components of the slit diaphragm have now been identified, the fundamental question remains how the different components are organized at this specialized cell-cell junction.
Zonula occludens (ZO) proteins are membrane-associated multidomain proteins usually localized at sites of intercellular junctions. They contain three PDZ domains, a SH3 domain, and a guanylate kinase domain (14). PDZ domains are proteinbinding modules that recognize short peptide motifs within their protein targets (15). In almost all cases the last three to five residues at the extreme carboxyl terminus of a transmembrane protein represent the target sequences. Genetic evidence from invertebrate systems demonstrates a role for ZO proteins in facilitating signal transduction, and evidence from vertebrate systems demonstrates a structural role in organizing transmembrane protein complexes (16,17). In podocytes, ZO-1 has been shown to specifically localize to the cytoplasmic surface of the slit diaphragms (18 -20). However, the function of ZO-1 at the slit and its binding partners have not been characterized.
We demonstrate now that ZO-1 specifically binds to the carboxyl terminus of Neph1. Neph1 and ZO-1 are co-localized in the kidney glomerulus. Interaction of ZO-1 with Neph1 alters the ability of Neph1 to stimulate signaling. Our findings suggest that the multiadapter protein ZO-1 serves as a scaffold for the organization of Neph1 molecules and/or provides a platform for the recruitment of signal transduction components to the dynamic protein complex at the slit.

MATERIALS AND METHODS
Reagents and Plasmids-Neph1, Neph2, and Neph3 have been described previously (12). Membrane-bound fusion proteins of the carboxyl-terminal cytoplasmic domain of Neph1 were generated using a pCDM8 cassette that contained the leader sequence of CD5 fused to the CH2 and CH3 domain of human IgG1 followed by the transmembrane region of CD7 (21). A full-length cDNA clone of ZO-1 was kindly provided by Dr. Anderson (Yale University School of Medicine, New Haven, CT). Truncations and mutations of ZO-1 and Neph1 were generated by standard cloning procedures. The Neph1 antiserum has been described (12); antibodies were obtained from Sigma (anti-FLAG M2), Santa Cruz Biotechnology, Inc. (anti-ZO-1 and anti-HA), Roche Molecular Bio-chemicals (anti-HA), and Upstate Biotechnology (anti-Tyr(P) 4G10).
Co-immunoprecipitation-Co-immunoprecipitations were performed as described previously (22). Briefly, HEK 293T cells were transiently transfected by the calcium phosphate method. After incubation for 24 h, cells were washed twice and lysed in a 1% Triton X-100 lysis buffer. After centrifugation (15,000 ϫ g, 15 min, 4°C) and ultracentrifugation (100,000 ϫ g, 30 min, 4°C) cell lysates containing equal amounts of total protein were incubated for 1 h at 4°C with the appropriate antibody followed by incubation with 40 l of protein G-Sepharose beads for ϳ3 h. The beads were washed extensively with lysis buffer, and bound proteins were resolved by 10% SDS-PAGE. Since native Neph1 and ZO-1 in kidney cortex may be associated with lipid rafts, the lysis buffer was supplemented with 20 mM CHAPS. Sufficient solubilization was monitored by Western blot of different fractions during the preparation. Before immunoprecipitation cellular lysates were extensively precleared by ultracentrifugation and absorption to protein G beads. All kidneys were freshly prepared and perfused in situ with ice-cold phosphate-buffered saline before lysis. Densitometric analysis FIG. 1. The cytoplasmic tail of Neph1 interacts with ZO-1. A and B, HEK 293T cells were transfected with the expression plasmids as indicated. Lysates were subjected to immunoprecipitation with anti-Myc (A) or anti-FLAG (B) antisera, resolved by SDS-PAGE, and immunoblotted for Neph1. Neph1 specifically interacts with the first 503 amino acids of ZO-1. C-E, the cytoplasmic, carboxyl-terminal domain of Neph1 or FAT was fused to human immunoglobulin and the transmembrane domain of CD7 (sIg.7.Neph1) and co-transfected into HEK 293T cells with FLAG-tagged ZO-1 constructs (F.ZO-1). After immunoprecipitation with protein G, ZO-1 is detectable in the precipitate containing Neph1 but not the control protein (sIg.7) lacking the cytoplasmic domain of Neph1 or a fusion protein with the cytoplasmic tail of FAT. Western blot analysis was performed using the FLAG-specific M2 monoclonal antibody. F, co-immunoprecipitation of Neph1 with ZO-1 from mouse kidneys. Mouse kidneys were perfused in situ with ice-cold phosphate-buffered saline, collected, and homogenized in a CHAPS-containing buffer. After extensive preclearing the lysate was subjected to immunoprecipitation with a control antibody (anti-HA, lane 1) or anti-ZO-1 antiserum (lane 2). Neph1 could only be detected in the anti-ZO-1 precipitate. Lane 3 shows Neph1 in the kidney lysate. HEK 293T cell lysates untransfected (lane 4) and transfected with a Neph1 cDNA (lane 5) served as control for antibody specificity. G, co-localization of ZO-1 and Neph1 in the adult mouse kidney. Double staining of ZO-1 (green) and Neph1 (red) shows a basement membrane-like staining, a staining pattern typical for the podocyte foot processes in glomeruli. An overlay technique using conventional immunofluorescence microscopy (upper panel) and confocal microscopy (lower panel) demonstrate co-localization of ZO-1 and Neph1 in a basement membrane-like appearance. WT, wild type; IF, immunofluorescence.
was performed on non-saturated radiographs using the NIH Image software package. 32 P Labeling and Phosphoamino Acid Analysis-After labeling of transfected HEK 293T cells with 32 P (0.5 mCi/ml for 6 h), Neph1 was immunoprecipitated with protein G. Immunoprecipitates were washed extensively, subjected to 10% SDS-PAGE, and transferred onto polyvinylidene difluoride membranes. After autoradiography membrane pieces containing the 32 P-labeled Neph1 were cut out and subjected to phosphoamino acid analysis.
Pull-down Assay-HEK 293T cells were transiently transfected with plasmid DNA as indicated. Cells were lysed in 1% Triton X-100, 20 mM Tris-HCl, pH 7.5, 50 mM NaCl, 50 mM NaF, 15 mM Na 4 P 2 O 7 , 2 mM Na 3 VO 4 , 1 mM EDTA, and protease inhibitors for 15 min on ice. Following centrifugation, the supernatant was incubated for 1 h at 4°C with 4 -8 g of recombinant purified glutathione S-transferase (GST) or GST⅐PDZ domain fusion protein prebound to glutathione-Sepharose beads (Amersham Biosciences). Bound proteins were separated by 10% SDS-PAGE, and precipitated proteins were visualized with anti-FLAG antibody. Equal loading of recombinant proteins was confirmed by Coomassie Blue staining of the gels.
Immunofluorescence-Frozen adult mouse kidneys were embedded in OCT, sectioned at 5 m, and fixed with ice-cold acetone. The sections were incubated with affinity-purified anti-Neph1 antiserum followed by anti-rabbit rhodamine red and anti-ZO-1 rat monoclonal antibody (Chemicon, Hofheim, Germany) followed by anti-rat fluorescein isothiocyanate (DAKO, Hamburg, Germany). An Axiophot 2 microscope (Zeiss, Jena) equipped with a CCD camera was used for image documentation. Confocal images were taken using a Zeiss laser scan microscope equipped with a 100ϫ oil immersion objective.
Luciferase Assay-HEK 293T cells seeded in 12-well plates were transiently transfected with a luciferase reporter construct, a ␤-galactosidase expression vector (kindly provided by C. Cepko), and vectors directing the expression of the proteins as indicated. The total DNA amount was 1.5-2.0 g/well. Cells were serum-starved for 12 h, harvested in cold phosphate-buffered saline, and lysed in 100 l of reporter lysis buffer (Applied Biosystems, Norwalk, CT) for 10 min at 4°C. Lysates were centrifuged at 14,000 rpm for 5 min to remove insoluble material. Luciferase activity was determined using a commercial assay system (Applied Biosystems) and normalized for ␤-galactosidase activity to correct for transfection efficiency. Equal expression of proteins was ensured by Western blot analysis.

RESULTS AND DISCUSSION
Recent genetic studies of human hereditary disease and genetically modified animals have led to the identification of new podocyte proteins and have highlighted the crucial role of proteins at the filtration slit for the integrity of the glomerular filtration barrier (23). However, it is unknown how these different protein components of the slit diaphragm are organized. Since the carboxyl-terminal tail of Neph1 contains a putative class 1 PDZ domain binding motif (scansite.mit.edu, (24)), we speculated that this motif could mediate interaction with PDZ domain proteins. ZO-1 has been the only PDZ domain protein demonstrated to localize to the cytoplasmic surface of the filtration slit. We therefore tested whether ZO-1 interacts with Neph1. Fig. 1A shows that Neph1 specifically co-immunoprecipitated with ZO-1. No interaction could be demonstrated for FAT, another slit diaphragm transmembrane protein containing a putative PDZ domain binding motif (Fig. 1E) (9). Binding of Neph1 could be localized to the first 503 amino acids of ZO-1, a region containing the three PDZ domains of ZO-1 (Fig. 1, B and D) and the cytoplasmic tail of Neph1 (Fig. 1C). FLAGtagged ZO-1-(1-503) appeared as a double band suggesting premature termination or post-translational modification of this construct. The interaction was verified for endogenous proteins from adult mouse kidney, demonstrating that this interaction occurs in vivo (Fig. 1F). Since slit diaphragm proteins are partially Triton X-100-insoluble in vivo, a special CHAPS-containing buffer was used to solubilize the glomeruli. Confirming the biochemical data, ZO-1 and Neph1 co-localized in the glomeruli of the native kidney (Fig. 1G). In addition to the SH3 domain and a guanylate kinase domain the multiadapter protein ZO-1 contains three PDZ domains. These PDZ domains are located within the first 503 amino acids, the region that confers binding to Neph1. We tested therefore which PDZ domain mediated the interaction with Neph1. Pull-down experiments revealed that the interaction specifically involved the first PDZ domain of ZO-1 ( Fig. 2A). Deletion of the last three amino acids of the carboxyl-terminal cytoplasmic tail of Neph1 (Thr-His-Val) completely abrogated binding (Fig. 2D). Interest-ingly mutation of the critical threonine at the Ϫ2 position of the PDZ domain binding motif to glutamate disrupted binding to ZO-1 (Fig. 2C). Since threonine-to-glutamate mutation may mimic phosphorylation of the threonine, this finding suggests that phosphorylation of the threonine at the Ϫ2 position regulates the interaction between ZO-1 and Neph1 as has been reported for other PDZ-dependent interactions (25)(26)(27). However, this hypothesis needs further clarification. The PDZ binding motif is highly conserved in all currently known Neph proteins (Fig. 2B), and as demonstrated in Fig. 2E, all Nephs bind to ZO-1 in transiently transfected HEK 293T cells; all interactions were mediated by the first PDZ domain of ZO-1 (data not shown). These data suggest that ZO-1 binding may be a general mechanism in the regulation of Neph proteins. FIG. 3. ZO-1 enhances tyrosine phosphorylation of the cytoplasmic tail of Neph1 and augments Neph1mediated AP-1 activation. A, HEK 293T cells were transfected with an AP-1-dependent luciferase construct and expression plasmids as indicated. Neph1 stimulated an ϳ4-fold increase in AP-1 activity, whereas ZO-1 expression had no effect. ZO-1 co-expression strongly augmented the Neph1-induced AP-1 activity (*, p Ͻ 0.05; n ϭ 3) but was without effect on the Neph1 expression level (lower panel). ZO-1 did not augment the AP-1 activation of the Neph1 mutant (T787E) lacking the ZO-1 interaction site. B, ZO-1 enhances tyrosine phosphorylation of the carboxyl terminus of Neph1. HEK 293T cells were transfected with the plasmids as indicated. Neph1 was precipitated, and precipitates were resolved by SDS-PAGE and blotted with anti-phosphotyrosine antibody (4G10, left panel) followed by anti-human IgG antiserum to confirm equal expression levels (right panel). C, densitometric analysis of non-saturated radiographs using the NIH Image software (**, p Ͻ 0.01, n ϭ 6). D, two-dimensional gel electrophoresis of Neph1 precipitates reveals that the increased tyrosine phosphorylation of the cytoplasmic tail of Neph1 is not a result of overlap from coprecipitating proteins but can be attributed to Neph1 itself. E and F, in vivo labeling with 32 P and phosphoamino acid analysis reveal that increased phosphorylation of immunoprecipitated Neph1 in the presence of ZO-1 (E) is a result of strongly augmented tyrosine phosphorylation and serine phosphorylation of the cytoplasmic tail of Neph1 (F, phosphoamino acid analysis). IEF, isoelectric focusing; WT, wild type; WB, Western blot; PY, phosphotyrosine; w/o, without.
It has been shown that ZO-1 may organize signal transduction, and multiple evidences demonstrate a structural role in organizing transmembrane protein complexes (16). We examined therefore whether ZO-1 expression modulates Neph1 signaling in HEK 293T cells. HEK 293T cells contain only very low levels of ZO-1 and Neph1 and represent an ideal model system to examine ZO-1 effects on Neph1 signaling. AP-1 activity, measured by luciferase assays, is a very sensitive parameter for the activation of several pathways including the c-Jun NH 2 -terminal kinase, p38, extracellular signal-regulated kinase 1/2, or phosphatidylinositol 3-kinase/AKT signaling cascades. Since ZO-1 can recruit signal transduction components including G proteins and kinases to cell-cell contacts and may interact with a variety of additional signal transduction components including Fyn, Grb-2, Crk, or glycogen synthase kinase-3␤ (scansite.mit.edu/), we analyzed the effect of ZO-1 on Neph1-mediated AP-1 activation (Fig. 3A). Co-expression of ZO-1 significantly accelerated the Neph1-mediated AP-1 activity, whereas ZO-1 alone had no effect (Fig. 3A). Augmentation of Neph1 signaling by ZO-1 requires direct binding of Neph1. Neph1-induced AP-1 activity was not modulated by ZO-1 after deletion of the PDZ binding motif in Neph1 or mutation of ZO-1 (data not shown). We have previously shown that the cytoplasmic tail of Neph1 is tyrosine-phosphorylated (12). Interestingly ZO-1 co-expression strongly augmented tyrosine phosphorylation of the carboxyl terminus of Neph1 (Fig. 3, B-F) but did not influence tyrosine phosphorylation of Neph1 T787E, a mutant that fails to interact with ZO-1 (data not shown). Of note, ZO-1 did not only enhance tyrosine phosphorylation but also augmented serine phosphorylation of the cytoplasmic tail of Neph1 as demonstrated by in vivo labeling and phosphoamino acid analysis (Fig. 3F). These findings suggest that ZO-1 organizes Neph1 complexes and helps to recruit signal transduction components to facilitate Neph1 signaling.
Recent work has highlighted the exquisite role of the slit diaphragm for the integrity of the glomerular filter (1,2,28,29). Several critical proteins including nephrin, Neph1, podocin, and CD2AP have been identified and were localized to the slit diaphragm (8, 30 -33). However, until now it is not clear how these different components are organized at the filtration slit. Our results suggest that the PDZ domain-containing multiadapter protein ZO-1 may help to cluster the transmembrane protein Neph1. This conclusion is based on two observations. First, ZO-1 specifically interacts with the carboxyl-terminal cytoplasmic tail of Neph1. This interaction, mediated by the first PDZ domain of ZO-1, was present in vivo and may be dynamically regulated by phosphorylation. Second, direct binding of ZO-1 to the carboxyl terminus of Neph1 dramatically alters the phosphorylation state of Neph1 and the ability to induce signal transduction. It is well established that PDZ domain proteins act as scaffolds for signaling complexes and serve to recruit signal transduction components (14,34). ZO-1 has been shown to regulate signaling involved in the control of cell polarity, the tightness of the paracellular seal, and transcriptional responses (35,36). Our findings extend the function of ZO-1 and suggest that ZO-1 is involved in the regulation of signaling by slit diaphragm proteins. It has been suggested that ZO-1 in addition to its signaling function interacts with the actin cytoskeleton and components of the paracellular seal (37). Although the functional implications of ZO-1/actin association have not yet been established in vivo ZO-1 could link Neph1 and its associated binding proteins to the actin cytoskeleton and contribute to the organization of the filtration slit. Given the importance of the submembranous actin cytoskele-ton for the maintenance of the podocyte architecture (1), it will be interesting to evaluate the exact role of ZO-1 for the organization of the podocyte actin cytoskeleton in animal models that selectively target the ZO-1-Neph1 interaction.