Visual Abstract
Abstract
Background Mutations in about 50 genes have been identified as monogenic causes of nephrotic syndrome, a frequent cause of CKD. These genes delineated the pathogenetic pathways and rendered significant insight into podocyte biology.
Methods We used whole-exome sequencing to identify novel monogenic causes of steroid-resistant nephrotic syndrome (SRNS). We analyzed the functional significance of an SRNS-associated gene in vitro and in podocyte-like Drosophila nephrocytes.
Results We identified hemizygous missense mutations in the gene TBC1D8B in five families with nephrotic syndrome. Coimmunoprecipitation assays indicated interactions between TBC1D8B and active forms of RAB11. Silencing TBC1D8B in HEK293T cells increased basal autophagy and exocytosis, two cellular functions that are independently regulated by RAB11. This suggests that TBC1D8B plays a regulatory role by inhibiting endogenous RAB11. Coimmunoprecipitation assays showed TBC1D8B also interacts with the slit diaphragm protein nephrin, and colocalizes with it in immortalized cell lines. Overexpressed murine Tbc1d8b with patient-derived mutations had lower affinity for endogenous RAB11 and nephrin compared with wild-type Tbc1d8b protein. Knockdown of Tbc1d8b in Drosophila impaired function of the podocyte-like nephrocytes, and caused mistrafficking of Sns, the Drosophila ortholog of nephrin. Expression of Rab11 RNAi in nephrocytes entailed defective delivery of slit diaphragm protein to the membrane, whereas RAB11 overexpression revealed a partial phenotypic overlap to Tbc1d8b loss of function.
Conclusions Novel mutations in TBC1D8B are monogenic causes of SRNS. This gene inhibits RAB11. Our findings suggest that RAB11-dependent vesicular nephrin trafficking plays a role in the pathogenesis of nephrotic syndrome.
The glomerular filter of the kidney is three-layered, consisting of a fenestrated endothelium, the glomerular basement membrane, and the podocytes that form the slit diaphragm. Disorders of the glomerular filter commonly involve the podocyte and manifest with edema, hypoalbuminemia, and severe proteinuria, a clinical triad that characterizes the nephrotic syndrome. Steroid-resistant nephrotic syndrome (SRNS) commonly entails declining renal function, and represents the second most frequent cause of ESRD in patients manifesting before 25 years of age.1 Mutations in about 50 different genes have been identified as monogenic causes of SRNS, including proteins of the slit diaphragm complex and actin regulators, but also factors of CoQ10 biosynthesis, nucleoporins, or members of the KEOPS complex.2–43 Many of these genetic causes of SRNS are rare; however, the discovery of these single gene mutations as molecular causes of SRNS contributed significantly to the understanding of the complex pathogenesis of nephrotic syndrome and podocyte biology.
There is mounting evidence supporting an essential role of vesicular trafficking via endocytosis for the function of the glomerular filter.44,45 Disrupting components of the endocytic machinery in mouse resulted in podocyte dysfunction with severe proteinuria.46–48 Nephrin, an essential component of the slit diaphragm, undergoes endocytic trafficking.49–55 Recently, we discovered mutations of the RAB5 interactors GAPVD1 and ANKFY1 as novel causes of nephrotic syndrome and the first endosomal regulators in the pathogenesis of human nephrotic syndrome.41 However, the functional role of endocytosis for the slit diaphragm is still unclear. In particular, the mechanistic role of the various aspects of the endocytic pathway for slit diaphragm formation and maintenance remains elusive. We now report hemizygous mutations of the endosomal regulator TBC1D8B, discovered by whole-exome sequencing (WES) in five individuals with nephrotic syndrome. Our analysis in vitro and in the Drosophila model supports a novel role of RAB11-dependent vesicular trafficking in the pathogenesis of the nephrotic syndrome.
Methods
Study Approval
Approval for human subjects research was obtained from the University of Michigan, University of Freiburg and the Boston Children’s Hospital Institutional Review Boards. All participants or their guardians provided written informed consent.
Study Participants
After obtaining informed consent, clinical data and blood samples were collected from individuals with nephrotic syndrome. Clinical data were acquired using an established questionnaire. The diagnosis of nephrotic syndrome was made by (pediatric) nephrologists, on the basis of standardized clinical and renal histologic criteria. Renal biopsy specimens were evaluated by renal pathologists.
Homozygosity Mapping, Whole-Exome Resequencing, and Mutation Calling
Homozygosity mapping, whole-exome resequencing, and mutation calling were performed as described previously.23
Plasmids, siRNAs, Cell Culture, and Transfection
Murine full-length Tbc1d8b cDNA was subcloned by PCR from full-length murine cDNA (GenBANK BC147581.1; BioCAT GmbH). Human TBC1D8B was subcloned by PCR from cDNA representing the shorter isoform 2 (GenBank BC122564; GE Dharmacon). Truncation constructs were generated by PCR. Primers are shown in Supplemental Table 1.
The following expression vectors were used: pRK5-N-Myc, pCDNA6.2-C-GFP, and pCDNA6.2-N-GFP. The mutations identified in individuals with nephrotic syndrome were introduced into the cDNA constructs by site-directed mutagenesis via Gibson assembly (New England Biolabs). The following constructs were obtained from Addgene: pSpCas9(BB)-2A-GFP (PX458) (#48138), pX459V2.0-eSpCas9 (#108292), GFP-RAB11 WT (#12674), and GFP-RAB7 (#61803) NPHS1 and RAB5 cDNA constructs have been described elsewhere.41
The TBC1D8B-specific and control siRNAs were purchased from GE Dharmacon.
Overexpression experiments were performed in HEK293T cells or immortalized human podocytes that were a gift from Dr. Moin Saleem (University of Bristol, Bristol, UK).
HEK293T cells were maintained in DMEM, supplemented with 10% FBS, 50 IU/ml penicillin, and 50 μg/ml streptomycin. Podocytes were maintained in RPMI plus GlutaMAX-I (Gibco) supplemented with 10% FBS, 50 IU/ml penicillin/50 μg/ml streptomycin, and insulin-transferrin-selenium X.
Plasmids and siRNAs were transfected into HEK293T cells at 37°C or podocytes grown at the permissive temperature of 33°C using Lipofectamine 2000 (Invitrogen) or polyethyleneimine (Polyplus-transfection or Sigma).
Nephrin Trafficking Assay
MDCK strain II cells (wt MDCK) were a gift from Enrique Rodriguez-Boulan (Weill Cornell Medical College, NY) and were maintained in DMEM supplemented with 5% FCSFCS, at 37°C and 5% CO2 in 10 cm culture dishes, and passaged every 2–3 days.
For growing polarized monolayers, cells were cultured on Transwell filters (#3401; from Corning Costar) for 4 days. Transfections with a plasmid encoding a nephrin construct (pCAD4-HA-nephrin-mCherry) comprising 4× conditional aggregation domains (CADs), furin cleaving site, HA-tag, nephrin, and mCherry (pCAD4-HA-nephrin-mCherry) were carried out using lipofectamine 2000 (Thermo Fisher Scientific). D/D-solubilizer (Clontech/Takara) was used to induce endoplasmic reticulum (ER) release at a concentration of 5 μM, and 100 µg/ml of cycloheximide was added during release to prevent further protein synthesis. Staining of nephrin arriving at the cell surface was carried out with anti-HA antibodies (Covance) conjugated to Cy5 that were applied to the apical or basolateral side of live cells.
For immunofluorescence, cells were fixed with 4% formaldehyde for 15 minutes at room temperature. After excising filters with a scalpel, they were mounted in DABCO-medium and imaged using Nikon A1R confocal microscope. Quantification of cell surface arrival of nephrin was done as described previously, with a custom-written MATLAB program.56,57
Immunoblotting, Immunoprecipitation, Pull-Down Assay, and Immunofluorescence Staining
Immunoblotting, immunoprecipitation, and immunofluorescence staining were performed as described previously.58 Briefly, HEK293T were lysed and precleared using rec-Protein A-Sepharose 4B Conjugate (Life Technologies) overnight. Then, equal amounts of protein were incubated with EZview Red Anti-c-Myc Affinity Gel (Sigma-Aldrich). Coimmunoprecipitation experiments were performed in three independent experiments. For chloroquine treatment, cells were exposed to culture medium containing 80 µM chloroquine for 24 hours. Immunoblotting was performed using mouse anti-TBC1D8B (SC-376637; Santa Cruz Biotechnology), rabbit anti-RAB11 (#5589; Cell Signaling Technology), mouse anti–c-Myc (sc-40; Santa Cruz Biotechnology), rabbit anti–c-Myc (sc-789; Santa Cruz Biotechnology), mouse anti-GFP (sc-9996; Santa Cruz Biotechnology), rabbit anti-GAPDH (14C10/2118; Cell Signaling Technology), mouse anti-actin (ab20272; Abcam), rabbit anti-LC3B (2775; Cell Signaling Technology), and rabbit anti-GFP (sc-8334; Santa Cruz Biotechnology).
Immunofluorescence of TBC1D8B was performed with rabbit anti-TBC1D8B (ab121780; Abcam) or mouse anti-TBC1D8B (SC-376637). Other antibodies used were rabbit anti-RAB11 (#5589S; Cell Signaling Technology), and mouse anti-Myc (9E10; Developmental Studies Hybridoma Bank).
Fluorescence images were obtained with a Zeiss LSM 880 laser scanning microscope, including the application of Airyscan technology for super-resolution microscopy.
Drosophila Studies
The Drosophila melanogaster stable RNAi stocks Tbc1d8b RNAi (#32929) and Rab11 RNAi (#42709) were obtained from the Bloomington Drosophila Stock Center (BDSC). Prospero-GAL459 or Dorothy-GAL4 (# 6903; BDSC) with or without tub-GAL80ts (#7018; BDSC) were used to control expression in garland cell nephrocytes. RAB11 overexpression was achieved using wild-type GFP-Rab11 (# 8506; BDSC). CRISPR/Cas9-mediated loss of function was generated in modification of an established protocol,60 using Dorothy-GAL4 to direct Cas9-expression (#58985) in nephrocytes. The CRISPR gRNA construct targeting CG7324 was generated by introducing two gRNAs via PCR and Gibson assembly into the pCFD4 plasmid (Addgene; #49411). The DNA was injected into flies expressing phiC31 integrase under vasa promoter with an attP landing site in 51C by the fly facility of the Department of Genetics at the University of Cambridge (Cambridge, UK). RNAi crosses were raised at 30°C. Flies expressing gRNAs and a Cas9 variant were raised at 25°C as higher temperatures induced unspecific toxicity by Cas9 expression alone (data not shown).
Fluorescent tracer uptake in nephrocytes was performed as previously described.61 Briefly, nephrocytes were dissected in PBS and incubated with FITC-albumin (Sigma) for 30 seconds. After a fixation step of 5 minutes in 8% paraformaldehyde cells were rinsed in PBS and exposed to Hoechst 33342 (1:1000) for 20 seconds and mounted in Roti-Mount (Carl Roth). Cells were imaged using a Zeiss LSM 880 laser scanning microscope. Quantitation of fluorescent tracer uptake was performed with ImageJ software. The results are expressed as a ratio to a control experiment with GFP RNAi that was done in parallel.
For immunohistochemistry, nephrocytes were dissected, fixed for 20 minutes in PBS containing 4% paraformaldehyde, and stained according to the standard procedure. The following primary antibodies were used: rabbit anti-sns62 (1:500, gift from S. Abmayr) and guinea pig anti-Kirre63 (1:200, gift from S. Abmayr). For imaging, a Zeiss LSM 880 laser scanning microscope was used. Image processing was done by ImageJ and GIMP software.
For transmission electron microscopy nephrocytes were dissected and fixed in 4% formaldehyde and 0.5% glutaraldehyde in 0.1 M cacodylate buffer at pH 7.4. Transmission electron microscopy was carried out using standard techniques.
Statistical Analyses
Paired t test was used to determine the statistical significance between two interventions. ANOVA followed by Dunnett’s correction (unless otherwise indicated) was used for multiple comparisons (GraphPad Prism software). Asterisks indicate significance as follows: *P<0.05; **P<0.01; ***P<0.001. A statistically significant difference was defined as P<0.05. Error bars indicate SD.
Results
Novel Mutations in TBC1D8B Cause SNRS
The known monogenic causes of SRNS explain only a fraction of idiopathic cases. We performed WES to discover novel mutations. After excluding relevant mutations in known SRNS genes, we identified hemizygous mutations of the gene TBC1 Domain Family Member 8B (TBC1D8B) in five individuals from five different families with nephrotic syndrome (Figure 1, A–F, Supplemental Figure 1, A–G, Table 1). All patients were male, and the available information was compatible with an X-linked inheritance. One heterozygous female conductor exhibited borderline proteinuria (B931, Figure 1, E and I, Supplemental Figure 1, B and C). Such information was not available for the other families. In two families (J12190/18 and B3482), we detected mutations introducing premature stop codons (c.1030C>T, p.Arg344* and c.1383G>A, p.Trp461*, respectively), while three families showed hemizygous missense mutations (c.2338A>T, p.Thr780Ser, c.1316T>G, p.Phe439Cys, and c.190C>T, p.Arg64Cys). The altered amino acid residues are well conserved in evolution (Figure 1, G–I, Table 1). Nephrotic syndrome was described as steroid-resistant in all but one patient (A2563). Extrarenal symptoms were not reported in any of these patients. The histology was described as FSGS in three cases (B931, A2563, A3803, Figure 1J, Table 1) and mesangioproliferative GN (MesPGN) in one case (J12190/18). No biopsy result was available for B3482. Supporting the significance of our findings, mutations of TBC1D8B were also reported most recently in two families with SRNS and X-linked inheritance.64
WES identifies recessive mutations in TBC1D8B in five families with nephrotic syndrome. (A) Schematic of TBC1D8B cDNA with the corresponding protein, including its functional domains. Arrows indicate the position of five hemizygous mutations of TBC1D8B that were identified by WES in individuals with nephrotic syndrome from five families (A2563, A3803, B3482, B931, and J12190/18). (B–F) Shown are Sanger chromatograms of the respective regions of TBC1D8B in which the individual mutation of each of the five patients is located. (G) Alignment of TBC1D8B amino acid sequences for Homo sapiens, Mus musculus, Gallus gallus, Xenopus tropicalis, Danio rerio, Ciona intestinalis, Caenorhabditis elegans, Drosophila melanogaster, and Saccharomyces cerevisiae demonstrates conservation of the residue Arginine 64. (H) Alignment of amino acid sequence of TBC1D8B for H. sapiens, M. musculus, X. tropicalis, and D. rerio, demonstrates conservation of the residue Leucine Phenylalanine 439. (I) Alignment of amino acid sequence of TBC1D8B for H. sapiens, M. musculus, G. gallus, X. tropicalis, and D. rerio indicates conservation of the residue Threonine 780. (J) Renal histology (periodic acid–Schiff base staining) of patient B931_21 with the Thr780Ser mutation of TBC1D8B shows segmental glomerulosclerosis.
Mutations of TBC1D8B in five families with nephrotic syndrome.
TBC1D8B Specifically Binds Active RAB11A and RAB11B and the Mutations from Patients with Nephrotic Syndrome Affect Interaction with Endogenous RAB11
TBC1D8B is a member of a family of more than 40 TBC-proteins that share the eponymous Tre-2-Bub2-Cdc16 (TBC) domain.65 This domain commonly confers a functional role as a GTPase-activating protein (GAP) for specific Rab-GTPases, the master regulators of vesicular trafficking including endocytosis. TBC1D9B, closely related to TBC1D8B, has been described as a GAP specific for RAB11A66 that plays a role in endocytic recycling, autophagy, and exocytosis. We performed coimmunoprecipitation assays, comparing the binding of TBC1D8B to RAB11A, the early endosomal RAB5A, and the late endosomal RAB7A. We observed strongest binding toward RAB11A (Figure 2A, quantitation Figure 2B). Rab proteins cycle between an active GTP-bound and an inactive GDP-bound state. GAP proteins promote the conversion to the inactive GDP-bound state through catalyzing GTP hydrolysis, demonstrating specific binding to the active form (Figure 2C). To evaluate such specific binding, we generated constitutively active and dominant negative variants of three Rab proteins that are involved in endocytic recycling: RAB4A (Q72L and S27N), RAB11A (Q70L and S25N), and RAB11B (S25N and Q70L). We performed coimmunoprecipitation, using these variants together with TBC1D8B, and quantified the relative band density of the respective dominant negative and constitutively active Rab protein. We consistently observed a binding ratio >1 (active/inactive Rab) for both RAB11A and RAB11B, but not RAB4A (Figure 2D, quantitation Figure 2E). We introduced TBC1D8B mutations discovered in patients with nephrotic syndrome (W461*, F439C, T780S, and R64C) into the corresponding amino acid residues of the murine full-length Tbc1d8b (W460*, F438C, T779S, and R64C). The mutation p.Arg344* was not investigated in light of the early premature stop codon preceding Trp461. Analyzing the binding affinity to endogenous RAB11 via coimmunoprecipitation in HEK cells, we observed a significant reduction for mutant proteins, except Tbc1d8bF438C (Figure 2F, quantitation Figure 2G). Mutations of TBC1D8B that cause nephrotic syndrome thus affect the interaction with endogenous RAB11.
TBC1D8B specifically binds to active forms of RAB11A and RAB11B and patient derived mutations of TBC1D8B affect binding to endogenous RAB11. (A) Upon overexpression and coimmunoprecipitation with anti-Myc antibody in HEK293T cells, GFP-tagged RAB11A precipitates with Myc-TBC1D8B whereas GFP-tagged RAB5A and RAB7A show weaker binding. (B) Quantitation of densities from (A) depicts a significantly weaker affinity of GFP-RAB5A and GFP-RAB7A toward Myc-TBC1D8B compared with GFP-RAB11A. Densitometry results from (A) are expressed as ratio of Rab protein/TBC1D8B (n=4, P<0.001 for RAB11 versus RAB5, P<0.001 for RAB11 versus RAB7 and P>0.5 for RAB5 versus RAB7). Sidak post hoc analysis was used to correct for multiple comparisons.). (C) Schematic showing a Rab protein shuttling between active (GTP-bound) and inactive (GDP-bound) states. GAP or TBC proteins predominantly bind to the active Rab protein (GTP-bound) to catalyze GTP hydrolysis, promoting the inactive state. (D) Overexpression and coimmunoprecipitation of GFP-tagged RAB11A, RAB11B, and RAB4A dominant negative (dom. neg.) and constitutively active (const. act.) constructs together with Myc-TBC1D8B. RAB11A and RAB11B constructs each interact with TBC1D8B. The constitutively active forms (†) show a stronger affinity toward TBC1D8B than the dominant negative forms (*). Binding affinity of RAB4 constructs is weaker. (E) Quantitation of density from precipitates analogous to (D) normalized to respectively precipitated TBC1D8B protein shown as a ratio of the respective constitutively active form and the dominant negative form. Ratios are consistently >1 for RAB11A and RAB11B but not RAB4, indicating stronger binding of the active Rab protein for RAB11. (F) Murine Tbc1d8b cDNA constructs that reflect the mutations from patients with nephrotic syndrome exhibit reduced binding affinity toward endogenous RAB11, except Tbc1d8bF438C (after correcting for the amount of TBC1D8B). The RAB11 antibody cannot discriminate between endogenous RAB11A and RAB11B. Please note that the mildly reduced interaction for the mutation F439C was not confirmed in other experiments [see quantitation in (G)]. (G) Quantitation of densities from (F) confirms a significantly reduced affinity of Tbc1d8b mutants to GFP-RAB11B except F438C. Densitometry results from (F) were expressed as endogenous RAB11/Tbc1d8bwild-type/mutant (n=3–5, P<0.01 for R64C, P<0.01 for T779S, P>0.05 for F438C, *P<0.05 for W460).
Silencing of TBC1D8B Upregulates RAB11-Dependent Autophagy and Exocytosis Suggesting a Function as GAP for RAB11
To evaluate endogenous expression of TBC1D8B in podocytes and HEK293T cells, we performed immunoblotting, and used transient siRNA transfection and CRISPR/Cas9-mediated loss of function.67 Podocytes and HEK293T cells exhibited endogenous expression of the long, full-length isoform of TBC1D8B that was reduced upon both loss-of-function strategies (Figure 3A, Supplemental Figure 2C). To study the subcellular localization of endogenous TBC1D8B protein, we tested the conditional CRISPR/Cas9-mediated loss of function in individual cells marked by GFP.67 We did not observe a consistent reduction of the antibody signal in cells with TBC1D8B loss of function, using two available antibodies (Supplemental Figure 3, A–C). Overexpression constructs of TBC1D8B, including murine Tbc1d8b constructs representing the patient mutations, predominantly localized to the cytosol (Supplemental Figure 3, D–I), but did not demonstrate overt colocalization with overexpressed or endogenous RAB11 protein (Supplemental Figure 2, A–B’’). The localization of the endogenous protein remains unclear.
Silencing TBC1D8B increases autophagy and exocytosis, suggesting a role as an RAB11-GAP. (A) Transient, siRNA-mediated silencing of TBC1D8B in human immortalized podocytes and HEK293T cells is demonstrated by immunoblot. This indicates endogenous expression of TBC1D8B in both cell lines and confirms the efficiency of the siRNAs. (B) Immunoblotting with anti-LC3B reveals an increase of basal autophagy in podocytes and HEK293T cells upon silencing TBC1D8B. Anti-GAPDH signal serves as loading control. (C) Quantitation of densities from podocyte data in (B) confirms a significant increase of the signal from anti-LC3B expressed as a ratio to the loading control (n=3–4, P<0.05 for siRNA-1 and P<0.01 for siRNA-2). (D) Immunoblotting with anti-LC3B indicates a higher difference between control treatment and chloroquine exposure for TBC1D8B siRNAs compared with control siRNA. Anti-GAPDH signal serves as loading control. Representative blot from three consecutive experiments is shown. (E) Western blot reveals increased delivery of GFP to the supernatant (above) upon siRNA-mediated silencing of TBC1D8B and overexpression of constitutively active RAB11B in HEK293T while the amount of intracellular GFP (lysates) is comparable. (F) Quantitation of densities from (E) confirms a significant increase of the GFP secretion for two independent siRNAs directed against TBC1D8B. Density results are expressed as ratio of GFP signal in the supernatant to GFP within the lysates (n=4–5, P<0.05 for siRNA-1 and P<0.01 for siRNA-2).
Promotion of autophagosome maturation is an established cellular function of RAB11.68–70 To confirm a role of TBC1D8B as an RAB11-GAP, we studied basal autophagy upon silencing of TBC1D8B in HEK293T cells and podocytes. Using LC3B/GAPDH-ratio as a read-out, immunoblotting indicated an increase of autophagy in both cell lines (Figure 3, B and C, Supplemental Figure 2D). To evaluate autophagic flux, we tested the effect of lysosomal inhibition via chloroquine. The difference between control and chloroquine treatment was more pronounced upon silencing of TBC1D8B, suggesting a higher autophagic flux (Figure 3D). As autophagy might be induced independent of RAB11, we tested an effect of TBC1D8B siRNA on exocytosis, another cellular process that requires RAB11 function71 without direct connection to autophagy. To study basal exocyst activity, we generated a secretory form of GFP by introducing an IFNA2-derived signal peptide into the N terminus of GFP. Upon transient transfection, HEK293T cells constitutively secreted GFP into the surrounding medium (Supplemental Figure 2D). Both, overexpression of constitutively active RAB11B and silencing of TBC1D8B increased secretion of GFP (Figure 3E, quantitation Figure 3F, Supplemental Figure 2, E and F). Thus, basal activity of both autophagy and exocytosis were increased upon knockdown of TBC1D8B. Taken together, these findings suggest a disinhibition of endogenous RAB11 activity being well compatible with a function of TBC1D8B as an RAB11-specific GAP.
TBC1D8B Interacts with the Slit Diaphragm Protein Nephrin and Regulates its Trafficking
The slit diaphragm protein nephrin is subject to endocytic trafficking.72 We hypothesized that TBC1D8B plays a role in vesicular trafficking of nephrin as the underlying pathogenesis for glomerular dysfunction. Performing coimmunoprecipitation experiments in HEK293T cells, we found that Myc-nephrin precipitated with human GFP-TBC1D8B (Figure 4A) and murine GFP-Tbc1d8b (Supplemental Figure 4A). Conversely, we observed that the GFP-tagged intracellular domain (ICD) of nephrin precipitated with Myc-TBC1D8B (Supplemental Figure 4B). The N-terminal half of the nephrin ICD was sufficient to bind Myc-TBC1D8B, whereas the C-terminal ICD failed to bind (Figure 4B). Truncation mapping thus identifies an interacting domain of 76 amino acids on the nephrin C-terminal tail (Figure 4C). We evaluated binding of mutant Tbc1d8b by coimmunoprecipitation and observed a reduced binding to Myc-nephrin for all mutants (Figure 4D, quantitation Figure 4E). The mutations of TBC1D8B that cause nephrotic syndrome thus impair the interaction with nephrin. In podocytes, overexpressed GFP-TBC1D8B and Myc-nephrin colocalized completely within vesicles (Figure 4, F–F’’). We confirmed colocalization of both proteins in HEK cells and partially in MDCK cells (Supplemental Figure 4, C–D’’).
The N-terminal cytosolic domain of nephrin interacts with TBC1D8B and both proteins co-localize in cultured cell lines. (A) Upon overexpression in HEK293T cells and coimmunoprecipitation, GFP-tagged TBC1D8B precipitates with Myc-tagged nephrin. (B) Upon overexpression in HEK293T cells and coimmunoprecipitation, the GFP-tagged N-terminal section of the ICD of nephrin (aa 1084–1160) precipitates with Myc-tagged TBC1D8B, whereas the C-terminal section of the ICD (aa 1160–1241) shows no interaction. (C) A schematic of truncation constructs of nephrin indicates their ability to interact with TBC1D8B by “+” (interaction) or by “–” (lack of interaction). The amino acid sequence of the interacting domain is noted below. (D) Murine Tbc1d8b cDNA constructs that reflect the mutations from patients with nephrotic syndrome exhibit reduced binding affinity toward Myc-nephrin (correcting for the amount of Myc-nephrin). (E) Quantitation of densities from (D) confirms a significantly reduced affinity of Tbc1d8b mutants to Myc-nephrin. Densitometry results are expressed as Tbc1d8bwild-type/mutant/Myc-nephrin (n=5, P<0.001 for F438C, P<0.001 for R64C, *P<0.001 for W460, and P<0.001 for T779S). (F) Overexpressed GFP-nephrin and Myc-TBC1D8B (human) colocalize in immortalized human podocytes in vesicles (see also enlarged insets). Scale bar represents 10 µm. (G) Schematic of the principle of the nephrin trafficking assay and the nephrin cDNA construct (pCAD4-HA-nephrin-mCherry) that was applied. CADs and an HA-tag are introduced into the nephrin N terminus, whereas an mCherry-tag resides in the C terminus. Without induced ER release, the expressed nephrin forms CAD-mediated clusters and is retained within the ER (left). Addition of the membrane-permeable D/D solubilizer causes dispersal of the clusters causing a synchronized ER release. Although the CADs are removed by furin cleavage in the Golgi, the double-tagged nephrin continues trafficking. Protein reaching the apical surface can be visualized by extracellular HA-staining, as nephrin is a type 1 transmembrane protein. (H) Before an ER release is induced, an MDCK cell transfected with pCAD4-HA-nephrin-mCherry [described in (F)] lacks extracellular HA signal, whereas the C-terminal mCherry indicates a perinuclear localization (upper panel). There is no indication of leakiness of the desired ER retention. Upon induction of ER release extracellular HA-staining becomes positive, indicating an increasing amount of tagged nephrin protein is subject to trafficking toward the apical surface (middle and lower panels). Scale bar represents 10 µm. (I) MDCK cells coexpressing pCAD4-HA-nephrin-mCherry either with GFP (control, upper panel), wild-type GFP-Tbc1d8b (middle panel), or the mutant Tbc1d8b-T779S (derived from patient with SRNS, lower panel). Apical delivery of pCAD4-HA-nephrin-mCherry is impaired by expression of wild-type Tbc1d8b but not the mutant protein. Scale bar represents 10 µm. (J) Quantitation of results from (I). Shown is the fluorescence intensity from the extracellular HA-staining in ratio to total nephrin (mCherry). Overexpression of wild-type Tbc1d8b but not mutant Tbc1d8b significantly reduces apical delivery of nephrin (>50 cells per group; P<0.05 for Tbc1d8b wild-type and P>0.05 for T779S).
To study biosynthetic trafficking of nephrin directly, we utilized an approach on the basis of conditional aggregation domains (CAD) that has been used for other proteins.56,57 Adapting this approach for nephrin, we introduced CADs to the N terminus of nephrin (pCAD4-HA-nephrin-mCherry, Figure 4G). CADs are expected to induce retention of newly synthesized protein within the ER by formation of aggregates. Such aggregates can be dispersed by adding a membrane permeable small molecule (D/D solubilizer), which entails rapid, synchronized ER release of the protein and consecutive removal of the CAD in the Golgi. The presence of two tags (HA-tag and mCherry) eventually permits separate detection of nephrin at the surface by extracellular staining (HA) and the total protein (mCherry, Figure 4G). We chose MDCK cells that form an epithelial monolayer with apico-basal polarity for this assay because these cells had previously been used for this approach.56,57 Studying nephrin trafficking this way, we observed no nephrin on the cell surface before the induction of ER release (Figure 4H, upper panels), indicating successful ER retention. Upon exposure to D/D solubilizer, we observed trafficking of nephrin toward the apical cell surface over time (Figure 4H, lower panels). This suggests that nephrin travels initially to the apical surface. Using this novel assay for nephrin trafficking, we studied the effect of overexpressed Tbc1d8b within this system. Interestingly, overexpression of murine full-length Tbc1d8b reduced the apical delivery of nephrin compatible with an inhibition of RAB11-dependent trafficking, but one of the patient-derived mutations failed to have a similar effect in these cells (Figure 4I, quantitation Figure 4J). These findings indicate that TBC1D8B regulates nephrin trafficking and support a pathogenetic role of Rab11-dependent vesicular trafficking of nephrin in nephrotic syndrome. Basolateral delivery of nephrin was minimal in all conditions (Supplemental Figure 4, E and F).
Loss of Function of the Drosophila Ortholog of TBC1D8B Impairs Nephrocytes and Results in Mistrafficking of Fly Nephrin
To validate a role of TBC1D8B in nephrin trafficking in vivo, we used the Drosophila model that harbors the podocyte-like nephrocytes. Within these cells, the ortholog of nephrin forms autocellular slit diaphragms across membrane invaginations. Nephrocytes have been established as a model for monogenic forms of nephrotic syndrome.61 By sequence analysis we identified the uncharacterized Drosophila gene CG7324 as the ortholog of human TBC1D8B (Figure 5A). This gene encodes a protein of 1256 amino acids that contains two GRAM domains and one TBC domain like the human TBC1D8B. We introduce the term Tbc1d8b for the Drosophila ortholog CG7324.
Silencing the Drosophila ortholog of TBC1D8B in nephrocytes affects slit diaphragm localization and nephrocyte function. (A) Amino acid sequence of human TBC1D8B is 29% identical to the Drosophila ortholog CG7324, which shares the two GRAM domains and the TBC domain. Scale bar represents 5 µm throughout the figure. (B) Shown is FITC-albumin endocytosis after exposure for 30 seconds as an established assay of nephrocyte function. Silencing the TBC1D8B ortholog by RNAi using prospero-GAL4 (upper panels) or a conditional CRISPR/Cas9-mediated approach (lower panels) in nephrocytes significantly reduces uptake of FITC-albumin compared with the respective controls. Nuclei are marked by Hoechst 33342 in blue here and throughout the figure. (C) Quantitation of results from (B) in ratio to a control experiment performed in parallel (n=5–7 per genotype, P<0.01 for Tbc1d8b-RNAi and P<0.001 for Tbc1d8b-gRNA/Cas9). Sidak post hoc analysis was used to correct for multiple comparisons. (D–D’’) Equatorial cross section of a control garland cell nephrocyte (pros-GAL4/+) costained for the nephrin ortholog Sns (green) and the KIRREL/NEPH1 ortholog Kirre (red). Slit diaphragm proteins localize at the cell periphery in a fine line. Inset depicts a magnification from the area marked by a white box, and regular spaced dots corresponding to individual slit diaphragms are discernible. Sns and Kirre are restricted to the plasma membrane. Scale bar represents 5 µm throughout the figure, unless otherwise indicated. (E–E’’) Surface section from control garland cell nephrocytes reveals the typical fingerprint-like staining pattern. Inset depicts subcortical section of the same cell with absence of slit diaphragm protein. (F–F’’) Equatorial cross section of garland cell nephrocyte expressing Tbc1d8b RNAi shows appearance of fine linear protrusions of slit diaphragm protein from the membrane toward the interior of the cell (arrowheads, see also magnified inset). (G–G’’) Surface section from garland cell nephrocyte expressing Tbc1d8b RNAi reveals the typical fingerprint-like staining pattern but individual slits seem brighter. Inset depicts subcortical section of the same cell with appearance of fine lines of slit diaphragm protein (arrowheads). (H) Electron microscopy (EM) image from a section through the surface of a control nephrocyte reveals regular slit diaphragms (black arrowheads) bridging the membrane invaginations called labyrinthine channels. Scale bar represents 200 nm. (I) EM image from a section through the surface of a garland cell nephrocyte expressing Tbc1d8b RNAi demonstrates regular slit diaphragms on the surface (black arrowheads) but also formation of additional, ectopic slit diaphragms deeper within the cell (red arrowheads). Scale bar represents 200 nm. (J) EM image from a section through the surface of a garland cell nephrocyte with CRISPR/Cas9-mediated loss of Tbc1d8b confirms the findings with Tbc1d8b RNAi. Scale bar represents 200 nm. (K–K’’) Equatorial cross section of a control garland cell nephrocyte (Dot>Cas9/+) costained for the nephrin ortholog Sns (green) and the KIRREL/NEPH1 ortholog Kirre (red) reveals a regular staining pattern. (L–L’’) Surface section from control garland cell nephrocytes (Dot>Cas9/+) reveals the typical fingerprint-like staining pattern. Inset shows subcortical section with absence of slit diaphragm protein. (M–M’’) Equatorial cross section of garland cell nephrocyte with conditional CRISPR/Cas9-mediated loss of Tbc1d8b demonstrates appearance of extensive linear and circular lines of slit diaphragms protruding from the surface but also without apparent connection to the membrane (see also magnified inset). Individual cells lack all slit diaphragm protein on a segment of the membrane (arrowhead). (N–N’’) Surface section from garland cell nephrocyte with conditional CRISPR/Cas9-mediated loss of Tbc1d8b reveals a reduction of lines of slit diaphragm proteins with irregular spacing and gaps (arrowheads). Inset depicts subcortical section of the same cell with appearance of extensive linear and circular lines of slit diaphragm proteins.
To study loss of function of Tbc1d8b within this invertebrate podocyte model, we expressed Tbc1d8b RNAi within nephrocytes using prospero-GAL4. Studying the uptake of FITC-albumin, an established read-out of nephrocyte function,61 we observed a significant functional impairment of nephrocytes compared with control conditions (Figure 5B, quantitation Figure 5C). To confirm this finding, we generated a conditional CRISPR/Cas9-mediated loss of Tbc1d8b in nephrocytes. Combining nephrocyte-restricted Cas9- expression by Dorothy-GAL4 with stable ubiquitous expression of a tandem Tbc1d8b gRNA, we noted a reduction of nephrocyte function that slightly exceeded the effect of the RNAi (Figure 5B, quantitation Figure 5C). We stained the slit diaphragm proteins Sns (ortholog of nephrin) and Kirre (ortholog of NEPH1) in nephrocytes expressing Tbc1d8b RNAi, and observed protrusions of these proteins from the surface in a fine linear pattern that also occurred solitarily (insets, Supplemental Figure 5, A–B’’). These protrusions were absent in control nephrocytes (compare Figure 5, F–F’’ and D–D’’). With a lower penetrance, we further observed patches lacking slit diaphragms entirely upon expression of Tbc1d8b RNAi (Supplemental Figure 5, A–B’’). Typically, the classic fingerprint-like staining pattern of slit diaphragm proteins on the surface was preserved upon expression of Tbc1d8b RNAi, whereas individual lines appeared brighter (compare Figure 5, panels E and G). The fine linear protrusions of slit diaphragm protein from the surface were visible subcortically when Tbc1d8b RNAi was expressed but absent from subcortical control sections (compare insets Figure 5, E and G). To understand the ultrastructural basis of the protrusions of slit diaphragm protein from the membrane, we performed electron microscopy and observed the formation of additional, ectopic slit diaphragms deeper within the membrane invaginations (Figure 5I, control Figure 5H). These ectopic slit diaphragms obviously corresponded to our observations using immunofluorescence (Figure 5, F–G’’). To confirm our findings in an independent loss-of-function approach, we stained Sns/Kirre in nephrocytes with CRISPR/Cas9-mediated loss of Tbc1d8b compared with control conditions (Cas9 expression without gRNA, Figure 5, K–L’’). Upon loss of Tbc1d8b, we observed a phenotype that was more pronounced but matched our findings using the Tbc1d8b RNAi (Figure 5, J and M–N’’). The more severe phenotype upon CRISPR/Cas9-mediated loss of function compared with RNAi expression matched the stronger reduction of FITC-albumin endocytosis and suggested an incomplete knockdown using RNAi. Taken together, these findings indicate mistrafficking of fly nephrin with mislocalization of slit diaphragms upon loss of Drosophila Tbc1d8b.
Loss of Function of Rab11 Impairs Trafficking of Slit Diaphragm Proteins in Nephrocytes
Our findings suggest a functional role of RAB11 for trafficking of nephrin. To evaluate such a role of Drosophila Rab11 for trafficking of slit diaphragm proteins in nephrocytes, we silenced Rab11 acutely by modifying the GAL4-dependent expression using the temperature-sensitive Gal80ts that suppresses GAL4 at nonpermissive temperatures. Acute Rab11 silencing resulted in localized loss of slit diaphragms (Figure 6, A–A’’) and shortened lines of slit diaphragm protein on the surface (Figure 6, B–B’’), whereas nephrocytes from animals raised at nonpermissive temperatures showed regular slit diaphragms (Supplemental Figure 5, C–C’’). Prolonged expression of Rab11 RNAi caused extensive loss of slit diaphragm protein from the membrane (Figure 6, C–C’’) with appearance of puncta of nephrin on the surface (insets Figure 6, D–D’’) and clusters of Kirre protein between nephrocytes in absence of fly nephrin (Figure 6, D’–D’’). To study a mild gain of function similar to the effect of a defective Rab11-GAP, we overexpressed GFP-tagged Drosophila Rab11 in nephrocytes and observed appearance of ectopic vesicles of fly nephrin and fine protrusions of fly nephrin from the membrane (Figure 6, E–E’’). Thus, overexpression of Rab11 partially phenocopies loss of Tbc1d8b. However, overexpression of wild-type Rab11 produced an overtly milder phenotype than lack of Tbc1d8b (compare Figure 5F’ to Figure 6E’). This may be explained by additional Rab11-independent functions of Tbc1d8b, low efficiency of the exogenous wild-type Rab11 expression, or divergent compensatory regulation. Taken together, these findings indicate that slit diaphragm formation in nephrocytes requires Rab11 and they are compatible with a function of Tbc1d8b as an Rab11-GAP in Drosophila.
Rab11 is required for slit diaphragm formation in Drosophila. (A–A’’) Equatorial cross-section of garland cell nephrocyte acutely expressing Rab11 RNAi after shift to permissive temperature (30°C) for 2 days. Cells are costained for the nephrin ortholog Sns (green) and the KIRREL/NEPH1 ortholog Kirre (red), revealing localized gaps of slit diaphragm protein on the cell surface (arrowhead, see also magnified insets with rarefied puncta corresponding to misspaced slit diaphragms). Scale bar represents 5 µm throughout the figure unless otherwise indicated. (B–B’’) Surface section of same cell as (A) shows localized thinning and shorter lines of slit diaphragm proteins (see also magnified inset). (C–C’’) Equatorial cross section of garland cell nephrocyte expressing Rab11 RNAi for 4 days after shift to permissive temperature (30°C). Cells are costained for the nephrin ortholog Sns (green) and the KIRREL/NEPH1 ortholog Kirre (red), revealing extensive loss of slit diaphragms from the membrane (arrowhead, see also magnified insets with rarefied puncta corresponding to misspaced slit diaphragms). (D–D’’) Surface section of same background and staining as (C) reveals local clustering of Kirre without presence of Sns (arrowhead) and localized thinning and shorter lines of slit diaphragm protein (see also magnified inset). Lines of slit diaphragm proteins become shortened increasingly to become puncta or miss entirely (see also enlarged inset). (E–E’’) Shown is an equatorial cross section of a garland cell nephrocyte expressing wild-type Rab11-GFP that localizes in a fine vesicular pattern particularly at the periphery of the cell. Staining the ortholog of nephrin (red) demonstrates appearance of additional fly nephrin vesicles and fine linear protrusions of slit diaphragm protein similar to the observations with Tbc1d8b RNAi (arrow in enlarged inset).
Discussion
Here, we report the identification of five mutations of the gene TBC1D8B in patients with nephrotic syndrome from five different families. Analyzing interaction of TBC1D8B with constitutively active RAB11A and RAB11B compared with their dominant negative variants, we observed specific binding for the active, GTP-bound forms of RAB11A and RAB11B. Loss of TBC1D8B increased basal autophagy, autophagic flux, and exocyst activity, suggesting a disinhibition of endogenous RAB11 function and supporting a role of TBC1D8B as a GAP protein for RAB11. We demonstrate TBC1D8B is an interaction partner of the slit diaphragm protein nephrin and both overexpressed proteins colocalize within different immortalized cell lines. Mutations derived from patients with nephrotic syndrome impair binding of TBC1D8B to endogenous RAB11 and nephrin. Overexpression of TBC1D8B affected trafficking of nephrin in vitro. Studying the loss of Tbc1d8b in the Drosophila model, we observed impaired fly nephrin trafficking within the podocyte-like nephrocytes in vivo. Silencing Rab11 in nephrocytes resulted in appearance of slit diaphragm proteins independent from each other, suggesting that the protein complex may be formed during RAB11-dependent transport.
Our findings are corroborated by a most recent report in two families64 and the discovery of mutations of TBC1D8B in five unrelated families with nephrotic syndrome conclusively defines TBC1D8B as a novel monogenic cause of nephrotic syndrome. Our results in nephrocytes confirm a role of TBC1D8B for glomerular function that was most recently reported in fish.64 Beyond that, the Drosophila model facilitated precise study of the subcellular localization of fly nephrin upon loss of function of Tbc1d8b and allowed acute silencing of Rab11. Thus, we could show a role of Tbc1d8b and Rab11 for fly nephrin trafficking in vivo to identify a possible mechanistic link between Rab11 regulation and glomerular dysfunction.
Surprisingly, the two truncating mutations (p.Arg344* and p.Trp461*) were observed hemizygously in two or one individuals within the gnomAD control population respectively. The incomplete penetrance of these nonsense mutations may be explained by a genetic compensation response.73,74 The compensatory upregulation of another RAB11-GAP might only be induced by a complete loss of function of TBC1D8B but not in missense mutations. Patients with mutations of TBC1D8B typically presented with FSGS in renal biopsy, while MesPGN was described in one case. MesPGN, a glomerular damage pattern, frequently showed progression to FSGS in repeat biopsies of children with SRNS.75
TBC1D8B binds to the active form of RAB11A and RAB11B indicating that TBC1D8B functions as a GAP for both variants of RAB11. Our data implicate RAB11-dependent vesicular trafficking in the pathogenesis of nephrotic syndrome, and delineate a role of TBC1D8B for trafficking of the slit diaphragm protein nephrin as the mechanistic basis for nephrotic syndrome associated with TBC1D8B mutations. We previously implied regulators of RAB5 in nephrotic syndrome.41 This Rab protein functions at early endosomes, while RAB11 regulates endocytic recycling and exocytosis. Localized cycling between exocytosis and endocytosis may provide plasticity and robustness for the glomerular filter potentially similar to the mechanisms at neuronal synapses.76 A direct role of autophagy within the pathogenesis of nephrotic syndrome associated with TBC1D8B is conceivable but will require further investigation.
Trafficking of nephrin in podocytes in vivo may deviate from our observations in MDCK cells, but it is possible that nephrin is delivered initially to the apical surface before eventually being stabilized within the slit diaphragm complex at the apico-basal border. Nephrin trafficking does not seem to occur constitutively, but as a regulated process. Together with our findings in Drosophila, our data support such a regulatory role for TBC1D8B. The mutations of TBC1D8B caused no extrarenal manifestations, suggesting that tight regulation of nephrin trafficking by TBC1D8B/RAB11 is required for the establishment and maintenance of the glomerular filter, whereas (compensatory) function of other RAB11-GAPs may suffice in other tissues.
In conclusion, our findings define TBC1D8B as an RAB11-GAP and together with a previous report64 as a novel cause of hereditary nephrotic syndrome. This supports RAB11-dependent trafficking of nephrin as a novel pathogenetic mechanism.
Disclosures
Dr. Bergmann is an employee of Bioscientia/Sonic Healthcare and holds a part-time faculty appointment at the University of Freiburg. Dr. Hildebrandt is a cofounder of Goldfinch-Bio and receives royalties from Claritas. Dr. MacArthur reports personal fees from Goldfinch Bio, outside the submitted work. All of the remaining authors have nothing to disclose.
Funding
This research was supported by grants from the Deutsche Forschungsgemeinschaft to Dr. Hermle (HE 7456/3-1) and the National Institutes of Health to Dr. Hildebrandt (DK076683), and to the Yale Center for Mendelian Genomics (U54HG006504). Ms. Kampf was supported by the MOTI-VATE program of the Medical Faculty of the University of Freiburg. The Broad CMG was funded by a grant from the National Human Genome Research Institute (UM1HG008900), the National Eye Institute, and the National Heart, Lung, and Blood Institute. A. Onuchic-Whitford acknowledges support from the T32 Ruth L. Kirschstein Institutional National Research Service Award (DK007527). Dr. Schrezenmeier is participant in the BIH-Charité Clinician Scientist Program funded by the Charité–Universitätsmedizin Berlin and the Berlin Institute of Health. Dr. Römer acknowledges the support by the German Research Foundation (grant RO 4341/2-1), the Excellence Initiative of the German Research Foundation (EXC 294), and the Ministry of Science, Research and the Arts of Baden-Württemberg (Az: 33-7532.20). Dr. Bergmann acknowledges support from the Deutsche Forschungsgemeinschaft Collaborative Research Centre (KIDGEM 1140) and the Federal Ministry of Education and Research (01GM1515C).
Supplemental Material
This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019040414/-/DCSupplemental.
Supplemental Figure 1. Additional information concerning patients with mutations of TBC1D8B.
Supplemental Figure 2. TBC1D8B does not colocalize with RAB11 but affects RAB11 function.
Supplemental Figure 3. Localization of TBC1D8B protein in cultured cell lines and secretory GFP upon CRISPR/Cas-mediated loss of TBC1D8B.
Supplemental Figure 4. TBC1D8B and nephrin trafficking.
Supplemental Figure 5. Tb1d8b and Rab11 in Drosophila.
Supplemental Table 1. Primer sequences.
Acknowledgments
Dr. Bergmann, Dr. Hermle, Dr. Laricchia, Dr. Lifton, Dr. MacArthur, Dr. Mane, Dr. Onuchic-Whitford, Dr. Rehm, and Dr. Schneider generated total genome linkage analysis, performed exome capture and massively parallel sequencing, and performed whole-exome evaluation and mutation analysis. Ms. Chen, Ms. Gerstner, Dr. Hermle, Ms. Kampf, Dr. Römer, Dr. Schneider, and Dr. Thünauer performed cDNA cloning, protein purification, tracer endocytosis and immunofluorescence, and subcellular localization studies in cell lines by confocal microscopy. Ms. Chen, Ms. Gerstner, Dr. Hermle, Ms. Kampf, and Dr. Schneider performed coimmunoprecipitation and the experiments in Drosophila. Dr. Helmstädter performed electron microscopy. Dr. Amar, Dr. Berdeli, Dr. Hermle, Dr. Hildebrandt, Dr. Loza Munarriz, Dr. Müller, and Dr. Walz recruited patients and gathered detailed clinical information for the study. Dr. Hermle conceived of and directed the study with support from Dr. Hildebrandt. Dr. Hermle wrote the manuscript. The manuscript was critically reviewed by all the authors.
We are grateful to the families and study individuals for their contribution. We thank the Yale Center for Mendelian Genomics and the Broad Center for Mendelian Genomics for whole exome sequencing analysis. We thank Dr. R. Nitschke, Life Imaging Centre, University of Freiburg, for help with confocal microscopy. We thank Severine Kayser for technical assistance with electron microscopy. We thank the Developmental Studies Hybridoma Bank for antibodies.
Footnotes
L.L.K., R.S., and L.G. contributed equally to this work.
Published online ahead of print. Publication date available at www.jasn.org.
- Copyright © 2019 by the American Society of Nephrology
References
In this issue
