mTOR Regulates Endocytosis and Nutrient Transport in Proximal Tubular Cells

Abstract

Renal proximal tubular cells constantly recycle nutrients to ensure minimal loss of vital substrates into the urine. Although most of the transport mechanisms have been discovered at the molecular level, little is known about the factors regulating these processes. Here, we show that mTORC1 and mTORC2 specifically and synergistically regulate PTC endocytosis and transport processes. Using a conditional mouse genetic approach to disable nonredundant subunits of mTORC1, mTORC2, or both, we showed that mice lacking mTORC1 or mTORC1/mTORC2 but not mTORC2 alone develop a Fanconi-like syndrome of glucosuria, phosphaturia, aminoaciduria, low molecular weight proteinuria, and albuminuria. Interestingly, proteomics and phosphoproteomics of freshly isolated kidney cortex identified either reduced expression or loss of phosphorylation at critical residues of different classes of specific transport proteins. Functionally, this resulted in reduced nutrient transport and a profound perturbation of the endocytic machinery, despite preserved absolute expression of the main scavenger receptors, MEGALIN and CUBILIN. Our findings highlight a novel mTOR-dependent regulatory network for nutrient transport in renal proximal tubular cells.

Keywords: albuminuria; endocytosis; epithelial transport; mTOR; proteomics; proximal tubule.

Figure 1.

Proximal tubular deletion of mTORC1 leads to a Fanconi like syndrome. (A–C) Schematic of recombination strategy and site of inducible Pax8rtTA × TetOCre–mediated (A) Raptor, (B) Rictor, and (C) double knockout within the tubular system. (D–F) Demonstration of knockout efficacy by Western blot of (D) RAPTOR and downstream target p-S6P/S6P in Rap^ΔTubule, (E) RICTOR and the downstream target p-NDRG1/NDRG1 in Ric^ΔTubule, and (F) RAPTOR and RICTOR and downstream targets p-S6P/S6P and p-NDRG1/NDRG1 in RapRic^ΔTubule. (G) Urinary glucose (gluc) and phosphate (PO₄³⁻) excretions in RapRic^ΔTubule were increased. (H and I) Coomassie staining of SDS-PAGE prepared from urine samples loaded after normalization to urinary creatinine concentration from (H) Rap^ΔTubule and (I) RapRic^ΔTubule. (J) Urinary albumin excretion of Rap^ΔTubule and RapRic^ΔTubule. (K) Urinary neutral amino acid excretion of Rap^ΔTubule, Ric^ΔTubule, and RapRic^ΔTubule. ACR, albumin-creatinine-ratio; Con, control; crea, creatinine. *P value versus control <0.05; **P value versus control <0.01.

Figure 2.

Deletion of mTORC1 leads to a reduced fluid-phase as well as receptor-mediated endocytosis and blocks intracellular trafficking. (A) Fluid-phase and (B) receptor-mediated endocytosis is shown by the uptake of horseradish peroxidase or Alexa555-coupled lactoglobulin (red stain), respectively, injected 5 minutes before perfusion-fixation. Reduced uptake is found in Rap^ΔTubule and RapRic^ΔTubule. (C) Cell by cell analysis of receptor-mediated endocytosis in Rap^ΔTubule × mT/mG mice. Cre-positive cells (green stain) show reduced endocytic uptake of Alexa555-coupled lactoglobulin (red vesicular staining) in comparison with Cre-deficient cells. (D) Quantification of Alexa555-coupled lactoglobulin of Rap^ΔTubule, Ric^ΔTubule, and RapRic^ΔTubule normalized to the respective control mice. Significantly reduced uptake is shown for Rap^ΔTubule and RapRic^ΔTubule animals. (E) Assessment of endogenous albumin content in Rap^ΔTubule, Ric^ΔTubule, and RapRic^ΔTubule. Note the increased number and size of albumin-containing vesicles. In immunofluorescence images, kidney sections were counterstained with Alexa647 phalloidin (blue stain) to mark actin filaments of the BBM. Con, control. Scale bar, 20 μm. **P<0.01.

Figure 3.

mTOR deletion leads to ultrastructural alterations in PT cells. (A) Periodic acid–Schiff–stained sections of control (Con), Rap^ΔTubule, Ric^ΔTubule, and RapRic^ΔTubule showing reduced BBM length in mutants compared with control mice. (B) Semithin sections counterstained with methylene blue from Con, Rap^ΔTubule, Ric^ΔTubule, and RapRic^ΔTubule showing intact morphology of epithelial cells. (C) Electron microscopy images of Con, Rap^ΔTubule, Ric^ΔTubule, and RapRic^ΔTubule showing reduced number and length of BBM microvilli and reduced number of clathrin-coated vesicles as well as recycling endosomes in otherwise normal cellular structures in mutant compared with control animals. Note the increase in early and late endosomes in mutant cells compared with Con in C. Scale bar, 20 μm and 2 μm.

Figure 4.

Proteomic analyses of Rap^ΔTubule mice shows reduced abundance and phosphorylation of both tubular transport proteins and regulators of endocytosis. (A) Hierarchical clustering of identified proteins and the samples on the basis of the protein LFQ intensities. (B) Volcano plot of the proteomic analysis of Rap^ΔTubule mice compared with the control mice. Proteins beyond the curved lines were considered to be significantly changed (red, increased; blue, decreased). (C) Analysis of over–represented gene ontology (GO) terms in the decreased and increased protein populations compared with the unchanged protein population (Fisher exact test; size of the squares represents number of condensed terms). (D) Cumulative histogram of all quantified proteins (black) and proteins with the uniprot keyword secreted (red). Subgroups of this protein population with small molecular weight are depicted in magenta. (E) Hierarchical clustering of identified phosphorylation site intensity. (F) Volcano plot of the intensity of phosphorylation sites in the Rap^ΔTubule mice compared with the control mice. Phosphorylation sites beyond the curved lines were considered to be significantly changed. (G) Analysis of over–represented GO terms in the significantly changed phosphorylation site population compared with the unchanged population (Fisher exact test). (H) Comparison of logarithmized expression ratios (Rap^ΔTubule to control) for proteins and their respective phosphorylation sites. Density is color coded. (I) Position-weighted matrix of phosphorylation motifs present in the increased and decreased phosphorylation site populations. GOCC, go-term cellular component; GOMF: go-term molecular function; LFQ, label free quantification; TTA: trans-tubular-absorption.

Figure 5.

Reduced expression of transport proteins and regulators of endocytosis can be found in proximal tubular cells of Rap^ΔTubule mice. For better illustration, Rap^ΔTubule × mT/mG mice were used for analysis of expression levels of various proteins by cell to cell analysis of Cre-positive (green stain) compared with wild-type cells. (A) B⁰AT1, (B) y⁺Lat1, (D) DOCK8, and (E) BCL-x_L showed strongly reduced expression levels in Cre-positive cells compared with wild-type cells. Borders between cells types are marked with white bars. (C) In the case of 4F2hc, Alexa555 lactoglobulin was used to distinguish between Cre-positive cells and wild-type cells, and borders are marked by white bars. Cre-positive cells with reduced uptake of Alexa555 lactoglobulin also showed reduced basolateral expression level of 4F2hc. (F) Western blots of B⁰AT1, y⁺Lat1, 4F2hc, DOCK8, and BCL-x_L from control (Con) and Rap^ΔTubule (left panel) and densitometric evaluation (right panel) confirm immunohistochemical results. (G) RNAi-mediated knockdown of Gbf-1 (CG8487; gartenzwerg) was generated in Drosophila, and RFP-ANP (red) fluorescence uptake was measured in nephrocytes (green). Quantification is presented in right panel. (H and I) Transmission electron microscopy images of Drosophila from (H) control and (I) Gbf-1 knockdown are shown. In comparison with control nephrocytes with a proper formation of the endocytotic machinery, Gbf-1 knockdown shows loss of clathrin vesicles and early, late, and recycling endolysomes, whereas the formation of aberrant vesicular structures is increased. Scale bar, 20 μm in A–E; 2 μm in H; 0.2 μm in H, inset; 0.5 μm in I; 0.25 μm in I, inset. *P value versus control <0.05; **P value versus control <0.01. RFP-ANP, red-fluorescent-protein atrial natriuretic peptide; RNAi, RNA interference.

Figure 6.

Putative mTORC1–dependent regulation of proximal tubular transport and endocytosis. (A) mTORC1 seems to regulate apical and basolateral PT transport proteins. mTORC1 could directly phosphorylates SGLT2 and NBC1 to increase reabsorption of glucose and sodium and excretion of sodium and bicarbonate. mTORC1 might positively influence the expression level of the amino acid transporter B⁰AT1 and y⁺LAT-4F2hc and the phosphorylation status of b^0,+AT1-rBAT to reabsorb filtered amino acids and secrete them into the bloodstream. (B) mTORC1 could regulate PT endocytosis, membrane biogenesis, vesicle trafficking, and cell polarity. mTORC1 likely controls apical membrane biogenesis and cell polarity by affecting expression level of DOCK8, apical clathrin–coated pit and vesicle formation, and endocytosis by affecting the expression levels of BCL-x_L, SNX8, and RAB10 and phosphorylation level of GBF-1. Also, it controls proper endoplasmic reticulum (ER) function and Golgi integrity again by influencing the expressions and phosphorylation levels of RAB10 and GBF-1, respectively.