Participation of OCRL1, and APPL1, in the expression, proteolysis, phosphorylation and endosomal trafficking of megalin: Implications for Lowe Syndrome

Abstract

Megalin/LRP2 is the primary multiligand receptor for the re-absorption of low molecular weight proteins in the proximal renal tubule. Its function is significantly dependent on its endosomal trafficking. Megalin recycling from endosomal compartments is altered in an X-linked disease called Lowe Syndrome (LS), caused by mutations in the gene encoding for the phosphatidylinositol 5-phosphatase OCRL1. LS patients show increased low-molecular-weight proteins with reduced levels of megalin ectodomain in the urine and accumulation of the receptor in endosomal compartments of the proximal tubule cells. To gain insight into the deregulation of megalin in the LS condition, we silenced OCRL1 in different cell lines to evaluate megalin expression finding that it is post-transcriptionally regulated. As an indication of megalin proteolysis, we detect the ectodomain of the receptor in the culture media. Remarkably, in OCRL1 silenced cells, megalin ectodomain secretion appeared significantly reduced, according to the observation in the urine of LS patients. Besides, the silencing of APPL1, a Rab5 effector associated with OCRL1 in endocytic vesicles, also reduced the presence of megalin's ectodomain in the culture media. In both silencing conditions, megalin cell surface levels were significantly decreased. Considering that GSK3ß-mediated megalin phosphorylation reduces receptor recycling, we determined that the endosomal distribution of megalin depends on its phosphorylation status and OCRL1 function. As a physiologic regulator of GSK3ß, we focused on insulin signaling that reduces kinase activity. Accordingly, megalin phosphorylation was significantly reduced by insulin in wild-type cells. Moreover, even though in cells with low activity of OCRL1 the insulin response was reduced, the phosphorylation of megalin was significantly decreased and the receptor at the cell surface increased, suggesting a protective role of insulin in a LS cellular model.

Keywords: APPL1; GSK3; insulin; lowe syndrome; megalin; ocrl1; proteolysis; renal disease.

FIGURE 1

Megalin is down-regulated in OCRL1 and APPL1 KD proximal tubule cells. (A) Schematic representation of the full-length megalin with its four ligand-binding domains and the mini receptor with the fourth ligand-binding domain tagged with HA-epitope (mMeg). (B–H) LLC-PK1 cells stably expressing mMeg (mMeg-LLC-PK1) were infected with non-target shRNA (shControl), shRNA directed against OCRL1 (shOCRL1) or APPL1 (shAPPL1) lentiviruses. (B) Endogenous megalin and mMeg were analyzed in whole-cell lysates of control and OCRL1 silenced cells by western blot. (C) Quantification of mMeg protein levels corrected with tubulin levels in control and OCRL1 KD cells. Data are expressed as the means ± SEM of N = 5 independent experiments (t-test, **p < 0.01). (D) Quantitative PCR was used to analyze mRNA levels of megalin. Three independent assays were performed, and the average ±SD was plotted on the graph. (E) Metabolic labeling of control or OCRL1 silenced cells with [³⁵S]-methionine/cysteine for 4 h mMeg was immunoprecipitated (IP) and analyzed by autoradiography. Additionally, an aliquot of the whole-cell lysate was used to total mMeg by western blot (WB). (F) Endogenous megalin and mMeg were analyzed in whole-cell lysates of control and APPL1 silenced cells by western blot. (G) Quantification of mMeg protein levels corrected with tubulin levels in control and APPL1 KD cells. Data are expressed as the means ± SEM of five independent experiments (t-test, **p < 0.01). (H) Control or silenced cells were treated with CHX (100 μM), and the expression of mMeg, Precursor-mMeg, and actin were determined by western blot at indicated times. (I) Graph corresponds to mMeg protein levels corrected with actin levels in control and silenced cells.

FIGURE 2

Characterization of megalin intracellular and secreted fragments. Immunofluorescence of LLC-PK1 cells stably expressing (A) mMeg or (B) Meg0 to detect amino-terminal (N-term) or carboxy-terminal (C-term) domains of the receptor with anti-HA (green) and anti-Megalin (red) antibodies, respectively. The images were acquired by epifluorescence microscopy. Scale bar, 10 µm. (C) Whole-cell lysates of wt LLC-PK1 and mMeg-LLC-PK1 cells were analyzed by western blot using anti-HA to detect mMeg or its N-term fragments. Symbols correspond to a non-specific band (&) and the low molecular weight band exclusive to the mMeg expressing cell lysate (*). (D) Whole-cell lysates of wt LLC-PK1 cells were incubated with conditioned media of either wt or mMeg -LLC-PK1 cells, analyzed by western blot using an anti-HA antibody, which recognizes the N-terminal portion of mMeg. Symbols correspond to the non-specific band (&) and low molecular weight band found in the cell lysate of wt LLC-PK1 cells incubated with conditioned media from mMeg-expressing LLC-PK1 cells (*). (E) mMeg-LLC-PK1 cells were grown to confluence for 48 h. The conditioned medium was concentrated and then immunoprecipitated with an anti-HA antibody. Protein samples, immunoprecipitated media, and cell lysates were analyzed with an anti-HA antibody to detect megalin ectodomain (N-term) and with an anti-megalin cytoplasmic domain polyclonal antibody to detect megalin carboxy-terminal fragment (C-term). Tubulin was used as a loading control. (F) mMeg expressing cells were incubated with vehicle (Control) or 10 μM MMPi (GM6001, general matrix metalloprotease inhibitor) for 48 h and replenished every 12 h. The immunoprecipitated media and cell lysates were analyzed by western blot to detect mMeg and tubulin as load control.

FIGURE 3

Cell surface megalin distribution and extracellular domain release decrease under OCRL1 and APPL1 silencing. (A) mMeg-LLC-PK1 cells, non-permeabilized or permeabilized, control, or OCRL1-silenced were incubated with an anti-HA-488-conjugated antibody and analyzed by flow cytometry. The results are plotted as the ratio of expression levels observed in non-permeabilized (surface) versus permeabilized cells (total). Data are expressed as the means ± SEM of N = 3 independent experiments (t-test, *p < 0.05). (B) Conditioned media of control or OCRL1 silenced mMeg-LLC-PK1 cells were processed by anti-HA immunoprecipitation (Media IP). Samples of IP and cell lysates were analyzed by western blot with an anti-HA antibody to detect the megalin ectodomain (N-term). (C–H) MDCK cells stably expressing mMeg (mMeg-MDCK) were infected with non-target shRNA (shControl), shRNA directed against OCRL1 (shOCRL1) or APPL1 (shAPPL1) lentiviruses. (C) Protein levels of OCRL1 were analyzed in whole-cell lysates of control and OCRL1 silenced cells by western blot. (D) Flow cytometry analyzes mMeg surface localization in control and OCRL1 silenced cells. The plot shows the surface vs. total ratio of expression levels. Data are expressed as the means ± SEM of N = 3 independent experiments (t-test, *p < 0.05). (E) Analysis of immunoprecipitated conditioned medium and cell lysates of control or OCRL1 silenced cells by western blot with an anti-HA antibody. The plot shows the levels of mMeg in the immunoprecipitated media corrected by total. Data are expressed as the means ± SEM N = 3 independent experiments (t-test, *p < 0.05). (F) Protein levels of APPL1 were analyzed in whole-cell lysates of control and APPL1 silenced cells by western blot. (G) Flow cytometry analyzes mMeg surface localization in control and APPL1 silenced cells. The plot shows the surface vs. total ratio of expression levels. Data are expressed as the means ± SEM of N = 3 independent experiments (t-test, *p < 0.05). (H) Analysis of immunoprecipitated conditioned medium and cell lysates of control or APPL1 silenced cells by western blot with an anti-HA antibody. The plot shows the levels of mMeg in media IP corrected by total. Data are expressed as the means ± SEM N = 3 independent experiments (t-test, **p < 0.01).

FIGURE 4

Reduced mMeg phosphorylation in mMeg-MDCK OCRL1 silenced cells. (A) Control or OCRL1 silenced cells were lysed and analyzed by western blot with antibodies against phosphorylated GSK3ß (Ser9) and total GSK3ß. (B) Control or silenced for OCRL were treated with LiCl (50 mM) or NaCl (50 mM) as control. For phosphorylation assays, the cells were incubated with [³²P]-orthophosphate for 2 h at 37°C, followed by immunoprecipitation of megalin from cell lysates using an anti-megalin cytoplasmic domain. The immune complexes were analyzed by SDS-PAGE and visualized by autoradiography. Aliquots of whole-cell lysates were used to detect mMeg by western blot. (C) Graph shows the percentage of phosphorylated mMeg related to the total. Data are expressed as the means ± SEM of N = 3 independent experiments (ANOVA, ****p < 0.0001, **p < 0.01). (D) Control or silenced for APPL1 were treated with LiCl (50 mM) or NaCl (50 mM) as control. For phosphorylation assays, the cells were incubated with [³²P]-orthophosphate for 2 h at 37°C, followed by immunoprecipitation of megalin from cell lysates using an anti-megalin cytoplasmic domain. The immune complexes were analyzed by SDS-PAGE and visualized by autoradiography. Aliquots of whole-cell lysates were used to detect mMeg by western blot (E) Graph corresponds to the percentage of phosphorylated mMeg related to the total of control or APPL1 silenced cells. Data are expressed as the means ± SEM of N = 3 independent experiments (ANOVA, *p < 0.05).

FIGURE 5

Endosomal distribution of megalin and its phosphorylated and non-phosphorylated forms in control and OCRL inhibition conditions. LLC-PK1 were transfected with HA-tagged minimegalins (mMeg) wt, phosphomimetic (mMeg S/D) or non-phosphorylatable mMeg S/A. Then, cells were treated with 50 μM YU142670 or vehicle for 4 h, fixed and processed for fluorescence microscopy. Preparations were imaged with Confocal Nikon Timelapse. Scale Bars 10 μM. (A) Cells were stained with anti-EEA1 (green) and anti-HA (red). (B) At least 34 separated cells were analyzed with ImageJ to measure the Manders percentage between EEA1 endosomes and HA-tagged mMeg structures. ANOVA, ****p < 0.0001. (C) LLC-PK1 cells were co-transfected with mCherry-Rab11 and HA-tagged mMeg. Then, cells were treated with 50 μM YU142670 for 4 h, fixed and processed to detect mMeg (HA, red) and Rab11 (green). The color was changed for consistency with the other panels. (D) At least 12 separated cells were analyzed with ImageJ to measure the Manders percentage between Rab11-positive endosomes and mMeg structures. ANOVA, ***p < 0.001, *p < 0.05. (E) LLC-PK1 transfected with HA-tagged mMeg. Then, cells were treated with 50 μM YU142670 for 4 h, fixed and processed to detect mMeg (HA, red) and Rab7 (green). (F) At least 18 separated cells were analyzed with ImageJ to measure the Manders percentage between Rab7-positive endosomes and mMeg structures. ANOVA, **p < 0.01.

FIGURE 6

Insulin decreases megalin phosphorylation and increases megalin surface expression in control and LS cellular models. (A) mMeg-LLC-PK1 cells treated with NaCl (50 mM), LiCl (50 mM) or insulin (100 nM) were labeled [³²P]-orthophosphate for 2 h at 37°C, followed by immunoprecipitation of megalin with an anti-megalin cytoplasmic domain. The immune complexes were analyzed by SDS-PAGE and visualized by autoradiography. Aliquots of whole-cell lysates were used to detect mMeg by western blot. (B) Graph corresponds to the percentage of phosphorylated mMeg related to the total of NaCl, LiCl or insulin-treated cells. Data are expressed as the means ± SEM of N = 3 independent experiments (ANOVA, ****p < 0.0001, ***p < 0.001). Phosphorylation assays were performed in OCRL1 (C) or APPL1 (D) silenced cells in the presence or the absence of 100 nM insulin. Immunoprecipitated radiolabeled mMeg was evaluated by autoradiography. Total mMeg and GSK3ß (total and phosphorylated forms) were detected by western blot. (E,F) Control or (G,H) OCRL1 silenced cells were serum-starved for 2 h before the incubation with 100 nM insulin by the indicated period of time. Cells were biotinylated to determine the surface levels of mMeg. The whole lysates were used for total receptor levels. Samples were analyzed by western blot. The graphs shows surface vs. total ratio of mMeg expression levels relative to control time 0. N = 4; ANOVA, *p < 0.05, **p < 0.01.

FIGURE 7

Insulin signaling is decreased in LS cellular models. mMeg-LLC-PK1 (A–B) or HeLa (C–D) cells were serum starved for 4 h ± 50 μM YU1426670 and then incubated with 100 nM insulin ± 50 μM YU1426670 for indicated periods of time. (A,C) Cell lysates were analyzed by western blot to detect total and phosphorylated forms of AKT. (B). Graph corresponds to the protein levels of phosphorylated AKT corrected by total levels of AKT relative to time 0. N = 3; ANOVA, *p < 0.05.(D) Graph corresponds to the protein levels of phosphorylated protein corrected by total levels relative to control, time 0, AKT (upper) or GSK3β (lower). N = 3, ANOVA, ***p < 0.001, **p < 0.01, *p < 0.05. (E–F) HeLa control or OCRL1 silenced cells were serum starved for 4 h and then incubated with 100 nM insulin ± 50 μM YU1426670 for indicated periods of time. (E) Cell lysates were analyzed by western blot to detect total and phosphorylated forms of AKT and GSK3β. (F) Graph corresponds to the protein levels of phosphorylated protein corrected by total levels relative to control, time 0, AKT (upper) or GSK3β (lower). N = 3 ANOVA, ***p < 0.001, **p < 0.01, *p < 0.05.