Ubiquitylation of leptin receptor OB-Ra regulates its clathrin-mediated endocytosis

Abstract

Leptin receptors are constitutively endocytosed in a ligand-independent manner. To study their endocytosis, leptin receptors OB-Ra and OB-Rb were expressed in HeLa cells. Both receptor isoforms were ubiquitylated, internalized by clathrin-mediated endocytosis and transported to Hrs-positive endosomes after their internalization. Proteasome inhibitors inhibited OB-Ra but not OB-Rb internalization from the cell surface. OB-Ra ubiquitylation occurred on lysine residues K877 and K889 in the cytoplasmic tail, the mutation of which abolished OB-Ra internalization. Fusion of an ubiquitin molecule at the C-terminus of an OB-Ra construct defective both in ubiquitylation and endocytosis restored clathrin-dependent endocytosis of the receptor. The internalization of this constitutively mono-ubiquitylated construct was no longer sensitive to proteasome inhibitors, which inhibited OB-Ra endocytosis by blocking its ubiquitylation. Fusion of an ubiquitin molecule to a transferrin receptor deleted from its own endocytosis motif restored clathrin-mediated endocytosis. We propose that mono-ubiquitin conjugates act as internalization motifs for clathrin-dependent endocytosis of leptin receptor OB-Ra.

Figure 1

Clathrin-dependent endocytosis of leptin receptors. HeLa cells were treated with control or clathrin siRNA and transfected with plasmid encoding HA-tagged OB-Ra or OB-Rb. (A) Anti-clathrin (top) or anti-actin (bottom) immunoblot of cell extracts. (B) Internalization of OB-Ra and OB-Rb in clathrin-depleted cells. Cells were incubated for 10 min with an anti-HA MAb at 37°C. Internalized antibodies were detected with cy3-conjugated secondary antibody after fixation and permeabilization. The internalization of transferrin receptor was monitored in parallel with Alexa555-conjugated transferrin (Tf). (C) Quantification of the internalization in control (white bar) or clathrin-depleted (gray bars) cells. Data are expressed as percentage of cells that internalized receptors.

Figure 2

Characterization of leptin receptor-positive endosomes. HeLa cells (top rows) expressing HA-tagged OB-Ra or OB-Rb, and COS cells expressing HA-tagged OB-Ra (bottom rows) were allowed to internalize anti-HA MAb for 20 min and processed for double-label immunofluorescent detection of internalized MAbs together with endosomal markers, as indicated. EEA-1, early endosome antigen-1; TfR, transferrin receptor; Hrs, Hepatocyte growth factor-regulated tyrosine kinase substrate.

Figure 3

Inhibition of OB-Ra endocytosis by lactacystin. (A) Stabilization of OB-R by lactacystin. HeLa cells expressing HA-tagged OB-Ra or OB-Rb were incubated in the presence (+), or the absence (−) of lactacystin for 2 h, and their content in OB-Ra or OB-Rb was analyzed by immunoblot. The lower band is not specific. The migration of prestained molecular weight markers is indicated on the left side of the gel. (B) Impact of lactacystin treatment on OB-R internalization. HeLa cells expressing HA-tagged OB-Ra or OB-Rb were incubated in the presence, or the absence of lactacystin for 2 h. OB-R endocytosis was analyzed by a 20-min anti-HA antibody uptake. (C) Inhibition of OB-Ra internalization by lactacystin. HeLa cells expressing HA-tagged OB-Ra were biotinylated at 4°C and allowed to internalize for 15 min at 37°C, or kept on ice (4°C). Endocytosed OB-Ra was protected from glutathione reduction (+GSH), isolated using streptavidin beads and quantified by immunoblot using anti-HA, together with 20% of the proteins initially biotinylated and not submitted to glutathione reduction (−GSH). (D) Impact of lactacystin on steady-state OB-Ra localization. HeLa cells transiently expressing HA-tagged OB-Ra were incubated for 2 h in the presence (LC) or the absence (control) of lactacystin, and processed for immunofluorescence detection of OB-Ra.

Figure 4

Role of the lysine residues of the cytoplasmic domain in OB-Ra endocytosis. (A) Sequences of the cytoplasmic tails of OB-Ra and lysine-to-arginine mutants. The positions of lysine residues in the sequence of OB-Ra are indicated. (B) Internalization of OB-Ra mutants analyzed by antibody uptake. HeLa cells transiently expressing the indicated construct were allowed to internalize anti-HA MAb for 20 min and internalized antibody was detected by immunofluorescence. (C) Quantification of the impact of each mutation on endocytosis. The number of endosomes were counted in 120 randomly chosen cells transiently expressing each construct. In the graphic, the mean endosome number per cell is indicated as a measure of the relative internalization rate of each construct. Values not significantly different from each other by statistical analysis are indicated by a square bracket over the corresponding bars. (D) Internalization of OB-Ra mutants analyzed by cell surface biotinylation and glutathione cleavage resistance. The fraction of protein internalized during 15 min at 37°C was analyzed as in Figure 3C.

Figure 5

Role of ubiquitylation in OB-Ra endocytosis. (A) Ubiquitylation of OB-Ra and OB-Rb. Lysates from HeLa cells expressing OB-Ra or OB-Rb or from control cells were immunoprecipitated with an anti-HA antibody and then immunoblotted with anti-ubiquitin MAbs P4D1 or FK1. The migration of unconjugated receptors is indicated on the right side. Each membrane was reprobed with anti-HA MAb. The specificity of each MAb was assessed with an ubiquitin ladder. Note that P4D1, but not FK1, detects mono-ubiquitin. (B) Ubiquitylation of OB-Ra and lysine-to-arginine mutants. HA-tagged OB-Ra, 2R2K, 2K2R, 3KR, or 2KRK were expressed in HeLa cells, alone (−) or in combination with His₆-tagged ubiquitin (+). Ubiquitylated proteins were purified from cell lysates by pull-down on Ni-agarose beads, and the presence of OB-R was analyzed by immunoblotting. The migration of unconjugated and conjugated (braket) OB-Ra is indicated on the left side. (C) Impact of dynamin 2 K44A overexpression on OB-Ra ubiquitylation. HA-tagged OB-Ra and/or His₆-tagged ubiquitin were expressed by transient transfection in HeLa cells and GFP-tagged dynamin 2 K44A was expressed with an adenoviral vector. Control cells were transduced with an equivalent dose of adenoviral vector expressing GFP. Ubiquitylated OB-Ra was purified on Ni-agarose beads and analyzed by immunoblotting. The migration of unconjugated and conjugated (braket) OB-Ra is indicated on the left side. (D) Ubiquitin-dependent internalization of OB-Ra. Two new constructs were generated: an ubiquitin molecule was fused to the C-terminus of 4R (4R-Ub) and all lysine residues of OB-Rb were replaced with arginine (OB-Rb12R). OB-Ra, 4R, 4R-Ub, OB-Rb or OB-Rb12R were transiently expressed in HeLa cells and their internalization analyzed by antibody uptake. (E) Relative internalization rates of OB-Ra and 4R-Ub analyzed by cell surface biotinylation and glutathione cleavage resistance. The amounts of protein internalized at 37°C during the indicated time were analyzed as in Figure 3C, together with 20% of the biotinylated material (surface). (F) Clathrin-mediated endocytosis of 4R-Ub. Internalization of 4R-Ub was analyzed in cells treated with control or clathrin siRNA as described in the legend of Figure 1. As a control, the uptake of Alexa555-conjugated transferrin (Tf) was monitored in parallel.

Figure 6

Action of proteasome inhibitors on OB-Ra ubiquitylation. (A) Inhibition of OB-Ra ubiquitylation by lactacystin. The ubiquitylation of OB-Ra in the presence of lactacystin was analyzed as described in the legend of Figure 5. (B) Depletion of the cellular pool of ubiquitin by lactacystin. HeLa cells transfected with an expression vector for HA-tagged ubiquitin were incubated for 2 h in the presence of lactacystin or chloroquine. The cellular content of ubiquitin and ubiquitin conjugates was analyzed by immunoblotting with an anti-HA MAb. The band of unconjugated ubiquitin is indicated by an arrowhead. The upper part of the gel shows conjugated proteins. (C) Endocytosis of 4R-Ub in the presence of proteasome inhibitor. HeLa cells transiently expressing OB-Ra or 4R-Ub were incubated for 2 h in the presence or the absence of MG-132, as indicated. Endocytosis was analyzed as described in the legend of Figure 1.

Figure 7

Ubiquitin as a transferable motif in clathrin-mediated endocytosis. (A) Structure of the constructs. TfR, HA-tagged human transferrin receptor; TfRΔ3–59, HA-tagged human transferrin receptor deleted form its cytoplasmic domain; Ub-TfRΔ3–59, HA-tagged human transferrin receptor, the cytoplasmic domain of which was replaced with ubiquitin. (B) Absence of internalization of TfRΔ3–59 monitored with a 20-min MAb uptake. (C) Clathrin-mediated internalization of Ub-TfRΔ3–59. HeLa cells were treated with control or clathrin siRNA and then transiently transfected with plasmid encoding TfR or Ub-TfRΔ3–59. The cells were treated for 2 h with 50 μg/ml cycloheximid and incubated with anti-HA MAb for 10 min at 37°C, to label cell surface and internalized material. Anti-HA MAbs were detected with a cy3-conjugated secondary antibody after fixation and permeabilization. (D) Quantification of the internalization of TfR and Ub-TfRΔ3–59 in control (white bar) or clathrin-depleted (gray bars) cells. Data are expressed as percentage of cells that internalized receptors.