PMC

Model for role of TBC1D14 and Rab11 mediating vesicular transport from the RE to the expanding phagophore. Under fed conditions, Rab11-positive REs function in recycling to the plasma membrane. Upon amino acid starvation, Rab11 mediates vesicle formation from the RE directed to forming autophagosomes. This process is negatively regulated by TBC1D14, which functions as an effector for Rab11. TBC1D14 dissociates from Rab11-positive REs in starvation and accumulates on the Golgi complex. Overexpression of TBC1D14 causes tubulation of REs, accumulation of the ULK1 complex and Rab11 on REs, and inhibition of vesicular transport from the RE.

TBC1D14 interacts with ULK1. (A) HEK293A cells transfected with GFP-TBC1D14 were fixed and stained 24 h later with the anti-ULK1 antibody. (B) HEK293A cells transfected with GFP-TBC1D14 were starved, fixed, and stained 24 h later with the anti-LC3 antibody. (C) HEK293A cells were starved for 2 h and then stained with anti-TBC1D14 and anti-LC3 antibodies. (A–C) Insets in merge show the colocalization of ULK1 and GFP-TBC1D14 on tubules. (D) GFP or GFP-tagged constructs encoding the full-length (FL; aa 1–693) or TBC domain (aa 224–669) of TBC1D14 were cotransfected into cells with myc-tagged ULK1 full-length (aa 1–1,051), aa 1–429, or C-terminal domain (CTD). Lysates were incubated with GFP-Trap resin, and bound proteins were detected by Western blotting. (E) GFP-tagged TBC1D14 as in D, plus GFP-tagged TBC1D14 aa 1–330, was cotransfected with myc-ULK1 (full length). (F) TBC1D14 (aa 224–330) is predicted to bind aa 429–828 of ULK1. Bars: (A–C, main images) 20 µm; (A–C, insets) 2 µm. IP, immunoprecipitation; MM, molecular mass; WB, Western blot.

Rab11 is required for TBC1D14-induced tubulation. (A) HEK293A cells were transfected with myc-TBC1D14 and, 24 h later, fed or starved, fixed, and stained with antibodies against TBC1D14 or Rab11. (B) HEK293A cells were transfected with the indicated siRNA and transfected 24 h later with GFP-TBC1D14. After 72 h, cells were fixed and labeled with an anti-Rab11 antibody. (C) Endogenous TBC1D14 and Rab11 partially colocalize in HEK293A cells and were fed or starved for 2 h in EBSS, anti-TBC1D14, and Rab11. Multiple z stacks were deconvolved using Huygens software. TBC1D14 can be found on Rab11-positive endosomes (arrowheads in insets). (A–C) The insets show areas of colocalization between TBC1D14 and Rab11 (A), GFP-TBC1D14 and Rab11 (B), and TBC1D14 and Rab11 (C). Asterisks indicate transfected cells. (D) Colocalization between Rab11 and TBC1D14 decreases upon starvation. Images were acquired as in C. Colocalization quantified using Imaris software (***, P < 0.0001 by two-way unpaired t test). Error bars indicate SEM (n = 21 cells for each condition). Bars: (A and B, main images) 10 µm; (C, main images) 5 µm; (A–C, insets) 2 µm.

REs are positive for ULK1 and mAtg9 and contribute membrane to autophagosomes. (A) HEK293A cells were starved for 2 h, fixed, and labeled with antibodies against ULK1, mAtg9, and TfnR. Inset i shows the triple-labeled merge. Insets ii and iii were pseudocolored: TfnR and mAtg9 are shown with ii, and ULK1 is shown in iii. The insets in i–iii indicate areas of colocalization of ULK1, mAtg9, and TfnR, mAtg9 and TfnR, and ULK1 and TfnR, respectively. Arrowheads indicate TfnR-, mAtg9-, and ULK1-positive puncta. Two enlarged images of REs are shown. (B) HEK293A cells were fed or starved for 2 h, fixed, and stained for ULK1 and TfnR. (C) Colocalization of TfnR with ULK1 was quantified from multiple z stacks and deconvolved using Huygens software using Imaris. No significant difference was detected (n = 20 cells). (D) mAtg9 is present on TfnR-positive REs in fed cells and redistributes upon starvation. HEK293A cells were fed or starved for 2 h in EBSS, fixed, and stained for mAtg9 and TfnR. (B and D) The insets indicate areas of colocalization of TfnR and ULK1 (B) and TfnR and Atg9 (D). (E) Quantification of TfnR with mAtg9 was performed as in C. ***, P < 0.0001; result of two-way unpaired t test; n = 18 cells. Error bars show SEM. Bars: (A, B, and D, main images) 10 µm; (A, i–iii; and A, i–iii, insets) 1 µm; (B and D, insets) 2 µm.

Autophagosomal membranes contain TfnR. (A and B) 2GL9 (A) or 2GL9 (B) cells transfected with myc-TBC1D14 were starved for 2 h in EBSS, fixed, and labeled with antibodies against TfnR (A) and TfnR and myc (B). Line scan profiles of GFP-LC3–positive autophagosomes were acquired using LSM software. Three representative mock-transfected cells are shown in A, and one myc-TBC1D14–expressing cell is shown in B. The enlarged areas show the colocalization between GFP-LC3 and TfnR. The arrows indicate the region that was analyzed by line scanning. Bars, 10 µm. (C) Colocalization of TfnR with GFP-LC3–positive autophagosomes is significantly reduced in myc-TBC1D14–overexpressing cells. Quantification was performed as in excluding the nuclear GFP-LC3 signal. **, P < 0.01 by two-way unpaired t test. Error bars show SEM. (D) Tfn-HRP is found in autophagosomes and appears in the intraluminal membrane space (arrows). Cells were serum starved for 1 h and incubated with Tfn-HRP for 2 h, fixed, incubated with DAB and H₂O₂, and embedded in Epon. Thin (70 nm) sections were imaged using a transmission electron microscope. Two representative images are shown with enlarged autophagosomes on the right. Bars, 1 µm.

TBC1D14 tubulates REs, which are required for autophagy. (A) HEK293A cells were transfected with GFP-TBC1D14, and 24 h later, Alexa Fluor 647–Tfn was internalized for 30 min. Cells were fixed and stained for Atg13. The arrows show regions of colocalization of Atg13 with Tfn on TBC1D14-induced tubules. (B) HEK293A cells were transfected with GFP-TBC1D14 and, 24 h later, fed or starved, fixed, and stained with an antibody against TfnR. The asterisks indicate transfected cells. (C) Colocalization of overexpressed GFP-TBC1D14 and TfnR was quantified using Imaris software (n = 10 cells fed and starved). (D) Ablation of Tfn-positive REs in 2GL9 cells loaded with HRP-labeled Tfn for 20 min at 37°C. Cells were washed, cooled to 4°C, and incubated for 60 min with DAB (control) or DAB and H₂O₂ to ablate Tfn-containing REs. Cells were incubated for 30 min at 37°C in fresh medium, starved for 2 h in EBSS, and fixed, and GFP-LC3 was imaged with a confocal microscope. (right) Quantification of GFP-LC3 puncta per cell. ***, P < 0.0001; result of two-way unpaired t test; n = 22 cells. Error bars indicate SEM. Bars, (A and B, main images; and D) 10 µm; (A and B, insets) 2 µm.

TBC1D14 binds Rab11, and inactive Rab11 inhibits autophagy. (A) Yeast two-hybrid screen to identify Rab-binding partners of TBC1D14. Yeast were transfected with the active (QA) Rab constructs (prey) and TBC1D14R472A (bait) and selected on plates lacking leucine and tryptophan (−L/−W) and medium lacking leucine, tryptophan, histidine, and adenine (QDO). Five colonies for each Rab were streaked on selective media. (B) HEK293A cells were transfected with GFP-tagged Rab11 constructs, 11aWT, 11aQ70A, 11aN124I, and 11bN124I and, 24 h later, fed or starved for 2 h. Western blot of a representative experiment with anti-GFP, anti-LC3, and antiactin and quantification of LC3-II/actin. n ≥ 3; one-way ANOVA followed by Bonferroni posttest analysis comparing to the vector control are starved with leupeptin (leu) versus Rab11aN124I starved with leupeptin: *, P < 0.05; starved with leupeptin versus Rab11bN124I starved with leupeptin: **, P < 0.01. (C) GST-FIP3 (aa 300–759) immobilized on glutathione–Sepharose beads was incubated with HEK293A lysate from cells cotransfected with either GFP or GFP-Rab11bQ70A and empty vector (EV), myc-TBC1D14, or myc-TBC1D14RAQA. Bound proteins were detected by Western blotting with anti-GFP (Rab11bQ70A) or myc (TBC1D14) antibodies. Endogenous Rab11b was detected with the anti-Rab11b antibody. Ponceau staining is the loading control. (D) HEK293A cells were transfected with myc-TBC1D14 and GFP-Rab11b constructs (WT, Q70A, and S25N). Quantification of pull-down of myc-TBC1D14 with GFP-Rab11bQ70A versus S25N (n = 3). ***, P < 0.0001 as result of two-way unpaired t test. Error bars indicate SEM. IP, immunoprecipitation; MM, molecular mass; WB, Western blot.

A screen of 38 human RabGAPs reveals that 11 inhibit autophagy. (A) An overexpression screen for RabGAPs, which are autophagy inhibitors. 38 myc-TBC domain–containing proteins were expressed in 2GL9 cells for 24 h, and then, cells were fed or starved in EBSS with leupeptin (leu) for 2 h. GFP–LC3-II was quantified by the Odyssey system with antibodies against LC3, actin, and myc. Black bars show levels of GFP–LC3-II after starvation normalized to the renilla vector control. Gray bars show the effect of overexpression of eight GAP activity–deficient RA mutants. The data shown are from single representative experiments. All myc-TBC transfections were repeated (n ≥ 2) and, if inhibitory or if performed with the RA mutant, repeated three or more times (n ≥ 3). ULK1ΔC, an inhibitory C-terminal truncation mutant () was included in each experiment. No catalytic arginine could be identified in TBC1D4, TBC1D9, or TBC1D7. Horizontal lines illustrate the value of the vector control (set to 1) and 60% value of the control. (B) Representative Western blots for myc, GFP-LC3, and actin from the 11 myc-TBC proteins in a screen that inhibited autophagy by 40% or more. The asterisk indicates a nonspecific band. MM, molecular mass; WB, Western blot.

TBC1D14 is a negative regulator of autophagy. (A) HEK293A cells were transfected in triplicate with vector, myc-TBC1D14WT (WT), kinase-inactive myc-ULK1K46I (ULK1 KI), or myc-TBC1D14R472AQ508A (RAQA). Cells were labeled for 18 h with [¹⁴C]valine, chased for 24 h, and then fed or starved for 2 h. LLPD was calculated as described (WT: **, P = 0.0025; RAQA: **, P = 0.0013; ULK1 kinase inactive: ***, P < 0.0001 by two-way unpaired t test compared with starved control values [n = 2]). (B) HEK293A cells were transfected and treated as in A, and p62 levels were detected by Western blotting and normalized to actin. ***, P < 0.001 by one-way analysis of variance (ANOVA) followed by Bonferroni posttest analysis comparing to the vector control. (C) HEK293A cells were transfected with control renilla vector and TBC1D14 constructs as in A and fed, starved, or starved with leupeptin (leu) for 2 h. A representative Western blot for myc, LC3, and actin is shown. Samples are separated for clarity. n = 3; **, P < 0.01 by one-way ANOVA followed by Bonferroni posttest. (D) HEK293A cells stably expressing mRFP-GFP-LC3 were transfected as in A and starved for 2 h. mRFP- and GFP-LC3 spots per cell were counted and analyzed using Imaris. Asterisks indicate myc-TBC1D14–transfected cells. Bar, 20 µm. n = 3; *, P < 0.05; result of two-way unpaired t test. (E) HEK293A cells mock or transfected with siRNA targeting TBC1D14 (catalog no. J-014032-12) for 24 h were starved with or without leupeptin for 2 h, and endogenous LC3 was measured as in B (n = 5). *, P < 0.05 by one-way ANOVA followed by Bonferroni posttest analysis and mock EBSS with leupeptin versus siTBC1D14 EBSS with leupeptin. A representative blot for TBC1D14, actin, and LC3 is shown. Error bars indicate SEM. MM, molecular mass; WB, Western blot.

Overexpression of TBC1D14 and Rab11N124I delays Tfn recycling, but only Rab11 is required for Tfn recycling. (A) HEK293A cells were transfected with GFP-TBC1D14, and 24 h later, Alexa Fluor–transferrin (Tfn) and Alexa Fluor–EGF were applied to the cells for 30 s before fixation (0 min). Alternatively, Tfn and EGF were internalized for 15 min at 37°C and followed by a 5- or 45-min chase. Asterisks indicate GFP-TBC1D14–transfected cells. Boxes are enlarged in bottom images. (B) HEK293A cells were transfected with renilla control, myc-TBC1D14, myc-TBC1D14RAQA, or GFP-Rab11aN124I. After 24 h, cells were serum starved for 2 h and fed ¹²⁵I-Tfn for 20 min at 37°C and then chased in serum-free medium for 10, 25, or 40 min. The supernatant and cells containing recycled ¹²⁵I-Tfn were collected, and ¹²⁵I-Tfn was counted. Error bars indicate SEM (n = 3); two-way ANOVA followed by Bonferroni posttests; Renilla vector control at 45 and 60 min versus TBC1D14WT: ***, P < 0.001; versus TBC1D14RAQA: ***, P < 0.001; versus Rab11aNI: ***, P < 0.001. (C) Alexa Fluor 647–Tfn was internalized for 30 min into cells treated with RISCfree, siRNA against TBC1D14, or Rab11a and b and chased for 30, 60, and 90 min. At each time point, the cells were fixed, and the amount of Alexa Fluor 647 retained was analyzed by FACS. Representative FACS plots are shown with quantification of the geometric mean. (D and E) Cells were treated with siRNAs as in C and starved for 2 h in ESBB or EBSS plus Bafilomycin A (BafA). LC3 lipidation was quantified normalized to tubulin. A representative experiment is shown for LC3, tubulin (tub), and Rab11b. Bars: (A, main images) 10 µm; (A, insets) 2 µm. MM, molecular mass; WB, Western blot.