Cargo-selective regulation of clathrin-mediated endocytosis by AMP-activated protein kinase

Abstract

The cell surface abundance of many proteins is controlled by clathrin-mediated endocytosis (CME). CME is driven by the assembly of clathrin and other proteins on the inner leaflet of the plasma membrane into clathrin-coated pits (CCPs). Regulation of CCP dynamics allows for control of the function of specific cell surface proteins, impacting a range of cellular outcomes. AMP-activated protein kinase (AMPK) becomes activated upon metabolic insufficiency and facilitates cellular adaptation to nutrient stress. Here, we examined how AMPK regulates CME and the cell surface membrane traffic of β1-integrin. We find that AMPK controls CCP dynamics and regulates the abundance of the endocytic adaptor protein Dab2 within CCPs in a manner that requires the GTPase Arf6, thus selectively promoting the CCP recruitment and internalization of β1-integrin. This study reveals a signaling pathway for cargo-selective metabolic regulation of CME by AMPK that impacts the function of cell surface proteins such as integrins.

Keywords: Biochemistry; Cell biology.

Graphical abstract

Figure 1

AMPK activation reduces cell surface β1-integrin abundance (A) Representative immunoblot of ARPE-19 whole cell lysate treated with 100 μM A-769662 for 5 min, or 5 μM oligomycin for 20 min, or 5 mM 2-deoxyglucose (2-DG) for 20 min, or a vehicle control treatment. (B) Widefield epifluorescence images of fixed ARPE-19 cells treated as in panel A and labeled for cell surface β1-integrin (4B7 antibody clone), as described in STAR Methods. Scale bar represents 20 μm. (C) Mean fluorescence intensity of cell surface β1-integrin from a representative experiment (left) and shown as the mean ± SD from three independent experiments (right) that analyzed 30–50 individual cells per condition. ∗p < 0.05 [one-way ANOVA with Dunnett’s post-hoc test]. A.U., arbitrary units; R.U., relative units.

Figure 2

AMPK activation selectively regulates β1-integrin endocytosis (A) ARPE-19 cells simultaneously treated with 1 mg/mL anti-β1-integrin (antibody clone K20) and either 100 μM A-769662 or vehicle control were fixed at the indicated timepoints. β1-integrin recycling rate shown as a percentage of timed condition fluorescence intensity/total fluorescence intensity. (B) Spinning disk confocal images of fixed ARPE-19 cells pre-treated with 4 μM ikarugamycin (IKA) or vehicle control for 10 min, followed by treatment with 100 μM A-769662 or vehicle control for 5 min and antibody labeling of cell surface β1-integrin as described in STAR Methods. Scale bar represents 20 μm. (C) Mean fluorescence intensity of cell surface β1-integrin from a representative experiment (left) and shown as the mean ± SD from three independent experiments (right) that analyzed >50 individual cells per condition. ∗p < 0.05 [two-way ANOVA with a Fisher’s LSD test]. (D) ARPE-19 cells simultaneously treated with 1 mg/mL anti-β1-integrin (antibody clone K20) and either 100 μM A-769662 or vehicle control for 7.5 or 15 min at 37C. Shown are representative z series slices of spinning disk confocal images of fixed cells labeled for β1-integrin and early endosome marker EEA1. Scale bar represents 5 μm. (E) Manders’ overlap coefficient for β1-integrin in EEA1-labelled endosomes shown as the mean ± SD from seven independent experiments that analyzed 20–30 individual cells per condition. ∗p < 0.05 [two-way ANOVA with a Sidak post-hoc test]. A.U., arbitrary units; R.U., relative units.

Figure 3

AMPK activation regulates clathrin-coated pit size and dynamics (A) eGFP-CLCa-RPE cells treated with 100 μM A-769662 or vehicle control were fixed at the indicated timepoints and plasma membrane clathrin structures were visualized with TIRF-M. Scale bar represents 5 μm. (B) Automated detection and analysis of eGFP-clathrin puncta fluorescence intensity reveals CLS size as described in STAR Methods. Shown is a representative experiment (left) and the mean ± SD from three independent experiments, where for 0 min A-769662: k (cells) = 322 and n (CLSs) = 212852; for 10 min A-769662: k (cells) = 46 and n (CLSs) = 30415; for 15 min A-769662: k (cells) = 46 and n (CLSs) = 33068; for 30 min A-769662: k (cells) = 46 and n (CLSs) = 26592; for 60 min A-769662: k (cells) = 46 and n (CLSs) = 31398. ∗p < 0.05 [one-way ANOVA with Dunnett’s post-hoc test]. (C–G) Time-lapse TIRF-M imaging of eGFP-CLCa-RPE cells treated with 100 μM A-769662 (or not) were analyzed by automated detection and tracking of CLSs as described in STAR Methods to identify bona fide CCPs. Scale bar represents 5 μm. Shown is the mean ± SD from five independent experiments for (D) sCLS nucleation, (E) CCP initiation, (F) CCPs with lifetimes <15 s, and (G) eGFP-clathrin intensity within CCPs. Control: k (cells) = 36 and n (CLSs) = 22714; A-769662: k (cells) = 42 and n (CLSs) = 20262. ∗p < 0.05 [Mann-Whitney U test for D-F; two-way ANOVA with a Sidak post-hoc test]. A.U., arbitrary units; R.U., relative units.

Figure 4

AMPK activation selectively regulates cargo in clathrin structures (A and B) ARPE-19 cells treated with either (A) 1 μg/mL biotin-xx-Tfn or (B) 5 ng/mL biotin-xx-EGF in the presence of 100 μM A-769662 or vehicle control treatment were fixed at the indicated timepoints. Tfn and EGF internalization were quantified as described in STAR Methods. Shown are the means ± SD from three independent experiments. (C) eGFP-CLCa-RPE cells simultaneously treated with 1 mg/mL anti-β1-integrin (antibody clone K20) and either 100 μM A-769662 or vehicle control at 37C were fixed at the indicated timepoints and imaged by TIRF-M. Scale bar represents 5 μm. (D) Automated detection and analysis of β1-integrin intensity within clathrin structures is shown as a representative experiment (left) and as the mean ± SD from four independent experiments (right). Control, 2 min: k (cells) = 45 and n (CLSs) = 5708; A-769662, 2 min: k (cells) = 37 and n (CLSs) = 6785; Control, 5 min: k (cells) = 45 and n (CLSs) = 6532; A-769662, 5 min: k (cells) = 49 and n (CLSs) = 7820. ∗p < 0.05 [two-way ANOVA with a Fisher’s LSD test]. (E) eGFP-CLCa-RPE cells were treated simultaneously with A647-Tfn and 100 μM A-769662 or vehicle control treatment for 5 min, fixed, and imaged by TIRF-M. Scale bar represents 5 μm. (F) Automated detection and analysis of clathrin structures and A647-Tfn fluorescence intensity within CLSs are shown as a representative experiment (left) and the mean ± SD from four independent experiments (right). Control: k (cells) = 34 and n (CLSs) = 7229; A-769662: k (cells) = 37 and n (CLSs) = 8719. ∗p < 0.05 [Mann-Whitney U test]. A.U., arbitrary units; R.U., relative units.

Figure 5

AMPK activation selectively promotes Dab2 recruitment to clathrin structures (A–H) eGFP-CLCa-RPE cells treated with 100 μM A-769662 or vehicle control for 5 min were fixed and imaged by TIRF-M. Automated detection and analysis of the fluorescence intensity of (A-B) Dab2, (C-D) CALM, (E-F) epsin, and (G-H) AP2 in CLSs were determined by automated detection and analysis of clathrin structures and shown as the means ± SD from three independent experiments. Dab2, control: k (cells) = 147 and n (CLSs) = 47929; Dab2, A-769662: k (cells) = 170 and n (CLSs) = 66109; CALM, control: k (cells) = 38 and n (CLSs) = 15972; CALM, A-769662: k (cells) = 32 and n (CLSs) = 12545; epsin, control: k (cells) = 46 and n (CLSs) = 19642; epsin, A-769662: k (cells) = 39 and n (CLSs) = 17852; AP2, control: k (cells) = 38 and n (CLSs) = 15134; AP2, A-769662: k (cells) = 45 and n (CLSs) = 18130. ∗p < 0.05 [Mann-Whitney U test]. Scale bars represent 5 μm. (I) eGFP-CLCa-RPE cells transfected with Dab2 siRNA or non-targeting control siRNA were simultaneously treated with 1 mg/mL anti-β1-integrin (antibody clone K20) and either 100 μM A-769662 or vehicle control at 37C for 5 min. Cells were fixed and imaged by TIRF-M. Scale bar represents 5 μm. (J) Automated detection and analysis of β1-integrin intensity within clathrin structures is shown as a representative experiment (left) and as the mean ± SD from four independent experiments (right). Control siRNA, control: k (cells) = 49 and n (CLSs) = 9144; Control siRNA, A-769662: k (cells) = 52 and n (CLSs) = 9481; Dab2 siRNA, control: k (cells) = 54 and n (CLSs) = 8193; Dab2 siRNA, A-769662: k (cells) = 49 and n (CLSs) = 7498. ∗p < 0.05 [two-way ANOVA with a Tukey post-hoc test]. A.U., arbitrary units; R.U., relative units.

Figure 6

Arf6 associates with clathrin components in a GTP-dependent manner (A) Dot plot showing prey proteins identified with Arf6-WT-BirA∗, Arf6-Q67L-BirA∗ (GTP-bound), or Arf6-T44N-BirA∗ (GDP-bound) enriched over endogenous biotinylation (untransfected) and non-specific pan-cellular biotinylation (BirA∗ alone—BFDR ≤1%, SAINT). (B) RFP-CLCa-RPE cells transiently transfected Arf6-WT-eGFP, Arf6-Q67L-eGFP, or Arf6-T44N-eGFP were visualized by TIRF-M. Scale bar represents 5 μm. (C) The fluorescence intensity of Arf6-eGFP constructs in clathrin structures is shown as a representative experiment (left) and as the mean ± SD from three independent experiments (right). Arf6-WT: k (cells) = 54 and n (CLSs) = 20570; Arf6-T44N: k (cells) = 52 and n (CLSs) = 22237; Arf6-Q67L: k (cells) = 54 and n (CLSs) = 25351. ∗p < 0.05 [one-way ANOVA with Dunnett’s post-hoc test]. (D) Arf6-WT-eGFP-RPE cells treated with 1 μM doxycycline for 24 h were fixed, labeled with anti-Dab2 and anti-clathrin antibodies, and imaged by TIRF-M. Scale bar represents 5 μm. (E) Clathrin structures containing above-threshold Arf6-WT-eGFP intensity levels were determined by automated detection and the fluorescence intensity of Dab2 within structures that either contain Arf6 (Arf6+) or not (Arf6-) are shown as a representative experiment distribution (left) and as the mean ± SD from four independent experiments (right). k (cells) = 55, n (Arf6+ CLSs) = 10167 and (Arf6- CLSs) = 3694. ∗p < 0.05 [Mann-Whitney U test]. A.U., arbitrary units; R.U., relative units.

Figure 7

Arf6 is required for AMPK-dependent regulation of Dab2 recruitment to CCPs and β1-integrin internalization (A) Arf6-WT-eGFP-RPE cells were transfected with non-targeting control siRNA, then stimulated with 1 μM doxycycline for 24 h. These cells were then treated with 100 μM A-769662 or vehicle control for 5 min, fixed, labeled with an anti-clathrin antibody, and imaged by TIRF-M. Scale bar represents 5 μm. (B) Automated detection and analysis of Arf6-WT-eGFP intensity in clathrin structures is shown as the mean ± SD from four independent experiments. Control: k (cells) = 33 and n (CLSs) = 5137; A-769662: k (cells) = 44 and n (CLSs) = 6466. ∗p < 0.05 [Mann-Whitney U test]. The data presented here is a part of the experiment that is also shown in Figures 8A, 8B, and S7A, and the images shown in (A) are also shown as part of Figure 8A. (C) Arf6-WT-UltraID cells stimulated with 1 μM doxycycline (dox) for 24 h were treated with 50 μM biotin in the presence of 100 μM A-769662 or vehicle control for 15 min. Shown is a representative immunoblot of whole cell lysates (input) and biotinylated interactors isolated by streptavidin pulldown (IP) detected by western blotting using the indicated antibodies. (D) Relative abundance of AP2 normalized to Arf6-WT-UltraID expression shown as the mean ± SD from six independent experiments. ∗p < 0.05 [Wilcoxon signed-rank test]. (E) eGFP-CLCa-RPE cells transfected with Arf6 siRNA or non-targeting control siRNA were treated with 100 μM A-769662 or vehicle control for 5 min. Shown are TIRF-M images of fixed cells labeled with an anti-Dab2 antibody. Scale bar represents 5 μm. (F) Automated detection and analysis of Dab2 intensity in clathrin structures is shown as a representative experiment (left) and as the mean ± SD from five independent experiments (right). Control siRNA, control: k (cells) = 177 and n (CLSs) = 74005; control siRNA, A-769662: k (cells) = 166 and n (CLSs) = 76139; Arf6 siRNA, control: k (cells) = 184 and n (CLSs) = 89747; Arf6 siRNA, A-769662: k (cells) = 190 and n (CLSs) = 91736. ∗p < 0.05 [two-way ANOVA with a Sidak post-hoc test]. The data presented here is a part of the experiment that is also shown in Figures 8C, 8D, and S8C, and the images depicting control siRNA treated cells in Figure 7E are also shown as part of Figure 8C. (G) ARPE-19 cells transfected with Arf6 siRNA or non-targeting control siRNA were treated with 100 μM A-769662 or vehicle control for 5 min and antibody labeled for cell surface β1-integrin (antibody clone 4B7). Scale bar represents 20 μm. (H) Mean fluorescence intensity of cell surface β1-integrin is shown as a representative experiment (left) and as the mean ± SD from three independent experiments (right) that analyzed 30–50 individual cells per condition. ∗p < 0.05 [two-way ANOVA with a Fisher’s LSD test]. A.U., arbitrary units; R.U., relative units.

Figure 8

AMPK-dependent regulation of Dab2 and Arf6 recruitment to CCPs requires ArfGAP3 (A) Arf6-WT-eGFP-RPE cells transfected with ArfGAP3 siRNA or non-targeting control siRNA were stimulated with 1 μM doxycycline for 24 h, followed by treatment with 100 μM A-769662 or vehicle control for 5 min. Shown are TIRF-M images of fixed cells labeled with an anti-clathrin antibody. Scale bar represents 5 μm. (B) Automated detection and analysis of Arf6-WT-eGFP intensity in clathrin structures is shown as a representative experiment (left) and as the mean ± SD from four independent experiments (right). Control siRNA, control: k (cells) = 33 and n (CLSs) = 5137; control siRNA, A-769662: k (cells) = 44 and n (CLSs) = 6466; ArfGAP3 siRNA, control: k (cells) = 32 and n (CLSs) = 5716; ArfGAP3 siRNA, A-769662: k (cells) = 37 and n (CLSs) = 7442. ∗p < 0.05 [two-way ANOVA with a Šídák post-hoc test]. The data presented here is a part of the experiment that is also shown in Figures 8A, 8B, and S7A, and the images for control siRNA treated conditions shown in Figure 8A are also shown in Figure 7A. (C) eGFP-CLCa-RPE cells transfected with ArfGAP3 siRNA or non-targeting control siRNA were treated with 100 μM A-769662 or vehicle control for 5 min. Shown are TIRF-M images of fixed cells labeled with an anti-Dab2 antibody. Scale bar represents 5 μm. (D) Automated detection and analysis of Dab2 intensity within clathrin structures is shown as a representative experiment (left) and as the mean ± SD from five independent experiments (right). Control siRNA, control: k (cells) = 177 and n (CLSs) = 74005; control siRNA, A-769662: k (cells) = 166 and n (CLSs) = 76139; ArfGAP3 siRNA, control: k (cells) = 182 and n (CLSs) = 70963; ArfGAP3 siRNA, A-769662: k (cells) = 206 and n (CLSs) = 106193. ∗p < 0.05 [two-way ANOVA with a Sidak post-hoc test]. The data presented here is a part of the experiment that is also shown in Figures 7E, 7F, and S8C, and the images depicting control siRNA treated cells in Figure 8C are also shown as part of Figure 7E. (E) ARPE-19 cells transfected with ArfGAP3 siRNA or non-targeting control siRNA were treated with 100 μM A-769662 or vehicle control for 5 min and antibody labeled for cell surface β1-integrin (antibody clone 4B7). Scale bar represents 20 μm. (F) Mean fluorescence intensity of cell surface β1-integrin is shown as a representative experiment (left) and as the mean ± SD from three independent experiments (right) that analyzed 30–50 individual cells per condition. ∗p < 0.05 [two-way ANOVA with a Fisher’s LSD test]. A.U., arbitrary units; R.U., relative units.