FBP17 and CIP4 recruit SHIP2 and lamellipodin to prime the plasma membrane for fast endophilin-mediated endocytosis

Abstract

Endocytosis mediates the cellular uptake of micronutrients and the turnover of plasma membrane proteins. Clathrin-mediated endocytosis is the major uptake pathway in resting cells¹, but several clathrin-independent endocytic routes exist in parallel^2,3. One such pathway, fast endophilin-mediated endocytosis (FEME), is not constitutive but triggered upon activation of certain receptors, including the β₁ adrenergic receptor⁴. FEME activates promptly following stimulation as endophilin is pre-enriched by the phosphatidylinositol-3,4-bisphosphate-binding protein lamellipodin^4,5. However, in the absence of stimulation, endophilin foci abort and disassemble after a few seconds. Looking for additional proteins involved in FEME, we found that 20 out of 65 BAR domain-containing proteins tested colocalized with endophilin spots. Among them, FBP17 and CIP4 prime the membrane of resting cells for FEME by recruiting the 5'-lipid phosphatase SHIP2 and lamellipodin to mediate the local production of phosphatidylinositol-3,4-bisphosphate and endophilin pre-enrichment. Membrane-bound GTP-loaded Cdc42 recruits FBP17 and CIP4, before being locally deactivated by RICH1 and SH3BP1 GTPase-activating proteins. This generates the transient assembly and disassembly of endophilin spots, which lasts 5-10 seconds. This mechanism periodically primes patches of the membrane for prompt responses upon FEME activation.

Figure 1. FBP17 and CIP4 colocalize with Endophilin and mediate β1-AR uptake.

a, Colocalization of named EGFP-tagged BAR proteins on endogenous Endophilin spots in BSC1 cells. Histograms show the mean ± SEM from three independent biological experiments (n>150 puncta per construct). b, Colocalization of named endogenous BAR proteins on Endophilin spots at the leading edge of resting BSC1 cells. Histograms show the mean ± SEM from three independent biological experiments (n=150 puncta per condition). c, Intracellular accumulation of β1 adrenergic receptor (β1-AR) in Endophilin triple knocked-down (TKD), AP2 knocked-down (KD), Amphiphysin and Bin1 double knocked-down (DKD) or FBP17 and CIP4 DKD cells, treated with 10μM dobutamine for 30min. Counterstaining of the targeted proteins (red) validated the knock-downs in the cells imaged. Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition), normalized to control cells. d, β1-AR uptake into FEME carriers (cytoplasmic Endophilin-positive assemblies) following 4min addition of 10μM dobutamine in cells depleted for FBP17 and CIP4 (F+C DKD) or not. Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition). e, Pull-down using GST or GST-SH3 domains of Endophilin A2, FBP17, CIP4 and TOCA-1 and cell extracts expressing EGFP-tagged third intracellular loops (TIL) of β1-, β2-, β3-, α2a- or α2b-adrenergic receptors. Inputs correspond to 1 to 5% of the cell extracts. Unprocessed original scans are provided in Supplementary Fig. 6. All images are representative of at least 10 captures taken from three independent biological experiments for each condition. All experiments were repeated independently at least three times with similar results. Statistical analysis was performed by one-way ANOVA (a) or two-way ANOVA (b,c); NS, non significant P>0.99. Statistical source data are provided in Supplementary Table 2. Scale bars, 10 (c) and 5μm (b,d).

Figure 2. FBP17 and CIP4 prime cells for FEME.

a, colocalization of endogenous CIP4 and Endophilin in resting BSC1 cells. b, Colocalization of CIP4 on Endophilin spots at the plasma membrane but not on FEME carriers (10μM isoproterenol for 4min). c, Recruitment of endogenous CIP4 and Endophilin in resting cells depleted or not for FBP17 and CIP4 (F+C DKD) or Endophilin (Endo TKD). Histograms show the mean ± SEM from three independent biological experiments (n=150 cells per condition), normalized to control. d, Endogenous Endophilin cell surface spots and FEME carriers in cells overexpressing CIP4-EGFP (CIP4 OEx). Histograms show the mean ± SEM from three independent biological experiments (n>30 cells per condition), normalized to control. e, Proximity Ligation Assays between endogenous β1-AR and Endophilin in CIP4 OEx or control cells. Cells were pre-treated with 20nM GDC-0941 (PI3Ki) for 5min, before stimulation with 10μM dobutamine for 4min, as indicated. Histograms show the mean ± SEM from three independent biological experiments (n>30 cells per condition). f, β1-AR uptake into FEME carriers in cells overexpressing CIP4-Myc and treated as in e. Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition). g, Kymographs and till views from leading edge or ventral surface of resting cells expressing low levels of FBP17-EGFP and EndophilinA2-RFP and imaged at 0.5Hz (top) and 2Hz (middle). Right, percentage of maximum signals (means ± SEM, n=50 spots from three independent biological experiments). h, Kymographs from cells expressing low levels of β1-AR-EGFP or EndophilinA2-RFP and OEx or not CIP4-EGFP, treated with dobutamine and GDC-0941 as indicated and imaged at 2Hz. Kymographs are representative of 5 captures from three independent biological experiments. Histograms show the mean ± SEM from three independent biological experiments (n=5 cells per condition). All images are representative of at least 10 captures taken from three independent biological experiments for each condition. All experiments were repeated independently at least three times with similar results. Statistical analysis was performed by one-way ANOVA (e,f) or two-way ANOVA (c,d,h); NS, non significant P>0.99. Statistical source data are provided in Supplementary Table 2. Scale bars, 20 (a) and 5μm (b,c,d,f).

Figure 3. FBP17 and CIP4 recruit SHIP2 and Lpd.

a, Pull-down experiments with GST-tagged SH3 domains of the indicated proteins and EGFP-tagged SHIP2, Lpd or N-WASP (positive control). Inputs correspond to 1 to 5% of the cell extracts. Unprocessed original scans are provided in Supplementary Fig. 6. b-e, Recruitment of endogenous CIP4, SHIP2, Lpd or Endophilin in resting BSC1 cells depleted or not for FBP17 and CIP4 (F+C DKD), SHIP (SHIP1+2 DKD) or Lpd, as indicated. Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition), normalized to control. f, Recruitment endogenous SHIP2 or Lpd at the leading edge of resting cells overexpressing the indicated constructs. Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition), normalized to control. g, Kymograph from a cell expressing low levels of CIP4-EGFP and EndophilinA2-RFP, treated with AS19499490 (SHIP2i) as indicated and imaged at 2Hz. The kymograph is representative of 9 captures from three independent biological experiments. h, Proximity Ligation Assays between endogenous β1-AR and Endophilin in cells pre-treated or not with 10μM AS19499490 (SHIP2i) for 5min before stimulation with 10μM dobutamine for 4min, as indicated. i, β1-AR uptake into FEME carriers (cytoplasmic Endophilin-positive assemblies) in cells treated as in h. j, Histograms show the mean ± SEM (left and middle, n=30 cells per condition, right, n=5 cells per condition) of cells treated as in g-i, respectively. All images are representative of at least 10 captures taken from three independent biological experiments. All experiments were repeated independently at least three times with similar results. Statistical analysis was performed by one-way ANOVA (j, middle and right) or two-way ANOVA (e,f,j left); NS, non significant P>0.99. Statistical source data are provided in Supplementary Table 2. Scale bars, 20 (h) and 5μm (b,c,d,f,i).

Figure 4. GTP-loaded Cdc42 recruits FBP17 and CIP4 to the plasma membrane.

a-c, Recruitment of endogenous CIP4, SHIP2, or Endophilin in resting BSC1 cells depleted or not for Endophilin (Endo TKD), and treated with 10μM ML141 (Cdc42i 1) for 10min with as indicated. Images are representative of 10 captures from three independent biological experiments for each condition. d, Recruitment of endogenous CIP4, Lpd or Endophilin in resting cells overexpressing EGFP-tagged dominant negative (Cdc42-DN, T17N mutant) or constitutively active (Cdc42-CA, Q61L mutant) versions of Cdc42, as indicated. Focal planes were located at the bottom membrane or in the middle of the cells, as indicated. Images are representative of 6 captures from three independent biological experiments for each condition. e, Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition), treated as in a-d and normalized to the respective controls. Secramine (Cdc42i 2) was used at 10μM for the indicated time). f, scheme of the EGFP-tagged full-length or truncated versions of CIP4 used. g, h, Recruitment of EGFP-tagged full-length or truncated versions of CIP4 in cells depleted for endogenous FBP17, CIP4 and TOCA1 (TKD) and treated or not with 10μM ML141 for 10min (+Cdc42i), as indicated. Cells were immunostained for endogenous CIP4 (red) to control for the depletion in the cells imaged. Images are representative of at least 6 captures taken from three independent biological experiments for each condition. Histograms show the mean ± SEM (n>6 cells per condition). All experiments were repeated at least three times with similar results. Statistical analysis was performed by one-way ANOVA (g) or two-way ANOVA (e). Statistical source data are provided in Supplementary Table 2. Scale bars, 5μm.

Figure 5. Local recruitment of Cdc42GAPs terminates the priming cycle.

a, Recruitment of endogenous Endophilin in resting BSC1 cells overexpressing wild-type EGFP-tagged RICH1 or R288A mutant (GAP*), as indicated. Images are representative of 10 captures from three independent biological experiments. b, Recruitment of endogenous Endophilin and RICH1 in resting cells depleted or not for Endophilin (Endo TKD) or RICH1 and SH3BP1 (R+S DKD). Images are representative of 10 captures from three independent biological experiments. c, Histograms show the mean ± SEM from three independent biological experiments (n=30 cells per condition) treated as indicated and normalized to the respective controls. d, Kymograph from a R+S DKD cell expressing low levels of CIP4-EGFP and EndophilinA2-RFP and stimulated with dobutamine at the indicated times. The kymograph is representative of 5 captures from three independent biological experiments. e, Proximity Ligation Assays between endogenous β1-AR and Endophilin in cells R+S DKD depleted pre-treated or not with 20nM GDC-0941 (PI3Ki) for 5min before stimulation with 10μM dobutamine for 4min, as indicated. f, β1-AR uptake into FEME carriers (cytoplasmic Endophilin-positive assemblies) in cells treated as in e. g, Histograms show the mean ± SEM (left and middle, n=30 cells per condition, right, n=3 cells per condition) of cells treated as in d-f, respectively. (h) Model summarizing the priming cycle of FEME in resting cells: Step 1, active GTP-loaded Cdc42 recruits FBP17 and CIP4 through their REM domains. Step 2, FBP17 and CIP4 cluster 5’-phosphatases SHIP1 and 2 as well as Lpd through their SH3 domains. Lpd is further stabilized by Pi(3,4)P₂ locally produced by SHIP1/2, Step 3, Endophilin is recruited and concentrated by Lpd. From there, pre-enriched Endophilin mediates prompt FEME carrier formation upon cargo activation. In absence of cargo activation, the FEME priming complex disassembled (Step 4), upon local Cdc42 deactivation by the GAPs RICH1 and SH3BP1. All experiments were repeated at least three times with similar results. Statistical analysis was performed by one-way ANOVA (c, g middle and right) or two-way ANOVA (g left); NS, non significant P>0.99. Statistical source data are provided in Supplementary Table 2. Scale bars, 20 (e) and 5μm (a,b,f).