The activity of Sac1 across ER-TGN contact sites requires the four-phosphate-adaptor-protein-1

Abstract

Phosphatidylinositol-4-phosphate (PI4P), a phosphoinositide with key roles in the Golgi complex, is made by Golgi-associated phosphatidylinositol-4 kinases and consumed by the 4-phosphatase Sac1 that, instead, is an ER membrane protein. Here, we show that the contact sites between the ER and the TGN (ERTGoCS) provide a spatial setting suitable for Sac1 to dephosphorylate PI4P at the TGN. The ERTGoCS, though necessary, are not sufficient for the phosphatase activity of Sac1 on TGN PI4P, since this needs the phosphatidyl-four-phosphate-adaptor-protein-1 (FAPP1). FAPP1 localizes at ERTGoCS, interacts with Sac1, and promotes its in-trans phosphatase activity in vitro. We envision that FAPP1, acting as a PI4P detector and adaptor, positions Sac1 close to TGN domains with elevated PI4P concentrations allowing PI4P consumption. Indeed, FAPP1 depletion induces an increase in TGN PI4P that leads to increased secretion of selected cargoes (e.g., ApoB100), indicating that FAPP1, by controlling PI4P levels, acts as a gatekeeper of Golgi exit.

Figure 1.

The depletion of VAPs, ORP10, and ORP9 in combination with OSBP1, or of FAPP1 increases Golgi PI4P. (A) Immunodetection of PI4P with a monoclonal anti-PI4P antibody in control (CTRL), FAPP1-KD, ORP10-KD, and VAP-KD HeLa cells fixed and costained with anti–Golgin-97 antibody. Bar, 10 µm. (B and C) Quantification of Golgi PI4P levels (the ratio of PI4P and Golgin-97 fluorescence intensity at the Golgi complex; see Materials and methods) in cells transfected with the indicated single (B) or double (C) siRNAs. Means ± SD, three independent experiments, n > 200 cells per experiment; Student’s t test. Bar, 10 µm. (D) PI4P levels in control and FAPP1-KD HeLa cells and in WT and FAPP1-KO MEFs with (+FAPP1) or without GFP-FAPP1 overexpression. Top panel: Representative images of the MEFs. The arrowhead in a lower panel indicates a cell overexpressing GFP-FAPP1. Graph: Quantification of Golgi PI4P levels in control and FAPP1-KD HeLa cells and in WT and FAPP1-KO MEFs, with (+FAPP1) or without GFP-FAPP1 overexpression. Means ± SD, three independent experiments, n = 60–80 cells per experiment. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 in B–D; Student’s t test. (E) Distribution of the PI4P GFP-P4C probe in control, FAPP1-KD, ORP10-KD, and VAP-KD fixed HeLa cells. Bar, 10 µm. (F) The ratio between GFP-P4C fluorescence intensity in the Golgi and PM area. Means ± SD, three independent experiments, n = 30–43 cells/condition. (G and H) Evaluation of Golgi PI4P levels by mass assay. (G) Golgi fractions were isolated from control (CTRL), FAPP1-KD, and ORP10-KD HeLa cells (see Materials and methods). Western blot showing the distribution of a TGN protein (TGN46) and a PM protein (Na⁺/K⁺ ATPase) after sucrose gradient separation. Dashed rectangle, fractions enriched for Golgi markers. (H) Golgi fractions were incubated with ³²P-ATP and, where indicated, with recombinant PIP5K1C. ³²P-PI4,5P₂ produced by PIP5K1C incubated with 3 µM porcine brain PI4P (PI4P, right lane) was taken as a reference for PI4,5P₂. The amount of ³²P-PI4,5P₂ produced by PIP5K1C using PI4P present in Golgi fractions as a substrate was assessed by TLC and expressed as a percentage of control (graph). Means ± SD of three independent experiments.

Figure 2.

FAPP1 interacts with VAPs and Sac1 and promotes the in-trans 4-phosphatase activity of Sac1 in vitro. (A) HeLa cells transfected with FAPP1-BioID2 or Sac1-BioID2 constructs were treated with 50 µM biotin. Biotinylated proteins were used in a PD assay with streptavidin-conjugated beads, resolved by SDS-PAGE, and immunoblotted with the indicated antibodies. (B) In vitro PD of GST-FAPP1 proteins (FL, residues 1–100, or residues 97–300) with His-tagged forms of Sac1 and VAPA showing direct binding to Sac1 (residues 1–522 and 1–188) and to VAPA. (C) Top: In vitro PD of His-Sac1 (residues 1–188) with GST-tagged VAPA and VAPB proteins. Bottom: GST-FAPP1 PD assay shows simultaneous binding with Sac1 (1–522) and VAPA. The asterisk indicates cross-reaction of the anti-His-antibody with FL-FAPP1. (D) SIM–super resolution images of endogenous FAPP1 localizing at ERTGoCS. Right panels: Magnification of boxed area; arrowheads indicate interposition of FAPP1 (green) between the TGN (TGN46 in blue) and the ER (Cb5 in red), as highlighted in the orthogonal projection (bottom). Bars, 1 µm. (E) Immuno-EM image showing FAPP1 localizing at the ERTGoCS. FAPP1 was enriched at the TGN as compared with earlier Golgi compartments, as previously described (Godi et al., 2004). (F) In vitro 4-phosphatase activity of liposome-bound His6-Sac1 (1–522) in the presence of GST-FAPP1 (+FAPP1) or GST alone (without FAPP1) in the cis (solid line) or trans (dotted line) conformation (schematized above panel). The reaction was performed in 100 µl in the presence of 100 nM Sac1; 12.5 µl was taken at the indicated time points to measure phosphate release (see Materials and methods). Means ± SD, three technical replicates of a representative experiment, n = 3. (G) Effects of GST-FAPP1 or GST alone (w/o FAPP1) on liposome-bound His6-Sac1 (1–522) activity toward soluble diC8-PI4P. Means ± SD, three technical replicates of a representative experiment, n = 3. (H) Dose response of phosphate release in the trans conformation in the presence of increasing concentrations of GST-FAPP1. Means ± SD, three technical replicates of a representative experiment, n = 3 (see Materials and methods for details). (I) OSBP1 but not FAPP1 can extract PI4P from membranes. 3 µM GST-OSBP1-FL (red line), GST-FAPP1-FL (green line) or GST (blue line) was added to liposomes containing TopFluor-PI4P and DilC18 (see Materials and methods).

Figure 3.

The stabilization of ERTGoCS decreases PI4P at the TGN in a FAPP1-dependent manner. (A) Schematic representation of the opto-ERTGoCS construct. (B) The optogenetic approach used to visualize ERTGoCS. In the absence of light (left), the ER-fused SsrA domain is enclosed in the C-terminal helix of the LOV domain (iLID construct). Blue light activation (488 nm, right) induces a conformational change that releases the SsrA domain, allowing its dimerization with the TGN-fused SspB domain and, in turn, ERTGoCS stabilization. (C) Representative images of HeLa cells transfected with the opto-ERTGoCS construct and with the PI4P probe P4M (PI4P). Cells were kept in the dark (left panels) and pulsed with blue light (488 nm; middle panels), and then the blue light was removed (right panels). After blue light stimulation, Cb5 is recruited to the TGN46-GFP area (inset, middle panels) and P4M Golgi localization is reduced. After removal of blue light, Cb5 returns to the ER localization, and P4M intensity increases in the Golgi area. Bar, 10 µm. (D) Mean fluorescence intensity values ± SD of Cb5-mCherry (red) and P4M (gray) in the Golgi area (defined as TGN46-GFP localization; see Materials and methods) in control (left) and in FAPP1-KD cells (right). *, P < 0.05. (E) mCherry-LOV-SsrA-Cb5 localization and PI4P Golgi levels (detected by anti-PI4P antibody) in HeLa cells transfected with the opto-ERTGoCS construct in the dark (top panels), exposed to a blue-light pulse (middle panels) and after removal of light stimulus (bottom panels). Bar, 10 µm. (F) Quantification of the ratio of PI4P and Golgin-97 fluorescence intensity in the Golgi area expressed as a percentage of maximum. Three independent experiments, n = 50. ***, P < 0.001; Student’s t test. (G and H) Schematic representation of constructs used to stabilize the ERTGoCS with rapamycin. (I) PI4P distribution in control, VAP-KD, FAPP1-KD, and VAP-KD+FAPP1-KD cells expressing TGN46-FRB-HA and mCherry-T2A-FKBP-Cb5 reporter proteins upon short (2 min rapamycin, middle panel) and prolonged (15 min rapamycin, right panel) stabilization of ERTGoCS. Inserts show Cb5 localization. Arrowheads indicate ERTGoCS formation. Bar, 10 µm. (J) Quantification of Golgi PI4P levels (solid lines) and number of cells forming ERTGoCS (dashed lines) in control, VAP-KD, FAPP1-KD, and VAP-KD+FAPP1-KD cells treated with 200 nM rapamycin for the indicated times. PI4P was detected using a monoclonal anti-PI4P antibody, and PI4P levels were calculated as the ratio between PI4P and Golgin-97 fluorescence intensity at the Golgi complex. PI4P values are expressed as a percentage of values in cells at time 0 (not exposed to rapamycin). ERTGoCS values are expressed as a percentage of cells showing colocalization of Cb5 with TGN46. Means ± SD, three independent experiments; n = 60–100 cells. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; Student’s t test. (K) Golgi PI4P levels (solid lines) and number of cells forming ERTGoCS (dashed lines) in HeLa WT and VAP-KO cells treated with 200 nM rapamycin for the indicated times. PI4P levels were calculated as the ratio between PI4P and Golgin-97 fluorescence intensity at the Golgi complex. PI4P values are expressed as a percentage of values at time 0 (cells not exposed to rapamycin). ERTGoCS values are expressed as a percentage of cells showing colocalization of Cb5 with TGN46. Means ± SD, three independent experiments; n = 60–100 cells; Student’s t test.

Figure 4.

FAPP1 negatively controls ApoB100 export from the TGN in a PI4P-dependent manner. (A) Immunoblot of the medium and cell lysates from Mock, FAPP1-KD, or ORP10-KD cell cultures using the anti-ApoB100 antibody (see Materials and methods). (B) Quantification of the Western blot shown in A. Secretion of ApoB100 (ratio between the amount in the medium and in the cell lysates) expressed as a percentage of control (CTRL). Means ± SD, n = 3. *, P < 0.05. (C) ApoB100 secretion evaluated by ELISA from control (CTRL) or FAPP1-KD cells, expressed as the amount (ng) of ApoB100 in the medium normalized for the protein content. *, P < 0.05. n = 3. (D) ApoB100, albumin, and α1-anti-trypsin secretion evaluated by pulse-chase. Control or FAPP1-KD cells were incubated as described in Materials and methods. Secretion is expressed as a ratio between the amount of labeled cargo in the medium and the total labeled cargo (medium plus cell lysate). *, P < 0.05. n = 3. (E) Control, FAPP1-KD, and ORP10-KD HepG2 cells were stained with anti-PI4P (green) and anti–Golgin-97 (red) antibodies. Bar, 10 µm. (F) Quantification of PI4P levels shown in E. n = 3; ***, P < 0.001. (G) Golgi-to-PM transport of ApoB100 in control and FAPP1-KD HepG2 cells. Representative images of the distribution of ApoB100 (anti-ApoB100 staining) under steady-state conditions, after a temperature block at 20°C, and after 15 and 60 min of the release from the 20°C temperature block. Insets, Golgin-97 staining. Bar, 10 µm. (H) Golgi-to-PM transport of ApoB100 in control, FAPP1-KD, ORP10-KD, FAPP1+PI4KIIIβ-KD, and ORP10+PI4KIIIβ-KD HepG2 cells. Quantification of Golgi emptying was performed by measuring the ratio between the ApoB100 signal and the Golgin-97 signal at the indicated times after release of the temperature block. Data are means ± SD expressed as a percentage of the ApoB100 signal in the Golgi compared with the signal at time 0 (i.e., at the end of the temperature block). n > 100; three independent experiments. **, P < 0.01; ***, P < 0.001; Student’s t test. (I) Control and FAPP1-KD HepG2 cells were treated as described in G, and analyzed after 15 min of the release of the 20°C block. Cells were stained for ApoB100 (gray), Golgin-97 (G97 inset), and COPI (blue). Bar, 10 µm. The arrowheads indicate ApoB100-positive peripheral structures that are probably post-Golgi carriers since they lack a COPI coat.

Figure 5.

Models of the distribution of modes of action of Sac1 and of the distribution of ERTGoCS and PI4P among Golgi stacks. (A) Two possible modes of action of Sac1 at the ERTGoCS. The width of ERTGoCS ranges from 5 to 20 nm (Venditti et al., 2019). We envisage that the in-trans activity of Sac1 may occur only at the tighter contact sites and only when FAPP1 is present. In the cis mode, Sac1 acts on PI4P that is transferred from the TGN to the ER by OSBP1 in exchange for cholesterol. In this mode, Sac1 can also operate at contact sites that have a greater distance between the opposing membranes since OSBP1 can in principle shuttle between the TGN and the ER. (B) Two possible scenarios of ERTGoCS and PI4P distribution among Golgi stacks. I: The TGN of every Golgi stack can engage in ERTGoCS at any given time. PI4P levels are kept equal and at a medium level in the TGN of all the stacks. ApoB100 export also occurs from the TGN of every stack but at a low rate. II: Only TGN from a subset of Golgi stacks can engage in ERTGoCS. PI4P levels are heterogeneous in the TGN of different stacks, some with high levels of PI4P (devoid of ERTGoCS) and others with low PI4P levels. ApoB100 export occurs exclusively and at a high rate from a TGN with high PI4P levels.