PI(4,5)P2 controls plasma membrane PI4P and PS levels via ORP5/8 recruitment to ER-PM contact sites

Abstract

Phosphatidylinositol 4,5-bisphosphate (PI(4,5)P₂) is a critically important regulatory lipid of the plasma membrane (PM); however, little is known about how cells regulate PM PI(4,5)P₂ levels. Here, we show that the phosphatidylinositol 4-phosphate (PI4P)/phosphatidylserine (PS) transfer activity of the endoplasmic reticulum (ER)-resident ORP5 and ORP8 is regulated by both PM PI4P and PI(4,5)P₂ Dynamic control of ORP5/8 recruitment to the PM occurs through interactions with the N-terminal Pleckstrin homology domains and adjacent basic residues of ORP5/8 with both PI4P and PI(4,5)P₂ Although ORP5 activity requires normal levels of these inositides, ORP8 is called on only when PI(4,5)P₂ levels are increased. Regulation of the ORP5/8 attachment to the PM by both phosphoinositides provides a powerful means to determine the relative flux of PI4P toward the ER for PS transport and Sac1-mediated dephosphorylation and PIP 5-kinase-mediated conversion to PI(4,5)P₂ Using this rheostat, cells can maintain PI(4,5)P₂ levels by adjusting the availability of PI4P in the PM.

Figure 1.

PM interaction of ORP5/8 determines level of PI4P in the PM. (A) PI4P metabolism at ER-PM contact sites (left) and a linear domain structure of ORP5/8 (right). After being synthesized by PI4KA, PI4P is either phosphorylated by PIP5K to PI(4,5)P₂ in the PM or transported to the ER. In the ER, PI4P is dephosphorylated by Sac1 phosphatase (left). Both ORP5 and ORP8 consist of a polybasic (PB) domain, a PH domain, an OSBP-related domain (ORD), and a transmembrane (TM) domain from N to C terminus (right). (B) Intracellular localization of ORP5, ORP8, and ORP8-PLCδ1-PH in live cells. HEK293-AT1 cells were transfected with GFP-tagged ORP5, ORP8, or ORP8-PLCδ1-PH and observed with confocal microscopy after 1 d. The same group of cells was observed on both the middle focal plane (left) and the bottom focal plane (right). Bars, 10 µm. (C) Quantitation of PM PI4P levels in live HEK293-AT1 cells with overexpression of ORP5, ORP8, or ORP8-PLCδ-PH by BRET analysis (see Materials and Methods for details). Grand means ± SEM are shown from three independent experiments performed in triplicate and normalized to the mean BRET value of mCherry-transfected (control) cells. Statistical significance was obtained with one-way ANOVA (P < 0.005). (D) Kinetics of PM-PI4P decrease by A1 treatment in BRET analysis. After measuring baseline BRET signal as shown in C, cells were treated with A1 (30 nM) and monitored for PI4P decrease. Grand means ± SEM are shown from three independent experiments performed in triplicate and normalized to the initial BRET value of mCherry-transfected (control) cells.

Figure 2.

PM binding of ORP5 depends on both PI4P and PI(4,5)P₂. (A) Kinetics of PI(4,5)P₂ resynthesis after AngII stimulation in ORP5- or ORP8-expressing cells. HEK293-AT1 cells transfected with mCherry-tagged ORP5, ORP8, or mCherry were analyzed with BRET using PM-anchored Venus and PLCδ-PH–fused luciferase. After baseline measurement, cells were treated with vehicle or AngII (100 nM). Relative change in PI(4,5)P₂ level by AngII (AngII/vehicle) was normalized to the mean values gained from baseline measurements (the baseline values showed a slight increase both with ORP5 and ORP8 expression as shown in the insert). Grand means ± SEM are shown from three independent experiments performed in triplicate. (B) Quantitation of PI4P changes after PM recruitment of FKBP12-fused ORP5/8. HEK293-AT1 cells were transfected with mCherry or mCherry-tagged FKBP-ORP5 (FK-ORP5) or -ORP8 (FK-ORP8) in addition to PM-anchored FRB (PM2-FRB). PM PI4P levels were quantitated with BRET analysis using PM-anchored Venus and P4M(2x)-fused luciferase. After baseline measurement, cells were treated with DMSO or rapamycin (100 nM) for recruiting FK-ORP5 or -ORP8 to the PM. Relative PI4P level (rapamycin/DMSO) was normalized to mean value of baseline measurement. Grand means ± SEM are shown from three independent experiments performed in triplicate. (C) ORP5-PM contacts in PM PI4P- or PI(4,5)P₂-depleted cells. COS-7 cells were transfected with GFP-tagged ORP5 and mCherry-tagged FKBP-PJ, -PJ-Sac, -PJ-Dead, or FKBP-INPP5E. Lyn₁₁-FRB-iRFP was transfected as a recruiter. 1 d after transfection, cells were imaged by time-lapse TIRF microscopy, inducing acute depletion of PI4P, PI(4,5)P₂, or both with the addition of 1 µM rapamycin at time 0. Representative images from the time-lapse 0, 2.5, 5, 7.5, and 10 min of rapamycin treatment (left). The graph shows normalized GFP-ORP5 fluorescence intensity in the evanescent field for 28–34 cells imaged across three independent experiments (means ± SEM). Bars, 10 µm.

Figure 3.

Extra PI(4,5)P₂ production increases the PM engagement of ORP8. (A) Representative live cell images showing intracellular localization of ORP8 under overexpression of PIP5Kβ. HEK293-AT1 cells were cotransfected with GFP-tagged ORP5 and wild-type or kinase-dead (dead) mRFP-tagged PIP5Kβ. After 1 d of transfection, cells were observed with confocal microscopy. Bars, 10 µm. (B) Comparison of amino acid sequences of N-terminal PH domains between ORP5 and ORP8. ORP5 and ORP8 share strictly defined PH domains (blue box). Conserved polybasic residues (blue) precede strict PH domains in both ORP proteins. (C) Representative live cell images showing localization of PH domains with or without adjacent polybasic residues. HEK293-AT1 cells were transfected with GFP-tagged ORP8-PH, ORP8-ePH, ORP5-PH, or ORP5-ePH. After 1 d, localization of PH domains was observed with confocal microscopy. Bars, 10 µm. (D) Representative live-cell images displaying effect of PIP5K overexpression in PM engagement of ORP8-ePH. HEK293-AT1 cells were cotransfected with GFP-tagged ORP8-ePH and mRFP-tagged PIP5Kβ (wild-type or kinase-dead). After 1 d, cells were observed with confocal microscopy. Bars, 10 µm. (E) Dependence of the Förster resonance energy transfer of ORP8 PH domain tryptophan residues to dansyl-labeled liposomes (monitored by dansyl fluorescence intensity) on the concentration of the PH domains. Solid and open circles, 1–284 and 91–284 constructs of the ORP8 PH domain. Liposomes contained 5 mol % of PI(4,5)P₂ (black), 5 mol % of PI4P (red), or 2.5 mol % of PI4P plus 2.5 mol % of PI(4,5)P₂ (blue). Squares, control liposomes bearing the same net charge from the negatively charged PS but no phosphoinositides. One representative result is shown from three separate experiments that gave identical results but differed in absolute fluorescent intensity values.

Figure 4.

Both PI4P and PI(4,5)P₂ bind to the PH domain of ORP8. (A) The interaction with PI4P/PI(4,5)P₂ induced specific changes in the NMR spectra of ORP8. (B and C) We observed both significant reduction of signal intensity (B) and changes in positions of signals (C) in the 2D ¹⁵N/¹H HSQC spectra of ¹⁵N-labeled ORP8 PH domain upon the addition of the ligand. The graphs represent simple differences in signal positions and intensities between the free and ligand-bound protein obtained from a single experiment. (D) The PH domain adopts the canonical fold, and both PI4P and PI(4,5)P₂ bind to the same interface as suggested by highlighting the relative changes observed in the NMR spectra on the 3D protein structure. (E and F) The NMR-data driven models of the PH/PI4P (E) and PH/PI(4,5)P₂ (F) complexes suggest that the phosphate groups of PI4P/PI(4,5)P₂ are stabilized by an extensive network of electrostatic interactions with the positively charged side chains of Arg¹⁵⁸, Arg²⁰¹, Lys²¹¹, and Arg²⁴⁶. PI(4,5)P₂ is additionally stabilized by an interaction with Lys²⁰⁴ and Ser²⁰³.

Figure 5.

PI4P transport is controlled by PI(4,5)P₂ levels in the PM. (A) Acute depletion of PM PI(4,5)P₂ achieved by PM recruitment of INPP5E. The cartoon illustrates that PM-recruited 5ptase catalyzes dephosphorylation of PI(4,5)P₂ to PI4P in the PM (top). Changes in PM PI(4,5)P₂ levels were quantitated by BRET analysis before and after recruitment of 5ptase (bottom). HEK293-AT1 cells were cotransfected with mCherry-tagged FK-INPP5E and PM2-FRB before BRET analysis. Control cells were transfected with FK-PJ-dead plasmid instead of FK-INPP5E construct. After baseline measurement, cells were treated with DMSO or rapamycin for recruiting INPP5E to the PM. Relative PI(4,5)P₂ levels were normalized to the initial BRET value of control cells. Grand means ± SEM are shown from three independent experiments performed in triplicate. (B) Rate of PI4P clearance controlled by PM PI(4,5)P₂ levels. The cartoon illustrates that PM PI4P clearance after A1 treatment decreases when the ORP5/8-mediated transport is switched off as the 5ptase is recruited to the PM (top). PM PI4P clearance (after A1 addition) was monitored with or without PM PI(4,5)P₂ depletion in BRET analysis (bottom). HEK293-AT1 cells were cotransfected with PM2-FRB and FK-INPP5E together with the BRET construct before BRET measurement. Control cells were transfected with FK-PJ-dead instead of FK-INPP5E. After baseline measurement, cells were treated with rapamycin for recruiting FK-PJ-dead (control) or -INPP5E to the PM and subsequently with A1 (30 nM) as indicated by the arrows. Grand means ± SEM are shown from three independent experiments performed in triplicate and normalized to the initial BRET value of control cells. The gray area shows the time period for which the rate of decline was calculated in each experiment. These rate values are shown in the column diagram relative to the rates calculated in the controls (***, P < 0.005; n = 3). (C) Representative BRET experiment showing rate of PI4P clearance from the PM under dose-dependent overexpression of myc-tagged PIP5Kβ. HEK293-AT1 cells were transfected with 0 ng (blue), 0.5 ng (green), 1 ng (red), 4 ng (orange), and 16 ng (purple) DNA per well of myc-PIP5Kβ before BRET analysis. Means ± SD are shown from triplicate experments, normalized to the initial BRET value of the cells with no myc-PIP5Kβ (blue). (D) Rate of PI4P clearance correlated with initial PM PI4P levels (square symbols). Two more BRET experiments were conducted by transfecting 0, 2, 6, and 12 ng PIP5Kβ DNA per well (circles) or 0, 4, 8, and 16 ng of the myc-PIP5Kβ DNA per well (triangles), also performed in triplicate. After treatment of A1, PI4P reduction rate (slope designated in y-axis) was calculated from the highest levels (initial PI4P designated in x-axis) for 1,000-s. Relative reduction rates were normalized to initial PI4P levels with no myc-PIP5Kβ in each experiment. (E) Effect of PIP5Kβ overexpression on PM PI4P levels responding to AngII stimulation. HEK293-AT1 cells were transfected with pcDNA3.1(HA) vector or myc-tagged PIP5Kβ. Change in PM PI4P levels was analyzed with BRET. After baseline measurement, cells were simulated with AngII (100 nM). Grand means ± SEM are shown from three independent experiments performed in triplicate and normalized to the initial BRET value of vector-transfected cells. (F) Effect of PIP5Kβ overexpression in PM PI(4,5)P₂ levels responding to AngII stimulation. BRET analysis was conducted as described in E except that PM PI(4,5)P₂ levels were monitored. Grand means ± SEM are shown from three independent experiments performed in triplicate and normalized to the initial BRET value of vector-transfected cells.

Figure 6.

Loss of ORP5/8 increases both PI4P and PI(4,5)P₂ levels in the PM. (A) PM PI4P levels in ORP5-, ORP8-, or ORP5/8-depleted cells. HEK293-AT1 cells were transfected with control, ORP5-, ORP8-, or both ORP5- and ORP8-silencing siRNAs for 2 d and subjected to BRET analysis monitoring PM PI4P. Three independent experiments performed in triplicate were each normalized to control. Grand means ± SEM are shown. Statistical analysis was performed by two-way ANOVA (**, P < 0.005; ns, not significant). (B) Same experiments as in A except that PM PI(4,5)P₂ levels were assessed by BRET measurements. (C) Effect of knockdown of ORP5/8 on PI4P levels in PIP5Kβ-expressing cells. HEK293-AT1 cells were transfected with the indicated siRNAs as in A. 1 d after siRNA transfection, cells were transfected with pcDNA3.1(HA) or myc-tagged PIP5Kβ (2 ng/well). 1 d after DNA transfection, cells were subjected to BRET measurement for PM PI4P. Statistical analysis from three independent experiments performed in triplicate was obtained as described for A. (D) Representative live-cell confocal images showing changes in PI4P distribution in response to PIP5Kβ expression in cells treated with control or ORP5/8-targeting siRNAs. HEK293-AT1 cells were transfected with indicated siRNAs. 1 d after siRNA transfection, cells were cotransfected with GFP-tagged P4M(2x) and mRFP-tagged PIP5Kβ (50 ng/well). 1 d after DNA transfection, cells were observed with a confocal microscope. Note the presence of some PI4P in the PM even in the PIP5K-expressing cells in the ORP5/8-depleted cells. Bars, 10 µm. (E) Kinetics of PM PI4P decrease after A1 treatment (left) in cells expressing PIP5Kβ with or without depletion of ORP5/8. Experiment was as described in C except for treatment of cells with A1 (30 nM) after a control measurement period. The rate of PM PI4P decrease was calculated for the time period indicated by the gray area (D). Grand means ± SEM are shown from three independent experiments performed in triplicate. Values were normalized to the rate value calculated for the vector-transfected cells treated with control siRNA. Statistical analysis was obtained with two-way ANOVA (**, P < 0.005).

Figure 7.

Molecular interaction between ORP5 and ORP8 affects their PM localization. (A) Representative live-cell confocal images of cells expressing full-length ORP5 and ORP8 (left) and the cartoon of their structural features (right). Abbreviations are the same as used in Fig. 1, and the yellow box designates the acidic N-terminal stretch present in ORP8 but not ORP5. HEK293-AT1 cells were cotransfected with mCherry-tagged ORP5 and GFP-tagged ORP8 for 1 d for confocal analysis. Note the lack of punctate ORP5 signal in the ORP8-expressing cell. (B) Representative images showing cells coexpressing GFP-ORP5 and mCherry-tagged ORP8 in which the PH domain was replaced by an FKPB12 module (mCherry-FK-ORP8, purple box). mCherry-FK-ORP8 expressing cells show more ORP5 showing up in the ER, but still preserve some PM puncta. (C) Representative images showing cells coexpressing full-length GFP-ORP5 and mCherry-tagged ORP8 N-terminally truncated to the same point defined as the start of ePH as described in Fig. 3. Note that the truncated ORP8 is less effective in preventing ORP5 PM localization and shows punctate localization itself. Bars, 10 µm.

Figure 8.

PIP5Kβ-expressing cells show reduced PM PS levels. (A) Representative live-cell image displaying PS distribution under overexpression of wild-type PIP5Kβ. HEK293-AT1 cells were cotransfected with GFP-tagged Lact-C2 and mRFP-tagged PIP5Kβ and subjected to confocal microscopy 1 d after transfection. (B) Representative live-cell image displaying PS distribution under overexpression of kinase-dead PIP5Kβ. HEK293-AT1 cells were cotransfected with GFP-tagged Lact-C2 and mRFP-tagged PIP5Kβ-dead and subjected to confocal microscopy 1 d after transfection. Bars, 10 µm. (C) Quantitation of PM PS levels in wild-type or kinase-dead PIP5Kβ-expressing cells with BRET analysis. HEK293-AT1 cells were transfected with pcDNA3.1(HA) vector, myc-PIP5Kβ, or myc-PIP5Kβ-dead before BRET experiment. PM PS levels were analyzed by measuring mean emission intensity of PM-anchored Venus per Lact-C2-fused luciferase in the presence of coelenterazine h for 7 min. Grand means ± SEM are shown from three independent experiments performed in triplicate. Statistical significance was obtained with two-way ANOVA (***, P < 0.001; **, P < 0.005). (D) Dynamic range of PM PS levels quantitated with BRET. HEK293-AT1 cells were transfected with mCherry empty vector, and PM PS levels were analyzed by measuring mean intensity of PM-anchored Venus per Lact-C2-fused luciferase. After monitoring steady state under DMSO treatment, Ionomycin (10 µM) was treated to deplete PS in the inner leaflet of the PM resulting from PS externalization. Grand means ± SEM are shown from three independent experiments performed in triplicate.