Optogenetic control of phosphoinositide metabolism

Abstract

Phosphoinositides (PIs) are lipid components of cell membranes that regulate a wide variety of cellular functions. Here we exploited the blue light-induced dimerization between two plant proteins, cryptochrome 2 (CRY2) and the transcription factor CIBN, to control plasma membrane PI levels rapidly, locally, and reversibly. The inositol 5-phosphatase domain of OCRL (5-ptase(OCRL)), which acts on PI(4,5)P(2) and PI(3,4,5)P(3), was fused to the photolyase homology region domain of CRY2, and the CRY2-binding domain, CIBN, was fused to plasma membrane-targeting motifs. Blue-light illumination (458-488 nm) of mammalian cells expressing these constructs resulted in nearly instantaneous recruitment of 5-ptase(OCRL) to the plasma membrane, where it caused rapid (within seconds) and reversible (within minutes) dephosphorylation of its targets as revealed by diverse cellular assays: dissociation of PI(4,5)P(2) and PI(3,4,5)P(3) biosensors, disappearance of endocytic clathrin-coated pits, nearly complete inhibition of KCNQ2/3 channel currents, and loss of membrane ruffling. Focal illumination resulted in local and transient 5-ptase(OCRL) recruitment and PI(4,5)P(2) dephosphorylation, causing not only local collapse and retraction of the cell edge or process but also compensatory accumulation of the PI(4,5)P(2) biosensor and membrane ruffling at the opposite side of the cells. Using the same approach for the recruitment of PI3K, local PI(3,4,5)P(3) synthesis and membrane ruffling could be induced, with corresponding loss of ruffling distally to the illuminated region. This technique provides a powerful tool for dissecting with high spatial-temporal kinetics the cellular functions of various PIs and reversibly controlling the functions of downstream effectors of these signaling lipids.

Fig. 1.

Rapid PI(4,5)P₂ dephosphorylation produced by blue light-induced recruitment of a 5-ptase to the plasma membrane of COS-7 cells. (A) Schematic drawing depicting constructs used to induce and detect PI(4,5)P₂ dephosphorylation. (B) (Upper) Confocal micrographs showing the localization of mCh-CRY2-5-ptase_OCRL before, during, and 10 min after illumination with 20 × 300-ms blue-light pulses. (Scale bar: 10 μm.) (Lower) Kymograph drawn along the dashed white line in the upper panel illustrating the nearly instantaneous (within seconds) plasma membrane recruitment of the 5-ptase. Pictures in the upper row are from the time-points indicated below the kymograph. (C) (Upper) Confocal micrographs showing the localization of mCh-CRY2-5-ptase_OCRL and iRFP-PH_PLCδ1 before, during, and 16 min after blue-light illumination. (Scale bar: 10 μm.) (Lower) Kymographs drawn along the dashed white lines in the fields shown in the upper panel. Pictures in the upper row are from the time-points indicated below the kymograph. (D) Dynamics of plasma membrane-associated fluorescence for mCh-CRY2PHR-5pase_OCRL (green) and iRFP-PH_PLCδ1 (red) before, during, and after blue-light illumination (n = 12 cells). (E) Confocal micrographs of the peripheral region of a cell expressing both mCh-CRY2-5pase_OCRL (green) and iRFP-PH_PLCδ1 (red) before and 10 s after a single 100-ms blue-light pulse delivered locally (blue square). Note that the effect of illumination occurs only in the adjacent plasma membrane region. (Scale bar: 5 μm.)

Fig. 2.

PI changes produced by blue light-induced 5-ptase recruitment to the plasma membrane and arrest of clathrin-mediated endocytosis. Both illumination and image recording of COS-7 cells were carried out by TIRFM. (A) Images of cells expressing mCh-CRY2-5-ptase_OCRL (Upper) and RFP-PH_PLCδ1 (Lower) before, during, and 10 min after exposure to a train of 30 × 300-ms blue-light pulses delivered through the evanescent field. (Scale bar: 10 μm.) (B) Schematic drawing illustrating blue-light delivery through TIRF illumination and selective dephosphorylation of PI(4,5)P₂ on the ventral cell membrane. (C) TIRFM recordings of plasma membrane RFP-PH_PLCδ1 fluorescence following blue-light exposure of cells coexpressing CIBN-CAAX and CRY2-5-ptase_OCRL (black) or catalytically inactive CRY2-5-ptase_OCRL(D523G) (blue). Data are presented as means ± SEM for 44 WT and 17 D523G cells. (D) TIRFM recording of plasma membrane RFP-PH_PLCδ1 fluorescence of a cell subjected to three 30 × 200-ms sequential illumination pulse trains showing reproducible PI(4,5)P₂ dephosphorylation. (E) Scatterplot showing the drop in plasma membrane RFP-PH_PLCδ1 fluorescence from cells coexpressing CIBN-CAAX and CRY2-5-ptase_INPP5E, CRY2-5-ptase_{INPP5E(D556A)}, CRY2-5-ptase_OCRL, or CRY2-5-ptase_OCRL(D523G) (n = 16–40 cells). (F) The t_1/2 for RFP-PH_PLCδ1 plasma membrane dissociation following recruitment of CRY2-5-ptase_OCRL or CRY2-5-ptase_INPP5E (n = 32 and 28 cells, respectively). (G) The t_1/2 for RFP-PH_PLCδ1 reassociation with the plasma membrane (n = 32 and 28 cells). (H–J) Dual-color TIRFM recordings from single cells expressing CRY2-5-ptase_OCRL, CIBN-CAAX, and pairs of fluorescent proteins as indicated during exposure to 30 × 200-ms blue-light pulses. Note that data points for GFP imaging were collected only during blue-light illumination. (K) Average change in plasma membrane fluorescence for the data presented in H–J (n = 8–40 cells). (L–N) TIRFM analysis of cells expressing CRY2-5-ptase_OCRL, CIBN-CAAX, and μ2-mCherry before, during, and after exposure to a 10-min train (200 ms, 5-s interpulse intervals) of blue-light pulses. (L) μ2-mCherry fluorescence reflecting individual clathrin-coated pits (fluorescence is shown in black for clarity) in a small cell region before, during, and 20 min after the illumination. (Scale bar: 3 μm.) (M) Representative kymograph. (N) Average number (± SEM) of μ2-mCherry spots (red) and newly formed spots (black) (n = 6 cells).

Fig. 3.

Blue light-induced recruitment of a 5-phosphatase to the plasma membrane rapidly decreases KCNQ2/3 current. (A) (Top) Confocal micrographs show subcellular distribution of mCh-CRY2-5-ptase_OCRL 10 s before and 150 s after blue-light illumination. Simultaneous measurement of normalized mCh-CRY2-5-ptase_OCRL fluorescence intensity at the plasma membrane (Middle) and whole-cell KCNQ2/3 currents (Bottom) from a tsA-201 cell following sustained, global illumination of the cell with 458-nm blue light. Fluorescence intensity was normalized to initial intensity (F/F₀). (Inset) KCNQ2/3 current traces before (black) and after (blue) illumination. The histogram summarizes the percentage inhibition of KCNQ2/3 currents following mCh-CRY2-5-ptase_OCRL recruitment (n = 9). (B) (Top) Confocal micrographs show subcellular distribution of mCh-CRY2-5-ptase_OCRL 10 s before and 150 s and 600 s after blue-light illumination. Simultaneous confocal measurement of normalized mCh-CRY2-5-ptase_OCRL fluorescence intensity at the plasma membrane (Middle) and KCNQ2/3 currents (Bottom) following a 1-s blue-light pulse. (C) (Top) Confocal micrographs show subcellular distribution of PH_PLCδ1-RFP 10 s before and 60 s after blue-light illumination. (Middle) Confocal time series measurements of normalized PH_PLCδ1-RFP fluorescence intensity in the cytoplasm (solid red line) and plasma membrane (dashed red line) following recruitment of CRY2-5-ptase_OCRL with a 1-s pulse of blue light. (Bottom) Whole-cell KCNQ2/3 current recordings in response to the same stimulus. (D) (Top) Confocal micrographs show subcellular distribution of PH_PLCδ1-RFP 10 s before and 40 s after blue-light illumination. Measurement of fluorescence intensity of PH_PLCδ1-RFP (Middle) at the plasma membrane and whole-cell KCNQ2/3 current (Bottom) following the recruitment of CRY2-5-ptase_OCRL(D523G) with a 2-s pulse of blue light. (E) (Left) Confocal micrographs of a tsA-201 cell expressing CRY2-5-ptase_OCRL, CIBN-CAAX, KCNQ2/3, and PH_PLCδ1-RFP 10 s before, 20 s after local (area of local stimulation = 2 × 5 μm; blue arrowheads), and 100 s after global illumination of the cell with blue light. (Right) Concurrent monitoring of normalized cytoplasmic PH_PLCδ1-RFP fluorescence intensity (Upper) and KCNQ current (Lower) following local and global illumination with blue light.

Fig. 4.

Local perturbation of actin dynamics induced by focal blue light-dependent recruitment of a 5-ptase. (A) Confocal micrographs of a COS-7 cell expressing mCh-CRY2-5-ptase_OCRL (green), CIBN-CAAX, and iRFP-PH_PLCδ1 (red) before and 10 s after exposure to a single, locally delivered 100-ms blue-light pulse (blue square). (B) Kymographs drawn along the two white lines in A, either close to (a) or far from (b) the site of focal blue-light illumination. The red arrow indicates a peripheral ruffle. (C) Quantification of the change in plasma membrane RFP-PH_PLCδ1 fluorescence close to (green) or far from (red) the site of focal blue-light illumination as shown in the drawing. Data shown are from 10 cells (means ± SEM). (D and E) TIRFM analysis of a COS-7 cell expressing CIBN-CAAX, CRY2-5-ptase_OCRL, and RFP-PH_PLCδ1 before, during, and after exposure to a train (30 × 200 ms) of blue-light pulses delivered through the evanescent field. (D) TIRFM images (Upper) and kymograph drawn along the white dotted line (Lower). Note the appearance of RFP-PH_PLCδ1–positive ruffles during the recovery phase (red arrows). At time point 2 min there is a change in the position of the cell causing a downward shift in the kymograph of approximately 1 μm. (E) TIRFM recordings of RFP-PH_PLCδ1 fluorescence (average of eight cells). Lines represent fluorescence change for the corresponding color-coded areas in D, Left. (F) Confocal micrographs of RFP-PH_PLCδ1 fluorescence in a PC-12 cell coexpressing CRY2-5-ptase_OCRL and CIBN-CAAX before and after exposure to a locally delivered 100-ms blue-light pulse (blue square). Black-and-white signal has been inverted for clarity. (Scale bars: 10 μm.)

Fig. 5.

Global and local membrane ruffling produced by light-induced PI(3,4,5)P3 synthesis. (A) Schematic drawing depicting constructs used to induce blue light-mediated plasma membrane recruitment of the endogenous catalytic p110α-subunit of PI3-kinase using the iSH2 region of the regulatory p85α-subunit as bait. (B) Confocal micrographs of a COS-7 cell expressing CIBN-CAAX and mCh-CRY2-iSH2. Images were taken before, during, and 10 min after the cells were exposed to a 5-min train (60 × 200 ms) of blue-light pulses. (Scale bar: 10 μm.) (C) Confocal micrographs showing a portion of a COS-7 cell expressing CIBN-CAAX, CRY2-iSH2, and RFP-PH_Akt before, during, and 10 min after cell-wide blue-light illumination. (Scale bar: 5 μm.) (D) Confocal micrograph of a COS-7 cell expressing CIBN-CAAX, CRY2-iSH2, and RFP-PH_Akt (RFP signal is shown in black). (Scale bar: 10 μm.) The cell was subjected to sequential focal illumination at two opposite sites as indicated by small blue circles. (E) Magnifications of the two boxed areas in D (a and b) at the times indicated. (F) Kymographs drawn along the blue and red lines in E. (G) Diagrams showing the distribution of RFP-PH_Akt–positive membrane ruffles in the absence or presence of focal blue-light illumination at the site indicated. Data shown are pooled from five cells.

Fig. P1.

Blue-light–induced control of phosphoinositides. (A) The system is based on the light-dependent heterodimerization between a phosphoinositide-metabolizing enzyme (PI enzyme) fused to CRY2 and the membrane-targeted CIBN-domain of CIB1. (B) Recruitment of an inositol 5-phosphatase to the plasma membrane of a fibroblast resulted in PI(4,5)P₂ dephosphorylation. (C) A focal 100-ms blue-light pulse (blue square) induced local recruitment of the 5-phosphatase (right image, 10 s after illumination) and local PI(4,5)P₂ loss with enhanced membrane ruffling at the opposite cell pole. (D) Light-dependent recruitment of PI 3-kinase to the plasma membrane triggered PI(3,4,5)P₃ formation and membrane ruffling.