Plasma Membrane Phosphatidylinositol 4,5-Bisphosphate ... - Cell Press

Article

Plasma Membrane Phosphatidylinositol 4,5-Bisphosphate Regulates Ca2+-Influx and Insulin Secretion from Pancreatic b Cells Graphical Abstract

Authors Beichen Xie, Phuoc My Nguyen, ek, Antje Thonig, Alenka Guc Sebastian Barg, Olof Idevall-Hagren

Correspondence [email protected]

In Brief Insulin secretion from pancreatic b cells is regulated by glucose. Xie et al. used optogenetic manipulations to show that the concentration of the lipid PI(4,5)P2 in the plasma membrane of pancreatic b cells regulate insulin secretion by controlling the cytosolic Ca2+ concentration. Loss of PI(4,5)P2 suppressed both glucose-induced Ca2+-influx and insulin secretion.

Highlights d

d

d

PI(4,5)P2 regulation of b cells was studied by optogenetic synthesis and depletion Plasma membrane PI(4,5)P2 supports voltage-dependent Ca2+-influx in b cells Loss of PI(4,5)P2 suppresses glucose-induced Ca2+-influx and insulin secretion

Xie et al., 2016, Cell Chemical Biology 23, 816–826 July 21, 2016 ª 2016 Elsevier Ltd. http://dx.doi.org/10.1016/j.chembiol.2016.06.009

Please cite this article as: Xie et al., Plasma Membrane Phosphatidylinositol 4,5-Bisphosphate Regulates Ca2+-Influx and Insulin Secretion from Pancreatic b Cells, Cell Chemical Biology (2016), http://dx.doi.org/10.1016/j.chembiol.2016.06.009

Cell Chemical Biology

Article PlasmaMembranePhosphatidylinositol4,5-Bisphosphate Regulates Ca2+-Influx and Insulin Secretion from Pancreatic b Cells ek,1 Antje Thonig,1 Sebastian Barg,1 and Olof Idevall-Hagren1,* Beichen Xie,1,2 Phuoc My Nguyen,1,2 Alenka Guc 1Department

of Medical Cell Biology, Uppsala University, Biomedical Centre, Husargatan 3, Box 571, 75123 Uppsala, Sweden author *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chembiol.2016.06.009 2Co-first

SUMMARY

Insulin secretion from pancreatic b cells is regulated by the blood glucose concentration and occurs through Ca2+-triggered exocytosis. The activities of multiple ion channels in the b cell plasma membrane are required to fine-tune insulin secretion in order to maintain normoglycemia. Phosphoinositide lipids in the plasma membrane often gate ion channels, and variations in the concentration of these lipids affect ion-channel open probability and conductance. Using light-regulated synthesis or depletion of plasma membrane phosphatidylinositol 4,5-bisphosphate (PI[4,5]P2), we found that this lipid positively regulated both depolarization- and glucose-triggered Ca2+ influx in a dose-dependent manner. Small reductions of PI(4,5)P2 caused by brief illumination resulted in partial suppression of Ca2+ influx that followed the kinetics of the lipid, whereas depletion resulted in marked inhibition of both Ca2+ influx and insulin secretion.

INTRODUCTION Phosphoinositide lipids (PIs) are found in all cellular membranes where they play key roles as regulators of protein localization and function. Reversible phosphorylation of the inositol head group generates seven different PIs, with phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate (PI[4,5]P2) being the most abundant PIs in the plasma membrane. Many fundamental processes, such as endo- and exocytosis, cell motility, and receptor signaling require the presence of PI(4,5) P2 in the plasma membrane (Di Paolo and De Camilli, 2006; Balla, 2013; Martin, 2015). In excitable cells, such as neuroendocrine cells, PI(4,5)P2 additionally acts as a regulator of ion-channel conductivity, where the polar head group of the lipid interacts with transmembrane segments of ion channels to stabilize the open conformation and facilitate conductance (Hille et al., 2015). Multiple ion channels are regulated by PI(4,5)P2, including voltage-dependent Ca2+ channels (VDCC) and ATP-gated K+ channels (KATP channels). Experimental reduction of PI(4,5)P2 in the plasma membrane induces varying degrees of suppres-

sion of conductance through the channels, ranging from 10% to complete inhibition depending on the channel type, experimental model system, and method used for PI(4,5)P2 depletion (Rohacs et al., 2003; Shyng and Nichols, 1998; Suh et al., 2006, 2010). Most studies of ion-channel regulation by PI(4,5) P2 are limited to electrophysiological characterization of changes in channel currents, and it is still largely unknown how the channel gating affects changes in intracellular Ca2+ concentration ([Ca2+]i) and Ca2+-triggered exocytosis. Insulin secretion from the electrically excitable pancreatic b cells is metabolically regulated. As glucose enters the cells, it is rapidly metabolized to ATP. This nucleotide closes KATP channels in the plasma membrane, resulting in depolarization and opening of primarily L-type VDCCs. The resulting influx of Ca2+ is the major trigger of insulin granule exocytosis (Tengholm and Gylfe, 2009). The presence of PI(4,5)P2 in the b cell plasma membrane has been found to have a hyperpolarizing effect by decreasing the KATP-channel sensitivity to ATP (Lin et al., 2005). The L-type VDCCs are also regulated by PI(4,5)P2 in other cell types (Suh et al., 2010). It is not clear whether such regulation is of importance for b cell electrical activity and insulin secretion. The concentrations of multiple PIs, including PI(4,5)P2, are altered during glucose stimulation of b cells, and these changes are tightly related to changes in [Ca2+]i (Hagren and Tengholm, 2006; Thore et al., 2004, 2005; Wuttke et al., 2010). However, the precise role of this lipid in b cell function is not clear, and there are studies demonstrating both stimulatory and inhibitory effects on insulin secretion (Huang et al., 2011; Olsen et al., 2003; Tomas et al., 2010; Zhang et al., 2009). We employed recently developed optogenetic modules that allow acute light-driven modulation of plasma membrane PI(4,5)P2 to define its role in the regulation of b cell function. PI(4,5)P2 depletion resulted in strong suppression of voltagedependent Ca2+ influx and of insulin secretion. Ca2+ influx was suppressed or augmented already at moderately lowered or elevated PI(4,5)P2 concentrations, respectively, indicating that conditions that affect plasma membrane PI(4,5)P2 might have profound effects on b cell secretory capacity. RESULTS Plasma Membrane PI(4,5)P2 and Cytosolic Ca2+ Concentrations Correlate The concentration of plasma membrane PI(4,5)P2 is regulated by ATP availability in permeabilized b cells (Thore et al., 2007). To

816 Cell Chemical Biology 23, 816–826, July 21, 2016 ª 2016 Elsevier Ltd.


A

E

B

C

F

D

G

Figure 1. Ca2+ and PI(4,5)P2 Are Interdependently Regulated (A) TIRF microscopy recordings of PH-PLCd1-GFP fluorescence from an MIN6 cell exposed to 30 mM K+ and 1 mM FCCP. (B) Quantification of the response in (A). Data from 26 cells are presented as means ± SEM. (C) TIRF microscopy recordings of PH-PLCd1-GFP fluorescence in an a-toxin permeabilized MIN6 cell exposed to the indicated Ca2+ buffers. (D) Quantification of the response in (A). Data from ten cells are presented as means ± SEM. (E) TIRF microscopy recording of R-GECO (red) and PH-PLCd1-GFP (green) fluorescence from an MIN6 cell exposed to 100 mM CPA, 30 mM K+, and 100 mM carbachol. (F) TIRF microscopy recording of R-GECO (red) and PH-PLCd1-GFP (green) fluorescence from an M1-receptor-overexpressing MIN6 cell exposed to 100 mM CPA, 30 mM K+, and 100 mM carbachol. (G) Quantification (means ± SEM) of TIRF microscopy recordings of R-GECO (top) and PH-PLCd1-GFP (bottom) fluorescence change in response to the indicated treatments. Data from 47 (control) and 22 (M1R overexpression) cells. ***p < 0.001 compared with K+/CPA, ##p < 0.01 compared with K+, #p < 0.05 compared with K+, {p < 0.001 compared with control (CPA/K+/carb).

test if the same is true in intact cells, we examined the effect of blocking mitochondrial ATP production by the proton ionophore FCCP on plasma membrane PI(4,5)P2. When clonal MIN6 b cells expressing the PI(4,5)P2-binding protein GFP-PH-PLCd1 were exposed to FCCP, this caused an immediate dissociation of GFP-PH-PLCd1 from the plasma membrane, seen as a 32% ± 5% decrease in GFP fluorescence near the plasma membrane when examined with total internal reflection fluorescence (TIRF) microscopy (Figures 1A and 1B). PI(4,5)P2 has also been shown to be regulated by Ca2+. To directly test this, we used a-toxin-permeabilized MIN6 b cells expressing GFP-PH-PLCd1 and the Ca2+-indicator R-GECO. TIRF microscopy revealed dose-dependent dissociation of GFP-PH-PLCd1 from the plasma membrane that was maximal at 50 mM Ca2+ (Figures 1C and 1D). Consistent with previous studies (Thore et al., 2004), depolarization of intact MIN6 cells with 30 mM K+ caused voltage-dependent Ca2+ influx that was accompanied by a small reduction in plasma membrane PI(4,5)P2 (Figure 1A). In addition to regulation by ATP and Ca2+, PI(4,5)P2 is consumed after receptor-triggered activation of phospholipase C (PLC). When carbachol was added to activate PLC, this resulted in further loss of plasma membrane PI(4,5)P2 and, surprisingly, was associated with 21% ± 3% (n = 47, p < 0.001) reduction in the depo-

larization-induced Ca2+ influx. Similar results were obtained in the presence of the sarco-ER Ca2+-ATPase (SERCA) inhibitor cyclopiazonic acid (CPA), which prevented carbachol-induced release of Ca2+ from the ER (Figures 1E and 1G). Both the carbachol-induced PI(4,5)P2 hydrolysis and the accompanying suppression of Ca2+ influx was more pronounced in MIN6 cells overexpressing PLC-coupled M1-receptors for carbachol (Figures 1F and 1G). Together, these observations indicate that plasma membrane PI(4,5)P2 is a positive regulator of the cytoplasmic Ca2+ concentration in b cells. Light-Induced Reduction of Plasma Membrane PI(4,5)P2 To gain better temporal control over plasma membrane PI(4,5)P2 we took advantage of a recently developed optogenetic technique that enabled recruitment of a PI(4,5)P2 50 -phosphatase module (GFP-CRY2-OCRL) to CIBN-CAAX in the plasma membrane through blue-light-dependent interactions between CRY2 and CIBN (Opto-5ptase) (Idevall-Hagren et al., 2012; Kennedy et al., 2010) (Figure 2A). Blue-light illumination of MIN6 or INS1 b cells expressing these modules, together with the PI(4,5)P2 biosensor PH-PLCd1-mRFP, caused immediate binding of GFP-CRY2-OCRL to the plasma membrane, followed by PH-PLCd1-mRFP dissociation (Figures 2B and 2C). Using TIRF Cell Chemical Biology 23, 816–826, July 21, 2016 817


A

B

C

D

E

F

G

H

I

J

K

Figure 2. Optogenetic Control of Plasma Membrane PI(4,5)P2 (A) Principle of blue-light-induced PI(4,5)P2 dephosphorylation. (B) Confocal micrographs of MIN6 cells expressing mCh-CRY2-OCRL, iRFP-PH-PLCd1, and CIBN-CAAX (not shown) in the absence of blue light. Magnifications show the time course of light-induced recruitment of mCh-CRY2-OCRL to the plasma membrane and the corresponding loss of iRFP-PH-PLCd1. (C) mCh-CRY2-OCRL and iRFP-PH-PLCd1 fluorescence change at the plasma membrane following illumination. (D) TIRF microscopy recording from an MIN6 cell expressing CIBN-CAAX, GFP-CRY2-OCRL, and RFP-PH-PLCd1 during blue-light illumination of varying durations (blue bars). Bottom trace shows mRFP-PH-PLCd1 fluorescence during continuous imaging and the top trace shows the GFP-CRY2-OCRL fluorescence at the interruption of blue-light illumination (circle) with extrapolated dotted lines. (E) Schematic drawing of the principle for TIRF microscopy imaging and manipulation of islets. (F and G) TIRF microscopy images (top) and recordings (bottom) from two cells within a mouse islet infected with Ad-CIBN-CAAX, Ad-GFP-CRY2-OCRL, and Ad-RFP-PH-PLCd1 during repeated blue-light illumination. (H) Scatterplot showing the correlation between evanescent wave blue-light illumination time and RFP-PH-PLCd1 dissociation from the plasma membrane. Each dot represents one cell (n = 9–20 cells per time point). The data have been fitted to a single exponential function. (I) TIRF microscopy recordings from an MIN6 cell exposed to blue-light evanescent wave illumination of varying duration. Stimulations have been aligned to the first frame of illumination. (J and K) Quantification (means ± SEM) of the RFP-PH-PLCd1 fluorescence decrease in response to illumination (J) and time to recovery following interruption of the illumination (K) in MIN6 cells (n = 40 cells). The response to 100 mM carbachol is shown for comparison. See also Figure S1.

microscopy, blue light was delivered at the cell-coverslip interface through the low intensity evanescent field. This resulted in light-dose-dependent binding of GFP-CRY2-OCRL to the plasma membrane and the corresponding loss of PH-PLCd1mRFP from the same compartment (Figures 2D and 2H). Brief illumination (0.4 s) resulted in PH-PLCd1-mRFP dissociation 818 Cell Chemical Biology 23, 816–826, July 21, 2016

that was comparable with PLC activation (19% ± 2% fluorescence decrease) (Figure S1), whereas maximal response required 10 s illumination and corresponded to 62% ± 4% (n = 40, p < 0.001) drop in PH-PLCd1-mRFP fluorescence. Recovery of baseline PI(4,5)P2 was also time dependent and took 4.2 ± 0.6 min following interruption of illumination after maximal

Light

3.5

1.0

2.5

0.8

2.0

0.6

1.5

0.4

1.0

D 1.2 1.0 0.8 0.6 0.4 0.2 0.0

rs = -0.82 P