Choline Is an Intracellular Messenger Linking Extracellular Stimuli to IP3-Evoked Ca2+ Signals through Sigma-1 Receptors

Abstract

Sigma-1 receptors (Sig-1Rs) are integral ER membrane proteins. They bind diverse ligands, including psychoactive drugs, and regulate many signaling proteins, including the inositol 1,4,5-trisphosphate receptors (IP₃Rs) that release Ca²⁺ from the ER. The endogenous ligands of Sig-1Rs are unknown. Phospholipase D (PLD) cleaves phosphatidylcholine to choline and phosphatidic acid (PA), with PA assumed to mediate all downstream signaling. We show that choline is also an intracellular messenger. Choline binds to Sig-1Rs, it mimics other Sig-1R agonists by potentiating Ca²⁺ signals evoked by IP₃Rs, and it is deactivated by metabolism. Receptors, by stimulating PLC and PLD, deliver two signals to IP₃Rs: IP₃ activates IP₃Rs, and choline potentiates their activity through Sig-1Rs. Choline is also produced at synapses by degradation of acetylcholine. Choline uptake by transporters activates Sig-1Rs and potentiates Ca²⁺ signals. We conclude that choline is an endogenous agonist of Sig-1Rs linking extracellular stimuli, and perhaps synaptic activity, to Ca²⁺ signals.

Keywords: Ca(2+); G-protein-coupled receptor; IP(3) receptor; Sigma-1 receptor; bradykinin; choline; intracellular messenger; neurotransmitter; phospholipase C; phospholipase D.

Graphical abstract

Figure 1

Choline Is an Agonist of Sig-1Rs (A) Clusters of Sig-1Rs anchored at MAMs are thought to dissociate into monomers when they bind a Sig-1R agonist, freeing Sig-1Rs to interact with their targets, within and beyond MAMs. The targets include IP₃Rs. (B) Specific binding of [³H](+)-pentazocine (5 nM) in the presence of choline and related compounds using membranes from Neuro-2A cells stably expressing Sig-1R-GFP (mean ± SEM; n = 5, with 3 replicates for each). Specific binding of ³H-pentazocine was 90% ± 3% of total binding (mean ± SEM; n = 3) for membranes from cells overexpressing Sig-1R, and 13% ± 5% for mock-transfected cells. (C) Choline metabolism (structures from http://www.hmdb.ca). (D) NG108-15 cells were incubated (2 hr, 37°C) with PRE-084 (25 μM) or BD1047 (25 μM), and then, in the continuous presence of the Sig-1R ligands, loaded with Fluo-8 by incubation with Fluo-8 AM in HEPES-buffered saline (HBS) (30 min, 20°C, with a further 30 min to allow de-esterification of Fluo-8). BAPTA (2.5 mM) was then added to chelate extracellular Ca²⁺ before addition of bradykinin (10 μM). Results show typical responses as means of 3 replicates. (E) Summary results (mean ± SEM; n = 5, each with 3 replicates) show peak increases in [Ca²⁺]_i (Δ[Ca²⁺]_i) evoked by bradykinin. ^∗p < 0.05 for maximal responses relative to control, one-way ANOVA with Dunnett’s test. (F) Pooled results (mean ± SEM; n = 20; as percentages of matched control response) for all bradykinin concentrations. The asterisk (^∗) denotes 95% confidence intervals that exclude 100%. See also Figure S1A.

Figure 2

Choline Potentiates IP₃-Evoked Ca²⁺ Release by Stimulating Sig-1Rs (A) Ca²⁺ signals recorded from Fura-2-loaded NG108-15 cells after microinjection (∼1% cell volume) of IP₃ (pipette concentration, 0.5 μM), (+)SKF-10047 (SKF, 100 μM), or choline (100 mM). Results (n = 6 cells) show untransfected cells or after transfection with scrambled shRNA or Sig-1R shRNA, each tagged with red fluorescent protein (RFP). (B) Summary (mean ± SD; n = 6) shows peak [Ca²⁺]_i. ^∗p < 0.05, ANOVA with Bonferroni test, relative to matched stimuli in untransfected cells. The effects of pre-incubating cells with BD1047 (25 μM, 15 min) are also shown. (C) Similar analysis of the effects of microinjected IP₃ (pipette concentration, 0.5 μM) or acetylcholine, betaine, or phosphocholine (pipette concentration, 100 mM for each), alone or in combination. (D) Summary (mean ± SD; n = 6) shows peak [Ca²⁺]_i. ^∗p < 0.05, ANOVA with Bonferroni test, relative to IP₃ alone. (E) Western blot (WB) of Sig-1R after transfection of NG108-15 cells with scrambled or Sig-1R shRNA, each tagged with RFP. Tagged shRNAs were used to allow identification of transfected cells in microinjection experiments. Hence, WB from cell populations probably over-estimates Sig-1R expression in functional analyses of micro-injected cells treated with Sig-1R shRNA. Sig-1R expression was reduced to 50% ± 12% of control levels by the shRNA treatment (mean ± SD; n = 3). (F) WB showing detectable expression of Sig-1R in MCF7 cells only after transfection with Sig-1R-GFP. Typical of 4 blots. M_r markers (kDa) are shown. (G) Ca²⁺ signals recorded from Fura-2-loaded MCF7 cells after microinjection as described for (C). Results (n = 6 cells) are from control cells or after transfection with GFP or Sig-1R-GFP. (H) Summary (mean ± SD; n = 6) results show [ΔCa²⁺]_i. ^∗p < 0.05 for maximal responses relative to matched untransfected cells, one-way ANOVA with Dunnett’s test. See also Figures S1B and S1C.

Figure 3

Sig-1R and PLD Contribute to Ca²⁺ Signals Evoked by Agonists of GPCRs (A) Typical pseudocolor images show peak Ca²⁺ signals (F₃₄₀/F₃₈₀) evoked by ATP (50 μM) in Fura-2-loaded NG108-15 cells transfected with control shRNA or shRNA to PLD1 and PLD2, or Sig-1R, each tagged with RFP. Calibration code (F₃₄₀/F₃₈₀) and scale bar (20 μm) apply to all panels. (B) Time course of response to ATP (bar; n = 6). (C) Summary (mean ± SD; n = 6) shows Δ[Ca²⁺]_i evoked by ATP. ^∗p < 0.05, ANOVA with Bonferroni test, relative to untransfected cells. (D) Δ[Ca²⁺]_i evoked by bradykinin in populations of NG108-15 cells. Histogram (which shares the y axis) compares responses to bradykinin (10 μM) after treatment with scrambled or Sig-1R shRNA. Results are means ± SEM; n = 3 with duplicate determinations. ^∗p < 0.05, Student’s t test. (E) WB shows effects of indicated shRNA, each tagged with RFP, on expression of PLD1 and PLD2 in NG108-15 cells. M_r markers (kDa) are shown. Results, typical of 3 WBs, underestimate knockdowns in the cells used for Ca²⁺ measurements, which used only cells shown to be transfected by expression of RFP (see A). (F and G) Intracellular concentrations of choline (F) and IP₃ (G) during stimulation of NG108-15 cells with ATP (50 μM, bar) show the effects of shRNA for PLD1 and PLD2. Results show means ± SD; n = 6. (H) GPCRs that activate PLC and phospholipase D (PLD) initiate two parallel signaling pathways that converge at IP₃Rs. IP₃ from PLC directly activates IP₃R. Choline from PLD activates Sig-1R, which potentiates IP₃-evoked Ca²⁺ release.

Figure 4

CTL1-Mediated Choline Uptake Potentiates IP₃-Evoked Ca²⁺ Signals (A) NG108-15 cells were incubated in HBS alone or with 3 mM choline for the indicated times before adding bradykinin (1 μM) and immediately recording the increase in [Ca²⁺]_i. Results (mean ± SEM; n = 3 with duplicate determinations) show Δ[Ca²⁺]_i evoked by bradykinin. (B) Summary results (mean ± SEM; n = 3) show bradykinin-evoked Δ[Ca²⁺]_i after incubation with the indicated choline concentrations (105 min). (C) WB showing effects of the indicated siRNA (for CTL1) or shRNA (for Sig-1R) and their scrambled counterparts on expression of CTL1 and Sig-1R in NG108-15 cells. M_r markers (kDa) are shown. (D) Summary results (mean ± SD; n = 5) show CTL1 expression in cells treated with the indicated siRNA expressed as a percentage of the matched cells treated with scrambled siRNA. (E) Summary results (mean ± SEM; n = 5 plates with 2 replicates) show the effects of 10 mM choline on bradykinin-evoked Ca²⁺ signals. ^∗p < 0.05, ^∗∗p < 0.01, one-way ANOVA with Dunnett’s test, relative to control (B and E). (F) Ca²⁺-mobilizing GPCRs stimulate PLC and PLD, with consequent formation of IP₃ and choline. Although we have not resolved how GPCRs stimulate PLD in NG108-15 cells, signals evoked by both PLC and parallel pathways are known to stimulate PLD. IP₃ stimulates IP₃R, while choline binds to Sig-1Rs, causing them to potentiate IP₃R activity. Metabolism of IP₃ and choline terminates their signaling. Hence, GPCRs regulate IP₃Rs through two parallel, but converging, pathways. Import of extracellular choline by transporters, including the widely expressed CTL1, can also deliver choline to Sig-1Rs. (G) Acetylcholine (ACh) released at cholinergic terminals can activate post- and pre-synaptic receptors, before its rapid hydrolysis to choline by acetylcholinesterase (AChE). Hence, synaptic activity is rapidly followed by a substantial local increase in choline concentration. Transporters (red circles) in the cholinergic terminal (CHT1) and neighboring cells (CTL1-5 and OCT) can import the choline, which will then stimulate Sig-1Rs, providing cells with a paracrine reporter of recent synaptic activity. See also Figure S2.