IP3 Receptors Preferentially Associate with ER-Lysosome Contact Sites and Selectively Deliver Ca2+ to Lysosomes

Abstract

Inositol 1,4,5-trisphosphate (IP₃) receptors (IP₃Rs) allow extracellular stimuli to redistribute Ca²⁺ from the ER to cytosol or other organelles. We show, using small interfering RNA (siRNA) and vacuolar H⁺-ATPase (V-ATPase) inhibitors, that lysosomes sequester Ca²⁺ released by all IP₃R subtypes, but not Ca²⁺ entering cells through store-operated Ca²⁺ entry (SOCE). A low-affinity Ca²⁺ sensor targeted to lysosomal membranes reports large, local increases in cytosolic [Ca²⁺] during IP₃-evoked Ca²⁺ release, but not during SOCE. Most lysosomes associate with endoplasmic reticulum (ER) and dwell at regions populated by IP₃R clusters, but IP₃Rs do not assemble ER-lysosome contacts. Increasing lysosomal pH does not immediately prevent Ca²⁺ uptake, but it causes lysosomes to slowly redistribute and enlarge, reduces their association with IP₃Rs, and disrupts Ca²⁺ exchange with ER. In a "piston-like" fashion, ER concentrates cytosolic Ca²⁺ and delivers it, through large-conductance IP₃Rs, to a low-affinity lysosomal uptake system. The involvement of IP₃Rs allows extracellular stimuli to regulate Ca²⁺ exchange between the ER and lysosomes.

Keywords: Ca(2+); IP(3) receptor; concanamycin A; endoplasmic reticulum; genetically encoded Ca(2+) sensor; lysosome; membrane contact site; proximity ligation assay; store-operated Ca(2+) entry.

Graphical abstract

Figure 1

Inhibition of Lysosomal V-ATPase Potentiates Cytosolic Ca²⁺ Signals Evoked by IP₃Rs (A) Bright-field and wide-field fluorescence images of HEK cells loaded with LysoTracker Red (100 nM, 10 min) with or without CcA (1 μM, 1 hr). Images are typical of three experiments. (B) Fluo 8-loaded HEK cells were treated with CcA (1 μM, 1 hr) in HBS before addition of 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA) (2.5 mM) to chelate extracellular Ca²⁺ and then CCh (1 mM) to stimulate IP₃ formation. Typical traces show mean ± SD from three wells in one experiment. (C and D) Summary results show effects of CcA on basal [Ca²⁺]_c (C) and peak increase in [Ca²⁺]_c (Δ[Ca²⁺]_c) evoked by CCh (D). Results show paired individual values (each from three determinations) and the mean (n = 7, line). ^∗p < 0.05, paired Student’s t test. (E) Expression of mRNA for ATP6V0C relative to GAPDH in cells treated with non-silencing siRNA (NS) or siRNA for ATP6V0C. Mean ± SEM, n = 6. ^∗p < 0.05, paired Student’s t test. (F) TIRFM images show effects of siRNAs in HEK cells expressing TPC2-GFP or stained with LysoTracker Red (100 nM, 10 min). Images are typical of three experiments. (G) Effects of siRNA on basal [Ca²⁺]_c (n = 6, each with three determinations). (H) Effects of siRNA on Δ[Ca²⁺]_c evoked by CCh alone or after CcA (1 μM, 1 hr) (mean ± SEM, n = 5, each with three determinations). ^∗p < 0.05, two-way ANOVA with Bonferroni test. (I) Results and Figure S1 demonstrate that lysosomes (LY) selectively sequester Ca²⁺ released from ER through IP₃Rs, but not Ca²⁺ entering the cell through SOCE. See also Figure S1.

Figure 2

IP₃Rs Selectively Deliver Ca²⁺ to Some Lysosomes (A) Wide-field fluorescence images of HeLa cells expressing TPC2-RFP with Ly-GG or Cy-GG. (B and C) Recordings from HeLa cells expressing Cy-GG (B) or Ly-GG (C) showing responses to histamine (100 μM) in Ca²⁺-free HBS and then ionomycin (10 μM) with 2 mM CaCl₂ (to saturate the sensor). Results (F/F_max, where F_max is response after ionomycin) show responses of a single tracked lysosome (C) or a similarly sized cytosolic region of interest (ROI) (B) (see Video S1). (D and E) Summary results (mean ± SEM) show basal fluorescence (D) (3 experiments with 198 ROIs for Cy-GG and 83 tracks for Ly-GG) and peak fluorescence signals evoked by histamine (E) for Cy-GG (3 experiments with 198 ROIs) and Ly-GG (4 experiments with 83 lysosome tracks). ^∗p < 0.05, Student’s t test. (F and G) Distribution of peak F/F_max values for Cy-GG (F) and Ly-GG (G) in cells stimulated with histamine. Distribution of Ly-GG fluorescence values is significantly different from a normal distribution (Kolmogorov-Smirnov normality test, p = 0.0018), whereas Cy-GG fluorescence is consistent with a normal distribution (p > 0.1). (H) Peak responses from tracked regions for Cy-GG and Ly-GG for cells in which SOCE was evoked by restoration of extracellular Ca²⁺ (2 mM) to cells treated with thapsigargin (1 μM, 15 min) in Ca²⁺-free HBS. Mean ± SEM from at least three experiments (30 tracks for Ly-GG and 45 ROIs for Cy-GG). (I and J) Distribution of peak F/F_max values for Cy-GG (I) and Ly-GG (J) after photolysis of ci-IP₃ in Ca²⁺-free HBS (n = 59 ROIs from 4 dishes for Cy-GG and 49 tracks from 3 dishes for Ly-GG). (K) Summary results (mean ± SEM). ^∗p < 0.05, Student’s t test. (L) Effects of U73122 (10 μM, 20 min) on peak Ly-GG signals evoked by histamine (100 μM) or photolysis of ci-IP₃. Mean ± SEM, n = 3–4. ^∗p < 0.05, Student’s t test, relative to control. See also Figure S1 and Video S1.

Figure 3

Lysosomes Sequester Ca²⁺ Released by All IP₃R Subtypes (A) Basal [Ca²⁺]_c in HEK cells expressing only the indicated IP₃R subtypes and treated with bafilomycin A₁ (Baf A₁, 1 μM, 1 hr). Mean ± SEM. n = 3, each with three determinations. (B–G) Effects of Baf A₁ (1 μM, 1 hr) on Ca²⁺ release evoked by CCh in HEK cells expressing only the indicated IP₃R subtypes. Mean ± SEM, n = 6. The code in (B) applies also to (C)–(G). Similar results from WT cells are shown in Figure S1B. (H) Summary results show the potentiating effect of Baf A₁ on the peak CCh-evoked Ca²⁺ signal. Results (mean ± SD) show the increase in amplitude of Ca²⁺ signal in the presence of Baf A₁ as a percentage of the control response. WT, wild-type. See also Figure S1.

Figure 4

IP₃Rs Associate with ER-Lysosome Contacts (A) TIRFM images of HeLa cell expressing mTurquoise-ER and LAMP1-mCherry showing that most lysosomes associate with ER and maintain contact as they move (yellow arrows). Images are shown at 8.4-s intervals (see Video S2). (B) TIRFM images of EGFP-IP₃R1-HeLa cell expressing markers for lysosomes (mTurquoise-LAMP1) and ER (mCherry-ER). Merged image shows overlay of EGFP-IP₃R1 and lysosomes. (C) Time-lapse TIRFM images (10-s intervals from Video S4) of boxed region in (B) show distribution of EGFP-IP₃R1, lysosomes, and ER. Arrows show examples of mobile lysosomes (yellow) and those that remain immobile for sustained periods (white). Immobile lysosomes coincide with IP₃R puncta. Mobile IP₃R puncta are not visible in these images because of the long capture intervals (3.3 s). (D and E) Kymograms (3.3-s intervals) of boxed regions in (C) show a lysosome that is stationary for a prolonged period adjacent to an EGFP-IP₃R1 punctum (D, from white box in C), whereas another lysosome pauses only briefly at adjacent ER (E, from yellow box in C). See also Figures S2, S3, and S6, and Videos S2, S3, S4, S5, S6, S7, and S8.

Figure 5

PLA Analyses Show IP₃Rs at ER-Lysosome MCSs (A) PLA uses primary antibodies that recognize proteins in ER or lysosome (LY) membranes. Complementary oligonucleotides conjugated to secondary antibodies hybridize only if they are close to each other. Ligation of the hybridized strands then allows rolling circle amplification (RCA) of the oligonucleotide and incorporation of the red fluorescent probe. (B) Images of HEK cells with (WT) and without (KO) IP₃Rs from PLA analyses of VAMP-associated protein A (VAP-A) proximity to Rab7. Confocal maximum intensity Z-projections show PLA spots (red) and nuclei (gray). Effects of CcA (1 μM, 1 hr) and omission of either primary antibody (Ab) are shown. (C) Summary results (mean ± SD, n = 15–25 cells from three experiments [two experiments for single-antibody controls]) show number of PLA spots/cell. ^∗p < 0.05, one-way ANOVA with Dunnett’s test, relative to WT control. (D) Confocal section of PLA in EGFP-IP₃R1 HeLa cells shows VAP-A proximity to LAMP1 (red) and endogenously tagged IP₃R1 (green). Boxed areas are shown enlarged to illustrate the coincidence of EGFP-IP₃R puncta with PLA spots (ER-lysosome MCS) (Manders’s split coefficient, 0.70 ± 0.21, n = 18 cells). (E) Confocal maximum intensity Z-projections of PLA in EGFP-IP₃R1 HeLa cells using antibodies to GFP and LAMP1 to show their proximity (red spots). Nucleus is shown in gray. Effects of CcA (1 μM, 1 hr) and of performing the same PLA in wild-type (WT) HeLa cells without EGFP-IP₃R1 are also shown. (F) Summary PLA results using Rab7 (left panel) or LAMP1 (right) to identify lysosomes and either GFP (from EGFP-IP₃R1) or VAP-A to identify ER (shown by bars above the histograms). Mean ± SD, n = 25–97 cells from 3–5 experiments. ^∗∗∗p < 0.001, Student’s t test for CcA-treated relative to matched control EGFP-IP₃R1-HeLa cell. (G) Most lysosomes (LY) are closely associated, aided by tethers, with ER at MCS. Small clusters of IP₃Rs associate with these MCS, but IP₃Rs are not required for their assembly. WT, wild-type. See also Figures S2 and S6 and Videos S2, S3, S4, S5, S6, S7, and S8.

Figure 6

IP₃Rs Are Not Required for ER-Lysosome Contacts (A) TIRFM images of HEK cells with (WT) and without IP₃Rs (KO), expressing LAMP1-mCherry and EGFP-ER. (B) Enlargements of boxed region in (A) show associations of lysosomes and ER. (C) Time series (3-s intervals) of boxed regions in (B) show dynamics of ER-lysosome interactions (from Videos S9 and S10). (D) HEK cells were treated with bafilomycin A₁ (Baf A₁, 1 μM, 1 hr) in HBS before addition of BAPTA (2.5 mM) and then thapsigargin (1 μM). Mean ± SD from three wells in one experiment. (E and F) Summary results show peak thapsigargin-evoked Ca²⁺ release in WT (E) and KO cells (F) as paired values (each with three determinations) and mean (n = 4, line). ^∗p < 0.05, paired Student’s t test. See also Figures S5 and S6 and Videos S9 and S10.

Figure 7

Increasing Lysosomal pH Slowly Redistributes Lysosomes and Attenuates Ca²⁺ Handling (A) Lysosomes of HEK cells were loaded with fluorescein-dextran (pK_a = 6.4). Wide-field images show fluorescence recorded at pH-sensitive (λ_ex = 488 nm) and -insensitive (λ_ex = 425 nm) wavelengths before (0 min) and after CcA (1 μM, 1 hr). Images are typical of three experiments. (B) Summary results (mean ± SEM, n = 3) show time course of lysosomal pH changes after CcA as fluorescence ratios (R = F₄₈₈/F₄₂₅), which increase as pH increases. R₀ is R determined before CcA. (C) Effects of CcA (1 μM) using LysoTracker red (50 nM, 1 hr) fluorescence, which declines as lysosomal pH increases. Mean ± SEM, n = 4–10. ^∗p < 0.05, one-way ANOVA with Dunnett’s post hoc test, relative to t = 0. (D) Parallel analysis of CcA (1 μM) effects on peak increase in [Ca²⁺]_c evoked by CCh (1 mM) in Ca²⁺-free HBS. Results (mean ± SEM, n = 4–9, with three determinations) show increase in peak Ca²⁺ signal in the presence of CcA relative to that in its absence, as percentage of response evoked by CCh alone. (E) Confocal z stack of HEK cells expressing LAMP1-mCherry before and after CcA (1 μM, 1 hr). Nuclei stained with NucBlue. The appearance of larger lysosomes in the cell periphery (white arrows) following CcA (1 μM, 1 hr) was clearly observed in four of six cells. (F) Effect of CcA (1 μM, 1 hr) on distribution of lysosome sizes (reported as Feret diameter, see STAR Methods). Results are from 721 (control) and 617 lysosomes (CcA-treated) from 4 cells in 4 independent experiments. Inset shows enlargement of the largest size category. ^∗p < 0.05, Student’s t test. A similar analysis using lysosomes identified by an endocytosed fluorophore is shown in Figure S7. (G and H) Effects of CcA (1 μM, 1 hr) on the distance between each lysosome and the nearest IP₃R punctum. Because these distances are reported as centroid-centroid separations, they can be smaller than the diffraction limit of the microscope. Results from four cells, with ∼20 × 20 μm analyzed in each, show each measurement and mean (line) (G) and frequency distributions (H). ^∗∗∗p < 0.001, two-tailed Student’s t test. (I) Delivery of Ca²⁺ through IP₃Rs or unidentified “leak channels” into MCSs provides a low-affinity Ca²⁺ uptake system in lysosomes with the high local [Ca²⁺] required for its activity. The ER, with its high-affinity Ca²⁺ pump (SERCA), accumulates Ca²⁺ from the cytosol and delivers it at high local concentration to the surface of lysosomes through large-conductance IP₃Rs. ER behaves as an ATP-powered piston to concentrate Ca²⁺ around the lysosomal uptake system. (J) Dissipating lysosomal pH gradient does not immediately prevent lysosomal Ca²⁺ uptake, but slowly disrupts junctions within which it occurs. See also Figure S7.