Fig 1.
Deletion of RYR2 reshapes the [Ca2+]in transient and impairs insulin secretion in response to tolbutamide.
A) Tolbutamide (200 μM) stimulates a biphasic rise in [Ca2+]in in INS-1 cells. Preincubation with ryanodine (100 μM) selectively inhibits the initial peak in the [Ca2+]in response. B) RyR2KO cells stimulated with tolbutamide show selective loss of the rapid peak of [Ca2+]in compared to control INS-1 cells. Experiments in A and B were performed in a 96-well format, over two minutes. Tolbutamide (200 μM) was injected after a 15 sec baseline was recorded. Each point is the mean of triplicate values shown ± SE. C) Example traces from single cell [Ca2+]in imaging of a control INS-1 cell, an RyR2KO cell, and an IRBITKO cell over 5 minutes. Tolbutamide was applied at 60 seconds. D) Time to greatest peak analysis of INS-1, RyR2KO, and IRBITKO cells. The time to peak was longer for both RyR2KO and IRBITKO cells compared to INS-1 cell (**P < 0.01; ****P < 0.0001 One-way ANOVA with Tukey’s post-hoc test). The time to peak was longer in RyR2KO compared to IRBITKO cells (***P < 0.001) One-way ANOVA with Tukey’s multiple comparisons test. INS-1n = 24; RyR2KO: n = 28; IRBITKO: n = 39. E) Area under the curve analysis (AUC) of the [Ca2+]in response to tolbutamide. The AUC was significantly reduced in IRBITKO cells compared to RyR2KO cells (*, P < 0.05), and 1 μM xesto reduced AUC in RyR2KO and IRBITKO cells only (****, P < 00001; **, P < 0.01) One-way ANOVA with Tukey’s multiple comparisons test. INS-1: n = 25; INS-1 + xesto: n = 40; RyR2KO: n = 28; RyR2KO + xesto: n = 32; IRBITKO: n = 37; IRBITKO + xesto: n = 30. F) Deletion of RyR2 or IRBIT reduces tolbutamide-stimulated insulin secretion compared to INS-1 cells, and reduces insulin secretion stimulated by tolbutamide + ESCA (####P < 0.0001). Basal insulin secretion is reduced in RyR2KO cells compared to INS-1 cells (####, P < 0.0001). In all cases, tolbutamide stimulated an increase in insulin secretion over basal (2.5 mM glucose; ****, P < 0.0001) and tolbutamide + ESCA stimulated an increase in insulin secretion over tolbutamide alone (****, P < 0.0001). n = 9 for all conditions. Two-way ANOVA with Tukey’s multiple comparisons test.
Fig 2.
Deletion of RYR2 and IRBIT reduces stimulated and basal PLC activity.
A) Example traces from single-cell [Ca2+]in imaging of a control INS-1 cell, an RyR2KO cell, and an IRBITKO cell over 3 minutes. Bethanechol was applied at 55–60 seconds. B) Quantification of the increase in [Ca2+]in stimulated by the muscarinic agonist bethanechol (AUC). 50 μM bethanechol stimulates a greater increase in [Ca2+]in in IRBITKO and RyR2KO cells than in INS-1cells (****, P < 0.0001), and a greater increase in IRBITKO cells than in RyR2KO cells (****, P < 0.0001). Stimulation of [Ca2+]in by 1 μM or 5 μM bethanechol is not different between the three cell lines, but in each case, is different from that stimulated by 50 μM bethanechol (####, P < 0.0001). Atropine (100 μM) significantly reduced the [Ca2+]in response to 50 μM bethanechol in all cell lines (####, P < 0.0001) INS-1: 50 μM + Atro n = 175; 50 μM n = 143; 5 μM n = 153, 1 μM n = 132. Ry2RKO: 50 μM + Atro n = 164; 50 μM n = 293; 5 μM n = 159, 1 μM n = 111. IRBITKO: 50 μM + Atro n = 106; 50 μM n = 195; 5 μM n = 126, 1 μM n = 102. Outlier analysis was performed using the ROUT method, with Q = 1%. Outliers were removed before statistical analysis. C) IP1 assay of basal and stimulated PLC activity. Basal (2.5 mM glucose) and 7.5 mM glucose-stimulated PLC activity was reduced in RyR2KO and IRBITKO cells compared to INS-1 cells (***, P < 0.001; ****, P < 0.0001). Glucose-stimulated IP1 accumulation was greater in IRBITKO cells than in RyR2KO cells (##, P < 0;01). In contrast, carbachol (500 μM) stimulated IP1 accumulation was significantly reduced in RyR2KO cells compared to both control cells and IRBITKO cells (****, P < 0.0001). Data are shown as mean ±SD. INS-1: 2.5 Glc n = 16, 7.5 Glc n = 17, Carb n = 15; RyR2KO: 2.5 Glc n = 13, 7.5 Glc n = 8, Carb n = 14; IRBITKO: 2.5 Glc n = 11, 7.5 Glc n = 11, Carb n = 12. Data for INS-1 cells and RyR2KO cells were previously published in [11], and are shown for comparison with data from IRBITKO cells. D) IP1 accumulation stimulated by carbachol in INS-1, RyR2KO, and IRBITKO cells is inhibited by 100 μM atropine and 100 μM 2-APB (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). Data are shown as mean ± SD. Two-way ANOVA with Tukey’s multiple comparisons test. INS-1 cells: basal; n = 8; Carb; n = 8; Carb + Atr; n = 3, Carb + 2-ABP; n = 3. RyR2KO cells: basal, n = 8; Carb, n = 8; Carb + Atr, n = 4; Carb + 2-APB, n = 3. IRBITKO cells: basal, n = 8; Carb, n = 7; Carb + Atr, n = 3, Carb + 2-APB, n = 3.
Fig 3.
Modulation of PIP2 levels by PLC activity in INS-1 and RyR2KO cells.
A) Micrographs of fixed INS-1, RyR2KO, and IRBITKO cells stained with antibodies to PIP2 and IgG-k binding protein conjugated to CFL 488. Scale bars = 50 μm. B) Quantification of fluorescence intensity of PIP2 immunostaining in control INS-1 cells, RyR2KO cells, and IRBITKO cells. PIP2 staining was greater in RyR2KO cells and IRBITKO cells compared to control INS-1 cells (****, P < 0.0001). INS-1 cells: n = 1294; RyR2KO cells: n = 1377; IRBITKO cells n = 1650. Data are shown as mean ±SD. One-way ANOVA with Dunnett’s multiple comparisons test. Data for control INS-1 cells and RyR2KO cells were previously published in [11] and are shown for comparison with data from IRBITKO cells. C) Time lapsed TIRFm images of GFP-C1-PLCdelta-PH localization to the plasma membrane upon stimulation with 100 μM carbachol in a control INS-1 cell. Scale bar = 10 μm. D) Example traces showing decreases in plasma membrane GFP-C1-PLCdelta-PH fluorescence intensity in response to carbachol in living INS-1 and RyR2KO cells. E) Quantification (AUC) of the decrease in plasma membrane GFP-C1-PLCdelta-PH fluorescence in control INS-1 cells and RyR2KO cells in the presence or absence of 100 μM 2-APB. Carbachol stimulated a greater decrease in PH-PLCδ-GFP fluorescence intensity in control INS-1 cells than in RyR2KO cells (***, P < 0.001). 2-APB significantly inhibited this decrease in control INS-1 cells (****, P < 0.0001), but not in RyR2KO cells. INS-1 cells + Carb: n = 27; RyR2KO cells + Carb: n = 25; INS-1 cells + Carb + 2-APB: n = 18; RyR2KO + Carb + 2-APB: n = 13. One-way ANOVA with Tukey’s multiples comparisons test.
Fig 4.
SOCE is diminished in the absence of RYR2, but not IRBIT.
Representative experiments showing activation of SOCE in control INS-1 cells A), in RyR2KO cells B), and in IRBITKO cells C). ER Ca2+ stores were depleted with thapsigargin in the absence of extracellular Ca2+, and SOCE initiated by increasing extracellular Ca2+ to 2.5 mM. Each point is the mean of three replicates and is shown ± SE. D) Quantification of SOCE (AUC). The SOCE Ca2+ integral in significantly reduced in RyR2KO cells compared to either control INS-1 cells or IRBITKO cells (***, P < 0.001). 2-APB (100 μM) significantly reduced SOCE in control INS-1 cells and IRBITKO cells, but not in RyR2KO cells (****, P < 0.0001). Two-way ANOVA with Tukey’s multiple comparisons test. INS-1 cell: n = 6 separate experiments; INS-1 cells + 2-APB: n = 6 separate experiments; RyR2KO cells: n = 3 separate experiments; RyR2KO cells + 2-APB: n = 3 separate experiments. E) Immunoblot for STIM1 from cell lysates prepared from control INS-1, RyR2KO, and IRBITKO cells. Representative of three separate experiments. F) Quantification of STIM1 immunoblots from INS-1, RyR2KO, and IRBITKO cell lysates. The intensity of STIM1 bands was normalized to that of actin in each replicate (n = 3 for all cell lines). No significant difference in the STIM1/actin ratios among the three cell lines was detected (One-way ANOVA).
Fig 5.
Deletion of RyR2 or IRBIT increases Cav current density.
A) Example barium current ensembles, recorded from the indicated cell types, elicited by stepping to voltages ranging from -70 mV to +50 mV for 100 ms from a holding potential of -80 mV. B) Barium current-voltage relationship of complied data from the indicated cell types. Data are shown as mean ± SE. C) Voltage-dependence of activation (V1/2 act.) was not different between INS-1 (-9.3 ± 1.2 mV; n = 11) and RyR2KO (-7.5 ± 1.0 mV; n = 24), or IRBITKO cells (-13.7 ± 1.2 mV; n = 19) (One-way ANOVA). D) Peak barium current (IBa) density of RyR2KO and IRBITKO cells was greater than that of control INS-1 cells (*, P < 0.05; Welch’s unpaired t-test). INS-1 cells: n = 11; RyR2KO cells: n = 22; IRBITKO cells: n = 19. E) Peak calcium current (ICa) density was RyR2KO cells was great than that of control INS-1 cells (P < 0.05; Welch’s unpaired t-test). INS-1 cells: n = 12; RyR2KO cells: n = 16. F) The fraction of current blocked by 5 μM nifedipine wasn’t different between INS-1 and RyR2KO cells, or between INS-1 and IRBITKO cells (Welch’s unpaired t-test). INS-1 cells: n = 14; RyR2KO cells: n = 21; IRBITKO cells: n = 8.
Fig 6.
Increased plasma membrane PIP2 levels may contribute to the increased Cav channel activity in RyR2KO cells.
A) Whole-cell membrane capacitance of INS-1, RyR2KO, and IRBITKO cells measured in extracellular solution containing 10 mM BaCl2. (*, P < 0.05, **, P < 0.01, One-way ANOVA with Dunnette’s post-hoc test). INS-1: n = 13 cells; RyR2KO: n = 24 cells; IRBITKO: n = 19 cells. B) Whole-cell membrane capacitance of INS-1 and RyR2KO cells measured in extracellular solution with 10 mM CaCl2. (**, P < 0.01, Welch’s unpaired t-test). INS-1: n = 14 cells; RyR2KO: n = 17 cells. C) Cortical f-actin levels are reduced in RyR2KO cells compared to controls. Fluorescence intensity of phallodin-CF405 was detected by confocal microscopy, and quantified using line scans of stained cells. Scale bars = 10 μm. D) Comparison of phalloidin-CF405 fluorescence intensity in RyR2KO cells and INS-1 cells, normalized to INS-1 cells. Cortical CF405 fluorescence was significantly lower in RyR2KO cells compared to control INS-1 cells (****, P < 0.0001; un-paired t-test). INS-1 cells: n = 10; RyR2KO cells: n = 10. E) Diagram of the pseudojanin (PJ) lipid phosphatase system [34]. Rapamycin induces dimerization of FK506 binding protein 12 (FKBP), and fragment of mTOR that binds rapamycin (FRB). FRB is fused to the plasma membrane-localizing peptide lyn11, and dimerization drives FKBP-PJ fusion to the plasma membrane. PJ thus localized, rapidly degrades PIP2 in the plasma membranes. Cells expressing these constructs can be identified by detection of red fluorescent protein (RFP) emission. F) Rapamycin perfusion (1 μM) reduces current compared to baseline in RyR2KO cells expressing PJ but not in INS-1 cells expressing PJ, or in either cell line expressing the phosphatase-dead PJ construct PJ-D. (****, P < 0.0001; one sample t-test). INS-1 cells + PJ: n = 13; INS-1 cells +PJ-D: n = 8; RyR2KO cells + PJ: n = 14; RyR2KO cells + PJ-D: n = 4.
Fig 7.
Deletion of RyR2 inhibits activation of SK channels during glucose stimulated electrical activity.
A) Example trains of glucose-stimulated (18 mM) action potentials in INS-1 and RyR2KO cells. The SK channels blocker apamin (1 μM) was added at the time indicated by the arrows. B) Glucose-stimulated action potentials were more frequent in RyR2KO cells than in control INS-1 cells. Addition of apamin increased action potential frequency 2-fold in control INS-1 cells, but had no effect on action potential frequency in RyR2KO cells. (****, P < 0.0001; One-way ANOVA with Tukey’s multiple comparisons test) INS-1 control: n = 9 cells; RyR2KO control: n = 15 cells; INS-1 + apamin: n = 7 cells; RyR2KO + apamin: n = 10 cells C) Afterhyperpolarization (AHP) amplitude of glucose-stimulated action potentials is reduced in RyR2KO cells compared to control INS-1 cells. Apamin reduces AHP amplitude in control INS-1 cells, but not in RyR2KO cells. (****, P < 0.0001; One-way ANOVA with Tukey’s multiple comparisons test). Inset- overlay of action potentials from a control INS-1 cells (blue) and an RyR2KO cells (red) illustrates the difference AHP amplitude. INS-1 control: 9310 action potentials from 8 cells; RyR2KO control: 21,241 action potentials from 14 cells; INS-1 + apamin: 5371 action potentials from 6 cells; RyR2KO + apamin: 7018 action potentials from 6 cells. D) Model for RyR2 control of glucose-stimulated action potential frequency. In control cells, Ca2+ release from RyR2 induced by Ca2+ influx from voltage-gated Ca2+ channels actives SK channels during glucose stimulation. In the absence of RyR2, the enhanced Ca2+ influx (i.e. increased Cav channel current density and increased action potential frequency) isn’t able to activate SK channels. However, the reduced SOCE observed in RyR2KO cell could also account for the deficit in SK channel activation in these cells.
Fig 8.
Model for regulation of Ca2+ dynamics in pancreatic β-cells.
RyR2 is regulated by influx of Ca2+ via L-type Cav channels [3, 5], which can also directly stimulate rapid insulin release [67]. RyR2 regulates insulin secretion either directly via Ca2+ release from the ER or indirectly via regulation of insulin content and transcript level [11]. RyR2 plays a key role in regulating SOCE via a yet unknown mechanism. Maintenance of SOCE is critical for basal and stimulated PLC activity. Reduced PLC activity in the absence of RyR2 increases plasma membrane PIP2 levels, and increases Cav channel activity. RyR2 also directly or indirectly (via regulation of SOCE) regulates activation of SK channels. Finally, RyR2 regulates IP3R-mediated Ca2+ release, and perhaps many other cellular processes, via regulation of IRBIT levels [11].