Multivesicular exocytosis in rat pancreatic beta cells

Abstract

Aims/hypothesis: To establish the occurrence, modulation and functional significance of compound exocytosis in insulin-secreting beta cells.

Methods: Exocytosis was monitored in rat beta cells by electrophysiological, biochemical and optical methods. The functional assays were complemented by three-dimensional reconstruction of confocal imaging, transmission and block face scanning electron microscopy to obtain ultrastructural evidence of compound exocytosis.

Results: Compound exocytosis contributed marginally (<5% of events) to exocytosis elicited by glucose/membrane depolarisation alone. However, in beta cells stimulated by a combination of glucose and the muscarinic agonist carbachol, 15-20% of the release events were due to multivesicular exocytosis, but the frequency of exocytosis was not affected. The optical measurements suggest that carbachol should stimulate insulin secretion by ∼40%, similar to the observed enhancement of glucose-induced insulin secretion. The effects of carbachol were mimicked by elevating [Ca(2+)](i) from 0.2 to 2 μmol/l Ca(2+). Two-photon sulforhodamine imaging revealed exocytotic events about fivefold larger than single vesicles and that these structures, once formed, could persist for tens of seconds. Cells exposed to carbachol for 30 s contained long (1-2 μm) serpentine-like membrane structures adjacent to the plasma membrane. Three-dimensional electron microscopy confirmed the existence of fused multigranular aggregates within the beta cell, the frequency of which increased about fourfold in response to stimulation with carbachol.

Conclusions/interpretation: Although contributing marginally to glucose-induced insulin secretion, compound exocytosis becomes quantitatively significant under conditions associated with global elevation of cytoplasmic calcium. These findings suggest that compound exocytosis is a major contributor to the augmentation of glucose-induced insulin secretion by muscarinic receptor activation.

Fig. 1

Small and large release events detected in single rat beta cells expressing P2X₂Rs. a,b ATP-induced P2X₂R current transients from rat beta cells infused with 0.2 (a) or 2 (b) μmol/l [Ca²⁺]_i. Three sweeps have been superimposed in each panel. c,d Histograms of charge (Q) of ATP-induced P2X₂R currents recorded at 0.2 (c) and 2 (d) μmol/l [Ca²⁺]_i. Insets in c,d show events with Q-values <10 pC. e Cumulative distribution (N/∑N) of the integrated current (Q) of P2X₂R current spikes measured from rat beta cells infused with 0.2 μmol/l (dashed line; 696 events) and 2 μmol/l (continuous line; 800 events) [Ca²⁺]_i. Note inflection of curve obtained with the higher [Ca²⁺]_i at ∼100 pC representing the subset of large events (arrow). f,g Relationship between the integrated P2X₂R-mediated current (Q) and duration of the release events measured as t _10–90% (f) and halfwidths (g) of the current spikes recorded at 0.2 μmol/l (black circles) and 2 μmol/l (white circles) [Ca²⁺]_i

Fig. 2

Large P2X₂R events correlate with large increases in membrane capacitance in single rat beta cells. a,b Representative examples of parallel measurements of P2X₂R currents (I _m) and membrane capacitance (ΔC _m) for small (a) and large (b) events. Note that the currents rise and fall monotonically. Activation of P2X₂Rs results in large increases in membrane conductance that ‘leaks’ into the capacitance signal, and these parts of the experimental record have been removed. c As in a,b, but showing an example where a step increase in capacitance was associated with a series of discrete superimposed current spikes (arrows) reflecting exocytosis of individual granules. d Relationship between the integrated P2X₂R-mediated current (Q) and the associated capacitance increase (ΔC _m) in rat beta cells. The regression coefficient (r) was 0.79 (p < 0.001). e Relationship between size of capacitance steps (ΔC _m) and time to reach the new plateau value recorded in non-transfected cells but conditions otherwise as in a–d. Inset shows an example of an exocytotic event with an amplitude of 18 fF (simultaneous fusion of five or six granules). Note that this analysis only includes large events and that events <4 fF (reflecting single exocytotic events) cannot be resolved during whole-cell recordings. For display purposes, individual data points have been grouped according to ΔC _m in d and e (n = 48 and 47, respectively)

Fig. 3

Carbachol-induced compound exocytosis in single rat beta cells. a Fura-2 measurements of [Ca²⁺]_i in an intact rat islet stimulated with 20 mmol/l glucose and 20 μmol/l carbachol (CCh) as indicated by horizontal lines. b,c Fluo-5 recording of [Ca²⁺]_i (F [in arbitrary units, a.u.]) in a rat beta cell exposed to 20 mmol/l glucose before and after addition of 20 μmol/l carbachol as indicated (b) and confocal (‘2.5D’) image of the same isolated rat beta cell showing the fluorescence increase (ΔF) evoked by application of 20 μmol/l carbachol (c) obtained by subtracting image immediately before addition of carbachol from that obtained at the peak (representative of five different cells). The dotted line in c indicates the circumference of the beta cell. d ATP release in rat beta cells (measured as activation of P2X₂Rs) under control conditions (10 mmol/l glucose alone) when cell was held at −70 mV and after moderate depolarisation to −40 mV. Note stimulation of exocytosis seen as emergence of current transients. e,f ATP release measured at −40 mV in rat beta cells exposed to 10 mmol/l glucose in the absence (e) or presence (f) of 20 μmol/l carbachol. Several sweeps have been superimposed. g Cumulative distribution (N/ΣN) of charge (Q) recorded in the presence of 10 mmol/l glucose alone (dashed line; 836 events) and in the presence of 20 μmol/l carbachol (continuous line; 1,234 events). h As in g but after pretreatment of the cells with 20 μmol/l of the myristoylated CAMKII autoinhibitory peptide for 2 h (glucose: 448 events; glucose + carbachol: 603 events)

Fig. 4

Optical imaging of exocytosis in intact rat islets. a Two-photon excitation microscopy of an intact pancreatic islet perfused with an extracellular solution containing the fluorescent polar tracer SRB (0.7 mmol/l) and 20 mmol/l glucose. Scale bar: 3 μm. The three arrows indicate exocytotic events. b,c Example of a typical single event (b) and a large event (c) in the presence of 20 mmol/l glucose and 20 μmol/l carbachol. Scale bar: 1 μm. Images were taken at the indicated times (relative to the leftmost image). Note that the large event in c appears within the cell and is connected to the exterior by a stalk (dim fluorescence, arrow). d Histogram of diameters (D) of the fluorescent spots reflecting exocytosis in the presence of glucose. The inset shows the distribution on an expanded abscissa. A Gaussian curve with a mean value of 0.23 μm has been superimposed (102 events, 8 islets). e As in d but showing distribution after addition of 20 μmol/l carbachol (100 events, 8 islets). f The same data plotted as the cumulative distribution (N/∑N) of the diameters of exocytotic events evoked by 20 mmol/l glucose alone (dashed line) and 20 mmol/l glucose + 20 μmol/l carbachol (continuous line). Note inflection at ∼0.5 μm on distribution of events collected in the presence of carbachol (arrow). g,h Pie charts summarising relative frequency of single-vesicle (white), sequential (black) and multivesicular (hatched) exocytosis in beta cells within intact rat islets exposed to 20 mmol/l glucose with (h) or without (g) 20 μmol/l carbachol. The data were collected in 13 different islets and based on a total observation period of >3,100 s for each condition. i Insulin secretion measured in the presence of 1 or 20 mmol/glucose during a 1 min incubation. Carbachol (20 μmol/l; Carb.) was included as indicated. *p < 0.005 vs 20 mmol/l glucose alone; †p < 0.001 vs 1 mmol/l glucose

Fig. 5

Kinetics of exocytosis in beta cells within intact rat pancreatic islets. a,b Examples of fluorescent intensity changes associated with events recorded in the presence of 20 mmol/l glucose alone (a) and in the simultaneous presence of glucose (20 mmol/l) and carbachol (20 μmol/l) (b) as indicated. c Examples of changes in fluorescence associated with the larger events observed in the presence of carbachol. Traces in a–c are shown using the same timebase and ordinate scales. d Relationship between diameter (D) of exocytotic events and rise time (t _10–90%) recorded in the presence of 20 mmol/l glucose alone (white circles) and after supplementation of medium with 20 μmol/l carbachol (black squares). For clarity, the individual data points have been grouped according to size. e Cumulative distribution (N/ΣN) of the rise times (t _10–90%) of the fluorescence increases associated with exocytosis observed in the presence of 20 mmol/l glucose alone (71 events in 7 islets; dashed line) and 20 mmol/l glucose plus 20 μmol/l carbachol (76 events in six islets; continuous line). f Montage of images during an example of sequential exocytosis in a rat beta cell exposed to 20 mmol/l glucose and 20 μmol/l carbachol. Images were taken at the indicated times (relative to the one in the upper left corner). Scale bar: 1 μm. The three circles in the lower right corner indicate the position of the 1st, 2nd and 3rd release events (expanded ×2.5) shown in the montage. Note that they occur in close proximity to each other but not in exactly the same place. g Time course of SRB fluorescence increase of events 1–3 shown in f. a.u., arbitrary units

Fig. 6

Multivesicular exocytosis detected by confocal imaging in isolated rat beta cells. a 3-D reconstruction of exocytotic events observed in rat beta cells captured during 30 s exposures to FM1-43FX in the presence of 20 mmol/l glucose alone. b Histogram of diameters/lengths of FM1-43FX-labelled structures in cells stimulated with 20 mmol/l glucose alone. A Gaussian curve with a peak at 0.60 μm has been superimposed on distribution. c Cumulative distribution of structures labelled by FM1-43FX in the presence of glucose alone (dashed line; n = 140 in 70 cells) and when combined with carbachol (continuous line; n = 176 in 40 cells). d As in a but showing an example of a large event observed in the simultaneous presence of 20 mmol/l glucose and 20 μmol/l carbachol. e As in b but showing histogram of diameters/lengths of FM1-43FX-labelled structures in cells stimulated with the combination of 20 mmol/l glucose and 2 μmol carbachol

Fig. 7

Ultrastructural evidence for multigranular structures in beta cells captured by 3-D and two-dimensional electron microscopy in the presence of glucose or glucose plus the cholinergic agonist carbachol in intact islets. a Schematic representation of a 3-D block of a serially sectioned rat pancreatic islet. b Serial sections of a single beta cell exposed to the cocktail of glucose (20 mmol/l) and carbachol (20 μmol/l; 5 min). Images illustrate a group of homotypically fused granules with connected cores. Note that the captured multivesicular complex is approximately one granule diameter (blue double-headed arrow) from the plasma membrane (PM, red dotted line). Section thickness: 50 nm. Scale bar: 0.4 μm. c The percentage of homotypically fused granules relative to the total number of granules observed within the cytosol of four complete beta cells using the 3-D electron microscopic images in 20 mmol/l glucose alone (n = 3,338 granules) and when medium was supplemented with 20 μmol/l carbachol (n = 4,580 granules); **p < 0.01. d,e Transmission electron microscopy (two-dimensional) examples of two individual multigranular structures in a rat beta cell exposed to 20 mmol/l glucose in the presence (d) and absence (e) of 20 μmol/l carbachol. Note that in e the multivesicular structure aligns below the plasma membrane (cf. Fig. 6d). The frequency of multivesicular structures was 0.18% (seven compound and 2,712 single events) and 0.37% (16 compound and 4,094 single events) in the absence and presence of carbachol, respectively. Data were from ≥25 cells for both experimental conditions. The plasma membrane (PM) is indicated by the dotted red lines and arrow heads. Scale bars: 0.5 μm in d and 0.2 μm in e. A few mitochondria (m) are indicated