Imaging of evoked dense-core-vesicle exocytosis in hippocampal neurons reveals long latencies and kiss-and-run fusion events

Abstract

Evoked neuropeptide secretion in the central nervous system occurs slowly, but the basis for slow release is not fully understood. Whereas exocytosis of single synaptic vesicles in neurons and of dense-core vesicles (DCVs) in endocrine cells have been directly visualized, single DCV exocytic events in neurons of the central nervous system have not been previously studied. We imaged DCV exocytosis in primary cultured hippocampal neurons using fluorescent propeptide cargo and total internal reflectance fluorescence microscopy. The majority of Ca(2+)-triggered exocytic events occurred from immobile plasma-membrane-proximal DCVs in the cell soma, whereas there were few events in the neurites. Strikingly, DCVs in the cell soma exhibited 50-fold greater release probabilities than those in neurites. Latencies to depolarization-evoked fusion for DCVs were surprisingly long, occurring with an average time constant (tau) of 16 seconds for DCVs in the soma and even longer for DCVs in neurites. All of the single DCV release events exhibited rapid fusion-pore openings and closures, the kinetics of which were highly dependent upon Ca(2+) levels. These ;kiss-and-run' events were associated with limited cargo secretion. Thus, the slow evoked release of neuropeptides could be attributed to very prolonged latencies from stimulation to fusion and transient fusion-pore openings that might limit cargo secretion.

Fig. 1.

ANF-EGFP marks a subgroup of DCVs in cultured hippocampal neurons. (A) Single ANF-EGFP-labeled DCVs visualized by TIRF microscopy. Left panel: vesicular localization of ANF-EGFP in both the soma and neurites of hippocampal neurons. Right panel: fluorescent beads 100 nm and 400 nm in diameter visualized under the same TIRF settings. (B) Immunocytochemical characterization of ANF-EGFP puncta. Row 1: all ANF-EGFP-containing vesicles were positive for the DCV marker SgII. Row 2: all ANF-EGFP-containing vesicles contained the Ca²⁺ sensor SytI. Row 3: ANF-EGFP-containing vesicles were distinct from monoaminergic VMAT2-positive vesicles. Row 4: most ANF-EGFP-containing vesicles also contained immunoreactive BDNF. Fluorescence line scans are shown in insets. These indicate that all ANF-EGFP-containing vesicles (green) also contained SgII and SytI (red). All images were visualized by TIRF microscopy. Scale bars: 1 μm.

Fig. 2.

Time course of the elevation of intracellular Ca²⁺ level by K⁺ stimulation revealed by Fluo-4 imaging. Cells were stimulated at 90 mM K⁺ (red) or 30 mM K⁺ (blue).

Fig. 3.

Exocytosis of ANF-EGFP-containing DCVs in neuronal soma and neurites. (A) Comparison of release probabilities for vesicles in soma and neurites. In the soma, about one third of the plasma-membrane-proximal vesicles underwent exocytosis in 100 seconds, whereas less than 1% in the neurites could release. A total of 1091 vesicles were analyzed in soma and 1344 vesicles in neurites (^***P<0.001). (B) Depolarization-evoked DCV events exhibit long latencies. Events in the soma and neurites were binned in 10-second intervals and plotted by percentage. Soma events were fitted with a single exponential curve (broken line) to calculate a time constant, τ (15.6 seconds). Neurite events did not fit a single exponential curve but were evident at long times. (C) Single exocytosis events in neuronal soma. (Ci) A representative example of DCV exocytosis in the soma. The time interval between two consecutive points was 0.1 seconds. The time between the fluorescence baseline and the fluorescence peak (fluorescence rise time) was designated as t_r. Release of content was calculated based on the fluorescence change (ΔF) between 2 seconds before and 15 seconds after exocytosis. (Cii) A DCV in the soma underwent two consecutive exocytic events. Shown below the curves are images of the corresponding DCV at a time interval of 0.2 seconds. Start points of the events are indicated by arrows. Scale bars: 500 nm. (D) A representative example of DCV exocytosis in the neurites. The time interval between two consecutive points was 0.2 seconds. Shown below the curve are images of the corresponding DCV at a time interval of 0.2 seconds. The start point of the exocytic event is indicated by an arrow. Scale bar: 500 nm. (E) Analysis of the content released during exocytosis for DCVs in soma and neurites. Content release was determined as the loss of fluorescence (ΔF) by subtraction of the value 2 seconds before an event from that obtained 15 seconds after the event. Gaussian fitting revealed two populations of events in the soma (solid lines) and one population in the neurites (broken line). (F) Comparison of the fluorescence rise time (t_r) of the events in soma and neurites (^***P<0.001). A total of 332 fusion events in soma and 48 events in neurites were analyzed in B, E and F.

Fig. 4.

Exocytosis of BDNF-EGFP-containing DCVs in neuronal soma and neurites. Exocytosis of BDNF-EGFP in the soma (A) and neurites (B) was imaged using methods similar to those used in Fig. 3. Exocytosis of BDNF-EGFP-containing DCVs was not accompanied by a fluorescence cloud or by a significant loss of fluorescence. Curves shown are averages of ten individual events with standard deviations indicated. Shown below the curves are images of representative DCVs at 1.0-second time intervals. Start points of the exocytic events, occurring at various times following depolarization, are indicated by arrows.

Fig. 5.

DCV exocytosis in neuronal soma under conditions of strong and weak depolarization. (A) Comparison of release probabilities at high [total number of analyzed vesicles (n)=1091] and low (n=1525; ^***P<0.001) K⁺ stimulation. (B) A representative example of DCV exocytosis in soma under low K⁺ (30 mM) stimulation. The time interval between two consecutive points is 0.1 seconds. (C) Comparison of the fluorescence rise time (t_r) of events at high- and low-K⁺ stimulation (^***P<0.001). (D) Analysis of the content released during exocytosis for DCVs in soma at high (90 mM) and low (30 mM) K⁺ stimulation. The fluorescence loss (ΔF) from a vesicle was calculated from values 2 seconds before and 15 seconds after an event. Gaussian fitting revealed two populations under both stimulation conditions. At low-K⁺ stimulation (broken lines), a larger subset of DCVs exhibited extensive content release compared with DCVs stimulated at high K⁺ (solid lines). A total of 332 fusion events under 90 mM K⁺ stimulation and 90 events under 30 mM K⁺ stimulation were analyzed in C and D.

Fig. 6.

L-type Ca²⁺ channels mediate DCV exocytosis in hippocampal neurons. Release probabilities in the neuronal soma in response to 90 mM K⁺ were determined after treatment with the indicated Ca²⁺-channel inhibitors. Verapamil at 100 μM (specific for L-type Ca²⁺ channels) completely blocked evoked DCV exocytosis [total number of analyzed vesicles (n)=803], whereas 10 μM ω-conotoxin MVIIC (N- and Q-type specific, n=734) and 200 nM ω-agatoxin TK (P-type specific, n=712) were without effect (P>0.5 by t-test).