Catecholamine exocytosis during low frequency stimulation in mouse adrenal chromaffin cells is primarily asynchronous and controlled by the novel mechanism of Ca²⁺ syntilla suppression

Abstract

Adrenal chromaffin cells (ACCs), stimulated by the splanchnic nerve, generate action potentials (APs) at a frequency near 0.5 Hz in the resting physiological state, at times described as ‘rest and digest’. How such low frequency stimulation in turn elicits sufficient catecholamine exocytosis to set basal sympathetic tone is not readily explained by the classical mechanism of stimulus–secretion coupling, where exocytosis is synchronized to AP-induced Ca²⁺ influx. By using simulated action potentials (sAPs) at 0.5 Hz in isolated patch-clamped mouse ACCs, we show here that less than 10% of all catecholaminergic exocytosis, measured by carbon fibre amperometry, is synchronized to an AP. The asynchronous phase, the dominant phase, of exocytosis does not require Ca²⁺ influx. Furthermore, increased asynchronous exocytosis is accompanied by an AP-dependent decrease in frequency of Ca²⁺ syntillas (i.e. transient, focal Ca²⁺ release from internal stores) and is ryanodine sensitive. We propose a mechanism of disinhibition, wherein APs suppress Ca²⁺ syntillas, which themselves inhibit exocytosis as they do in the case of spontaneous catecholaminergic exocytosis.

Introduction

In the well-established, classical mechanism of stimulus–secretion coupling, secretion is initiated by depolarization, caused physiologically by action potentials (APs) in many cell types. In neurons and neuroendocrine cells this depolarization induces the opening of plasmalemmal voltage-dependent Ca²⁺ channels (VDCCs), which generate nano- or microdomains of relatively high intracellular calcium concentration ([Ca²⁺]_i) in the vicinity of docked, primed vesicles (Neher & Sakaba, 2008). Due to the rapid rise and fall of [Ca²⁺]_i within these domains, the exocytic machinery is quickly and transiently activated, causing fusion of vesicles with the plasma membrane to be highly synchronized with the AP (Chow 1994; Voets et al. 1999). This classical mechanism readily accounts for synchronous exocytosis.

Yet it is known that in many cases APs elicit neurotransmitter or hormone release in two phases: a short burst of synchronous exocytosis followed by a sustained asynchronous one (Goda & Stevens, 1994; Zhou & Misler, 1995). Previously the focus has been on synchronous exocytosis, but the importance of the asynchronous phase is becoming more evident (Glitsch, 2008). Our current understanding of asynchronous exocytosis presents us with an uncertain picture, consisting of a wide array of mechanisms, based largely on Ca²⁺ influx from an external source with vesicle proteins as the target (Smith et al. 2012; Chung & Raingo, 2013). In the face of this uncertainty, it is worthwhile to consider whether there are unrecognized asynchronous mechanisms of exocytosis linked to stimulation. We hasten to make clear that this report does not call into question the long-standing and meticulously documented classical mechanisms of synchronized transmitter release based on Ca²⁺ influx through VDCCs. However, here we present evidence that another, additional mechanism is involved in the case of asynchronous exocytosis at low frequency (0.5 Hz) but nevertheless physiological stimulation.

The mechanism we present for asynchronous exocytosis results from a series of studies on the role of ryanodine-sensitive internal Ca²⁺ stores which we have carried out in recent years and on which we build further here. They involve the study of both neuroendocrine terminals and chromaffin cells. These began with work on hypophyseal terminals of hypothalamic neurons (De Crescenzo et al. 2004), where we found quantal, focal Ca²⁺ release events via ryanodine receptors (RyRs) from intracellular Ca²⁺ stores which were similar to Ca²⁺ sparks in muscle cells (Cheng et al. 1993). We designated these as Ca²⁺ syntillas (scintilla, Latin for ‘spark’ from a nerve terminal, generally a SYNaptic structure) (Fig.1B). We demonstrated in mice, using a knock-in mutation, that the type 1 ryanodine receptor (RyR1) was involved in the regulation of syntillas in these nerve terminals (De Crescenzo et al. 2012).

Figure 1. Detection of catecholamine exocytosis and two sources of cytosolic Ca²⁺ in mouse ACCs.

A, representative sAP and the elicited Na⁺ current (I_Na) and Ca²⁺ current (I_Ca) in a freshly isolated mouse chromaffin cell at a holding potential of −80 mV. sAPs were composed of a three step ramp as follows (start potential (mV), end potential (mV), duration (ms)): −80, 50, 2.5; 50, −90, 2.5; −90, −80, 2.5. B, representative Ca²⁺ syntilla arising from ryanodine-sensitive intracellular stores imaged at 50 Hz with Fluo-3 Ca²⁺ indicator dye from a freshly isolated mouse ACC and rendered on a pseudo-colour scale as change in fluorescence over baseline (ΔF/F₀). Scale bar, 1 μm. The image of the entire ACC was fitted with a black mask for background contrast. C, representative amperometric records of catecholamine release from individual vesicles with and without stimulation by sAPs at 0.5 Hz from the same ACC. (Small hash marks occurring regularly at 0.5 Hz on amperometric traces during stimulation are artifacts indicating the onset of an sAP.) D, individual amperometric event types magnified. SAFs at left indicate ‘kiss and run’ exocytosis, while spikes (middle) can represent full fusion or ‘kiss and run’. Some spikes are preceded by a foot (right). An artifact is shown in the current trace of the spike on the right, which indicates the onset time of an sAP.

We also found similar events in mouse adrenal chromaffin cells (ACCs) (ZhuGe et al. 2006) due in this case to the opening of type 2 ryanodine receptors (RyR2s), and again we designated them syntillas as the ACCs are neurosecretory cells. In the ACCs type 2 RyRs are the dominant type with relatively few type 3, which are perinuclear, and essentially no type 1, as was shown both with analysis of mRNAs and with specific antibodies to the RyRs. In both preparations, nerve terminals and ACCs, Ca²⁺ syntillas are readily recorded in the absence and presence of extracellular Ca²⁺ and do not depend on Ca²⁺ influx through VDCCs. Furthermore, the syntillas do not directly trigger exocytosis in either preparation, as demonstrated by simultaneous recording of amperometric events and Ca²⁺ syntillas at the same location (ZhuGe et al. 2006; McNally et al. 2009).

As exocytosis of catecholamines can be studied with great temporal precision at the level of individual exocytotic vesicles using amperometry of catecholamines (i.e. without use of false transmitter), we studied the effects of syntillas on exocytosis in freshly isolated mouse ACCs of the type used herein. We found that in these cells there is spontaneous exocytosis –in both the presence (Lefkowitz et al. 2009) and the absence (ZhuGe et al. 2006) of extracellular Ca²⁺. Strikingly we found that this spontaneous exocytosis was increased when syntillas were blocked. This block could be effected by inhibiting syntillas in either of two ways. First, ryanodine at blocking concentrations (100 μm; Xu et al. 1998) blocked syntillas, as was directly confirmed with high resolution imaging (ZhuGe et al. 2006; Lefkowitz et al. 2009), and increased exocytosis. Second, thapsigargin acting on sarcoendoplasmic reticulum calcium transport ATPase (SERCA) pumps decreased syntilla frequency by partially emptying the intracellular Ca²⁺ stores and decreasing syntilla frequency. Hence the effect does not appear to be due to a non-specific effect of either agent as they acted by different mechanisms and on different proteins. Moreover, the degree of syntilla block correlated negatively with spontaneous catecholamine release (Lefkowitz et al. 2009). That is, syntilla suppression increased spontaneous exocytosis. As we calculated that a syntilla provides sufficient Ca²⁺ to cause exocytosis if it occurs in the region of a docked, primed vesicle we concluded that a syntilla releases Ca²⁺ into a microdomain different from one which houses these vesicles. This effect of syntillas was indeed surprising given that Ca²⁺ in the syntilla microdomain exerts the opposite effect of that due to Ca²⁺ in the VDCC microdomain.

Given their inhibitory role in spontaneous exocytosis (i.e. exocytosis in the absence of APs), we hypothesized that Ca²⁺ syntillas could play a role in the physiology of elicited exocytosis, especially the asynchronous phase as its timing is only loosely coupled to an AP. Here we examine exocytosis caused by low level physiological stimulation generated by APs at a frequency of 0.5 Hz, a frequency documented to be the physiological state popularly termed ‘rest and digest’ (Guyton & Hall, 2006). We report three major findings: (1) at low frequency stimulation less than 10% of all catecholaminergic exocytosis is synchronized to an AP; (2) the asynchronous phase of exocytosis does not require Ca²⁺ influx; and (3) we report a novel addition to the mechanism of stimulus–secretion coupling in ACCs wherein APs suppress Ca²⁺ syntillas. By this suppression of an inhibition, that is a disinhibition, exocytosis occurs.

Methods

Patch-clamp recording and preparation of mouse ACCs

Tight-seal, whole cell recordings on ACCs, freshly dissociated from adult male Swiss Webster mice as described previously (ZhuGe et al. 2006), were performed with a HEKA EPC10 amplifier (HEKA Electronics, Lambrecht, Germany) on the same day as isolation. Mice (6–8 weeks) were killed by cervical dislocation in accordance with the IACUC guidelines at the University of Massachusetts Medical School. Patch pipette solution (mm) was: 0.05 K₅fluo-3 or 0.025 K₅fura-2 (Molecular Probes, Eugene, OR, USA), 135 KCl, 2 MgCl₂, 30 Hepes, 4 Mg-ATP and 0.3 Na-GTP (pH 7.3). Bath solution comprised (mm): 135 NaCl, 5 KCl, 10 Hepes, 10 glucose, 1 MgCl₂ and 2.2 CaCl₂ (pH 7.2); Ca²⁺-free: 135 NaCl, 5 KCl, 10 Hepes, 10 glucose, 0.2 EGTA and 1 MgCl₂ (pH 7.2).

Amperometry

Catecholamine release was detected from individual cells using carbon fibre electrodes with a tip diameter of 5.8 μm (ALA Scientific Instruments, Westbury, NY, USA), as described before (ZhuGe et al. 2006). Only cut fibres with intrinsic noise < 0.5 pA were used. Amperometric signals were monitored with a VA-10 amplifier (NPI Electronic, Tamm, Germany), filtered at 0.5 kHz, digitized at 1 kHz with a Digidata 1200B acquisition system, and acquired with Patchmaster software from HEKA. Amperometric spikes were identified and analysed using the Mini Analysis program (Synaptosoft, Decatur, GA, USA). Each event was visually inspected to exclude artifacts from the analysis. The root mean square (RMS) noise in acquired traces was typically <0.25 pA as determined by Mini Analysis. The detection threshold for an event was set to 2.5 times the baseline RMS. Overlapping events were rare, and were excluded from analysis.

Analysis of stand alone foot events (SAFs) and spikes

In Table1 SAFs were separated from spikes based on criteria somewhat similar to those used by Wang et al. (2006), where an index of event shape was used to evaluate the ‘rectangularity’ of a putative SAF. To qualify as an SAF an event had to meet the criteria of an amplitude less than 2.5 pA and a ratio of full-width at half-height to event duration greater than 0.25. Event durations for spikes and SAFs are defined as the duration between the time when the event signal exceeds, and the time when it returns to, the detection threshold amplitude. For the analyses of SAFs and spikes comparing asynchronous to spontaneous events we approximated stimulated recordings to represent asynchronous exocytosis, as the majority of amperometric events in records from 0.5 Hz stimulation are asynchronous (i.e. >90% when uncorrected for the underlying spontaneous component) (see Results).

Table 1.

Kinetic and charge parameters of amperometric SAFs and spikes

	SAFs			Spikes
	Amplitude (pA)	Duration (ms)	Charge (pC)	Amplitude (pA)	Rise time (ms)	Charge (pC)
Pre	1.51 ± 0.14	53.60 ± 7.22	0.036 ± 0.006	7.38 ± 1.38	11.60 ± 1.15	0.133 ± 0.016
0.5 Hz	1.39 ± 0.09	53.95 ± 5.39	0.046 ± 0.007	5.86 ± 1.09	13.55 ± 1.05	0.160 ± 0.023
P-value	0.463	0.97	0.36	0.391	0.217	0.449

The kinetic parameters of stand alone foot events (SAFs) and spikes are largely unaffected by low frequency stimulation with simulated action potentials. Statistical comparisons were made with a two sample t test and charge values were first log-transformed.

Recording protocols

Fluo-3 Ca²⁺ imaging and amperometry

Once in whole cell configuration we waited until the Fluo-3 reached equilibrium and fluorescence was stable (about 2 min). We recorded two 4 s image sequences in a row (200 images separated by 20 ms, with an exposure time of 10 ms). Single 4 s recordings were made thereafter over time as indicated in each experiment. Amperometric recordings were made in 1 or 2 min segments sequentially, and the data were binned into intervals as shown in the figures.

Simulated action potentials (sAPs)

Patched cells with access resistances less than 20 MΩ and leak current below 30 pA were selected for stimulation experiments where they received trains of sAPs at 0.5 Hz. sAP waveforms consisted of a three step ramp as follows (start potential (mV), end potential (mV), duration (ms)): (1) −80, 50, 2.5; (2) 50, −90, 2.5; (3) −90, −80, 2.5. This waveform evoked Ca²⁺ and Na⁺ currents statistically identical to native APs (Figs 1A and 2) and thus are considered functionally equivalent (Chan & Smith, 2001).

Figure 2. sAPs evoke Na⁺ and Ca²⁺ currents identical to native action potentials in freshly isolated mouse ACCs.

A (top), representative current trace generated from a train of sAPs delivered at 0.5 Hz for 2 min. (Bottom) Na⁺ current typically attenuates during the first 5–7 sAPs, while the Ca²⁺ current remains constant throughout the entire 2 min of stimulation (e.g. −208.1 ± 18.8 pA at the 5th sAP vs. −186.6 ± 15.7 pA at the last sAP; n = 22 cells). The current trace above has been expanded at the location of select sAPs. B, representative current traces elicited by an sAP after 2 min in the presence (bottom panel) and absence (top panel) of 5 μm nifedipine, a dihydropyridine known to selectively inhibit Cav1.2 (L-type) currents in mouse chromaffin cells (Perez-Alvarez et al. 2011). Nifedipine was prepared from a 1000× stock solution in DMSO and applied to the cell by exchanging the bath solution. C, 5 μm nifedipine reduced the starting Ca²⁺ current evoked by an sAP to 65.2 ± 7% vs. the vehicle (1:1000 dilution of DMSO) which on average did not, 101.2 ± 7% of the starting Ca²⁺ current (P = 0.012, n = 4). The effects of nifedipine did not wash off after exchanging the bath for 2 min with the normal external solution. The percentage of starting Ca²⁺ current after the vehicle wash was 98.3 ± 13% vs. after nifedipine wash, 59.8 ± 13% (P = 0.0885, n = 4).

Ryanodine experiments

Ryanodine stock was first prepared in DMSO at 100 mm. Just before the experiments, ryanodine was dissolved in the physiological solution at 1 : 1000 to reach the 100 μm concentration used. The cells were bathed in the 100 μm ryanodine solution in the dark for 30 min before recordings started.

Ca²⁺ syntilla and global [Ca²⁺] measurement

Fluorescence images using fluo-3 as a Ca²⁺ indicator were obtained using a custom-built, wide-field digital imaging system described previously (ZhuGe et al. 2006). To assess the properties of individual Ca²⁺ syntillas quantitatively, the signal mass approach was used, as conceptualized by Sun et al. (1998) and developed for wide-field microscopy of Ca²⁺ sparks by ZhuGe et al. (2000). The purpose of this approach is to obtain a measure of the total amount of Ca²⁺ (as opposed to concentration of Ca²⁺) released by a focal Ca²⁺ transient. Global [Ca²⁺]_i was measured by fluorescence with cell-impermeant fura-2 (25 μm) that was loaded into cells through the patch pipette and measured as previously described (Grynkiewicz et al. 1985; Becker & Fay, 1987; Drummond & Tuft, 1999).

Bar charts of arrival times after an sAP

For each cell, amperometric events were binned into 200 ms increments according to their latency from the last sAP during 0.5 Hz stimulation (i.e. ten 200 ms bins between each sAP). The number of events in each bin was then averaged across all cells and the data are reported as a bar chart of the average number of events per cell.

Statistical analyses

Statistical analyses and plots were performed in OriginPro 8.5 (Origin, Northampton, MA, USA). Syntilla frequency is reported as the mean ± SEM of individual 4 s records. In all other cases, data were first averaged per cell and are reported as mean ± SEM of all cells. Unless indicated differently in the legends, ANOVA for repeated measures was performed on syntilla and amperometric event frequencies and pairwise comparisons vs. pre-stimulation were made post hoc using Fisher's least significant difference test. Amperometric charge values were first log-transformed, then subjected to Shapiro–Wilk and Kolmogorov–Smirnov tests for normality. Statistical comparisons were made with a paired or unpaired t test as the log-transformed data were normally distributed with equal variance between populations. All other statistical tests are indicated in the legends. A Bonferroni correction was applied in cases of multiple comparisons. P-values less than 0.05 were considered significant. Asterisks indicate levels of significance: *P < 0.05, **P < 0.01, ***P < 0.001.

Results

Freshly isolated mouse ACCs respond to physiological stimulation with the same ionic currents previously shown in slice preparations

In ACCs one nerve impulse every other second is sufficient to maintain the basal sympathetic tone, a physiological state popularly designated as ‘rest and digest’ (Guyton & Hall, 2006). To mimic the sympathetic tone, freshly isolated ACCs were voltage-clamped and stimulated at 0.5 Hz with sAPs that evoked Na⁺ and Ca²⁺ currents (Figs 1A and 2) with mean peak currents of −338.9 ± 31.1 and −208.1 ± 18.8 pA, respectively (n = 22), similar to native APs in mouse adrenal slices (Chan & Smith, 2003). At the same time, high-speed wide-field microscopy was employed to examine single Ca²⁺ syntillas (Fig.1B), a second source of cytosolic Ca²⁺, which arise from intracellular stores and occur spontaneously as we have shown before (De Crescenzo et al. 2004). Simultaneously catecholamine release from single vesicles was monitored with carbon fibre amperometry (ZhuGe et al. 2006). As we (Lefkowitz et al. 2009) and others (Liu & Misler, 1998; Vardjan et al. 2007; Fang et al. 2013) have reported previously, spontaneous exocytic events occurred in the absence of stimulation and their frequency was increased by sAP stimulation at 0.5 Hz, as shown by the representative traces in Fig.1C. The changes in exocytosis due to stimulation will be considered quantitatively below. Representative traces of individual amperometric events, SAFs, spikes and spikes preceded by a foot are shown at higher temporal and spatial resolution in Fig.1D.

Asynchronous exocytosis is the dominant form of exocytosis during low frequency physiological stimulation

Typical amperometric responses synchronized with each sAP at 0.5 Hz are shown in Fig.3A (right) along with their controls, i.e. no stimulation (left). Bar charts of all data are shown in Fig.3B. The shading in Fig.3A and B (right panels) marks the first 200 ms after each sAP. Figure3C indicates the averaged rate of amperometric events, both spikes and SAFs. The P-values in each case result from a comparison to pre-stimulation, i.e. spontaneous rates. (Note that the data in Fig.3A are of the same sort as Fig.1C but with the amperometric events presented in terms of time of occurrence after the preceding sAP, to allow the visualization of synchronous versus asynchronous events.)

Figure 3. Spontaneous exocytosis and two phases of elicited exocytosis in response to 0.5 Hz sAP stimulation.

A, representative traces of amperometric events from two cells unstimulated (left) and then during stimulation with sAPs at 0.5 Hz for 120 s (right). The upper and lower sets of traces are from two separate cells. On the right the 120 s traces were divided into 60 segments of 2 s and overlaid, such that the onset of each trace is synchronized with the sAP as shown in the schematic above, i.e. 60 segments of 2 s where each starts at the initiation of an sAP. On the left the traces are similarly accumulated but in the absence of stimulation. (Note that the duration of the sAP in the schematic is longer than its actual duration, 7.5 ms (Fig.1A), for purposes of clarity and to indicate its form. The onset of the traces below the schematic begin at the peak of the sAP.) B, right, amperometric events in each 2 s segment were binned into 200 ms increments according to their latency from the last sAP during 0.5 Hz stimulation. The first bin (coloured overlay) contains events within 200 ms of an sAP which are considered as synchronized exocytosis (n = 22 cells, 1320 sAPs, 412 events). Left, control, pre-stimulation data from the same cells from each 2 s sweep were binned into 200 ms intervals beginning at the onset of each sweep, with no sAPs (177 events). C, effect of 0.5 Hz stimulation on asynchronous vs. synchronous release frequency. Events that occurred within 200 ms of an sAP (i.e. synchronous release events) increased from a spontaneous frequency of 0.07 ± 0.02 s⁻¹ (Pre) to 0.25 ± 0.05 s⁻¹ (P = 0.004), while events that occurred after 200 ms of an sAP (i.e. asynchronous events) more than doubled, compared to spontaneous frequency, to 0.15 ± 0.03 s⁻¹ (P = 0.008) (paired t tests corrected for multiple comparisons).

Similar to previous studies (Zhou & Misler, 1995; Fulop et al. 2005; Doreian et al. 2008), sAPs induced a burst of amperometric spikes well within 200 ms of the sAP (synchronous exocytosis) followed by a sustained increase (asynchronous exocytosis) (Fig.3B, right). We note that 200 ms is an upper limit for latency of synchronous exocytosis, with most studies estimating the latency for synchronized exocytosis being of the order of tens of milliseconds (Chow et al. 1992, 1994; Heinemann et al. 1994; Zhou & Misler, 1995; Haller et al. 1998). One study, however, shows that with a 20 ms depolarizing square pulse the synchronized burst persists to about 150 ms (Chow et al. 1996). Thus, we chose 200 ms as a cutoff for synchronized release to avoid counting any synchronous events as asynchronous. However, this choice conceals the magnitude of the initial synchronized burst, which is much more evident when the data are binned at 15 ms intervals as shown in Fig.4. Finally we note that at the stimulation frequency of 0.5 Hz employed here the majority of exocytosis occurs at latency greater than 200 ms and hence is asynchronous. If we assume that the amperometric events in the first 200 ms are due to both synchronous and spontaneous events (Fig.3B, shaded bin), then in the 2 s period after each sAP, only about 10% are due to synchronized exocytosis.

Figure 4. Amperometric latency histograms binned at 15 ms intervals reveal a synchronized burst phase.

A, composite amperometric latency histograms from 22 ACCs before stimulation and stimulated at 0.5 Hz with sAPs according to the schematic above. Right, amperometric events in each 2 s segment of a 120 s amperometric trace were binned into 15 ms increments according to their latency from the last sAP during 0.5 Hz stimulation (n = 22 cells, 1320 sAPs, 412 events). Latencies were defined as the time from the peak of the last sAP. A synchronized burst occurs within 60 ms of an sAP (red bars). Left, control, pre-stimulation data from the same cells from each 2 s segment starting at the beginning of a 120 s amperometric trace with no sAPs were binned into 15 ms intervals (177 events). B, effect of 0.5 Hz stimulation on asynchronous and synchronous vs. spontaneous release. The mean number of events per bin that occurred within 60 ms of an sAP (i.e. the synchronous burst) increased from 1.32 ± 0.11 (Pre or spontaneous) to 6.75 ± 2.25 (P = 4.78 × 10⁻¹²), while the mean number of events per bin that occurred after 60 ms of an sAP (i.e. asynchronous events) more than doubled, compared to the spontaneous condition, to 2.96 ± 0.1 (P = 3.99 × 10⁻¹⁶) (paired t tests corrected for multiple comparisons). C, amperometric events were similarly binned into 15 ms increments according to their latency from the last sAP during 0.5 Hz stimulation, but in a Ca²⁺-free external solution (n = 18 cells, 1080 sAPs, 295 events). Note that there is no burst phase.

Ca²⁺ influx is not required for asynchronous exocytosis

One explanation for the asynchronous release is that it is caused by residual Ca²⁺ from sAP-induced Ca²⁺ influx. If this were the case, then the asynchronous exocytosis should be lost in the absence of external Ca²⁺. As can be seen in Fig.5, where the experiments of Fig.3 were repeated in Ca²⁺-free EGTA-buffered solution, this is not the case. Moreover direct measurements of global cytosolic [Ca²⁺] by Fura-2 (Fig.7D, see below) when external Ca²⁺ is present show no change in the whole cell [Ca²⁺], which remained well below the threshold for exocytosis. That is, in no case did the level of global [Ca²⁺] exceed 150 nm during stimulation (see Fig.7D). In previous work we have shown that buffering the cytosolic [Ca²⁺] up to a concentration of 500 nm caused no increase in exocytosis in mouse ACCs (Lefkowitz et al. 2009), consistent with what had been found in ACCs from other species (Chow et al. 1992, 1994). Thus, neither an increase in [Ca²⁺] locally due to residual influx nor an increase in global [Ca²⁺] over time is responsible for sAP-induced asynchronous exocytosis. We also note that the ‘burst’ in the first 200 ms is absent in Ca²⁺-free external solution as expected because it depends on the classical mechanism involving depolarization-induced Ca²⁺ influx (Fig.4C).

Figure 5. 0.5 Hz sAPs increase exocytosis in the absence of Ca²⁺ influx.

A, experiment schematic. ACCs were patched in normal external solution (with Ca²⁺). The whole cell configuration was achieved after the chamber was rapidly exchanged (within 3 min) with 30–40 ml of Ca²⁺-free external solution. The ACC and internal solution were allowed to equilibrate for 5 min and then 2 min amperometric recordings were performed, first in the absence of stimulation, followed by simultaneous stimulation with sAPs at 0.5 Hz. B, representative traces of amperometric events from two cells unstimulated (left) and then during stimulation with sAPs at 0.5 Hz for 120 s (right). The upper and lower sets of traces are from two separate cells. On the right the 120 s traces were divided into 60 segments of 2 s and overlaid, such that the onset of each trace is synchronized with the sAP as shown in the schematic above, i.e. 60 segments of 2 s where each starts at the initiation of an sAP. On the left the traces are similarly accumulated but in the absence of stimulation. C, data from B binned in the same fashion and according to the same conventions as in Fig.2B. Amperometric events in each 2 s segment were binned into 200 ms increments according to their latency from the last sAP during 0.5 Hz stimulation. Right, the first bin (coloured overlay) contains events within 200 ms of an sAP, which are considered as synchronized exocytosis (n = 22 cells, 1320 sAPs, 412 events). Left, control, pre-stimulation data from the same cells from each 2 s sweep were binned into 200 ms intervals beginning at the onset of each sweep, with no sAPs (177 events). D, effect of 0.5 Hz stimulation on asynchronous vs. synchronous release frequency. Events within 200 ms of an sAP increase from 0.047 ± 0.02 s⁻¹ (Pre) to 0.176 ± 0.05 s⁻¹ (P = 0.043); events after 200 ms of an sAP increase to 0.169 ± 0.05 s⁻¹ (P = 0.042) (Bonferroni-corrected, paired sample t tests).

Figure 7. Low frequency stimulation by simulated APs suppresses syntillas and increases exocytosis.

A, 0.5 Hz stimulation completely suppresses syntillas within 2 min. Closed circles: syntilla frequency before (Pre) and during stimulation at 0.5 Hz: Pre (0.573 ± 0.07 s⁻¹) vs. 0–30 s (0.15 ± 0.06 s⁻¹), P = 1.55 × 10⁻⁶; vs. 30–60 s (0.033 ± 0.03 s⁻¹), P = 1.07 × 10⁻⁸; vs. 60–120 s (0 s⁻¹), P = 2.62 × 10⁻⁹ (N = 15 cells). Open circles: syntilla frequency in the absence of stimulation at 0 s (0.523 ± 0.2 s⁻¹), 120 s (0.545 ± 0.17 s⁻¹), 7 min (0.591 ± 0.19 s⁻¹, not shown) and 12 min (0.607 ± 0.14 s⁻¹, not shown) (n = 11 cells). B, 0.5 Hz stimulation causes a 3-fold increase in amperometric frequency over the same time course as syntilla suppression. Pairwise comparisons of amperometric frequency were made within each cell and the means were compared: Pre (0.067 ± 0.016 s⁻¹) vs. 0–30 s (0.111 ± 0.032 s⁻¹), P = 0.37; vs. 30–60 s (0.165 ± 0.047 s⁻¹), P = 0.044; Pre vs. 60–120 s (0.197 ± 0.051 s⁻¹), P = 0.008 (n = 22). C, 0.5 Hz stimulation for 2 min does not significantly alter quantal charge, Q, of amperometric events. The mean charge of all amperometric events before and during stimulation from the same 22 cells presented in Fig.1C: Pre vs. 0–30 s, P = 0.865; Pre vs. 30–60 s, P = 0.966; Pre vs. 60–120 s, P = 0.521. D, 0.5 Hz stimulation does not alter mean global [Ca²⁺]_i as detected by Fura-2 dye: pre (81.0 ± 13.4 nm) vs. 0.5 Hz stimulation during 0–30 s (85.6 ± 16.1 nm); 30–60 s (87.3 ± 17.2 nm); 60–120 s (86.1 ± 15.8 nm), P = 0.514, 0.484 and 0.483, respectively, paired t tests (P = 1 after correction for multiple comparisons) (n = 12 cells). A representative trace of the un-averaged global [Ca²⁺]_i is overlaid.

ACCs employ the ryanodine receptor, RyR2, in asynchronous exocytosis

What accounts for the asynchronous phase during 0.5 Hz stimulation if it is not tied to Ca²⁺ influx? In a previous set of studies we demonstrated that: (1) mouse ACCs had spontaneous exocytotic activity and spontaneous Ca²⁺ syntillas (ZhuGe et al. 2006; Lefkowitz et al. 2009); (2) the spontaneous exocytosis was increased when Ca²⁺ syntillas were inhibited by ryanodine (blocking RyRs) or thapsigargin and caffeine (blocking endoplasmic reticulum (ER) Ca²⁺ uptake pumps and emptying the ER Ca²⁺), thus acting on different molecular targets and via different mechanisms (Lefkowitz et al. 2009); and (3) the degree of enhanced exocytosis correlated with the decrease in syntilla frequency (Lefkowitz et al. 2009). We concluded that Ca²⁺ syntillas block spontaneous exocytosis. Thus, it was natural to ask whether modulation of Ca²⁺ syntillas might account for enhanced asynchronous exocytosis during stimulation.

If this were the case, then syntilla suppression by sAP stimulation should produce no further increase in exocytosis if syntillas were already blocked. To examine this, the ACCs were treated with 100 μm ryanodine, a concentration previously shown to suppress syntillas (ZhuGe et al. 2006; Lefkowitz et al. 2009) by blocking RyRs (Xu et al. 1998), for 30 min and then stimulated with sAPs at 0.5 Hz. Consistent with our previous study (Lefkowitz et al. 2009), ryanodine increased spontaneous catecholamine exocytosis (Fig.6C vs. Fig.3C, leftmost bar in each case). Moreover, 0.5 Hz stimulation failed to elicit additional increases in exocytosis (Fig.6A), particularly asynchronous exocytosis (Fig.6B and C). This suggests that the suppression of Ca²⁺ syntillas mediates sAP-induced asynchronous exocytosis. We were unable to detect a significant increase in synchronized exocytosis (Fig.6B, shaded bin and Fig.6C, middle bar), and it was not apparent even when the data were rebinned at 15 ms intervals (not shown).

Figure 6. Low frequency stimulation in the presence of ryanodine does not promote additional asynchronous exocytosis compared to the blockade of RyRs alone.

A, 0.5 Hz stimulation does not further increase amperometric frequency in the presence of 100 μm ryanodine: P = 0.66 Pre vs. 0–30 s; P = 0.40 Pre vs. 30–60 s; P = 0.66 Pre vs. 60–120 s (n = 14, paired t test). B, effect of ryanodine on asynchronous release. Data from A binned in the same fashion and according to the same conventions as in Fig.2B. C, no additional effect of 0.5 Hz stimulation on asynchronous or synchronous release frequency. Events within 200 ms of an sAP increased from 0.131 ± 0.04 s⁻¹ (Pre) to 0.185 ± 0.05 s⁻¹ (P = 0.311), while events after 200 ms of an sAP increased to 0.15 ± 0.04 s⁻¹ (P = 0.656) (paired sample t tests).

Syntilla suppression is caused by APs

We next examined the possible involvement of syntillas in the regulation of asynchronous exocytosis by direct measurement. To be consistent with the results presented above, stimulation via sAPs would have to suppress Ca²⁺ syntillas. This in turn presented a possible contradiction, as in cardiac and skeletal muscle, stimulation via APs causes an increase in spark frequency due to coupling between dihydropyridine receptors and RyRs (Cannell et al. 1995; Lopez-Lopez et al. 1995; Klein et al. 1996). Surprisingly then, we found that sAPs delivered at 0.5 Hz drastically reduced syntilla frequency within 30 s of the onset of stimulation, abolishing them within 2 min (Fig.7A). This stimulation also induced a 3-fold increase in frequency of amperometric events (Fig.7B), both spikes (0.0477 vs. 0.125 s⁻¹, P = 0.017) and SAFs (0.0136 vs. 0.0413 s⁻¹, P = 0.013), during 2 min of stimulation with no detectable change in their mean charge or kinetics (Fig.7C and Table1). There was an inverse relationship between the frequency of syntillas and amperometric events over the same time (Fig.7A vs. Fig.7B). These results, taken together with the results where 0.5 Hz stimulation was unable to elicit any further increase in exocytosis after ryanodine was used to block syntillas (Fig.6), provide support for the hypothesis that syntillas are an intermediary regulating asynchronous exocytosis.

Syntilla suppression does not require Ca²⁺ influx

How did the sAPs reduce the frequency of Ca²⁺ syntillas? There are two general classes of mechanism whereby dihydropyridine receptors (DHPRs) affect RyRs. In one case as in skeletal muscle, the mechanism depends only on depolarization, i.e. voltage-induced Ca²⁺ release from internal stores (VICaR) and in another, as in cardiac muscle the coupling depends on depolarization-induced Ca²⁺ entry, or Ca²⁺-induced Ca²⁺ release (CICR). When we repeated our experiments in a Ca²⁺-free, EGTA-buffered external solution, we again found sAPs at 0.5 Hz to effectively suppress syntilla frequency within 2 min of the stimulation (Fig.8A). That is, a necessity for calcium influx could be excluded altogether in the mechanism for syntilla suppression. Furthermore, the stimulation under the Ca²⁺-free condition caused a similar, approximately 3-fold increase in amperometric frequency, but which had a faster onset and began to fade during the last minute of stimulation (Fig.8B). Another difference in the Ca²⁺-free condition was that the charge of amperometric events increased slightly within the first 30 s of stimulation. Noted, however, that before stimulation the charge was low compared to when Ca²⁺ was present outside of the cell (compare the leftmost bar in Fig.7C to that in Fig.8C). Again we found an inverse relationship between the frequency of syntillas and amperometric events over the same period (Fig.8A vs. Fig.8B).

Figure 8. Syntilla suppression by 0.5 Hz sAPs increases exocytosis in the absence of Ca²⁺ influx.

A, 0.5 Hz stimulation effectively suppresses syntillas within 2 min. Syntilla frequency recordings before (Pre) and during stimulation: Pre (1.1 ± 0.14 s⁻¹) vs. 0–30 s (0.1 ± 0.08 s⁻¹), P = 8.42 × 10⁻¹⁰; vs. 30–60 s (0.1 ± 0.08 s⁻¹), P = 8.42 × 10⁻¹⁰; vs. 60–120 s (0.025 ± 0.025 s⁻¹), P = 1.84 × 10⁻¹⁰ (n = 10 cells). B, 0.5 Hz stimulation over the same time course as syntilla suppression increases amperometric frequency in the absence of Ca²⁺ influx: Pre (0.047 ± 0.02 s⁻¹) vs. 0–30 s (0.239 ± 0.1 s⁻¹), P = 0.016; vs. 30–60 s (0.211 ± 0.07 s⁻¹), P = 0.038; vs. 60–120 s (0.126 ± 0.03 s⁻¹), P = 0.312 (n = 18). C, quantal charge, Q, of amperometric events is significantly altered during the first 30 s of 0.5 Hz stimulation. The mean charge of events from the same 18 cells presented in B over the same time course: Pre (0.057 ± 0.01 pC) vs. 0–30 s (0.14 ± 0.04 pC), P = 0.019; vs. 30–60 s (0.129 ± 0.03 pC), P = 0.209; vs. 60–120 s (0.112 ± 0.03 pC), P = 0.139 (Student's t test).

Asynchronous events differ from spontaneous events in their frequency but not in their characteristics

As we previously found the same inverse relationship between syntillas and spontaneous exocytosis (Lefkowitz et al. 2009), we wondered if the asynchronous phase of exocytosis elicited by an AP may simply be the result of an increase in the same class of events as spontaneous exocytosis. If this were true, we would expect the individual events detected in stimulated and unstimulated conditions to be similar, with the former showing only an increase in frequency. Are the asynchronous amperometric events similar to spontaneous amperometric events, in total charge per event and other parameters, differing only in frequency?

In the presence of normal extracellular solution, comparing spontaneous to asynchronous events, there was no detectable difference in the mean charge or amplitude of either the SAFs or spikes, nor of the rate of rise of spikes or duration of SAFs (Table1). Thus, we were unable to detect a difference in spontaneous and asynchronous events. This finding is of importance when considering a mechanism for asynchronous exocytosis (see Discussion). (We note that as the amplitude and charge remained unchanged during stimulation, the increased frequency of asynchronous exocytosis reported here is not due to improved detection of larger events.) Finally the ratio of the frequency of SAFs to spikes was approximately 1:3 for both the spontaneous (SAFs/spikes, 0.014 ± 0.004:0.048 ± 0.008, n = 22) and the asynchronous groups (SAFs/spikes, 0.041 ± 0.009:0.125 ± 0.027, n = 22), suggesting no change in the mode of fusion.

Discussion

Our findings provide three new insights into stimulus–secretion coupling. First, during low frequency stimulation at 0.5 Hz with sAPs, catecholaminergic exocytosis in mouse chromaffin cells is predominantly due to asynchronous exocytic events, i.e. those occurring at a latency greater than 200 ms following sAP; the asynchronous exocytic frequency during this stimulation is about twice that of the spontaneous frequency (Fig.3B). Second, this asynchronous exocytosis does not require Ca²⁺ influx. Third, we present evidence that the asynchronous exocytic pathway is regulated through a novel mechanism wherein APs generated at a rate of 0.5 Hz suppress Ca²⁺ released from internal stores (i.e. Ca²⁺ syntillas). As Ca²⁺ entry into the syntilla microdomain normally inhibits spontaneous exocytosis, as we have demonstrated earlier (Lefkowitz et al. 2009), we propose that the suppression of syntillas by APs causes an increase in exocytosis (Fig.9).

Figure 9. Model for elicited exocytosis at low frequency, physiological stimulation through syntilla suppression.

APs elicit asynchronous or synchronous exocytosis by two separate pathways. During low frequency stimulation the major path for catecholamine secretion is the asynchronous one, accounting for about 90% of exocytosis. This path employs a disinhibition mechanism wherein APs inhibit syntillas, relieving their inhibition on exocytosis, which leads to an increase in the asynchronous phase. APs also elicit synchronized exocytosis, to a lesser extent, via a classical Ca²⁺ influx pathway (black arrow).

Our estimation of the fraction of exocytosis that is asynchronous is conservative

We attach a number of 90% to the fraction of exocytosis that is asynchronous at 0.5 Hz stimulation, but this may be a lower bound for two reasons. First, we obtained this number by subtracting the baseline or spontaneous level from the exocytic frequency in the first 200 ms after the sAP, a conservative cutoff for synchronized release that accommodates criteria used across multiple studies (see Results) to avoid counting any synchronized events as asynchronous. That puts 10% of the total increase in exocytosis in the first 200 ms interval. But, second, it is likely that there is also a component in that interval caused by the syntilla suppression mechanism. If we take that into account by subtracting the increase in amperometric event frequency in the subsequent 1800 ms, we get a value closer to 95%. However, we do not use this number because it involves an additional assumption, albeit one that is probably correct.

Asynchronous exocytosis is regulated by Ca²⁺ differently from synchronous exocytosis

During 0.5 Hz stimulation the classical mechanisms of stimulus–secretion coupling associated with synchronous exocytosis (i.e. Ca²⁺ influx based) do not apply to catecholamine release events that are only loosely coupled to an AP, asynchronous exocytosis. Unlike the synchronized phase, the asynchronous phase does not require Ca²⁺ influx. This is supported by our findings that (1) the asynchronous exocytosis could be increased by sAPs in the absence of external Ca²⁺ and (2) in the presence of external Ca²⁺, sAPs at 0.5 Hz increased the frequency of exocytosis without any significant rise in the global Ca²⁺ concentration, thus excluding the possibility that the exocytosis was increased by residual Ca²⁺ from sAP-induced influx. These results are not the first to challenge the idea that spontaneous or asynchronous release arises from the ‘slow’ collapse of Ca²⁺ microdomains, due to slow Ca²⁺ buffering and extrusion. For example, a decrease of Ca²⁺ buffers such as parvalbumin in cerebellar interneurons (Collin et al. 2005) and both GABAergic hippocampal and cerebellar interneurons (Eggermann & Jonas, 2012) did not correlate with an increase in asynchronous release. And in the case of excitatory neurons, it has been shown that Ca²⁺ influx is not required for spontaneous exocytosis (Vyleta & Smith, 2011).

Asynchronous exocytosis is regulated similarly to spontaneous exocytosis

The fact that the asynchronous amperometric events reported here were similar to spontaneous amperometric events in total charge per event and release parameters listed in Table1, differing only in frequency, is consistent with their belonging to the same population of vesicles as in spontaneous exocytosis. In turn this leads us to postulate that the mechanism of asynchronous release is simply a stronger activation of the mechanism that regulates spontaneous release. This idea is further supported by our finding that 0.5 Hz stimulation did not have any noticeable effect on the fusion pore, as measured by the ratio of SAFs to spikes and the mean duration of SAFs. In contrast, in ACCs the fusion pore has been shown to dilate with more intense stimulation associated with synchronous release (Fulop & Smith, 2006; Doreian et al. 2008; Fulop et al. 2008). Finally, the regulation of asynchronous exocytosis involves RyRs, particularly RyR2, which we have previously shown to regulate spontaneous exocytosis in ACCs. This conclusion comes from our finding that 0.5 Hz stimulation failed to elicit additional increases in asynchronous exocytosis after the exocytic frequency was already elevated by inhibition of the RyRs with blocking concentrations of ryanodine.

Syntilla suppression as a mechanism regulating asynchronous exocytosis

In our previous studies in ACCs, we found that spontaneous exocytosis could be increased if Ca²⁺ syntillas were suppressed by ryanodine (blocking RyRs) or a combination of thapsigargin and caffeine (blocking ER Ca²⁺ uptake pumps and emptying the ER Ca²⁺). We further demonstrated that the magnitude of the increased exocytosis correlated with decreasing syntilla frequency. That is, Ca²⁺ syntillas blocked spontaneous exocytosis. As the asynchronous exocytosis observed here did not require Ca²⁺ influx, and because the characteristics of the release events were similar to those of spontaneous exocytosis, we investigated the possibility that Ca²⁺ syntillas (i.e. the lack of Ca²⁺ syntillas) might account for the asynchronous exocytosis during stimulation. Indeed, we found that sAPs delivered at 0.5 Hz drastically reduced syntilla frequency while increasing the frequency of amperometric events 3-fold. That is, we uncovered an inverse relationship between the frequency of syntillas and amperometric events over time, similar to what we reported in our studies of spontaneous exocytosis.

The finding that sAPs suppressed Ca²⁺ syntillas surprised us, but at the same time resolved a paradox. In CICR, Ca²⁺ entry through VDCCs activates nearby RyR2s, causing quantal Ca²⁺ release from the ER, e.g. in the well-studied case of cardiac myocytes (Fabiato, 1983). Given that understanding, we predicted APs should increase syntillas, which serve to prevent spontaneous exocytosis. Yet, APs are classically known to increase exocytic output. AP-induced syntilla suppression explains this discrepancy. Furthermore our findings are consistent with an earlier study in which CICR was found only to a small extent in mouse ACCs (Rigual et al. 2002). However, that is not the entire story because CICR does come into play when cholinergic agonists are employed in certain experimental paradigms, as shown for example by the convincing study by Wu et al. (2010). (This is discussed in further detail below under ‘Implications’.)

How do our findings and mechanism compare with other studies?

Notably, our study is the first to describe a disinhibition mechanism to account for asynchronous exocytosis. In recent years a number of studies have put forth a variety of mechanisms to explain asynchronous exocytosis. These studies, however, describe mechanisms based for the most part on Ca²⁺ influx from outside a cell with vesicle proteins as the target. For example, some studies suggest that distinct Ca²⁺-sensing vesicle proteins regulate the synchronous and asynchronous release (e.g. synaptotagmin 1 and Doc2, respectively) based on differential sensitivity to Ca²⁺ influx (Walter et al. 2011; Yao et al. 2011). Others suggest that the determining factor lies in the distance of docked vesicles from the site of Ca²⁺ influx (Wadel et al. 2007). Few et al. (2012) have pointed out the possibility that delayed, long-lasting (500 ms) tail currents from VDCCs could contribute to asynchronous release. Still others suggest that VDCCs may play only a small role in asynchronous exocytosis, if any at all; instead, extracellular Ca²⁺ concentration ([Ca²⁺]_o) seems to be a determining factor and different ion channels and G-protein-coupled receptors may be involved (Smith et al. 2012). Not only is our study the first to describe a disinhibition mechanism in asynchronous exocytosis, but it is clear from the results in Ca²⁺-free extracellular solution that the mechanism does not involve Ca²⁺ influx.

There are a number of reasons why we might suspect the mechanism of disinhibition found here in ACCs to be a general one, extending to exocytosis in neurons. First, many neurons exhibit asynchronous release upon stimulation (Hefft & Jonas, 2005; Daw et al. 2009; Jiang et al. 2012). Second, RyRs are widely expressed throughout the brain (Giannini et al. 1995), with RyR2 being the most abundant isoform, the same isoform that dominates in the mouse ACCs used here (ZhuGe et al. 2006; Wu et al. 2010). And third, Ca²⁺ syntillas have been demonstrated in central nerve terminals (De Crescenzo et al. 2004, 2006, 2012; Ross, 2012), where we have already shown that they do not trigger exocytosis (McNally et al. 2009). Thus, regulation of Ca²⁺ syntillas could serve as a presynaptic mechanism to modulate synaptic strength, and stabilization.

Implications

Our findings raise a rich set of questions at the level of both physiology and molecular biology. Can syntilla suppression be activated by ACh, the physiological neurotransmitter? Physiologically, APs in ACCs are triggered by excitatory postsynaptic potentials mediated by muscarinic and nicotinic cholinergic receptors activated by ACh released from splanchnic nerve terminals (Douglas & Rubin, 1961; Douglas & Poisner, 1965). Furthermore, is ACh receptor stimulation directly involved in the activation of syntilla suppression or merely necessary to generate the excitatory postsynaptic potentials that cause APs? The answers to these questions require extensive and technically demanding experiments, wherein APs occurring at low frequency are reliably generated by a chemical stimulus. At this point we can also say that physiological ACh stimulation would have to be intermittent to induce syntilla suppression as prolonged, 10 s stimulations of nicotinic ACh receptors with 1,1-dimethyl-4-phenylpiperazinium iodide (DMPP) have been shown to induce Ca²⁺ influx which leads to CICR via RyRs in mouse ACCs (Wu et al. 2010).

What is the mechanism for the AP-induced suppression of Ca²⁺ syntillas? We know that the interaction is independent of Ca²⁺ influx, as syntillas are still suppressed by sAPs in the absence of extracellular Ca²⁺. Hence we must postulate a voltage-dependent mechanism, operating by a voltage-induced change within the membrane where the voltage change can be sensed. Two possibilities present themselves. First, because we know that the target must be the RyR2s through which syntillas arise, there could be a direct physical linkage between a voltage-dependent channel, probably a Ca²⁺ channel and RyR2. Such an interaction at present has been limited to RyR1 and the effect is to activate not inhibit RyR1. But there has been evidence to indicate an inhibitory interaction between a Ca²⁺ channel and one of the RyR isoforms, RyR1 (Zhou et al. 2006; Pouvreau et al. 2007). Presently, however, there is no clear indication of such an interaction between a channel and RyR2. The present study should motivate a search for such an interaction. Second, there may be a voltage-dependent enzyme in the membrane which generates a second messenger to shut down the RyR2 or act on another molecule which eventually leads to shut down of RyR2. Examples of such voltage-activated enzymes are limited (Murata et al. 2005), but they may be more widespread than recognized at present. A second question is how do Ca²⁺ syntillas inhibit asynchronous exocytosis? We have previously speculated on the involvement of a vesicular Ca²⁺ sensor which detects the syntilla and inhibits or limits the granule's ability to complex with the exocytotic machinery (Lefkowitz et al. 2009). Finally, what are the precise molecular entities that mediate these processes? The pursuit of answers to these questions promises to reveal heretofore unknown aspects of regulation of exocytosis and synaptic transmission.

Glossary

ACC

adrenal chromaffin cell

AP

action potential

CICR

Ca²⁺-induced Ca²⁺ release

ER

endoplasmic reticulum

RyR

ryanodine receptor

SAF

stand alone foot event

sAP

simulated action potential

VDCC

voltage-dependent calcium channel

Key points

• Although the importance of asynchronous exocytosis is becoming clearer, not enough is known about its roles and mechanisms.

• Here we describe the nature of exocytosis in mouse adrenal chromaffin cells during low frequency physiological stimulation, i.e. 0.5 Hz, providing new views.

• We report that less than 10% of all catecholaminergic exocytosis during low frequency stimulation is synchronized to a simulated action potential (sAP), i.e. the dominant phase is asynchronous.

• This asynchronous phase of exocytosis does not require Ca²⁺ influx, requires the ryanodine receptor, RyR2, and comprises exocytic events with characteristics similar to those of spontaneous events.

• We propose a novel mechanism of disinhibition wherein APs inhibit Ca²⁺ syntillas, relieving their inhibition of spontaneous exocytosis, which leads to an increase in the asynchronous phase of elicited exocytosis.

• The work has the specific physiological implication that basal sympathetic tone associated with the ‘rest and digest’ state is set in part by Ca²⁺ syntillas. Furthermore, there is evidence that this regulation of exocytosis by Ca²⁺ syntillas may be a general mechanism that extends to neurons.

Additional information

Competing interests

The authors declare no competing financial interests.

Author contributions

Conception and design of the experiments: J.J.L., V.D., K.D.B., K.E.F., J.V.W. and R.Z.G. Data collection: J.J.L and K.D. Analysis: J.J.L. Interpretation of data: J.J.L., V.D., K.E.F., J.V.W. and R.Z.G. Drafting the article: J.J.L., J.V.W. and R.Z.G. Revising it critically for important intellectual content: J.J.L., V.D., K.E.F., J.V.W. and R.Z.G. All authors have read and approved the final submission.

Funding

This study was supported by Grant HL21697 (to J.V.W.) from the National Institutes of Health; Grant 0835580D (to V.D.C.) from the American Heart Association; and US National Heart, Lung, and Blood Institute grants HL73875 and HL117104 (to R.Z.G.).