Kinetics of stimulus-coupled secretion in dialyzed bovine chromaffin cells in response to trains of depolarizing pulses

Stimulus-secretion coupling in bovine chromaffin cells was investigated with whole-cell patch-clamp recordings and capacitance detection techniques to monitor exocytosis in response to trains of depolarizing pulses. Two kinetically discrete modes of exocytotic responses were observed. In one mode, the first depolarization of a train elicited a large increase in membrane capacitance (Cm; mean approximately 70 fF). This secretory mode was characterized by small Ca2+ requirements, relative insensitivity to the pipette Ca2+ chelator concentration, and rapid depletion of the secretory response. This mode of stimulus-secretion coupling was labile and was seen only in response to the first and, occasionally, the second stimulus train of whole-cell recordings. The second type of exocytotic response persisted for the remainder of the whole-cell recordings and consisted of two distinct phases. During the earliest pulses of a stimulus train, Ca2+ entry did not evoke Cm increases. Instead, Cm responses were elicited by later pulses, despite diminished Ca2+ entry per pulse caused by Ca2+ channel inactivation. The secretory phase was initiated after a specific "threshold" amount of Ca2+ had entered the cell, which was determined by the concentration, but not the binding kinetics, of the Ca2+ chelator in the pipette. In both the early and the secretory phases, the response of the cell was proportional to cumulative Ca2+ entry, regardless of current amplitude, pulse duration, or number of pulses. Threshold-type secretory kinetics has been described previously in peptide-secreting neurohypophysial (NHP) nerve terminals (Seward et al., 1995). Secretory kinetics with minimal Ca2+ requirements has not been observed in that preparation. Chromaffin cells appear to possess a broader repertoire of stimulus-secretion coupling modes than NHP terminals.


Stimulus-secretion
coupling in bovine chromaffin cells was investigated with whole-cell patch-clamp recordings and capacitance detection techniques to monitor exocytosis in response to trains of depolarizing pulses. Two kinetically discrete modes of exocytotic responses were observed. In one mode, the first depolarization of a train elicited a large increase in membrane capacitance (C,; mean -70 fF). This secretory mode was characterized by small Ca'+ requirements, relative insensitivity to the pipette Ca'+ chelator concentration, and rapid depletion of the secretory response. This mode of stimulus-secretion coupling was labile and was seen only in response to the first and, occasionally, the second stimulus train of whole-cell recordings.
The second type of exocytotic response persisted for the remainder of the whole-cell recordings and consisted of two distinct phases. During the earliest pulses of a stimulus train, Ca"+ entry did not evoke C, increases. Instead, C, responses were elicited by later pulses, despite diminished Ca'+ entry per pulse caused by Ca*' channel inactivation.
The secretory phase was initiated after a specific "threshold" amount of Ca"+ had entered the cell, which was determined by the concentration, but not the binding kinetics, of the Ca*+ chelator in the pipette. In both the early and the secretory phases, the response of the cell was proportional to cumulative Ca*' entry, regardless of current amplitude, pulse duration, or number of pulses.
Threshold-type secretory kinetics has been described previously in peptide-secreting neurohypophysial (NHP) nerve terminals (Seward et al., 1995). Secretory kinetics with minimal Ca2+ requirements has not been observed in that preparation. Chromaffin cells appear to possess a broader repertoire of. stimulus-secretion coupling modes than NHP terminals.
Key words: calcium-secretion coupling; membrane-capacitance detection: exocytosis; large dense core-vesicles (LDCV); calcium chelators (EGTA, BAPTA); chromaffin cells Ca'+-secretion coupling in adrenal chromaffin cells is a complex cellular process that involves Ca'+-dependent preparatory and fusion steps and an extensive array of modulatory phenomena involving both second-messenger systems (Burgoyne and Morgan, 1993) and cytoskeletal elements (Cheek and Barry, 1993;Trifaro et al., 1993). The physiological stimulus for catecholamine secretion is acetylcholine release from the splanchnic nerve, which triggers bursts of action potential (AP) activity (Brandt et al., 1976;Ozawa and Sand, 1986;Zhou and Misler, 1995). As voltagegated Ca2+ channels are activated by each AP within a burst, bursts cause complex pulsatile Cazt elevations near the membrane as well as a slower global summation (Thayer and Miller, 1990;Yamada and Zucker, 1992;Nowycky and Pinter, 1993;Peng and Zucker, 1993;Stuenkel, 1994).
To avoid the complexities and uncertainties inherent in pulsatile Cazt entry, two groups have studied the kinetics of Ca2+stimulated exocytosis of large dense-cored vesicles (LDCVs) in response to flash photolysis of caged Ca2+ compounds (Heine-mann et al., 1993(Heine-mann et al., , 1994Neher and Zucker, 1993;Thomas et al., 1993). In both chromaffin cells and melanotrophs, flash photolysis of demeclocycline (DM)-nitrophen abruptly elevates intracellular Ca2+ concentration ([Ca"],) into the micromolar to hundreds of micromolar range and elicits membrane capacitance (C,) changes (AC,) that differ in speed and amplitude. Based on the distinct kinetic components of the C, response, both groups have produced models of Ca2+ secretion-coupling that are similar enough to be summarized together. The following are the essential features. (1) At rest, LDCVs exist in sets of functionally distinct "pools" or "states" in a cell. (2) Movement between pools is modeled as a series of simple chemical reactions. (3) The pools are arranged sequentially, leading from least prepared to most prepared and ready for exocytosis. (4) Movement between pools is reversible with the exception of the final exocytotic step. (5) The step(s) leading from the final I'esting state to final fusion has a third-or fourth-order dependence on [Ca2+],. Although the models are attractive in their simplicity, the flash photolysis experiments on which they are based have certain limitations. Among the most important of these limitations are the nonphysiologically high and prolonged Ca2+ elevations that are produced as well as cell dialysis in the absence of Mg-ATP. Two studies of rat chromaffin cells using capacitance (Horrigan and Bookman, 1994) and amperometric techniques (Zhou and Misler, 1995) that examined the effect of voltage-gated Cazt entry have reported exocytotic components that are not described by the above models.
In a previous report, we used stimulus-train protocols that mimic AP bursts to examine the coupling of LDCV release to pulsatile, voltage-gated Cazt entry in peptide-secreting neurohypophysial (NHP) terminals. We found that exocytosis exhibits a kinetic pattern, which we called "threshold" secretion (Seward et al., 1995). In this secretory mode, Ca2+ entry during initial depolarizations does not elicit secretion until a certain critical amount of summed Ca2+ entry occurs. Subsequent Ca2+ entry is responsible for the final steps leading to fusion and exocytosis. In this study, we perform similar experimental manipulations in bovine chromaffin cells. We find that chromaffin cells also respond with threshold-type secretion but have an additional kinetically distinct response that differs in its Ca2+ dependence.
Some of this work has been presented previously in abstract form (Seward and Nowycky, 1995).

Chromafin cell culture
Adult bovine chromaffin cells were prepared as described in Vitale et al. (1991). After digestion and purification on a Percoll gradient, cells were plated in 10% fetal calf serum and 90% Dulbecco's modified Eagle's medium on collagen-coated coverslips at a density of 1 X 10' cells/35 mm Petri dish. Most of the recordings were made between 36 and 72 hr after plating. The exceptions are recordings portrayed in Figure 6, which were carried out between 3 and 7 d after plating.

RESULTS
Ca2+-stimulated exocytosis was monitored with the C, detection technique in whole-cell recordings of cultured bovine chromaffin cells. Cells were stimulated with trains of short depolarizing pulses that mimic physiological bursts of APs (Brandt et al., 1976;Ozawa and Sand, 1986;Zhou and Misler, 1995). Two kinetically distinct patterns of C, responses were observed.
Two kinetically distinct patterns of AC,,, in single dialyzed cells Figure L4 illustrates a typical response to the first stimulus train after establishing the whole-cell configuration ("break-in"), and Figure 1B illustrates the response to the third stimulus train. The timing of individual pulses is indicated by gaps in the C, and conductance (G) traces. In Figure lAi, a large C,, jump (-92 fF) occurs during the first depolarizing pulse. This is followed by five progressively diminishing large jumps (Fig. l&i), for a total AC, of -310 fF. The responses to the remaining 14 depolarizing pulses consist of smaller, uniform steps for which the average amplitude is 12 fF. The C,,, trace elicited by the third train (Fig. lBi) does not have large jumps at the beginning of the train; instead, the average amplitude of the small steps for the first 6 pulses actually is lower than that for the last 14 pulses (7.7 vs 14.9 fF, Fig. 1Biii).
Individual jump amplitudes per pulse are plotted in Figure 1, Aiii and Biii, as a function of cumulative Ca2+ entry. This value is obtained by integrating the Ca2+ current evoked by each depolarizing pulse and then summing with previous Ca2+ entry during the stimulus train. The large C, jumps during the first train (Fig. l&ii) occur below 10 X lo7 total Ca2+ ions in this example. During the third train (Fig. lBiii), the larger average jumps begin after 10 X lo7 Cazt ions have entered the cell.
We will refer to the large C, jumps evoked by small amounts of Ca2+ entry as "docked" secretion, analogous with fast synapses and rapid responses in both melanotrophs and chromaffin cells (Neher and Zucker, 1993;Thomas et al., 1993); we will refer to the second type of exocytotic response as threshold secretion, analogous with similar responses in NHP nerve terminals (Seward et al., 1995). The remainder of the paper describes differences between the two secretory modes in terms of time-dependent run-down, Ca2+ requirement, and effect of Ca2+ chelators.
Disappearance of docked secretion and run-down of threshold secretion during cell dialysis In dialyzed bovine chromaffin cells, the characteristic pattern of large C, jumps during the beginning of the train (first 6 pulses of Fig. 1A) is seen only during the first and sometimes second stimulus trains after break-in, whereas the pattern illustrated in Figure 1B is obtained repeatedly during the whole-cell recording. Typically, the first stimulus train was delivered within -3 min of establishing the whole-cell configuration, i.e., after break-in and initiation of cell dialysis. Exchange of small molecules such as ethylene glycol his@-aminoethyl)ether-N,N,N',N'-tetra-acetic acid (EGTA) and BAPTA is calculated and measured to be complete within 20 set (Pusch and Neher, 1988;Neher and Augustine, 1992;Thomas et al., 1993). Thus, the presence of large C, jumps in response to small amounts of Ca2+ influx is not caused by insufficient diffusion of the exogenous Ca2+ chelators. This is supported further by the presence of both docked and "not-docked" kinetic responses within a single trace (Fig. L4). In several experiments, the first depolarizing train was delivered as much as 10 min after break-in; large-amplitude C, jumps still were obtained. In Figure 2, the AC,,, response per pulse is plotted for the first and second stimulus trains, which were given at 10 min ( Fig. 2A) and 13 min ( Fig. 2B) after establishing whole-cell recording. Large-amplitude jumps in response to small amounts of Ca2+ entry still are seen during the first train, but not in response to the second train. It appears that within the time frame of -10 min of whole-cell dialysis, docked vesicles do not "undock." However, the ability to replenish this secretory mode or vesicular pool is lost within a few minutes of whole-cell dialysis.
In contrast to the rapid disappearance of the docked secretory pattern in whole-cell recordings, the second kinetically distinct secretory pattern was relatively stable for prolonged recording and perfusion periods. Figure 3 illustrates a method for estimating the decline of secretory robustness over time. The amount of total AC,lCCa2+ for each train was measured in five cells that were stimulated every 3 min for up to 1 hr. These values were normalized to either the second or the third stimulating train, whichever train was the first that did not exhibit docked secretion. Secretory robustness declined monoexponentially with a time constant of 26 min.  F@ve 2. Docked secretory kinetics is observed during the first stimulus train even after prolonged cell dialysis. A, B, Plots of individual C,, jump amplitudes versus cumulative Ca2+ entry as in Figure 1. A, The first stimulus train was given -10 min after establishing the whole-cell configuration.
Prominent, large C, jumps are elicited by the first five depolarizations. B, The second stimulus train was given 3 min later, and all small C,,, responses are <20 fF. Cell B0585; 0.5 mM [BAPTA],.

Ca2+ dependence of docked versus threshold secretion
Docked secretion Because the large-amplitude C,,, jumps elicited at the beginning of a train were seen only during the first and/or second stimulus train of whole-cell recording experiments, we compared the Ca2+ requirements of the C, response to the first depolarizing pulse (C,,) of the first and third trains of a population of cells. Amplitude histograms for Cml jumps elicited by the first (Fig. 4A) and third (Fig. 4B) trains are shown for 37 cells dialyzed with 0.3 mM BAPTA. The average C,, of the first train was 69 -C 9 fF (mean 2 SEM), with 61% of responses >50 fF. The average C,, of the third train for the same cells is 23 -C 3 fF, with only 8% of responses >50 fF.
The relationship between C,, jump amplitude and amount of Ca2+ entry during the first depolarizing pulse is illustrated in Figure 4C. C,, amplitudes for the first (solid squares) and third During the first stimulating train, large C,, jumps averaging -70 fF are elicited by <5 X lo7 CaZt ions and pulse durations as short as 6 msec. Pulses with larger amounts of Cazt entry produced similar C,, amplitudes, suggesting that very small amounts of Ca2+ are sufficient to elicit maximal C,, responses for a single depolarizing pulse. During the third train, the same number of Cazt ions elicited maximal C,, responses of -20 fF amplitude. The much larger vertical error bars for C,, responses during the first train reflect the presence of both small and large amplitude jumps as seen in Figure 4A, whereas the tighter error bars of the C,, responses during the third train are attributable to the absence of large jumps (Fig. 4B).
A characteristic feature of the docked secretory mode is that the first pulse of a train almost always elicited the largest C, jump. The next several pulses of a stimulus train elicited progressively smaller responses, as seen in Figures 1A and 2A, giving the appearance of a limited vesicular pool that could be released by this secretory mode. The total amount of docked secretion in a single cell usually consisted of 200-500 fF; small C,,, jumps seen later in the train (Figs. ZAi, 2A) probably reflect the activation of threshold secretion.

Threshold secretion
During threshold secretion, the earliest pulses of a train elicit little or no C,, increases and, if pulse durations are brief, C, jumps late in the train are small and relatively uniform. Typical examples are seen in Figures 1B and 2B.
The relationship between AC, changes and Cazt entry during threshold secretory kinetics is illustrated in the experiment shown in Figure 5. The cell was stimulated with three stimulus trains that were identical except for pulse durations. The first and last Cazt current records for each stimulus train are shown superimposed in least total secretion (Fig. S&i), whereas the trains of 20 and 40 msec durations gave similar total C, responses. It is evident in Figure SAii-Cii that C, increases begin at different times in the train (second awow). Shorter pulse durations result in a greater number of pulses that fail to elicit detectable C, increases. Thus, for the 10 msec pulse train there is no response until about the 14th or 15th stimulus pulse, or -3.5 set after the first depolarizing stimulus, whereas for the 40 msec pulse train the response begins on the 5th or 6th stimulus pulse, or at -1.5 sec. C, changes elicited by trains with different stimulus parameters are consistently related to cumulative Ca2+ entry as shown in Figure Siii. It is apparent by cbmparing the three plots that the effect of Cazt entry during a stimulus train can be divided into two discrete phases. During each train, C, increases do not begin until after a critical amount of summed Ca*+ entry, which we call threshold. The lines drawn through the cumulative C, points during the second secretory phase are fit by linear regression, and the threshold value is taken as the intersection with the ordinate. In this experiment with 0.3 mM BAPTA, the threshold value is -20 X lo7 Ca2+ ions. Threshold secretory kinetics consists of two Ca*+-sensitive phases even when the amount of Ca2+ entry during a single pulse is very large. Figure 6A is a plot of cumulative C, increases versus cumulative Ca2+ entry from an experiment with 0.5 mM EGTA in the pipette in which the duration of depolarizing pulses ranged between 46 and 393 msec. During each train, C, increases did not begin until -55 X lo7 Ca2+ ions had entered, even though the number of ions during the first pulse varied from -6.7 to 40.2 X 10' for the shortest and longest pulse durations, respectively. Also, during the secretory phase C, increases are related to the amount of cumulative Ca*+ entry. For longer pulse durations, e.g., >72 msec, each depolarization elicits a relatively large am- plitude C, jump that is proportional to the amount of Ca*+ entry during that pulse.
The relationship between cumulative C, increase and cumulative Ca2+ entry during threshold secretory kinetics also is observed when Ca2+ influx per pulse is varied by protocols other than changing pulse durations. Figure 6B contains data from a cell in which [Ca'+], was increased from 5 to 10 mM. The pulse protocols were identical except for the number of pulses (10 pulses in 10 mM [Ca2+10, 20 pulses in 5 mM [Ca'+]J that were required to reach approximately the same amount of total Ca2+ entry. Under these conditions, the Cazt flux per pulse doubles at the higher [Ca'+],. Despite the differences in train duration (2.5 vs 5 set) and Ca2+ flux per pulse, the Ca'+-AC, dependence of the two traces coincides almost exactly. Thus, Ca2+-secretion coupling during threshold secretion is a function of cumulative Ca2+ entry rather than of any specific parameter of Ca2+ current amplitude, current duration, or the time or number of pulses required to reach a certain value of total Ca2+. Effects of Ca*+ chelators on docked versus threshold secretion In NHP terminals, low concentrations of Ca*+ chelators strongly influenced the secretory response by controlling the amount of Ca2+ entry necessary to reach threshold. We tested the effect of two chelators, EGTA and BAPTA, on the threshold secretory response, and we tested two concentrations of BAPTA on the docked secretory response.
Thresholds were estimated from the x-axis intercept of a line fitted by linear regression to the secretory phase in plots of cumulative C, changes versus cumulative Ca2+, as illustrated in Figure 5.  Figure 7A. The results presented in the on the average C,, jump amplitude. The inset in Figure 7A graph were obtained over a broad range of stimulus paradigms, illustrates the differences between Cazt requirements for elicincluding various pulse durations and current amplitudes. As in iting 70 fF of docked secretion versus initiating the secretory NHP terminals, low concentrations of buffers increased the phase in threshold secretion. The values for docked secretion threshold value for Cazt entry. EGTA, a chelator with slow represent the average Ca '+ ions required to elicit C,, in 0.1 binding kinetics, and BAPTA, a chelator with much faster binding mM BAPTA (n = 15) as well as the average of the first binned kinetics, had similar effects on threshold.
subset from Figure 4C (n = 25). Clearly, large jumps during We compared the effects of two concentrations of BAPTA on docked secretion occur in response to much less Ca2+ entry docked secretion. The data in Figure 7B Figure 4A.
transmission to Ca2+ chelators and indicates that the vesicles Lowering the BAPTA concentration by one-third had no effect are in relatively close vicinity to Ca2+ channels. To determine whether the threshold was the point at which the Ca2+ chelator was saturated, we monitored averaged cytosolic Ca2+ levels by including 0.2 mM Fura-red in the presence of 10 mM EGTA. Fura-red is chemically derived from BAPTA and has faster binding kinetics than EGTA but similar Cazt affinity as EGTA. During trains of pulses lasting several seconds, Ca2+ first binds to Fura-red but then quickly equilibrates with the much larger capacity provided by the EGTA. A recording of the C, trace (Fig. 8A) is compared with the fluorescent signal of the Ca2+ dye (Fig. 8B). C, increases begin well below the maximum Ca2+ signal, indicating that exogenous Ca2+ chelators are not saturated at the threshold level.

DISCUSSION
Chromaffin cells secrete catecholamines in response to bursts of APs that elicit complex pulsatile patterns of Ca2+ entry and diffusion. We have used trains of depolarizing pulses to examine the secretory response of dialyzed bovine chromaffin cells. In this study, we describe two kinetically distinct modes of Ca2+secretion coupling. The modes differ in Cazt dependence and in their sensitivity to dialysis in the whole-cell patch-clamp configuration.

Secretory run-down in chromaffin cells
Secretory responsiveness decays rapidly in permeabilized chromaffin cells compared with intact controls. The rate of decay, as well as the properties of the remaining secretory response, depends on the diameter of holes produced by the permeabilizing agent and the composition of the permeabilization medium (for review, see Sontag et al., 1991). The whole-cell patch-clamp recording mode can be viewed as another variant of permeabilization techniques in which the cell is dialyzed with a specific pipette solution through a single 0.5-to 2-pm-diameter hole. Threshold secretion declined slowly with an average time constant of 26 min. The slow, but still significant, run-down of threshold secretion may reflect the disappearance of or decline in critical secretory components during whole-cell recordings. The run-down described here differs from that in a previous study in which the secretory response was reported to "wash out" with a time constant of 1-4 min (see Fig. 8 in Augustine and Neher, 1992). Because the pipette solutions are similar in the two studies, the differences may arise from stimulus paradigms. In the previous study, lbng depolarizations (200-500 msec) were administered every 100 sec. With similar protocols, we also observed diminished secretory responses and less reproducibility (data not shown). In the present study, trains consisted of brief depolarizing pulses (5-100 msec) separated by 3-S min. The longest pulse durations were used only at high Ca*+ chelator concentrations.
Frequent and prolonged elevation of [Ca2+li to high levels may damage the secretory capability of chromaffin cells and obscure the time course of actual secretory run-down.
In contrast to threshold secretion, the second type of secretory kinetics is extremely labile and is observed only in response to the first and/or second stimulus train during whole-cell dialysis. The was 5.97 X lo7 2 9.14 (SEM). rapid loss of this kinetic component may reflect a requirement for some small metabolite or protein to promote this state. However, possible alternative explanations are discussed below.

Characteristics of threshold secretion
The properties of threshold secretion in chromaffin cells are similar to those observed in peptide-secreting NHP terminals (Lim et al., 1990;Seward et al., 1995). In both preparations, threshold secretion consists of two Ca*+-sensitive phases, the first of which is preparatory in nature and does not cause secretion. Exocytosis begins only after a critical threshold amount of Cazf entry that is governed by the concentration, but not species, of Ca*+ chelator. In the second phase, the final exocytotic step(s) appears to require elevated submembrane Ca*+, because most C, increases occur only when Ca*+ channels are open. The species of Ca*+ chelator, i.e., EGTA versus BAPTA, affects the secretory phase to a greater extent than the threshold phase (Seward et al., 1995). Thus, the two phases of threshold secretion appear to reflect two discrete Cazf-dependent steps within a single secretory mode. Finally, the critical parameter regulating both phases is cumulative Ca*+ entry during a stimulus train, rather than specific details of Ca*+ entry such as current amplitude or single-channel flux rates, which would produce different spatial and temporal gradients of submembrane Ca*+ concentration (Sala and Hernandez-Cruz, 1990;Nowycky and Pinter, 1993;Seward et al., 1995). Secretory responses for which kinetics resembles threshold secretion appear in all studies of bovine chromaffin cells stimulated with trains of depolarizing pulses (Augustine and Neher, 1992;von Ruden and Neher, 1993;Artalejo et al., 1994). Threshold-like secretion also is observed in melanotrophs (Thomas et al., 1990) and pancreatic p cells (Ammala et al., 1993). In pancreatic p cells, the Ca*+-dependent threshold occurs when average cytosolic [Ca*+], is 0.5 PM. As in NHP terminals and chromaffin cells, exocytosis is related to cumulative Ca*+ entry regardless of pulse protocol (Ammala et al., 1993). Threshold-like secretory behavior has been described in chicken II luteinizing hormone-releasing hormone (LHRH)-secreting sympathetic preganglionic fibers (Peng and Zucker, 1993) and may be a universal mechanism for secretion of LDCV.

Characteristics
of docked secretion In chromaffin cells, an additional form of Ca'+-secretion coupling is observed with strikingly different Ca2+ dependence from threshold secretion. In this secretory mode, vesicles are released in response to amounts of Ca2+ entry well below threshold and are relatively insensitive to different concentrations of Ca2+ chelators. We use the term "docked" as both analogous with fast synapses where, presumably because of the close proximity of vesicle fusion sites to Ca2+ channels, Ca2+ chelators have only weak effects on stimulus-induced secretion (Adler et al., 1991;von Gersdorff and Matthews, 1994), and following the nomenclature of Thomas et al. (1993) to describe the most rapid phase of release in melanotrophs.
Because of its labile nature, docked secretion is missed easily if the cell is depolarized before initiating capacitance measurements, and it has not been reported previously in studies of bovine chromaffin cells stimulated with depolarizing pulses. A kinetic response that resembles docked secretion was reported in pancreatic p cells (see Figs. 5, 6, 8 in Ammala et al., 1993) but it was not studied further.
Two types of secretory patterns have been reported in rat adrenal chromaffin cells (Horrigan and Bookman, 1994). These patterns have been ascribed to the existence of two vesicular pools: an immediately releasable pool (IRP) of -17 secretory vesicles of 35 fF, and a readily releasable pool (RRP), which is distinct and larger, consisting of -170 vesicles. The kinetics of exocytosis of IRP in some respects resembles docked secretion in that it proceeds at lower amounts of Ca2+ entry and is less sensitive to buffers than the release of RRP. However, the correspondence between the IRP release in rat cells and the docked secretion in bovine chromaffin cells is not exact. Differences include (1) the persistence of IRP exocytosis during whole-cell recordings in rat, (2) the smaller size of the IRP pool in rat cells, and (3) the tight correlation between pulse duration and C, responses during IRP secretion.
In experiments with flash photolysis of caged Ca2+ compounds, a series of kinetic responses is reported for both melanotrophs and bovine chromaffin cells (Neher and Zucker, 1993;Thomas et al., 1993;Heinemann et al., 1994). The earliest component is called "docked secretion" (Thomas et al., 1993) or secretion of a RRP of vesicles (also called "ultrafast";Neher and Zucker, 1993;Heinemann et al., 1994). This component may be the same as the docked secretion described here. However, it is difficult to be certain, because the initial flow of caged Ca2+ compound into the cell produces a Ca2+ transient as intracellular Mg2+ displaces Ca*+ from the DM-nitrophen.
It is likely that the large C, responses evoked by the Ca2+ transient (>lOOO fF in bovine chromaffin cells) (Neher and Zucker, 1993;Heinemann et al., 1994) include a majority of the labile docked pool. and perhaps we have not detected it because docked vesicles are released immediately.
The complexity of kinetic components presented here implies that a single adrenal chromaffin cell does not respond identically to electrical activity at all times. In the docked mode, a relatively rapid and reliable response might be elicited by a single or a few APs. In threshold mode, on the other hand, secretion would occur only during bursts of APs after a critical number of APs triggers this mode. The rules that govern the presence of one or the other secretory mode remain to be unraveled.