InsP3-mediated puffs are fundamental building blocks of cellular Ca2+ signalling, and arise through the concerted opening of clustered InsP3Rs (InsP3 receptors) co-ordinated via Ca2+-induced Ca2+ release. Although the Ca2+ dependency of InsP3Rs has been extensively studied at the single channel level, little is known as to how changes in basal cytosolic [Ca2+] would alter the dynamics of InsP3-evoked Ca2+ signals in intact cells. To explore this question, we expressed Ca2+-permeable channels (nicotinic acetylcholine receptors) in the plasma membrane of voltage-clamped Xenopus oocytes to regulate cytosolic [Ca2+] by changing the electrochemical gradient for extracellular Ca2+ entry, and imaged Ca2+ liberation evoked by photolysis of caged InsP3. Elevation of basal cytosolic [Ca2+] strongly increased the amplitude and shortened the latency of global Ca2+ waves. In oocytes loaded with EGTA to localize Ca2+ signals, the number of sites at which puffs were observed and the frequency and latency of puffs were strongly dependent on cytosolic [Ca2+], whereas puff amplitudes were only weakly affected. The results of the present study indicate that basal cytosolic [Ca2+] strongly affects the triggering of puffs, but has less of an effect on puffs once they have been initiated.
The InsP3R (InsP3 receptor) is a Ca2+-permeable channel expressed in the ER (endoplasmic reticulum) which is gated by the binding of the second messenger InsP3 and by cytosolic Ca2+ itself [1–7]. Ca2+ liberation occurs at discrete functional release sites, formed by clusters of InsP3R on the ER membrane. These participate in generating a hierarchy of cellular Ca2+ signals involving opening of single InsP3R channels [8,9], concerted release from several channels within a cluster  and global Ca2+ waves that propagate from cluster to cluster [11,12]. The positive-feedback mechanism of CICR (Ca2+-induced Ca2+ release), by which Ca2+ released from one InsP3R channel promotes opening of neighbouring channels, underlies these processes, and factors including cytosolic Ca2+ buffering, [InsP3] and basal cytosolic [Ca2+] determine the transition between local and global signalling patterns.
The role of cytosolic Ca2+ in modulating InsP3R channel gating has been extensively studied by single-channel recordings from InsP3R in excised nuclei and after reconstitution in lipid bilayers, revealing the well-known ‘bell-shaped’ curve of Ca2+ facilitation and inhibition [1,2,13,14]. However, less is known of how cytosolic Ca2+ modulates local InsP3-mediated signals in the intact cell, although imaging studies in Xenopus oocytes demonstrate a profound potentiation of global Ca2+ waves [13,15–17].
In the present study, we expressed Ca2+-permeable nAChRs (nicotinic acetylcholine receptor/channels) in the plasma membrane of Xenopus oocytes so as to experimentally regulate the basal cytosolic [Ca2+] concentration , and examined how elevations in cytosolic [Ca2+] affected the dynamics of local and global Ca2+ signals evoked by photoreleased InsP3. We show that an increased probability of triggering local Ca2+ release at puff sites underlies the strong augmentation of global InsP3-mediated Ca2+ waves, whereas puff amplitudes and durations were unaffected.
Oocyte preparation and expression of nAChRs
Xenopus laevis were purchased from Nasco International and the oocytes were surgically removed  following protocols approved by the UC Irvine Institutional Animal Care and Use committee. Stage V–VI oocytes were isolated and treated with collagenase (1 mg/ml collagenase type A1 for 30 min) to remove the follicular cell layers. At 1 day after isolation the oocytes were injected with a cRNA mixture for nAChR expression (α, β, γ and δ subunits at a ratio of 2:1:1:1; 50 nl at a final concentration of 0.1–1 mg/ml) and were then maintained in modified Barth's solution [88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3, 0.82 mM MgSO4, 0.33 mM Ca(NO3)2, 0.41 mM CaCl2, 5 mM Hepes and 1 mg/ml gentamicin (pH 7.4)] for 1–3 days at 16°C before use. Expression of nAChRs was evaluated using a voltage clamp to measure the currents evoked by 500 nM ACh: oocytes showing currents >1 μA at −80 mV were selected for the experiments.
Microinjection of oocytes
Intracellular microinjections were performed using a Drummond microinjector. Approximately 1 h before the Ca2+ imaging experiments, oocytes in Ca2+-free Barth's solution were injected with Fluo-4 dextran (high affinity; Kd=800 nM) to a final concentration of 40 μM, assuming equal distribution throughout a cytosolic volume of 1 μl and with caged Ins(1,4,5)P3,P4(5)-[1-(2-nitrophenyl)ethyl]ester (final concentration 8 μM). EGTA (final concentration 300 μM) was further injected for puff studies.
Ca2+ imaging and flash photolysis
Oocytes were voltage-clamped using a conventional two-microelectrode technique. The membrane potential was held at 0 mV during superfusion with a non-desensitizing concentration of ACh (100–500 nM) in Ringer's solution and was briefly stepped to −120 mV to strongly increase the electrical driving force for Ca2+ influx . Global Ca2+ signals were imaged at room temperature (18–20°C) by a custom-build confocal line scanner  interfaced to an Olympus inverted microscope IX 70, and fluorescence excitation was provided by the 488 nm line of an argon ion laser, with the laser spot focused by a ×40 oil-immersion objective [NA (numerical aperture)=1.35] and scanned along at a rate of 10 ms/50 μm line. To image puffs, we used wide-field fluorescence microscopy using an Olympus IX 71 inverted microscope equipped with a ×40 oil-immersion objective, a 488 nm argon ion laser for fluorescence excitation and an electron-multiplied CCD (charge-coupled device) camera (Cascade 128+; Roper Scientific) for imaging fluorescence emission (510–600 nm) at a frame rate of 500 s−1. Fluorescence was imaged from a 40 μm×40 μm (128×128 pixel) region within the animal hemisphere of the oocyte. Fluorescence measurements made by line-scan and camera imaging are expressed as a ratio (ΔF/F0) of the mean change in fluorescence (ΔF) at a pixel relative to the resting fluorescence at that pixel before stimulation (F0). The mean values of Fo were obtained by averaging over several scans/frames before stimulation. To calibrate the changes in ΔF/F0 values in terms of nanomolar increases of free [Ca2+] we determined maximal (Fmax) and minimal (Fmin) fluorescence values by injecting fluo-4 dextran-loaded oocytes (n=5) with 30 nl of 100 mM CaCl2 or 100 mM EGTA from a micropipette located close to the imaging site. After correcting for oocyte autofluorescence, the mean values were Fmax=8.52±1.16 and Fmin=0.857±0.024 relative to the resting fluorescence F0 before injection. We assumed a Kd value for fluo-4 dextran of 2400 nM, based on measurements of ~800 nM in free solution  and a roughly 3-fold reduction in affinity in the cytoplasmic environment . A fluorescence increase of ΔF/Fo=1 above baseline would then correspond to an increase of [Ca2+]cyt of about 360 nM. Photolysis of caged InsP3 was evoked by flashes of UV light (350–400 nm) from a mercury arc lamp, delivered through the microscope objective and adjusted to uniformly irradiate a circular region slightly larger than the imaging frame or scan line. Flash durations were set using a Uniblitz shutter and digital controller.
Fluo-4 dextran (high affinity; Kd ~800 nM), and caged InsP3 were purchased from Invitrogen. All other reagents were from Sigma–Aldrich.
Custom routines written in the IDL programming environment (Research Systems) were used for linescan image processing and measurements. MetaMorph (Molecular Devices) was used to process and measure data obtained from wide-field camera-based imaging. Further analysis and graphing was accomplished using Microcal Origin version 6.0 (OriginLab). Results are means±S.E.M. and significance was assessed by Student's t test.
Elevated basal cytosolic [Ca2+] enhances InsP3-evoked Ca2+ waves
In order to evoke cytosolic [Ca2+] elevations, Ca2+ influx was induced through nAChRs expressed in the oocyte plasma membrane. Oocytes were continuously superfused with Ringer's solution containing 1.8 mM Ca2+, together with a low non-desensitizing concentration (100–500 nM) of acetylcholine, and were voltage-clamped to control the electrochemical gradient for Ca2+ entry. The membrane potential was held at 0 mV to minimize Ca2+ influx, and was then stepped to more negative values to promote Ca2+ influx beginning 2.5 s before the delivery of a UV flash to photorelease InsP3 from a caged precursor loaded into the oocyte (Figure 1A). The resulting changes in fluo-4 fluorescence were imaged to compare InsP3-evoked Ca2+ responses evoked by identical UV flashes during cytosolic [Ca2+] elevation with control records when the voltage pulse was not applied.
Elevated basal cytosolic [Ca2+] enhances InsP3-induced Ca2+ waves
We first examined global Ca2+ signals evoked in oocytes that were not loaded with EGTA. The panels on the left-hand side of Figure 1(B) show representative linescan images of fluorescence changes evoked by photoreleased InsP3 without (upper panel) and with (lower panel) Ca2+ influx; the corresponding fluorescence profiles are presented on the right-hand side. We compared the latencies and peak amplitude of InsP3-evoked Ca2+ signals under resting cytosolic [Ca2+], and during Ca2+ influx that increased the basal fluorescence signal by a mean of 0.60±0.06 ΔF/Fo (six oocytes from three different frogs). Latencies (time from the UV flash to the initial rise in fluorescence) were significantly shorter during cytosolic [Ca2+] elevation (Figure 1C; control latency=284±14 ms and during Ca2+ influx latency=177±30 ms; P<0.05). The mean peak amplitude of Ca2+ waves was profoundly augmented by cytosolic Ca2+ elevation (Figure 1D; control ΔF/Fo=1.68±0.23 and during Ca2+ influx ΔF/F0=3.50±0.45; P<0.05, n=6). These results are consistent with previous observations showing facilitation of InsP3-evoked Ca2+ signals by cytosolic [Ca2+] [1–7].
Elevated basal cytosolic [Ca2+] promotes InsP3-evoked Ca2+ puffs
We next examined the effects of basal cytosolic [Ca2+] elevations on local Ca2+ puffs. For this purpose oocytes were loaded with EGTA (final intracellular concentration 300 μM) to suppress generation of Ca2+ waves by inhibiting inter-cluster diffusion of Ca2+ ions . We further employed wide-field fluorescence microscopy to image a 40μm×40 μm field of view with a fast (500 frames per s) electron-multiplied CCD camera, so as to sample many more puff sites than possible by one-dimensional linescan imaging. Figure 2 shows the experimental protocol (Figure 2A, a), and representative fluorescence traces monitored from small regions of interest centred on puff sites illustrating the responses evoked by photoreleased InsP3 at resting cytosolic Ca2+ (Figure 2A, b) and when Ca2+ was elevated by Ca2+ influx (Figure 2A, c and d). The photolysis flash was delivered 4 s after the onset of the hyperpolarizing pulse so as to allow cytosolic [Ca2+] to equilibrate, and puffs were then recorded for 6 s while the hyperpolarization was maintained. We varied the duration of the photolysis flash to evoke differing numbers of puffs at resting cytosolic [Ca2+]; ‘weak’ flashes (25–50 ms) were chosen to evoke on average about a single puff in the entire imaging field 40 μm×40 μm, and ‘strong’ flashes (50–100 ms) to evoke up to four puffs. Even the ‘strong’ stimulus was chosen to evoke responses well below the maximal, so as to avoid possible saturation effects when responses were further potentiated by Ca2+ influx.
Cytosolic [Ca2+]-dependent potentiation of InsP3-evoked Ca2+ puffs
Figure 2(B) shows a scatter plot of the relationship between the numbers of individual sites in the imaging field where puffs were observed during 6 s following photorelease of InsP3 as a function of the elevation of cytosolic [Ca2+] evoked by hyperpolarizing pulses. We express the [Ca2+] elevation in terms of fluorescence ratio change, without correction for oocyte auto-fluorescence (about 50% of resting fluo-4 fluorescence). On the basis of the calibration described in the Experimental section, an increase in ΔF/Fo of 0.1 corresponds to a rise in [Ca2+] of about 36 nM. With both weak and strong photolysis flashes the number of responding puff sites increased steeply with increasing basal cytosolic [Ca2+], with strong flashes giving greater numbers at any given basal [Ca2+]. Figure 2(C) shows mean data, grouped according to flash duration and whether basal cytosolic levels just before the photolysis flash were at or close to the resting level (ΔF/F0=0–0.1) or were appreciably elevated (ΔF/Fo>0.1).
Elevated cytosolic [Ca2+] shortens puff latency
Figures 3(A) and 3(B) show scatter plots of individual and mean latencies of puffs, grouped according to the cytosolic [Ca2+] elevation at the time of the photolysis flash. Puffs evoked by weak stimuli arose with relatively long (2–3 s) latencies, which tended to shorten with increasing cytosolic [Ca2+], but did not show a statistically significant correlation (Figure 3A). On the other hand, mean puff latencies were shorter with stronger photorelease of InsP3 (Figure 3B) and showed a marked dependence on cytosolic [Ca2+], reducing from 2133±200 ms at near resting level (ΔF/F0<0.1) to 1240±174 ms when the fluorescence was elevated to >0.1 ΔF/F0 during Ca2+ influx (P<0.01). Puff latencies followed roughly exponential distributions at both relatively low and high cytosolic [Ca2+] (Figures 3C and 3D respectively), with a markedly shorter time constant at higher [Ca2+].
Puff latencies shorten with increasing cytosolic [Ca2+]
Puff amplitudes are only weakly dependent on cytosolic [Ca2+]
Next, we analysed the effects of changes in cytosolic [Ca2+] on puff amplitudes. After pooling data across all different basal cytosolic [Ca2+] levels we found no significant difference in mean puff amplitudes evoked by weak or strong photorelease of InsP3 (Figure 4A; weak flash mean puff amplitude ΔF/Fo=0.43±0.04, n=56 and strong flash ΔF/Fo=0.42±0.03, n=155, P>0.05). Looking then at the effect of elevating cytosolic [Ca2+] levels, we observed little or no effect on the amplitudes of puffs evoked by weak photorelease (Figure 4B). On the other hand, puffs evoked by strong photorelease of InsP3 showed a significant increase in puff amplitude with higher elevations of cytosolic [Ca2+] (ΔF/Fo>0.2) (Figure 4C).
The amplitude of InsP3-evoked puffs is only weakly dependent on the basal cytosolic [Ca2+]
Puff durations are independent of basal cytosolic [Ca2+]
We had observed previously a prolongation of puff duration when puffs were evoked after loading ER Ca2+ stores by inducing a prior Ca2+ influx in oocytes transfected to overexpress SERCA (sarcoplasmic/ER Ca2+-ATPase), but not in control (non-expressing) oocytes. We now examined the effect of elevated [Ca2+]cyt on puff duration. Puffs evoked by strong photorelease of InsP3 were compared in the same imaging field at basal [Ca2+]cyt and during induction of Ca2+ influx. Figure 5(A) shows a scatter plot of durations of puffs [measured as FDHM (full duration at half-maximal amplitude)] against the latency of the puffs following the UV flash. No differences were apparent in puff durations between the control and Ca2+ influx records, and puff durations did not show any obvious systematic dependence on latency following the UV flash. Figure 5(B) further plots mean values of FDHM of control puffs and puffs during Ca2+ influx, showing no significant difference (P=0.64).
Duration of InsP3-evoked puffs is independent on the basal cytosolic [Ca2+]
The aim of the present study was to investigate how elevated basal cytosolic [Ca2+] would affect InsP3-evoked Ca2+ signals. We utilized the expression of Ca2+-permeable nicotinic receptor/channels in the plasma membrane as a means to evoke controlled entry of extracellular Ca2+ into the cell during hyperpolarizing voltage-clamped pulses. Consistent with previous observations [4,5], we confirmed that cytosolic [Ca2+] elevations powerfully facilitated InsP3-mediated Ca2+ waves in terms of increased peak amplitude and shortened latency (Figures 1A and 1B). We further investigated the effect of elevated basal [Ca2+]cyt on the local InsP3-mediated Ca2+ puffs that are the triggers and fundamental building blocks of Ca2+ waves, as well as serving signalling functions in their own right. Our results show that the numbers of puff sites that respond at a given [InsP3] are strongly potentiated in a graded manner with increasing [Ca2+]cyt, and that the mean latency of puffs markedly shortens. In contrast, puff amplitudes were little affected except at high [Ca2+]cyt and we observed no significant effects of [Ca2+]cyt on puff duration.
The effects we describe on InsP3-evoked Ca2+ liberation from the ER can be directly attributed to changes in basal [Ca2+]cyt, and not to any increase in Ca2+ store filling within the ER. We had previously utilized Ca2+ influx through nicotinic receptors as a means to increase ER Ca2+ loading, by applying a transient hyperpolarizing pulse and then allowing [Ca2+]cyt to subside to the resting level before examining responses to photoreleased InsP3. However, changes in puff properties were observed only when SERCA activity was accelerated by cADP ribose [20,24] or when SERCA 2b was overexpressed . With basal SERCA activity, no significant changes in puff triggering, kinetics or amplitude were apparent following even strong Ca2+ influx.
We have proposed that the puff is itself triggered by the stochastic opening of a single InsP3R channel within the cluster [25,26]. Factors that determine the occurrence of puffs thus include the number of channels present in the cluster and the open probability of each channel. The latter, in turn, is a function of the concentrations of InsP3 and Ca2+, acting as co-agonists to open the channel [3,7,27]. Concordant with this mechanism, increasing [InsP3] results in an increased frequency of puffs and a shortening of the latency to the first puff evoked at a site following photorelease of InsP3 [26,28]). Similarly, modest elevations of [Ca2+]cyt will increase the open channel probability at a given [InsP3], and hence increase the probability of puff triggering leading to a greater number of sites that generate puffs following photorelease of InsP3 and a shortening in mean latency of these puffs. Although gating of the InsP3R channel is biphasically regulated by [Ca2+], inhibition of the native Xenopus InsP3R arises only when [Ca2+] exceeds a concentration of several hundred micromolar  and thus would not be expected to be apparent in our experiments, where we estimate that the maximal Ca2+ influx (ΔF/Fo ~0.3) corresponded to an increase in [Ca2+]cyt of <100 nM.
Because the resting [Ca2+]cyt is very low, small elevations above this level will strongly potentiate puff triggering. On the other hand, once an initial ‘trigger’ channel opens, the Ca2+ flux passing through it will elevate the local free [Ca2+] at the puff site to much higher levels, predicted to reach a concentration of a few hundred micromolar at the mouth of the open channel and at least several micromolar at the neighbouring InsP3R within the cluster . This will effectively ‘swamp’ the effect of any smaller elevation of basal [Ca2+]. Once triggered, the puff thus becomes a self-regenerative process and its subsequent evolution is expected to be substantially independent of the preceding conditions; this is the probable explanation as to why we found little dependence of puff amplitudes and kinetics on basal [Ca2+]cyt.
The sensitization of global Ca2+ waves by elevated basal [Ca2+]cyt may similarly be explained by enhanced coupling between neighbouring release sites. Ca2+ waves propagate because Ca2+ released from one site diffuses to evoke CICR from adjacent sites [11,12], and this triggering will be facilitated if [Ca2+]cyt is already elevated. The results in Figure 1 were obtained using relatively weak photorelease of InsP3 that evoked only abortive Ca2+ waves, and basal [Ca2+] elevation promoted a more robust propagation by CICR resulting in strong potentiation of the spatially averaged Ca2+ signal. With stronger stimulation by InsP3 the amplitude of repetitive Ca2+ waves is not potentiated by Ca2+ influx , presumably because the more substantial Ca2+ release through InsP3R swamps any effect of the elevated basal [Ca2+], but wave velocities and frequency of repetitive spikes are increased .
InsP3-mediated Ca2+ signalling can function as a coincidence detector, whereby release of Ca2+ from intracellular stores is potentiated by extracellular Ca2+ entering through plasmalemmal ligand- or voltage-operated channels. This interaction may arise through two different mechanisms, operating on different timescales. Most directly, as we describe in the present paper, elevation of basal [Ca2+]cyt enhances the probability of triggering local and global [Ca2+] signals by binding to activating sites on the cytosolic face of the InsP3R. In addition, we have described a more circuitous mechanism, whereby extracellular [Ca2+] entering the cytosol is taken up by the action of SERCA pumps, leading to enhanced filling of ER Ca2+ stores . That, in turn, promotes Ca2+ puffs and waves, probably because increased Ca2+ flux through the InsP3R channel enhances CICR via the cytosolic activating sites on the InsP3R, and possibly also through luminal regulation of InsP3R function [30–32]. The direct action of Ca2+ influx on InsP3R is immediate and short lasting, depending on clearance rate from the cytosol. In contrast, potentiation via ER store filling is slower to develop, more persistent and subject to potential modulation by other messenger pathways, such as cADPR, that affect SERCA activity either directly or indirectly [20,33,34]. Interactions between these different modulatory mechanisms are likely to be of particular importance for Ca2+ signalling in neurons with regard to activity-dependent synaptic plasticity as well as gene expression and protein synthesis [35–37].
Michiko Yamasaki-Mann designed experiments, performed research, analysed the data and wrote the paper. Angelo Demuro performed research. Ian Parker designed experiments and wrote the paper.
This work was supported by the National Institutes of Health [grant number GM048071].