RCAN1 regulates vesicle recycling and quantal release kinetics via effects on calcineurin activity

Authors


Address correspondence and reprint requests to Dr Damien J. Keating, Department of Human Physiology, Flinders University, Sturt Rd, Bedford Park, Adelaide, South Australia 5042, Australia. E-mail: damien.keating@flinders.edu.au

Abstract

We have previously shown that Regulator of Calcineurin 1 (RCAN1) regulates multiple stages of vesicle exocytosis. However, the mechanisms by which RCAN1 affects secretory vesicle exocytosis and quantal release kinetics remain unknown. Here, we use carbon fibre amperometry to detect exocytosis from chromaffin cells and identify these underlying mechanisms. We observe reduced exocytosis with repeated stimulations in chromaffin cells over-expressing RCAN1 (RCAN1ox), but not in wild-type (WT) cells, indicating a negative effect of RCAN1 on vesicle recycling and endocytosis. Acute exposure to calcineurin inhibitors, cyclosporine A and FK-506, replicates this effect in WT cells but has no additional effect in RCAN1ox cells. When we chronically expose WT cells to cyclosporine A and FK-506 we find that catecholamine release per vesicle and pre-spike foot (PSF) signal parameters are decreased, similar to that in RCAN1ox cells. Inhibiting calcineurin activity in RCAN1ox cells has no additional effect on the amount of catecholamine release per vesicle but further reduces PSF signal parameters. Although electron microscopy studies indicate these changes are not because of altered vesicle number or distribution in RCAN1ox cells, the smaller vesicle and dense core size we observe in RCAN1ox cells may underlie the reduced quantal release in these cells. Thus, our results indicate that RCAN1 most likely affects vesicle recycling and quantal release kinetics via the inhibition of calcineurin activity.

Abbreviations used
AD

Alzheimer's disease

DS

Down syndrome

FKBP

FK506 binding protein

PB

phosphate buffer

PSF

pre-spike foot

RCAN1

Regulator of Calcineurin 1

WT

wild type

Regulator of Calcineurin 1 (RCAN1) is a chromosome 21 gene that is over-expressed in Down syndrome (DS) and Alzheimer's disease (AD) brains (Fuentes et al. 2000; Ermak et al. 2001). Human DS and AD brains demonstrate neuronal changes including reduced noradrenaline signalling, basal forebrain cholinergic neuron degeneration and enlarged endosomes (Lai et al. 1999). Mouse models of both disorders demonstrate reduced synaptic transmission, long-term potentiation and altered learning and memory (Oddo et al. 2003; Siarey et al. 2005). Identifying proteins common to both DS and AD which have roles in cell signalling or neurotransmitter release may enhance our understanding of the mechanisms underlying the neuronal changes observed in these disorders.

We have previously identified RCAN1 as a novel regulator of quantal release kinetics (Keating et al. 2008). These experiments were undertaken in mouse adrenal chromaffin cells using carbon fibre amperometry, a method which measures the number of individual vesicles undergoing exocytosis and the amount released from each vesicle. In addition, exocytotic events in these cells are often associated with a pre-spike ‘foot’ (PSF) signal that represents release of catecholamine during the formation of a stable fusion pore. We previously illustrated that altering RCAN1 expression reduces the number of exocytotic events occurring in chromaffin cells (Keating et al. 2008). We also observe that increased levels of RCAN1 reduce the amount released from individual vesicles and decrease PSF signal duration (Keating et al. 2008). These effects do not appear to be because of changes in vesicle loading or size, alterations in Ca2+ entry or the size of the readily releasable pool of vesicles.

The most well-characterized role of RCAN1 is as an inhibitor of calcineurin activity (Fuentes et al. 2000; Kingsbury et al. 2000; Rothermel et al. 2000). There are several potential mechanisms by which RCAN1 might regulate exocytosis and quantal release kinetics via its effects on calcineurin activity. Reduced calcineurin activity caused by RCAN1 over-expression could alter the phosphorylation and activity of a number of calcineurin targets involved in exocytosis and vesicle recycling, such as Munc18 (Craig et al. 2003), dynamin 1, amphiphysin 1/2 and synaptojanin (Cousin and Robinson 2001). Such an explanation focusses on the effect that relatively acute alterations in calcineurin activity could have on these processes. Chronic alterations in calcineurin activity may also have transcriptional effects via the well-characterized calcineurin/Nuclear factor of activated T-cells (NFAT) signalling node, which regulates pre-synaptic neurotransmitter release (Freeman et al. 2011). RCAN1 may also affect exocytosis and quantal release kinetics via direct mechanisms that are independent of its effect on calcineurin activity.

Here, we identify the role of calcineurin activity associated with the regulation of exocytosis by RCAN1. Increased RCAN1 expression reduces vesicle endocytosis and recycling in a repeated stimulation protocol and this effect is caused by the inhibition of calcineurin in cells over-expressing RCAN1. Chronic (3 days in vitro) calcineurin inhibition in wild-type (WT) cells replicates the changes in quantal release kinetics observed in RCAN1ox cells and such calcineurin inhibition has no additional effect on the amount of catecholamine released per vesicle in RCAN1ox cells. This treatment in RCAN1ox cells does further affect PSF signal kinetics, indicating that further calcineurin inhibition than that occurring in RCAN1ox cells may be required to fully affect early fusion pore kinetics. Electron microscopy confirms that RCAN1 over-expression does not affect vesicle distribution but does reduce vesicle size and possibly catecholamine loading in RCAN1ox chromaffin cells. Therefore, increased RCAN1 expression may inhibit calcineurin activity to negatively regulate the amount of vesicle content released as well as impacting on vesicle endocytosis and recycling.

Materials and methods

Chromaffin cell culture

Adrenal glands were taken from 6 to 8-week-old male WT and RCAN1ox mice and chromaffin cells were prepared as previously (Keating et al. 2008) as approved by the Flinders University Animal Welfare Committee. The animals were bred at the Flinders University Animal House Facility. Briefly, the adrenal medulla was dissected out in cold Locke's buffer (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO3, 5.6 mM glucose, 5.0 mM HEPES, pH 7.4) and then incubated with collagenase type A (Roche, Penzburg, Germany) in Locke's buffer at a concentration of 3 mg/mL in a shaking bath at 37°C. The collagenase was diluted further in cold Locke's buffer, cells pelleted and re-suspended in cell culture medium [Dulbecco's modified Eagle medium (DMEM) supplemented with 100 units/mL penicillin and 100 mg/mL streptomycin (Invitrogen, Carlsbad, CA, USA)], 2 mM l-glutamine (Invitrogen) and 10% FCS (JRH Biosciences, Lenexa, KS, USA) and filtered through nylon mesh. Cells were pelleted, re-suspended in supplemented Dulbecco's modified Eagle medium and plated on 35 mm culture dishes and incubated at 37°C with 5% CO2. Cells were maintained in primary culture for 3 to 4 days prior to experiments.

Carbon fibre amperometry

Catecholamine release from single chromaffin cells was measured using amperometry (Maritzen et al. 2008; Yu et al. 2008). A carbon fibre electrode (ProCFE; Dagan Corporation, Minneapolis, MN, USA) was placed on a chromaffin cell and +800 mV applied to the electrode under voltage clamp conditions. Current because of catecholamine oxidation was recorded using an EPC-9 amplifier and Pulse software (HEKA Electronic, Lambrecht/Pfalz, Germany), sampled at 10 kHz and low-pass filtered at 1 kHz. For quantitative analysis, files were converted to Axon Binary Files (ABF Utility, version 2.1; Synaptosoft, Decatur, GA, USA) and secretory spikes analysed (Mini Analysis, version 6.0.1; Synaptosoft) for a period of 1960s from the start of stimulation. The standard bath solution (Krebs buffer) contained 140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 5 mM d-glucose, 10 mM HEPES, pH 7.4. High K+-containing solution was the same as the control bath solution except that 70 mM K+ replaced an equimolar amount of NaCl. All solutions were applied to cells using a gravity perfusion system, the outlet of which was placed within 500 μm from the recorded cell. All experiments were carried out at 22–24°C. Acute and chronic calcineurin inhibition experiments were conducted using solutions containing FK-506 (5 μM) and cyclosporine A (5 μM) (Sigma, Castle Hill, Australia). Acute treatments involved incubating cells with inhibitors in Krebs buffer for 5 min at 22°C prior to stimulations, which were then carried out in the presence of the inhibitors. Chronic inhibition treatment involved incubating chromaffin cells, prepared the previous day, in cell culture medium containing the inhibitors for 3 days at 37°C with 5% CO2. Stimulations were performed using Krebs buffers containing the inhibitors.

Amperometry data analysis

Data analysis was performed as previously described (Keating et al. 2008). Briefly, amperometric spikes were selected for analysis of event frequency if spike amplitude exceeded 10 pA and overlapping peaks were included. For kinetic analysis of spikes and PSF signals, only those events which exceeded 20 pA or were not overlapping with other spikes were included. Only pre-spike foot features longer than 1 ms and > 2.5 times the RMS noise of the baseline in that recording were analysed as foot signals.

Electron microscopy

Adrenal tissue was fixed and stained using a modified chromaffin reaction (Tranzer and Richards 1976). Adrenal glands were fixed for 6 h at 4°C in 0.1M phosphate buffer (PB) pH 7.2, containing 3% glutaraldehyde. The tissue was then transferred to 0.1 M PB, pH 7.2 and kept overnight at 4°C. The samples were washed three times in 0.1 M PB before being post-fixed in 1% osmium tetroxide in 0.1 M PB, pH 7.2 for 1 h at 22°C. After multiple washes the tissue was stained with 2% aqueous uranyl acetate for 30 min. Following a series of dehydration steps with ethanol solutions (50–100%) and propylene oxide, the tissue was incubated for 1 h in a 1 : 1 mixture of propylene oxide: Durcupan resin (Sigma; Sigma-Aldrich Pty Ltd, Castle Hill, Australia). The samples were then embedded in pure resin in capsules and polymerization was carried out for 48 h at 60°C. An ultramicrotome (RMC Mechanical Advance Ultramicrotome, HD Scientific Supplies, Wetherill Park, Australia) was used to cut thick sections (1 μm), which were stained with 1% toluidine blue in 1% borax for light microscopic examinations. Ultrathin sections (80–100 nm) were cut using a diamond knife (Diatome, Hatfield, PA, USA) and sections were mounted on single-slot copper grids which were coated with 0.6–0.8% Butvar solution in chloroform. The sections were then stained with Reynolds' lead citrate solution (Reynolds 1963) and allowed to dry overnight. The ultra-thin sections were observed under transmission electron microscopy (JEOL 1200-EX transmission electron microscope; JEOL, Akishima, Japan) and images were taken at 30 000 magnification. Ultrastructural images were captured with a Megaview3 camera using the ITEM interface program (Olympus, Munster, Germany). Higher resolution montages of selected cells were compiled using the Multiple Image Alignment module in ITEM.

Chromaffin granule counting and statistics

Images were analysed using ImageJ software (NIH, Bethedsa, MD, USA). The number and spatial distribution of large dense-core vesicles (LDCV)s were evaluated on micrographs covering the entire cell's cytoplasmic region. LDCVs were identified by their electron-dense core and circular membrane. To measure vesicle localization, 200 nm-wide concentric zones starting from the cell membrane were defined within each cell and the number of dense core vesicles within each region were counted and presented as a percentage of the total number of vesicles. Dense core and vesicle diameter were measured from 50 vesicles per cell. Representative regions within the cell were selected by overlaying a numbered grid on each electron micrograph. A random number generator was then used to identify the LDCV containing regions to be analysed. All electron micrographs were analysed using ImageJ image analysis software (NIH).

Statistical analysis

For amperometry whole spike analysis, as data are not parametrically distributed we obtain the median value in each recording and present the final data as the mean of these median values (Maritzen et al. 2008). This data are then analysed using a Student's unpaired t-test. For PSF analysis, we often do not obtain sufficient data points to use the median value from each recording. Therefore, for these parameters, we obtain the mean of each individual value and gauge statistical differences between the means of different groups using the Mann–Whitney test. All data are presented as mean ± SEM. For electron microscopy analysis, data are represented as the average number of LDCVs per unit area ± SEM. Data for each genotype were compared using a Student's unpaired t-test. In all statistical tests, significance was obtained at < 0.05.

Results

RCAN1 over-expression does not alter vesicle localization in chromaffin cells

We have previously demonstrated changes in exocytosis in RCAN1ox chromaffin cells (Keating et al. 2008). We therefore used electron microscopy to analyse whether changes in exocytosis in RCAN1ox cells may be because of alterations in vesicle number and localization. The ultrastructural appearance of WT (Fig. 1a) and RCAN1ox (Fig. 1b) chromaffin cells appears similar and we find that the average number of vesicles per unit area is not different between WT (= 7 cells) and RCAN1ox cells (= 9 cells, Fig. 1c). We also assessed the distribution of vesicles with regard to their distance from the plasma membrane in these cells. We find no differences in the average distance from the membrane between WT and RCAN1ox cells (Fig. 1d). These results indicate that altered exocytosis in RCAN1ox chromaffin cells is not because of altered vesicle number or distribution. We additionally measured vesicle size in these cells and observe a significant decrease both in mean vesicle diameter (Fig. 1e, < 0.001) and dense core diameter (Fig. 1f, < 0.001). This data indicate that catecholamine loading into vesicles is negatively affected in RCAN1ox chromaffin cells.

Figure 1.

Regulator of Calcineurin 1 (RCAN1) over-expression does not alter vesicle number or localization. A representative electron micrograph of an adrenal medulla section from (a) wild type (WT) and (b) RCAN1ox mice clearly shows the large dense core vesicles throughout each chromaffin cell. The red line indicates the plasma membrane of a single chromaffin cell. (c) The average number of large dense core vesicles per unit area is not different nor is (d) the localization of vesicle from the plasma membrane. (e) Vesicle diameter and (f) dense core diameter are significantly smaller in RCAN1ox chromaffin cells indicating reduced catecholamine storage in RCAN1ox vesicles. = 3 animals per genotype and = 7 and 9 different cells from WT and RCAN1ox mice, respectively, ***< 0.001. Scale bar = 5 μm in a and b and 250 nm in inset images.

RCAN1 regulates vesicle recycling in chromaffin cells by inhibiting calcineurin

Calcineurin controls endocytosis primarily through regulation of the activity of the endocytic proteins known as the dephosphins (Cousin and Robinson 2001). As endocytosis and exocytosis are directly interlinked, we wished to know whether endocytosis and vesicle recycling defects might underlie reduced levels of exocytosis in RCAN1ox cells. To gauge this we implemented a repeated stimulations protocol, consisting of three 1-min stimulations with 5 min intervals between them that reflects release from vesicles that have been endocytosed in a dynamin-dependent manner after the initial stimulation (Elhamdani et al. 2001). This protocol induced consistent levels of exocytosis in WT cells with the first stimulation causing 39.8 ± 7.4 events, the second stimulation causing 45.8 ± 5.9 events and the third causing 38.4 ± 5.2 events (Fig. 2a–d, = 6 cells). When we acutely expose different cells to cyclosporine A and FK-506, the result is a decline (< 0.05) in the number of vesicles undergoing exocytosis during subsequent stimulations with the first stimulation causing 79.6 ± 11.6 events, the second stimulation causing 40.1 ± 5.4 events and the third causing 39.7 ± 5.2 events (Fig. 2e–h, = 16 cells). This indicates that calcineurin controls vesicle recycling in WT chromaffin cells and that acute exposure to known inhibitors of calcineurin activity causes a rundown of secretion with this repeated stimulation protocol. We repeated these experiments in RCAN1ox chromaffin cells. RCAN1ox cells display a rundown in exocytosis in the absence of any calcineurin inhibitors using this protocol with the first stimulation causing 68.7 ± 7.7 events, the second stimulation causing 42.8 ± 8.8 events and the third causing 31.9 ± 8.7 events (Fig. 3a–d, = 7 cells), indicative of a recycling defect when RCAN1 is over-expressed. Importantly, acute exposure to inhibitors of calcineurin in different RCAN1ox cells has no additional effect on vesicle recycling compared with RCAN1ox cells in the absence of these inhibitors with the first stimulation causing 62.8 ± 8.9 events, the second stimulation causing 38 ± 4.5 events and the third causing 19.5 ± 1.9 events (Fig. 3e–h, = 4 cells). These results therefore support the possibility that calcineurin activity controls vesicle recycling in chromaffin cells and that RCAN1 over-expression in RCAN1ox cells may inhibit vesicle recycling because of reduced activity of calcineurin.

Figure 2.

Acute calcineurin inhibition in wild type (WT) cells reduces exocytosis during repeated stimulations. Cells were stimulated three successive times for 1 min with 5 min rest between stimulations (Elhamdani et al. 2001). In WT cells, we observe similar levels of exocytosis between the first (a), second (b) and third (c) stimulations. (d) The average number of exocytotic events does not decrease with this protocol (= 6 cells). Undertaking these experiments (in a different cell to that shown in a, b and c) in the acute presence of the calcineurin inhibitors, cyclosporine A and FK-506 results in exocytotic events (e) that reduce in frequency with a second (f) and third (g) stimulation. (h) This reduction was significant upon the second and third stimulations (= 16 cells, *< 0.05). Typical amperometric traces from individual cells are depicted in a, b and c and from e, f and g. Scale bars represent 20 s and 50 pA.

Figure 3.

Exocytosis rundown occurs in RCAN1ox cells with no additional effect of acute calcineurin inhibition. Cells were stimulated 3 successive times for 1 min with 5 min rest between stimulations (Elhamdani et al. 2001). In RCAN1ox cells, we observe a gradual decline in exocytosis between the first (a), second (b) and third (c) stimulations. (d) The average number of exocytotic events significantly decreases with this protocol (= 7 cells, *< 0.05, **< 0.01). Undertaking these experiments (in a different cell to that shown in a, b and c) in the acute presence of the calcineurin inhibitors, cyclosporine A and FK-506 also results in exocytotic events (e) that reduce in frequency with the second (f) and third (g) stimulations. (h) This reduction was significant upon the second and third stimulations with no additional effect of cyclosporine A and FK506 observed (= 4 cells, *< 0.05, **< 0.01). Typical amperometric traces from individual cells are depicted in a, b and c and from e, f and g. Scale bars represent 20 s and 50 pA.

Acute inhibition of calcineurin in WT cells does not alter quantal release kinetics during repeated stimulations

Our previous studies identified a role for RCAN1 in the regulation of the amount of catecholamine released from individual vesicles (Keating et al. 2008). We found that a single stimulation in the acute presence of cyclosporine A and FK-506 did not change any aspect of quantal release or vesicle trafficking. However, we now find that vesicle recycling declines in WT cells in the acute presence of cyclosporine A and FK-506 during repeated stimulations and in RCAN1ox cells following repeated stimulations with or without cyclosporine A and FK506. This altered recycling could be caused by increased phosphorylation of exocytosis and endocytosis proteins that are dephosphorylated by calcineurin. Such proteins include Munc-18 and dynamin, the functions of which can also regulate fusion pore kinetics (Graham et al. 2002; Craig et al. 2003). We therefore wished to determine if quantal release kinetics are affected during this repeated stimulations protocol in the presence of cyclosporine A and FK-506. Thus, we analysed the spike kinetics from the experiments shown in Fig. 3, in which we acutely exposed cells to cyclosporine A and FK-506 for 10 min in WT cells and repeatedly stimulated cells in the continued presence of these drugs. We found this treatment had no effect on vesicle release kinetics such as amperometric spike area (Fig. 4a), rise time (Fig. 4b), half-width (Fig. 4c) or decay time (Fig. 4d). Therefore, short-term exposure to calcineurin inhibitors, even during repeated stimulations, does not affect quantal release kinetics and the amount of catecholamine released per vesicle in WT cells.

Figure 4.

Repeated stimulation of wild type (WT) chromaffin cells during acute calcineurin inhibition does not affect vesicle release kinetics. Although the acute inhibition of calcineurin resulted in reduced exocytosis with repeated stimulation, it does not affect amperometric spike (a) area, (b) rise time, (c) half-width or (d) decay time (= 16 cells) when comparing these parameters between the first (white bar), second (horizontal stripes) and third (vertical stripes) stimulation.

RCAN1 over-expression reduces quantal release kinetics possibly via the chronic inhibition of calcineurin activity

As calcineurin activity is thought to be chronically, rather than acutely, inhibited in RCAN1ox cells, we next investigated whether chronic exposure to cyclosporine A and FK-506 could explain the effect of RCAN1 over-expression on vesicle release in RCAN1ox chromaffin cells. We cultured chromaffin cells for 3 days in the presence or absence of cyclosporine A and FK-506 and compared the effect on amperometric spike kinetics in both WT and RCAN1ox cells. Chronic exposure to cyclosporine A and FK-506 reduced the total amount of catecholamine released per vesicle in WT cells (Fig. 5a). Therefore, the rise time (Fig. 5b), half-width (Fig. 5c) and decay time (Fig. 5d) of WT spikes were also changed during chronic exposure to cyclosporine A and FK-506. When we undertook these experiments in RCAN1ox cells we observed the reduced release from vesicles that we previously published (Keating et al. 2008) and importantly, when we chronically exposed RCAN1ox cells to cyclosporine A and FK-506 we observed no additional decrease in these spike parameters. Thus, it is possible that chronic inhibition of calcineurin activity reduces quantal release and that the reduced amount of catecholamine released from RCAN1ox vesicles is most likely because of chronic inhibition of calcineurin caused by RCAN1 over-expression.

Figure 5.

Chronic calcineurin inhibition may underlie the reduced quantal release kinetics observed in RCAN1ox chromaffin cells. Cells were treated for 4–6 days in culture with calcineurin inhibitors, cyclosporine A and FK-506 (5 μM, indicated by + CaN inhibition). In wild type (WT) cells, this resulted in significant reductions in (a) spike area, (b) rise time, (c) half-width and (d) decay time. However, this caused no change in these parameters in RCAN1ox cells. *< 0.05, **< 0.01 In WT = 11 control conditions and 17 under calcineurin inhibition. In RCAN1ox, = 11 in control and 10 under calcineurin inhibition.

The effect of calcineurin inhibitors on the earliest stages of fusion pore formation

In amperometric recordings, a PSF signal is often observed before a full secretory spike. This PSF represents the release of catecholamine through the developing fusion pore as it transitions to a stable, fully dilated pore. As we previously reported reductions in the PSF signal (Keating et al. 2008) when RCAN1 is over-expressed, we analysed changes in the PSF signal in WT and RCAN1ox cells when cells are chronically exposed to cyclosporine A and FK-506. In WT cells, we observe no change in PSF signal amplitude when cells are exposed to these drugs (Fig. 6a) but do observe significant decreases in PSF signal duration (< 0.001, Fig. 6b) and area (< 0.001, Fig. 6c). In RCAN1ox cells, chronic exposure to these drugs also caused a significant decrease in PSF signal duration (< 0.001) and area (< 0.01).

Figure 6.

Chronic calcineurin inhibition reduces aspects of the pre-spike foot (PSF) signal in both wild type (WT) and RCAN1ox cells. From the same experiments as those shown in Fig. 5, we observe that chronic calcineurin inhibition has (a) no effect on PSF amplitude in WT or RCAN1ox cells, but decreases both (b) PSF signal duration and (c) PSF signal area in WT and RCAN1ox cells. **< 0.01, ***< 0.001 In WT,= 207 spikes in control conditions and 285 under calcineurin inhibition. In RCAN1ox, = 287 spikes in control and 166 under calcineurin inhibition.

Discussion

We have previously shown that RCAN1 is a regulator of exocytosis and is involved in the regulation of both the amount of catecholamine released from vesicles and the number of vesicles undergoing fusion (Keating et al. 2008). In this study, we reveal further insights into the role of RCAN1 in the regulation of exocytosis. We firstly identify that over-expression of RCAN1 and the acute use of agents which inhibit calcineurin activity in WT cells both reduce vesicle recycling and re-use. Furthermore, the combination of calcineurin inhibitors and RCAN1 over-expression has no additive effect in our vesicle recycling assay indicating an overlapping mechanism of action. These effects on exocytosis/endocytosis are not caused by changes in vesicle number or distribution as evidenced from our electron microscopy studies. Short-term (minutes) exposure of cells to drugs known to inhibit calcineurin activity does not affect the amount of catecholamine released per vesicle during our repeated stimulation protocol in WT cells. However, when we chronically (days) expose WT cells to calcineurin inhibitors, as would occur if RCAN1 over-expression was inhibiting calcineurin in RCAN1ox cells, we find the amount of catecholamine released from single vesicles to be reduced to a similar level to that seen in RCAN1ox cells. Our finding that the chronic use of cyclosporine A and FK-506 does not have any additional effect on vesicle release in RCAN1ox cells suggests that RCAN1 over-expression in these cells reduces quantal vesicle release via the inhibition of calcineurin activity. Our EM analysis of vesicle and dense core size further illustrates that the underlying mechanism explaining reduced quantal release in RCAN1ox cells is decreased catecholamine loading into vesicles. Intriguingly, however, we still observe an effect of cyclosporine A and FK-506 in RCAN1ox cells on the PSF signal and as such, early stages of fusion pore formation may require further calcineurin inhibition to that occurring in RCAN1ox cells to be fully affected.

The progressive decrease in exocytosis observed with our repeated stimulation protocol in the presence of cyclosporine A and FK506 suggests that vesicle endocytosis and recycling are impaired when calcineurin activity is blocked. Over-expression of RCAN1 induces the same decrease in vesicle recycling. These results are similar to the effect of dynamin inhibition in chromaffin cells using the same protocol (Elhamdani et al. 2001) and indicate this protocol provides a measurement of vesicle recycling and re-use in chromaffin cells. The lack of any further effect with pharmacological calcineurin inhibition in RCAN1ox cells indicates that RCAN1 over-expression may be inhibiting calcineurin activity to reduce recycling and re-use in subsequent stimulations. A likely mechanism underlying this is inhibition of the dephosphins, a group of structurally diverse proteins that are essential for vesicle endocytosis and which are all dephosphorylated by calcineurin. These proteins include dynamin, amphiphysin 1 and 2, synaptojanin, AP180, Epsin and Eps15 (Cousin and Robinson 2001). It should be noted that the relative number of exocytosis events compared with the first stimulation is the important factor to be considered in these experiments. This provides an internally controlled comparison of the effect of RCAN1 and calcineurin activity on vesicle re-use in a single cell. All four experimental conditions were performed in different cell cultures with different carbon fibre probes. These differences therefore provide a source unknown levels of variance, making accurate comparisons of the average number of exocytosis events across experiments impossible. Our previously published work demonstrating that RCAN1ox cells have less exocytotic events that WT cells were obtained from tightly controlled experiments utilizing cells cultured in parallel and the pairing of electrodes to sequential WT and RCAN1ox cell recordings (Keating et al. 2008).

We also tested whether quantal release kinetics is affected by RCAN1 over-expression or exposure of the cell to cyclosporine A and FK506 during these repeated stimulation experiments. In these experiments, we hypothesized that reduced calcineurin activity, which was potentially causing changes in endocytosis, may also affect quantal release kinetics. The phosphorylation state of most exocytosis proteins, including SNARE proteins, affects their function, and subsequently the kinetics of vesicle release. For example, Munc18-1 binds to syntaxin 1 with higher affinity when dephosphorylated by calcineurin (Craig et al. 2003). Therefore, if calcineurin activity is decreased in RCAN1ox cells, Munc18-1 may bind to syntaxin 1 with lower affinity, resulting in unstable fusion pore formation and accelerated fusion pore kinetics. Acceleration of fusion pore kinetics as a result of decreased Munc18 phosphorylation has been previously demonstrated (Barclay et al. 2003) using Munc18 phosphomimetic mutations. The lack of effect on vesicle release kinetics by acute exposure to cyclosporine A and FK-506 indicates that short-term calcineurin inhibition, even during repeated stimulation, does not alter quantal release.

Our experiments in which we chronically expose WT cells to cyclosporine A and FK506 are the first example illustrating that calcineurin may control vesicle release kinetics. This effect may be because of changes in the phosphorylation state of Munc18 or other exocytosis proteins as discussed or, alternatively, longer term exposure to cyclosporine A and FK-506 may affect quantal catecholamine release because of changes in gene transcription that take longer than the total 10–30 min duration of the acute inhibition experiments. Calcineurin is able to affect transcription through the NFAT transcription pathway (Clipstone and Crabtree 1992). Loss of NFAT function in pancreatic β-cells causes hypoinsulinaemia because of down-regulation of insulin gene transcription and insulin synthesis (Heit et al. 2006). NFAT also regulates activity-dependent plasticity in Drosophila (Freeman et al. 2011) and NFAT-dependent transcription controls both morphological and electrophysiological properties of neurons (Schwartz et al. 2009). As amperometry measures only the release of oxidizable catecholamines, we cannot be sure if altered fusion pore kinetics underlie the reduced catecholamine release from individual vesicles or whether such changes are because of factors affecting the amount of catecholamine packaged inside each vesicle. Our electron microscopy results demonstrating a decrease in vesicle and dense core volume in RCAN1ox cells clearly point towards a defect in catecholamine loading in these cells and would argue against an effect on fusion pore kinetics. Global gene expression analysis may provide useful information regarding potential transcriptional changes occurring in RCAN1ox chromaffin cells and whether transcriptional changes affect various secretion-associated processes.

Although we observe no additional effect of chronic exposure to cyclosporine A and FK506 on single vesicle catecholamine release (spike area) in RCAN1ox cells, effects on the PSF signal do occur under these conditions. The PSF signal represents catecholamine flux through the transient, unstable fusion pore as it transitions to an open, stable full fusion pore (Zhou et al. 1996; Zhang and Jackson 2010). Our results present at least two intriguing possibilities. Firstly, RCAN1 may have a direct effect on the earliest stages of fusion pore formation independent of its role as a calcineurin inhibitor. There are currently no identified protein binding partners of RCAN1 which might explain such an effect. RCAN1 contains two proline-rich domains (Fuentes et al. 1995) which typically bind to SH3 regions. SH3 regions are contained in a number of proteins associated with vesicle trafficking, exocytosis and endocytosis including ITSN1 and amphiphysin 1 and 2 (Keating et al. 2006), but whether such interactions occur or even regulate fusion pore development is an area of future investigation. Secondly, calcineurin may not be completely inhibited in RCAN1ox cells and alterations in the PSF signal are only evident during exposure of the cell to cyclosporine A and FK506, which could be fully inhibiting calcineurin. There may therefore be differences in the calcineurin sensitivity associated with mechanisms underlying the amount released per vesicle (spike area) compared with those regulating the initial fusion pore development.

It must also be noted that the cyclosporine A and FK506 have effects other than calcineurin inhibition. To competitively bind to and inhibit calcineurin activity, cyclosporine A forms complexes with cyclophilin D, while FK506 complexes with FK506 binding protein (FKBP) (Liu et al. 1991). Cyclophilin D can also bind the mitochondrial permeability transition pore to affect mitochondrial function (Elrod et al. 2010), while FK506 reduces the peptidyl-prolyl isomerase activity of FKBP (Brecht et al. 2003). Thus, while calcineurin inhibition is the primary therapeutic effect of these immunosuppressants, it is possible that the changes in secretion we observe with these drugs could be independent of changes in calcineurin activity. However, given the lack of additive effect on vesicle recycling and quantal release kinetics when we over-express an endogenous calcineurin inhibitor (RCAN1) and add known calcineurin inhibitors, our primary conclusion is that these treatments affect secretion through the same mechanism of reduced calcineurin activity.

In summary, our results demonstrate that RCAN1 regulates cell communication at several stages. Endocytosis and vesicle re-use as well as vesicle release are regulated by RCAN1 as well as by pharmacological calcineurin inhibitors, likely through the same mechanism of action, that being reduced calcineurin activity. What the targets of calcineurin/RCAN1 action are in these instances remains to be identified. The reduction in vesicle size and catecholamine loading observed in our EM experiments would explain the decreased catecholamine release per vesicle in RCAN1ox cells. Catecholamine release through the developing fusion pore is additionally affected by RCAN1. This effect of RCAN1 may either be independent of any inhibitory action of RCAN1 on calcineurin or could be explained by a difference in the sensitivity to calcineurin activity of early fusion pore formation and quantal release/vesicle loading. Our data therefore reveal a role of RCAN1 in regulating exocytosis at several stages. Recently, published data illustrate that changes in hippocampal structure and function occur in these RCAN1ox mice brains (Martin et al. 2012). However, these changes may not be associated with alterations in calcineurin activity and involve instead changes directly associated with RCAN1 function. Therefore, while increased expression of RCAN1 may play a significant role in the cognitive and neuronal pathologies associated with Alzheimer's disease and Down syndrome, elucidating which of these are associated with the role of RCAN1 as an inhibitor of calcineurin needs to be clearly defined and may well vary for individual types of cells or neurons. This point is further highlighted by recent evidence demonstrating that the effects of RCAN1 on calcineurin are dependent on a number of factors including the degree of RCAN1 expression and the presence of other RCAN1-binding partners that effect the relationship between RCAN1 and calcineurin (Liu et al. 2009).

Acknowledgements

This study was supported by funding from the Australian National Health and Medical Research Council, Bio Innovation SA and the Australian Research Council. All authors declare no conflict of interest in this study.

Ancillary