- Email: [email protected]
Ca2+ clearance mechanisms in neurohypophysial terminals of the rat
Cell Calcium 37 (2005) 45–56
Ca2+ clearance mechanisms in neurohypophysial terminals of the rat Naoko Sasakia , Govindan Dayanithib , Izumi Shibuyaa,∗ a
Deparment of Physiology, University of Occupational and Environmental Health School of Medicine, Kitakyushu 807-8555, Japan b U-583 INSERM, L’Institut des Neurosciences de Montpellier, Hˆ opital St Eloi, F-34295 Montpellier Cedex 5, France Received 19 March 2004; received in revised form 14 June 2004; accepted 22 June 2004
Abstract The importance of intracellular calcium ([Ca2+ ]i ) in the release of vasopressin (AVP) and oxytocin from the central nervous system neurohypopyhysial nerve terminals has been well-documented. To date, there is no clear understanding of Ca2+ clearance mechanisms and their interplay with transmembrane Ca2+ entry, intracellular [Ca2+ ]i transients, cytoplasmic Ca2+ stores and hence the release of AVP at the level of a single nerve terminal. Here, we studied the mechanism of Ca2+ clearance in freshly isolated nerve terminals of the rat neurohypophysis using Fura-2 Ca2+ imaging and measured the release of AVP by radioimmuno assay. An increase in the K+ concentration in the perfusion solution from 5 to 50 mM caused a rapid increase in [Ca2+ ]i and AVP release. Returning K+ concentration to 5 mM led to rapid restoration of both responses to basal level. The K+ -evoked [Ca2+ ]i and AVP increase was concentration-dependent, reliable, and remained of constant amplitude and time course upon successive applications. Extracellular Ca2+ removal completely abolished the K+ -evoked responses. The recovery phase was not affected upon replacement of NaCl with sucrose or drugs known to act on intracellular Ca2+ stores such as thapsigargin, cyclopiazonic acid, caffeine or a combination of caffeine and ryanodine did not affect either resting or K+ -evoked [Ca2+ ]i or AVP release. By contrast, the plasma membrane Ca2+ pump inhibitor, La3+ , markedly slowed down the recovery phase. The mitochondrial respiration uncoupler, carbonyl cyanide 3-chlorophenylhydrazone (CCCP), slightly but significantly increased the basal [Ca2+ ]i , and also slowed down the recovery phase of both [Ca2+ ]i and release responses. In conclusion, we show in nerve terminals that (i) Ca2+ extrusion through the Ca2+ pump in the plasma membrane plays a major role in the Ca2+ clearance mechanisms of (ii) Ca2+ uptake by mitochondria also contributes to the Ca2+ clearance and (iii) neither Na+ /Ca2+ exchangers nor Ca2+ stores are involved in the Ca2+ clearance or in the maintenance of basal [Ca2+ ]i or release of AVP. © 2004 Elsevier Ltd. All rights reserved. Keywords: Hypothalamus; Posterior pituitary; Exocytosis; Vasopressin release
1. Introduction The hypothalamo-neurohypophyseal system has been the focus of much research since decades as it provides the rare opportunity to study, in a single system, the bioelectrical activity of individual neurosecretory cells and the mechanisms by which these cells release their products [1–3]. The magnocellular cell bodies, which are located in the hypothalamic supraoptic and paraventricular nuclei, synthesize the neuropeptides oxytocin and vasopressin, that are released from axon terminals at the neurohypophysis directly into the blood system to ultimately act as hormones. In the neural ∗
Corresponding author. Tel.: +81 93 691 7420; fax: +81 93 692 1711. E-mail address: [email protected] (I. Shibuya).
0143-4160/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ceca.2004.06.007
lobe, the mechanisms underlying depolarization–secretion coupling have been widely investigated and characterized [4–8]. Briefly, exocytosis results from a sequence of events that starts by the arrival of action potentials which promote the opening of voltage-sensitive Ca2+ channels, and includes cytoplasmic-free calcium concentration ([Ca2+ ]i ) increase as a key player in release [9,10]. The different Ca2+ -channel subtypes and their role in peptide release in neurohypophysial nerve terminal have also been characterized [7,11–17]. In most cell types, increase in [Ca2+ ]i results both from 2+ Ca entry mediated by voltage-gated Ca2+ channels [18,19] and from Ca2+ release from intracellular stores [20,21]. A large body of data is presently available in the literature regarding the mechanisms involved in intracellular Ca2+ buffering and restoration to basal level [22–24]. Such mech-
46
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
anisms include Ca2+ pumps in both intracellular compartments and the plasma membrane, mitochondrial Ca2+ uptake, plasma membrane Na+ /Ca2+ exchange under physiological conditions [25,26] or during aging [21,26,27]. Another wellknown neuronal Ca2+ store is endoplasmic reticulum (ER). The ER is multifunctional organelle regulating a wide range of neuronal functional responses [28–30]; a focal point for co-ordination of cellular activity [31]. The ER Ca2+ stores are shown to act as a Ca2+ -buffering system, i.e. ‘Ca2+ sink’, clearing cytoplasmic Ca2+ loads by SERCA pumping [18] and act as a source of Ca2+ for CICR phenomenon [29]. The Ca2+ -binding proteins were also isolated and purified from bovine neurohypophysis [32]. In neurohypophysial nerve terminals, it was reported that the Ca2+ -buffering mechanisms was sensitive to mitochondrial Ca2+ uptake but not or minimal Na+ /Ca2+ exchange is important in Ca2+ clearance mechanism [33], thus requiring an efficient Ca2+ buffering to allow sustainable release. The role of Na+ /Ca2+ exchange in the release of AVP from these nerve terminals, under basal and stimulated conditions was also demonstrated [34–36]. But recently in the same model, the presence of K+ -dependent Na+ /Ca2+ exchanger was reported [37] and it seems to play a role in Ca2+ clearance. Although the importance of intracellular Ca2+ stores in Ca2+ homeostasis has been well-documented in various tissues [26,38], controversial findings have been reported in the isolated nerve terminals [33,37,39]. These important discrepancies led us to re-investigate the exact mechanisms by which the neurohypophysial terminals regulate the intracellular Ca2+ and their buffering properties. Here we bring clear evidence that Ca2+ clearance mechanisms and hence efficient AVP release can be operated by cellular/membrane components other than Na+ /Ca2+ exchanger or Ca2+ stores. We further highlight the understanding of the fine interplay between transmembrane Ca2+ entry, intracellular [Ca2+ ]i transients, cytoplasmic Ca2+ stores, Ca2+ clearance mechanisms and AVP release at the level of isolated nerve terminals.
2. Material and methods 2.1. Preparation of isolated neurohypophysial nerve terminals
contains highly purified nerve terminals [5,6], was eventually resuspended in Locke buffer maintained at 37 ◦ C. Aliquots of 200 l of suspension were then plated on glass bottom culture dishes for [Ca2+ ]i measurements. 2.2. Drugs and solutions The Locke buffer described here was used both during dissection and as a control solution. It contained (mM): NaCl, 140; KCl, 5; MgCl2 , 1.2; CaCl2 , 2.2; glucose, 10; HEPES, 10; BSA, 0.02%, pH 7.25 adjusted with Tris–HCl. The osmolarity of all the solutions used in this study was maintained between 298 and 300 mOsm l−1 . To test the effects of extracellular Ca2+ , Ca2+ /EGTA buffer (see [4]) was used which contained (mM): ethylene glycol (bis-aminoethyl ether)N,N,N ,N -tetra acetic acid (EGTA), 2; NaCl, 140; KCl, 5; MgCl2 , 1.2; CaCl2 , 2.2; glucose, 10; HEPES, 10; pH 7.25 adjusted with Tris–HCl. In this buffer, the free Ca2+ concentration in the EGTA buffer was around 100 nM which corresponds to the resting [Ca2+ ]i typically measured in neurones. Buffer containing high K+ commonly used in this study contained (mM): NaCl, 90; KCl, 50; MgCl2 , 1.2; CaCl2 , 2.2; glucose, 10; HEPES, 10 at pH 7.25. For other K+ concentrations, KCl was added to the desired concentration and was adjusted with NaCl appropriately to bring osmolarity in the required range. For Na+ -free buffer, NaCl was replaced by appropriate concentrations of sucrose. Most of the standard chemicals were purchased from Sigma. Thapsigargin (TG), ryanodine, cyclopiazonic acid (CPA; Alomone labs, Jerusalem, Israel) and carbonyl cyanide 3-chlorophenylhydrazone (CCCP; Sigma), were first dissolved in DMSO (final concentration: 0.01%) and further diluted in working solution to appropriate concentrations. La3+ (Sigma) and Caffeine (Alomone labs) were directly dissolved in the working solution. Fura-2 AM (in DMSO solution) and Pluronic F-127 (30% stock in distilled water) were from Molecular Probes Inc., Eugene, OR, USA). 125 Ilabelled AVP was purchased from Amersham, France. All test substances for [Ca2+ ]i measurements were perfused in a 500 l chamber using a multi-channel Gilson peristaltic pump (1.5 ml/min). This method allowed fast and reliable exchange of the solution surrounding the cells. 2.3. [Ca2+ ]i measurements
Isolated nerve terminals were prepared from male Wistar rats (about 200 g) as previously described [5]. After animals were killed by decapitation with a guillotine following the guidelines laid down by the French/Japanese ethical committee, each pituitary was dissected out and transferred in Locke buffer (see the following). After removal of the anterior lobe and the pars intermedia under binocular control, the neural lobes were homogenized at 37 ◦ C in 1 ml of a solution containing (in mM): sucrose, 270; EGTA, 0.2; HEPES, 10; pH 7.25 adjusted with Tris. The dissociated neurohypophysis was first spun at 100 × g for 1 min, then the supernatant was further spun at 2400 × g for 4 min. The final pellet, which
The isolated nerve terminals plated on a cover glass attached to 35 mm culture dish were incubated 2.5 M Fura2 AM plus 0.02% Pluronic F-127 at room temperature for 1 h. The preparations were then washed with dye-free solution and kept at room temperature (about 23 ◦ C) until used. Peptidergic nerve terminals were identified using phase contrast optics. The arrangements for perfusion and measuring fluorescence are essentially described previously [40,41]. In brief, fluorescence was measured from fura-2-loaded nerve terminals in the perfusion chamber, which was positioned on the stage of an inverted microscope (IX-50; Olympus, Tokyo,
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Japan), using a Ca2+ -imaging system equipped with a 12-bit digital CCD camera (Quanticell/900; JEOL, Tokyo, Japan). Fluorescence intensities at 510 nm with excitation at 340 and 380 nm were recorded at an interval of 5 s. The [Ca2+ ]i in individual nerve terminal was calculated from the ratio of the fluorescence images measured with excitation at 340 nm to those at 380 nm using the equation of Grynkiewicz et al. [42]. Autofluorescence in the nerve terminals was negligible compared with the fluorescence in the fura-2 loaded terminals. Values of Rmin , Rmax and Fo /Fs (i.e. ratio of emitted fluorescence intensity at 380 nm excitation under Ca2+ limiting and Ca2+ saturating conditions) were determined using solutions of defined Ca2+ concentrations and were 0.3, 3.4 and 5.7, respectively. The Ca2+ Kd (=224) value for Fura-2 was taken from the paper by Grynkiewicz et al. [42]. Photobleaching was minimized by decreasing illumination intensity with neutral filters and by slowing the acquisition frame to a low rate. 2.4. Measurements of AVP release Release experiments were performed as described previously [5]. Briefly, for each experimental run, the isolated terminals from two neurohypophysis were equally distributed and loaded onto four filters, i.e. each filter receives about 25% of the total terminals. Following loading of isolated nerve terminals onto filters (0.45 m Acrodisc, LCP-VDF, Gelman Sciences, USA), they were perfused (Minipulse Peristaltic Pump, Gilson, France) for 45 min with Normal Locke at a flow rate of 50 l min−1 and the flow rate was increased to
47
100 l min−1 for 15 min. Collection of the perfusate over 5min periods started 60 min after loading the nerve terminals onto the filter. AVP content in each fraction was then determined by radioimmunoassay as already described [1,5] with necessary modifications [7]. The antisera were kindly given by Dr. John Bicknell, Babraham, UK. The AVP was raised in sheep. The final antibody dilution was 1:70,000. The crossreactivity of the AVP antiserum with OT was 0.001%. The sensitivity of the assay was 0.5 pg. The inter- and intra-assay coefficients of variation were 5–7%. The results represent the average AVP content of standard aliquots taken from the collected fractions arising from at least three separate groups of isolated nerve terminals. Internal control groups arising from the same pool of isolated nerve terminals were run for comparison to test groups. Differences in the total amount of AVP released between experiments resulted from differences in the amount of each preparation loaded onto the filters and from heterogeneity of secretary responsiveness between preparations. The mean evoked AVP release was calculated by subtracting the AVP release under basal conditions, i.e. prior to the stimulus from the peak amplitude observed after the application of the test substances. 2.5. Data analysis and statistics Results are expressed as mean ± S.E.M. Effects of test solutions on the basal [Ca2+ ]i level were expressed as percents of the mean [Ca2+ ]i value measured during 1 min just before
Fig. 1. Representative examples for fura-2 loaded isolated nerve terminals of rats and their response to 50 mM K+ . (A and B) Time courses of [Ca2+ ]i before, during and after 50 mM K+ challenge. The numbers (1–9) in the figures correspond to the number depicted in the images. The two round cells (see C) showed no response (B), whereas all nerve terminals showed robust increases in [Ca2+ ]i (A). (C) Images of fura-2 loaded nerve terminals taken in a phase contrast field (left most) and pseudo-color [Ca2+ ]i images before (a), during (b) and after (c) 50 mM K+ challenge (other three). The time when (a–c) were taken was shown with the arrows in (A) and (B). [Ca2+ ]i values are shown in the horizontal bar below.
48
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
the treatment, whereas the effects on the recovery phase time constant (τ) of [Ca2+ ]i after high K+ challenge were expressed as percents of the mean of responses measured prior and after the test in the same nerve terminal preparations. The Axograph software (version 4.0, Axon Instruments Inc., USA) was used for curve fitting. Statistical differences (P < 0.05) were determined by a Mann–Whitney’s U test when “n” was equal to or larger than 10, or by a Student’s t-test when “n” was smaller than 10.
3. Results 3.1. [Ca2+ ]i increases induced by high K+ Fig. 1 shows a typical recording of the [Ca2+ ]i increase induced by exposure to high K+ (50 mM) measured in an isolated nerve terminal. Increasing the K+ concentration in the perfusion solution from 5 to 50 mM for 30 s caused a fast and large increase in [Ca2+ ]i in all nerve terminals tested (Fig. 1A; mean K+ evoked [Ca2+ ]i increase, 902 ± 96 nM; resting [Ca2+ ]i , 59 ± 7 nM; n = 13). When viewed under
the microscope, healthy nerve terminals displayed irregular shapes and robust [Ca2+ ]i increases under high K+ exposure, therefore allowing easy identification (Fig. 1C). By contrast, the rare round cells observed in the preparation did not display any [Ca2+ ]i change under high K+ challenge (Fig. 1B). Upon returning the K+ concentration to 5 mM, [Ca2+ ]i in nerve terminals rapidly recovered (Fig. 1A and C). The amplitude of the [Ca2+ ]i increase (present study) or AVP release (see [4,5]) induced by exposure to a high K+ solution is concentration-dependent. As shown in Fig. 2A, exposing the nerve terminals to 30-s tests of increasing K+ concentration (10, 20, 30, 50 and 100 mM) evoked [Ca2+ ]i responses (Fig. 2B, respectively, 45 ± 6 nM; 119 ± 12 nM; 226 ± 26 nM; 649 ± 64 nM; 711 ± 70 nM; resting [Ca2+ ]i , 174 ± 17 nM; n = 25) displaying a single peak with progressively larger amplitude while the time course remained the same. Repetitive 30 s exposures to a constant high K+ concentration at intervals of 3–5 min caused comparable increases in [Ca2+ ]i (Fig. 2C; peak [Ca2+ ]i responses, respectively, 621 ± 47 nM; 630 ± 43 nM; 644 ± 50 nM; 650 ± 48 nM; resting [Ca2+ ]i , 68 ± 6 nM; n = 24) showing no significant inactivation.
Fig. 2. Increase in [Ca2+ ]i of isolated nerve terminals. (A) Traces represent changes in [Ca2+ ]i evoked by 30 s applications of increasing concentrations of K+ in a single neurohypophysial terminal. (B) The graph shows the dose–response relationship of evoked [Ca2+ ]i increase (mean ± S.E.M) as a function of K+ concentration. The resting [Ca2+ ]i concentration measured from each cell was subtracted from the peak amplitude evoked by each concentration of K+ . Data were obtained from 13 terminals, using the protocol described in (A). (C) Traces show changes in [Ca2+ ]i induced by 50 mM K+ exposures of increasing durations (30 s to 5 min) measured in a single neurohypophysial nerve terminal. Note the progressive change of [Ca2+ ]i increases from a single-phased response to a biphasic one.
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
The time course of the [Ca2+ ]i increase depends on the duration of high K+ exposure. Upon 50 mM K+ exposures of increasing duration (30 s to 5 min), the [Ca2+ ]i response progressively evolved from displaying an increase with a single peak to a biphasic increase that consisted of a rapid rising phase followed by a gradually declining plateau phase (Fig. 2C). Although the late phase declined in the presence of high K+ , we never observed a complete recovery of [Ca2+ ]i under high K+ , at least for the durations tested here. This result may result from inactivation of voltage-gated Ca2+ channels in the nerve terminals, that was already reported by others [7]. In the present study, we focus on the single-phased [Ca2+ ]i response, elicited by a 30 s-exposure to 50 mM K+ . 3.2. Effects of external Ca2+ removal The high K+ -induced [Ca2+ ]i increase or AVP release was abolished when external Ca2+ was reduced to 100 nM (see Section 2) from the bathing solution (Fig. 3), confirming that at least major part of the [Ca2+ ]i increase resulted from Ca2+ entry through voltage-gated Ca2+ channels in nerve ter-
49
minals. Reduction of external Ca2+ (from 2.2 to 100 nM) in the perfusion solution did not significantly modify the resting [Ca2+ ]i (Fig. 3A; 138 ± 16 nM versus 150 ± 16 nM; n = 17), indicating that there is no Ca2+ entry in resting conditions. The rate of [Ca2+ ]i recovery after K+ challenge was quantified by exponential fitting. Two exponentials were required to adequately fit the [Ca2+ ]i recovery phase (Fig. 3B). The first τ value ranged from 5.7 to 19.9 s and the second ranged from 313 to 2953 s. Since the latter showed much larger variation than the former, we focused on the time constant (τ) of the fast component because the first τ value in this study. Fig. 3C shows the AVP release responses in the presence of 2.2 or 100 nM external Ca2+ under basal or K+ -stimulated conditions. These results indicate that depolarizations-induced AVP release is dependent on the Ca2+ influx. 3.3. Effects of external Na+ replacement To examine the role of Na+ /Ca2+ exchange in intracellular clearance, Na+ in the perfusion solution was replaced by sucrose. As illustrated in Fig. 4A, substitution of Na+ had
Ca2+
Fig. 3. Effect of extracellular Ca2+ on K+ -induced [Ca2+ ]i and AVP release. (A) [Ca2+ ]i response to 50 mM K+ in the presence or absence of 2 mM external Ca2+ . The trace shows that 100 nM low Ca2+ -Locke buffer (indicated by open bar) totally suppressed the [Ca2+ ]i increase induced by a 30 s exposure to 50 mM K+ in neurohypophysial nerve terminals. The [Ca2+ ]i response rapidly recovered upon return to normal external Ca2+ . (B) An example of exponential curve fitting. The dotted line is exponential curve drawn with the time constant of 8.8 s. (C) The time course of AVP release from isolated nerve terminals. All data points are means (±S.E.) of five experiments.
50
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Fig. 4. Effect of external Na+ on basal and K+ -induced [Ca2+ ]i and AVP release. (A) Example of a recording showing that replacement of Na+ by sucrose in the bath had not effect neither on basal [Ca2+ ]i or on 50 mM K+ -induced [Ca2+ ]i increase. (B) Bar diagram represents the mean peak amplitude (percent of control) of the [Ca2+ ]i response observed in Na+ -free buffer (hatched column) and in control conditions after recovery (open column). All data points are means (±S.E.) of 10 experiments. (C) Fifty millimoles of K+ -induced AVP release profiles in the presence or absences of external Na+ . Data points are means (±S.E.) of five experiments.
no effect either on the basal [Ca2+ ]i (75 ± 11 nM versus 81 ± 13 nM; n = 10) or on the recovery phase after high K+ exposure (τ 10.32 ± 1.40 s versus τ 12. 07 ± 1.68 s; n = 10). The values of basal [Ca2+ ]i and τ in control versus Na+ free conditions obtained from 10 terminals are summarized in Fig. 4B. Similarly, the release of AVP under basal (34 ± 3 pg versus 40 ± 6 pg) or K+ -stimulated conditions (513 ± 33 pg versus 461 ± 43 pg; n = 4) was not affected by the removal of external Na+ . 3.4. Effects of drugs targeting Ca2+ stores To examine the role of Ca2+ stores, the response to high was observed in the presence of 1 M TG, or 30 M CPA, both of which are known to block Ca2+ pumps located on intracellular Ca2+ stores, or with a combination of 10 M ryanodine and 10 mM caffeine, which are known to deplete caffeine/ryanodine Ca2+ stores and thereby suppress subsequent Ca2+ release from such stores. Our data clearly show that neither TG (Fig. 5A; 151 ± 23 nM versus 142 ± 18 nM; n = 10), nor CPA (Fig. 5B; 303 ± 81 nM; n = 5) nor caffeine/ryanodine (298 ± 78 nM; n = 5) had a significant effect
on basal [Ca2+ ]i (297 ± 75 nM; n = 5) nor on the recovery phase after K+ challenge (Fig. 5A and B). The results of these experiments are summarized in Fig. 5C. In addition, basal AVP or K+ -evoked AVP was not affected in the absence or presence of TG or caffeine/ryanodine or CPA (Fig. 5D; basal AVP level, 63 ± 8 pg; K+ -evoked release, 285 ± 38 pg; K+ + TG, 287 ± 50 pg; K+ + caffeine/ryanodine, 291 ± 41 pg; K+ + CPA, 276 ± 30 pg; n = 4). We have also performed experiments using 10 mM caffeine alone in the presence or absence of external Ca2+ (2 mM). Two minutes continuous application of caffeine had no effect on the resting [Ca2+ ]i levels in all nerve terminals tested (n = 16). These terminals have responded only to high 50 mM K+ (data not shown).
K+
3.5. Effects of the mitochondrial uncoupler, CCCP To assess the role of mitochondrial Ca2+ uptake, the uncoupler of mitochondrial respiratory chain (CCCP) was used. CCCP (10 M) caused a small but significant increase in the resting [Ca2+ ]i (63 ± 10 nM versus 85 ± 13 nM; n = 10; P < 0.05) and significantly slowed down the recovery of [Ca2+ ]i (Fig. 6A; τ 9.07 ± 0.35 s versus τ 13.8 ± 1.12 s; n = 10; P<
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
51
Fig. 5. No effect of intracellular Ca2+ mobilizers on basal and K+ -induced [Ca2+ ]i increase or AVP release. (A) Trace showing a typical K+ (50 mM)-induced [Ca2+ ]i rise measured in the presence or absence of TG in an isolated nerve terminal. (B) In a different terminal, the effect of mixture of caffeine and ryanodine (Caf/Rya) or CPA was first on basal [Ca2+ ]i and then on K+ -induced [Ca2+ ]i increase. Note that none of the drugs did mobilize the Ca2+ from intracellular stores, as illustrated by the absence of any effect on basal [Ca2+ ]i . In addition, no effect of the drugs could be observed on the [Ca2+ ]i response to high K+ . (C) The bar graph represents the mean basal or evoked [Ca2+ ]i increase (expressed as percent of control) elicited by high K+ in absence (open column) and in presence (hatched column) of TG, Caf/Rya or CPA. Data are the means (±S.E.) of 5–10 experiments. (D) Basal or stimulated AVP release profiles of isolated nerve terminals in the presence or absence of above drugs. Data points are means (±S.E.) of five experiments.
0.05). Both these effects were rapidly reversible upon wash. The effects of CCCP on the basal [Ca2+ ]i and τ (n = 10) are summarized in Fig. 6B. Similarly, the resting AVP (47 ± 12 pg versus with CCCP, 164 ± 17 pg; n = 5; P < 0.05) or K+ evoked AVP release (K+ , 288 ± 26 pg versus K+ + CCCP, 802 ± 34 pg; n = 5; P < 0.05) was significantly altered by CCCP (Fig. 6C). 3.6. Effects of La3+ La3+ , a blocker of voltage-gated calcium channels [43], has also been reported to block the plasma membrane Ca2+ pump in some cell preparation [24,44]. To study further the effects of La3+ on nerve terminals, we first assessed its effects on basal and K+ -induced [Ca2+ ]i increase and AVP release responses. We observed that although La3+ (300 M) did not show significant effect on basal [Ca2+ ]i (125 ± 6 nM versus 121 ± 6 nM; n = 16) but it totally and reversibly blocked the high K+ -induced [Ca2+ ]i increase (Fig. 7A) and maintained the AVP release response due to increase in cy-
tosolic [Ca2+ ]I levels. These results strongly suggest that, as reported in other preparations, La3+ could inhibit voltagedependent Ca2+ channels in nerve terminals. Therefore, to assess the role of the plasma membrane Ca2+ pump without interfering with voltage-gated calcium channels, we only added La3+ (300 M) at the time high K+ was switched off from the bath (Fig. 7B). In all 10 terminals examined, La3+ nearly stopped the recovery of [Ca2+ ]i , preventing its return to the baseline for the duration of the exposure. This effect was reversible upon wash. The AVP release experiments also revealed that K+ -evoked response (432 ± 33 pg) was potentiated after La3+ application (673 ± 14 pg; n = 5; P < 0.05) as a consequence of sustained increase in [Ca2+ ]i (Fig. 7C).
4. Discussion In the present study we have investigated the cytosolic Ca2+ clearance mechanisms and Ca2+ regulation of AVP
52
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Fig. 6. Effect of CCCP on basal and K+ -induced [Ca2+ ]i and AVP responses. (A) The effect of the mitochondrial uncoupler, CCCP was assessed on the basal [Ca2+ ]i levels or the K+ -induced [Ca2+ ]i response. Profiles obtained from a single nerve terminal. (B) The bar graphs show the mean basal or stimulated [Ca2+ ]i response, expressed as percent of control, calculated from the peak amplitude responses (n = 10). Significant level: P < 0.05. (C) AVP release responses to CCCP. Data points are means (±S.E.) of five experiments.
release in freshly isolated rat neurohypophysial nerve terminals. The intracellular Ca2+ concentration ([Ca2+ ]i ) was monitored on single nerve terminal level. The [Ca2+ ]i displays complex fluctuations in response to many stimuli, and acts as signal for several neuronal functions [21]. The complex neuronal signaling involves activation of receptors, which leads to fluctuations in the [Ca2+ ]i that occurs different locations within the cell (see [45–47]). Increasing number of evidence has shown in many models that the Ca2+ signal is initiated by receptor- or voltage-gated opening of Ca2+ channels (see review [48]). It has been clear during recent years that the intracellular Ca2+ storage compartments such as mitochondria and the endoplasmic reticulum are also playing crucial roles in regulating Ca2+ signals and compartmentalized functions. These intracellular Ca2+ storage compartments are shown to initiate Ca2+ signals and also serve as a Ca2+ -buffering system. The Ca2+ -sensitive stores might act as a sink for increases in [Ca2+ ]i or might play an important role in neuronal aging [27].
4.1. The Ca2+ clearance mechanism of neurohypophysial nerve terminals The treatment that produced the largest effect on Ca2+ recovery phase after high K+ challenge was La3+ , the blocker of the plasma membrane Ca2+ pump. La3+ virtually arrested Ca2+ recovery. The result suggests that Ca2+ extrusion through the plasma membrane Ca2+ pump to the extracellular space may be the most critical mechanism for the Ca2+ clearance of the nerve terminal. La3+ is known to block other Ca2+ transport mechanisms such as Ca2+ channels and Na+ /Ca2+ exchange [49], however, the result that Na+ replacement with NMDG had little effect on the Ca2+ recovery phase suggest that the involvement of the Na+ /Ca2+ exchange is minor and that the arrest of Ca2+ recovery by La3+ was indeed due to the block of the Ca2+ pump. In contrast, in the same nerve terminals, it was claimed that Na+ /Ca2+ exchange is a major Ca2+ clearance mechanism [37] but our present results as well as the results from Stuenkel [33] on [Ca2+ ]i increase or AVP release have ruled out the above pos-
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
53
Fig. 7. Effect of La3+ on basal and K+ -induced [Ca2+ ]i and AVP responses. (A) The [Ca2+ ]i traces show the effect of La3+ under basal conditions followed by stimulation with high K+ . Note that the [Ca2+ ]i increase induced by high K+ was totally suppressed in the concomitant presence of La3+ in the bath. (B) To avoid voltage-gated calcium channels inhibition and allow proper assessment of the role of the plasma membrane Ca2+ -pump, La3+ was added to the bath at the end of high K+ stimulation. In these conditions, La3+ exposure almost prevented the high K+ -induced [Ca2+ ]i to recover. Similar results were obtained in 14 of nerve terminals. (C) AVP release profiles are shown under similar experimental conditions and (D) represents evoked AVP release. Data points are means (±S.E.) of five experiments.
sibility or any significant physiological role for Na+ /Ca2+ exchange. Another important and new finding in this study is concerning the role of mitochondria in the Ca2+ clearance mechanism. The mitochondrial uncoupler, CCCP, showed significant effect on the recovery phase as well as the basal [Ca2+ ]i and peptide release, suggesting only mitochondria plays an important role in the Ca2+ clearance of the nerve terminal both in basal (unstimulated) conditions. The significant effect of CCCP on the recovery phase suggests that mitochondria play a role also when Ca2+ overload occurred. In fact, electronmicrographs of isolated nerve terminal preparations as well as intact neurohypophysis have clearly showed the presence of mitochondria [2] (see also Fig. 8). CCCP, however, did not affect the peak [Ca2+ ]i level reached during the 50 mM K+ challenge. This suggests that mitochondria does not play a role as a Ca2+ sink during such rapid Ca2+ overload. It could be that Ca2+ uptake into mitochondria is not fast enough to counteract a massive [Ca2+ ]i increase caused by Ca2+ entry in response to 50 mM K+ . This interpretation may not be valid in conditions in which the fura-2 signal is saturated. Therefore, it should be confirmed with a lower affinity Ca2+ dye to make sure about the lack of differences in the amplitude of the [Ca2+ ]i signal dur-
ing the 50 mM K+ challenge in the presence or absence of CCCP. The lack of effect of the drugs acting on Ca2+ stores on the 2+ Ca -recovery phase as well as on basal [Ca2+ ]i and peptide release suggests that Ca2+ stores are absent or, if present, they play only minor physiological role in the nerve terminal. Surprisingly, using a signal mass approach to Ca2+ imaging in nerve terminals from mice, De Crescenzo et al. [39] has demonstrated an increase in [Ca2+ ]i due to caffeine in the absence of external Ca2+ , suggesting a possible presence of intracellular Ca2+ stores in the nerve terminals. However, such intracellular stores were not seen in the previous [33,35,50] and the present study from the neurohypophysial nerve terminals. One explanation for such contrasting results could be due to the use of high resolution camera to amplify the signal to noise ratio to observe small changes in the [Ca2+ ]i levels not only from the cytosol but also from the neurosecretory granules. Of interest, it should be noted that in a few terminals (about 10%) where the resting [Ca2+ ]i levels were very high (more than 400 nM), we observed an increase in [Ca2+ ]i due to caffeine (results not shown). In any case, our present results show that under our experimental conditions caffeine had no effect on AVP release and a previous report showed that caffeine had no effect on 45 Ca2+
54
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Fig. 8. Schematic representations of possible Ca2+ clearance mechanisms in nerve terminals of the neurohypophysis. Arrows suggests that increased cytosolic Ca2+ can be extruded via plasma membrane (PM) Ca2+ pump or Na+ /Ca2+ exchange, or stored by mitochondrial or intracellular compartments (via store Ca2+ pump). We propose that only PM Ca2+ pump and mitochondria play a major role in Ca2+ clearance of the nerve terminals (indicated by thick solid line), whereas Na+ /Ca2+ exchange or intracellular Ca2+ stores do not (indicated by thin dotted line). The superimposed background picture is an electron micrograph of isolated nerve terminals, where large size mitochondria can be seen.
efflux [36]. These results suggest that the external Ca2+ independent increase in [Ca2+ ]i due to caffeine is not crucial to acute AVP secretion under physiological conditions. This shows clear contrast with Ca2+ signaling mechanisms of soma and dendritic part of SON neurons, where TG-sensitive Ca2+ stores play major roles in both Ca2+ signaling and peptide secretion [50,51]. The present results also suggest that, although previous immunohistochemical evidence indicated that Ca2+ -buffering proteins play important role in the soma and dendritic part of SON neurons [52] they may be absent or do not function in the terminal. 4.2. The correlation between [Ca2+ ]i and peptide secretion In most of the experiments in the present study, there was a good correlation between [Ca2+ ]i and AVP secretion. This is consistent with previous findings that Ca2+ plays a key role in the stimulus-secretion coupling in the neurohypophysial nerve terminals [2,9]. However, there was one exception. When La3+ was used to block the plasma membrane Ca2+ pump, the recovery of [Ca2+ ]i was arrested, whereas peptide secretion slowly decreased to the basal level. The difference can be explained by ‘fatigue’ of exocytotic machinery. It has been shown that the [Ca2+ ]i response to high K+ was well-sustained but the peptide secretion to high K+ was more
transient [5]. The [Ca2+ ]i response in the presence of La3+ was maintained stably high for more than 2 min, whilst the [Ca2+ ]i response to high K+ tends to be more transient. The response to high K+ seems to be because of inactivation of voltage-gated Ca2+ channels. Thus, the protocol of the La3+ experiment in this study may be a good tool to study the fatigue mechanism of exocytosis in the nerve terminal. 4.3. The physiological significance of the Ca2+ clearance mechanism It has been shown that phasic bursting activity of AVP neurons is directly related to AVP secretion and that the interval between phasic bursts as well as the frequency of bursts play an important role in the peptide secretion (see [1,53]). The importance of intervals between bursting and the frequency of the bursts suggests importance and necessity of quick Ca2+ clearance of a large capacity for peptide secretion. The La3+ -sensitive Ca2+ extrusion via the plasma membrane Ca2+ pump as well as respiration-coupled Ca2+ uptake into mitochondria revealed in the present study may serve as such large capacity Ca2+ clearance mechanisms. The present results together with the previous reports for Ca2+ stores and Ca2+ -binding proteins in the somato-dendritic region of SON neurons [50–52] also suggest that there are considerable differences in the Ca2+
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
signaling/clearance mechanisms between nerve terminals and somato-dendritic regions of SON neurons. Although the precise Ca2+ clearance mechanism in the somato-dendritic region of SON neurons remains unclear, our preliminary results indicate that ER could be a potential candidate to play a crucial role in Ca2+ clearance mechanisms [54]. In conclusion, the present study showed that the major Ca2+ clearance mechanism of nerve terminals is Ca2+ extrusion via the Ca2+ pump in the plasma membrane, that mitochondria also plays some role, and that other mechanisms such as Ca2+ uptake into Ca2+ stores and Ca2+ extrusion via Na+ /Ca2+ exchange certainly are not present in the nerve terminal of the neurohypophysis (Fig. 8). Acknowledgement We are grateful to the Grant-in-aid from the Ministry of Education and Science of Japan (15590216). References [1] M. Cazalis, G. Dayanithi, J.J. Nordmann, The role of patterned burst and interburst interval on the excitation-coupling mechanism in the isolated rat neural lobe, J. Physiol. 369 (1985) 45–60. [2] J.J. Nordmann, G. Dayanithi, Release of neuropeptides does not only occur at nerve terminals, Biosci. Rep. 8 (1988) 471–483. [3] A. Dutton, R.E.J. Dyball, Phasic firing enhances vasopressin release from the neurohypophysis, J. Physiol. 290 (1987) 433–440. [4] M. Cazalis, G. Dayanithi, J.J. Nordmann, Requirements for hormone release from permeabilized nerve endings isolated from the rat neurohypophysis, J. Physiol. 390 (1987) 71–91. [5] M. Cazalis, G. Dayanithi, J.J. Nordmann, Hormone release from isolated nerve endings of the rat neurohypophysis, J. Physiol. 390 (1987) 55–70. [6] G. Dayanithi, N. Martin-Moutot, S. Barlier, D.A. Colin, M. KretzZaepfel, F. Couraud, J.J. Nordmann, The calcium channel antagonist omega-conotoxin inhibits secretion from peptidergic nerve terminals, Biochem. Biophys. Res. Commun. 156 (1988) 255–262. [7] G. Wang, G. Dayanithi, S. Kim, D. Hom, L. Nadasdi, R. Kristipati, J. Ramachandran, E.L. Stuenkel, J.J. Nordmann, R. Newcomb, J.R. Lemos, Role of Q-type Ca2+ channels in vasopressin secretion from neurohypophysial terminals of the rat, J. Physiol. 502 (1997) 351–363. [8] J.L. Branchaw, S.F. Hsu, M.B. Jackson, Membrane excitability and secretion from peptidergic nerve terminals, Cell Mol. Neurobiol. 18 (1998) 45–63. [9] D. Brethes, G. Dayanithi, L. Letellier, J.J. Nordmann, Depolarization-induced Ca2+ increase in isolated neurosecretory nerve terminals measured with fura-2, Proc. Natl. Acad. Sci. U.S.A. 84 (1987) 1439–1443. [10] G.J. Augustine, M.P. Charlton, S.J. Smith, Calcium action in synaptic transmitter release, Annu. Rev. Neurosci. 10 (1987) 633–693. [11] J.R. Lemos, M.C. Nowycky, Two types of calcium channels coexist in peptide-releasing vertebrate nerve terminals, Neuron 2 (1989) 1419–1426. [12] X. Wang, S.N. Treistman, J.R. Lemos, Two types of high-threshold calcium currents inhibited by omega-conotoxin in nerve terminals of rat neurohypophysis, J. Physiol. 445 (1992) 181–199. [13] G. Wang, J.R. Lemos, Effects of funnel web spider toxin on Ca2+ currents in neurohypophysial terminals, Brain Res. 663 (1994) 215–222.
55
[14] T.E. Fisher, C.W. Bourque, Calcium-channel subtypes in the somata and axon terminals of magnocellular neurosecretory cells, Trends Neurosci. 19 (1996) 440–444. [15] G. Wang, G. Dayanithi, R. Newcomb, J.R. Lemos, An R-type Ca2+ current in neurohypophysial terminals preferentially regulates oxytocin secretion, J. Neurosci. 19 (1999) 9235–9241. [16] R. Newcomb, B. Szoke, A. Palma, G. Wand, G. Dayanithi, X.-H. Chen, R. Cong, L. Urge, K. Tarczy-Homoch, J. Loo, D.J. Dooley, L. Nadasdi, J.R. Lemos, G. Miljanich, Novel class E calcium channel antagonist from Tarantula venom, C.N.S. Drug Rev. 6 (2000) 153–173. [17] Y. Yin, G. Dayanithi, J.R. Lemos, Ca2+ -regulated, neurosecretory granule channel involved in release from neurohypophysial terminals, J. Physiol. 539 (2002) 409–418. [18] D.D. Friel, R.W. Tsien, A caffeine- and ryanodine-sensitive Ca2+ store in bullfrog sympathetic neurones modulates effects of Ca2+ entry on [Ca2+ ]i , J. Physiol. 450 (1992) 217–246. [19] E.C. Toescu, Activity of voltage-operated calcium channels in rat cerebellar granule neurons and neuronal survival, Neuroscience 94 (1999) 561–570. [20] P.B. Simpson, J.T. Russell, Mitochondrial Ca2+ uptake and release influence metabotropic and ionotropic cytosolic Ca2+ responses in rat oligodendrocyte progenitors, J. Physiol. 508 (1998) 413–426. [21] A. Verkhratsky, E.C. Toescu, Calcium and neuronal ageing, Trends Neurosci. 21 (1998) 2–7. [22] Z. Ahmed, J.A. Connor, Calcium regulation by and buffer capacity of molluscan neurons during calcium transients, Cell Calcium 9 (1988) 57–69. [23] S.A. Thayer, R.J. Miller, Regulation of the intracellular free calcium concentration in single rat dorsal root ganglion neurones in vitro, J. Physiol. 425 (1990) 85–115. [24] E. Neher, G.J. Augustine, Calcium gradients and buffers in bovine chromaffin cells, J. Physiol. 450 (1992) 273–301. [25] E.C. Toescu, Intraneuronal Ca2+ stores act mainly as a ’Ca2+ sink’ in cerebellar granule neurones, Neuroreport 9 (1998) 1227–1231. [26] E.C. Toescu, Mitochondria and Ca2+ signaling, J. Cell Mol. Med. 4 (2000) 164–175. [27] E.C. Toescu, A. Verkhratsky, Neuronal ageing from an intraneuronal perspective: roles of endoplasmic reticulum and mitochondria, Cell Calcium 34 (2003) 311–323. [28] A. Verkhratsky, The endoplasmic reticulum and neuronal calcium signalling, Cell Calcium 32 (2002) 393–404. [29] A. Verkhratsky, O.H. Petersen, Neuronal calcium stores, Cell Calcium 24 (1998) 333–343. [30] M.J. Berridge, The endoplasmic reticulum: a multifunctional signaling organelle, Cell Calcium 32 (2002) 235–249. [31] M.D. Bootman, O.H. Petersen, A. Verkhratsky, The endoplasmic reticulum is a focal point for co-ordination of cellular activity, Cell Calcium 32 (2002) 231–234. [32] J.T. Russell, N.A. Thorn, Isolation and purification of calciumbinding proteins from bovine neurohypophyses, Biochim. Biophys. Acta. 491 (1977) 398–408. [33] E.L. Stuenkel, Regulation of intracellular calcium and calcium buffering properties of rat isolated neurohypophysial nerve endings, J. Physiol. 481 (1994) 251–271. [34] E.C. Toescu, J.J. Nordmann, Effect of sodium and calcium on basal secretory activity of rat neurohypophysial peptidergic nerve terminals, J. Physiol. 433 (1991) 127–144. [35] E.L. Stuenkel, J.J. Nordmann, Intracellular calcium and vasopressin release of rat isolated neurohypophysial nerve endings, J. Physiol. 468 (1993) 335–355. [36] E.L. Stuenkel, J.J. Nordmann, Sodium-evoked, calcium-independent vasopressin release from rat isolated neurohypophysial nerve endings, J. Physiol. 468 (1993) 357–378. [37] S.H. Lee, M.H. Kim, K.H. Park, Y.E. Earm, W.K. Ho, K+ -dependent Na+ /Ca2+ exchange is a major Ca2+ clearance mechanism in axon terminals of rat neurohypophysis, J. Neurosci. 22 (2002) 6891–6899.
56
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
[38] N. Solovyova, N. Veselovsky, E.C. Toescu, A. Verkhratsky, Ca2+ dynamics in the lumen of the endoplasmic reticulum in sensory neurons: direct visualization of Ca2+ -induced Ca2+ release triggered by physiological Ca2+ entry, Embo. J. 21 (2002) 622–630. [39] V. de Crescenzo, R. ZhuGe, C. Velazquez-Maarrero, L.M. Lifshitz, E. Custer, J. Carmichael, A. Lai, R.A. Tuft, K.E. Fogarty, J.R. Lemos, J.V. Walsh Jr., Ca2+ syntillas, miniature Ca2+ release events in terminals of hypothalamic neurons, are increased in frequency by depolarization in the absence of Ca2+ influx, J. Neurosci. 24 (2004) 1226–1235. [40] K. Tanaka, I. Shibuya, Y. Uezono, Y. Ueta, Y. Toyohira, N. Yanagihara, F. Izumi, T. Kanno, H. Yamashita, Pituitary adenylate cyclaseactivating polypeptide causes Ca2+ release from ryanodine/caffeine stores through a novel pathway independent of both inositol trisphosphates and cyclic AMP in bovine adrenal medullary cells, J. Neurochem. 70 (1998) 1652–1661. [41] I. Shibuya, K. Tanaka, Y. Hattori, Y. Uezono, N. Harayama, J. Noguchi, Y. Ueta, F. Izumi, H. Yamashita, Evidence that multiple P2X purinoceptors are functionally expressed in rat supraoptic neurones, J. Physiol. 514 (1999) 351–367. [42] G. Grynkiewicz, M. Poenie, R.Y. Tsien, A new generation of Ca2+ indicators with greatly improved fluorescence properties, J. Biol. Chem. 260 (1985) 3440–3450. [43] N.C. McNaughton, D. Bleakman, A.D. Randall, Electrophysiological characterisation of the human N-type Ca2+ channel II: activation and inactivation by physiological patterns of activity, Neuropharmacology 37 (1998) 67–81. [44] Y.B. Park, J. Herrington, D.F. Babcock, B. Hille, Ca2+ clearance mechanisms in isolated rat adrenal chromaffin cells, J. Physiol. 492 (1996) 329–346.
[45] R.J. Miller, Calcium signalling in neurons, Trends Neurosci. 11 (1988) 415–419. [46] R.J. Miller, The control of neuronal Ca2+ homeostasis, Prog. Neurobiol. 37 (1991) 255–285. [47] R.J. Miller, Neuronal Ca2+ : getting it up and keeping it up, Trends Neurosci. 15 (1992) 317–319. [48] K. Dunlap, J.I. Luebke, T.J. Turner, Exocytotic Ca2+ channels in mammalian central neurons, Trends Neurosci. 18 (1995) 89– 98. [49] M. Wakamori, M. Strobeck, T. Niidome, T. Teramoto, K. Imoto, Y. Mori, Functional characterization of ion permeation pathway in the N-type Ca2+ channel, J. Neurophysiol. 79 (1998) 622– 634. [50] M. Ludwig, N. Sabatier, P.M. Bull, R. Landgraf, G. Dayanithi, G. Leng, Intracellular calcium stores regulate activity-dependent neuropeptide release from dendrites, Nature 418 (2002) 85– 89. [51] R.C. Lambert, G. Dayanithi, F.C. Moos, P. Richard, A rise in the intracellular Ca2+ concentration of isolated rat supraoptic cells in response to oxytocin, J. Physiol. 478 (1994) 275–287. [52] Z. Li, C. Decavel, G.I. Hatton, Calbindin-D28k: role in determining intrinsically generated firing patterns in rat supraoptic neurones, J. Physiol. 488 (1995) 601–608. [53] M. Muschol, P. Kosterin, M. Ichikawa, B.M. Salzberg, Activitydependent depression of excitability and calcium transients in the neurohypophysis suggests a model of “stuttering conduction”, J. Neurosci. 23 (2003) 11352–11362. [54] Y. Komori, I. Shibuya, M. Kuba, Y. Ueta, G. Dayanithi, Ca2+ Clearance and somatodendritic vasopressin release of rat supraoptic neurons, Jpn. J. Physiol. 52 (2002) S197.
Ca2+ clearance mechanisms in neurohypophysial terminals of the rat Naoko Sasakia , Govindan Dayanithib , Izumi Shibuyaa,∗ a
Deparment of Physiology, University of Occupational and Environmental Health School of Medicine, Kitakyushu 807-8555, Japan b U-583 INSERM, L’Institut des Neurosciences de Montpellier, Hˆ opital St Eloi, F-34295 Montpellier Cedex 5, France Received 19 March 2004; received in revised form 14 June 2004; accepted 22 June 2004
Abstract The importance of intracellular calcium ([Ca2+ ]i ) in the release of vasopressin (AVP) and oxytocin from the central nervous system neurohypopyhysial nerve terminals has been well-documented. To date, there is no clear understanding of Ca2+ clearance mechanisms and their interplay with transmembrane Ca2+ entry, intracellular [Ca2+ ]i transients, cytoplasmic Ca2+ stores and hence the release of AVP at the level of a single nerve terminal. Here, we studied the mechanism of Ca2+ clearance in freshly isolated nerve terminals of the rat neurohypophysis using Fura-2 Ca2+ imaging and measured the release of AVP by radioimmuno assay. An increase in the K+ concentration in the perfusion solution from 5 to 50 mM caused a rapid increase in [Ca2+ ]i and AVP release. Returning K+ concentration to 5 mM led to rapid restoration of both responses to basal level. The K+ -evoked [Ca2+ ]i and AVP increase was concentration-dependent, reliable, and remained of constant amplitude and time course upon successive applications. Extracellular Ca2+ removal completely abolished the K+ -evoked responses. The recovery phase was not affected upon replacement of NaCl with sucrose or drugs known to act on intracellular Ca2+ stores such as thapsigargin, cyclopiazonic acid, caffeine or a combination of caffeine and ryanodine did not affect either resting or K+ -evoked [Ca2+ ]i or AVP release. By contrast, the plasma membrane Ca2+ pump inhibitor, La3+ , markedly slowed down the recovery phase. The mitochondrial respiration uncoupler, carbonyl cyanide 3-chlorophenylhydrazone (CCCP), slightly but significantly increased the basal [Ca2+ ]i , and also slowed down the recovery phase of both [Ca2+ ]i and release responses. In conclusion, we show in nerve terminals that (i) Ca2+ extrusion through the Ca2+ pump in the plasma membrane plays a major role in the Ca2+ clearance mechanisms of (ii) Ca2+ uptake by mitochondria also contributes to the Ca2+ clearance and (iii) neither Na+ /Ca2+ exchangers nor Ca2+ stores are involved in the Ca2+ clearance or in the maintenance of basal [Ca2+ ]i or release of AVP. © 2004 Elsevier Ltd. All rights reserved. Keywords: Hypothalamus; Posterior pituitary; Exocytosis; Vasopressin release
1. Introduction The hypothalamo-neurohypophyseal system has been the focus of much research since decades as it provides the rare opportunity to study, in a single system, the bioelectrical activity of individual neurosecretory cells and the mechanisms by which these cells release their products [1–3]. The magnocellular cell bodies, which are located in the hypothalamic supraoptic and paraventricular nuclei, synthesize the neuropeptides oxytocin and vasopressin, that are released from axon terminals at the neurohypophysis directly into the blood system to ultimately act as hormones. In the neural ∗
Corresponding author. Tel.: +81 93 691 7420; fax: +81 93 692 1711. E-mail address: [email protected] (I. Shibuya).
0143-4160/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ceca.2004.06.007
lobe, the mechanisms underlying depolarization–secretion coupling have been widely investigated and characterized [4–8]. Briefly, exocytosis results from a sequence of events that starts by the arrival of action potentials which promote the opening of voltage-sensitive Ca2+ channels, and includes cytoplasmic-free calcium concentration ([Ca2+ ]i ) increase as a key player in release [9,10]. The different Ca2+ -channel subtypes and their role in peptide release in neurohypophysial nerve terminal have also been characterized [7,11–17]. In most cell types, increase in [Ca2+ ]i results both from 2+ Ca entry mediated by voltage-gated Ca2+ channels [18,19] and from Ca2+ release from intracellular stores [20,21]. A large body of data is presently available in the literature regarding the mechanisms involved in intracellular Ca2+ buffering and restoration to basal level [22–24]. Such mech-
46
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
anisms include Ca2+ pumps in both intracellular compartments and the plasma membrane, mitochondrial Ca2+ uptake, plasma membrane Na+ /Ca2+ exchange under physiological conditions [25,26] or during aging [21,26,27]. Another wellknown neuronal Ca2+ store is endoplasmic reticulum (ER). The ER is multifunctional organelle regulating a wide range of neuronal functional responses [28–30]; a focal point for co-ordination of cellular activity [31]. The ER Ca2+ stores are shown to act as a Ca2+ -buffering system, i.e. ‘Ca2+ sink’, clearing cytoplasmic Ca2+ loads by SERCA pumping [18] and act as a source of Ca2+ for CICR phenomenon [29]. The Ca2+ -binding proteins were also isolated and purified from bovine neurohypophysis [32]. In neurohypophysial nerve terminals, it was reported that the Ca2+ -buffering mechanisms was sensitive to mitochondrial Ca2+ uptake but not or minimal Na+ /Ca2+ exchange is important in Ca2+ clearance mechanism [33], thus requiring an efficient Ca2+ buffering to allow sustainable release. The role of Na+ /Ca2+ exchange in the release of AVP from these nerve terminals, under basal and stimulated conditions was also demonstrated [34–36]. But recently in the same model, the presence of K+ -dependent Na+ /Ca2+ exchanger was reported [37] and it seems to play a role in Ca2+ clearance. Although the importance of intracellular Ca2+ stores in Ca2+ homeostasis has been well-documented in various tissues [26,38], controversial findings have been reported in the isolated nerve terminals [33,37,39]. These important discrepancies led us to re-investigate the exact mechanisms by which the neurohypophysial terminals regulate the intracellular Ca2+ and their buffering properties. Here we bring clear evidence that Ca2+ clearance mechanisms and hence efficient AVP release can be operated by cellular/membrane components other than Na+ /Ca2+ exchanger or Ca2+ stores. We further highlight the understanding of the fine interplay between transmembrane Ca2+ entry, intracellular [Ca2+ ]i transients, cytoplasmic Ca2+ stores, Ca2+ clearance mechanisms and AVP release at the level of isolated nerve terminals.
2. Material and methods 2.1. Preparation of isolated neurohypophysial nerve terminals
contains highly purified nerve terminals [5,6], was eventually resuspended in Locke buffer maintained at 37 ◦ C. Aliquots of 200 l of suspension were then plated on glass bottom culture dishes for [Ca2+ ]i measurements. 2.2. Drugs and solutions The Locke buffer described here was used both during dissection and as a control solution. It contained (mM): NaCl, 140; KCl, 5; MgCl2 , 1.2; CaCl2 , 2.2; glucose, 10; HEPES, 10; BSA, 0.02%, pH 7.25 adjusted with Tris–HCl. The osmolarity of all the solutions used in this study was maintained between 298 and 300 mOsm l−1 . To test the effects of extracellular Ca2+ , Ca2+ /EGTA buffer (see [4]) was used which contained (mM): ethylene glycol (bis-aminoethyl ether)N,N,N ,N -tetra acetic acid (EGTA), 2; NaCl, 140; KCl, 5; MgCl2 , 1.2; CaCl2 , 2.2; glucose, 10; HEPES, 10; pH 7.25 adjusted with Tris–HCl. In this buffer, the free Ca2+ concentration in the EGTA buffer was around 100 nM which corresponds to the resting [Ca2+ ]i typically measured in neurones. Buffer containing high K+ commonly used in this study contained (mM): NaCl, 90; KCl, 50; MgCl2 , 1.2; CaCl2 , 2.2; glucose, 10; HEPES, 10 at pH 7.25. For other K+ concentrations, KCl was added to the desired concentration and was adjusted with NaCl appropriately to bring osmolarity in the required range. For Na+ -free buffer, NaCl was replaced by appropriate concentrations of sucrose. Most of the standard chemicals were purchased from Sigma. Thapsigargin (TG), ryanodine, cyclopiazonic acid (CPA; Alomone labs, Jerusalem, Israel) and carbonyl cyanide 3-chlorophenylhydrazone (CCCP; Sigma), were first dissolved in DMSO (final concentration: 0.01%) and further diluted in working solution to appropriate concentrations. La3+ (Sigma) and Caffeine (Alomone labs) were directly dissolved in the working solution. Fura-2 AM (in DMSO solution) and Pluronic F-127 (30% stock in distilled water) were from Molecular Probes Inc., Eugene, OR, USA). 125 Ilabelled AVP was purchased from Amersham, France. All test substances for [Ca2+ ]i measurements were perfused in a 500 l chamber using a multi-channel Gilson peristaltic pump (1.5 ml/min). This method allowed fast and reliable exchange of the solution surrounding the cells. 2.3. [Ca2+ ]i measurements
Isolated nerve terminals were prepared from male Wistar rats (about 200 g) as previously described [5]. After animals were killed by decapitation with a guillotine following the guidelines laid down by the French/Japanese ethical committee, each pituitary was dissected out and transferred in Locke buffer (see the following). After removal of the anterior lobe and the pars intermedia under binocular control, the neural lobes were homogenized at 37 ◦ C in 1 ml of a solution containing (in mM): sucrose, 270; EGTA, 0.2; HEPES, 10; pH 7.25 adjusted with Tris. The dissociated neurohypophysis was first spun at 100 × g for 1 min, then the supernatant was further spun at 2400 × g for 4 min. The final pellet, which
The isolated nerve terminals plated on a cover glass attached to 35 mm culture dish were incubated 2.5 M Fura2 AM plus 0.02% Pluronic F-127 at room temperature for 1 h. The preparations were then washed with dye-free solution and kept at room temperature (about 23 ◦ C) until used. Peptidergic nerve terminals were identified using phase contrast optics. The arrangements for perfusion and measuring fluorescence are essentially described previously [40,41]. In brief, fluorescence was measured from fura-2-loaded nerve terminals in the perfusion chamber, which was positioned on the stage of an inverted microscope (IX-50; Olympus, Tokyo,
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Japan), using a Ca2+ -imaging system equipped with a 12-bit digital CCD camera (Quanticell/900; JEOL, Tokyo, Japan). Fluorescence intensities at 510 nm with excitation at 340 and 380 nm were recorded at an interval of 5 s. The [Ca2+ ]i in individual nerve terminal was calculated from the ratio of the fluorescence images measured with excitation at 340 nm to those at 380 nm using the equation of Grynkiewicz et al. [42]. Autofluorescence in the nerve terminals was negligible compared with the fluorescence in the fura-2 loaded terminals. Values of Rmin , Rmax and Fo /Fs (i.e. ratio of emitted fluorescence intensity at 380 nm excitation under Ca2+ limiting and Ca2+ saturating conditions) were determined using solutions of defined Ca2+ concentrations and were 0.3, 3.4 and 5.7, respectively. The Ca2+ Kd (=224) value for Fura-2 was taken from the paper by Grynkiewicz et al. [42]. Photobleaching was minimized by decreasing illumination intensity with neutral filters and by slowing the acquisition frame to a low rate. 2.4. Measurements of AVP release Release experiments were performed as described previously [5]. Briefly, for each experimental run, the isolated terminals from two neurohypophysis were equally distributed and loaded onto four filters, i.e. each filter receives about 25% of the total terminals. Following loading of isolated nerve terminals onto filters (0.45 m Acrodisc, LCP-VDF, Gelman Sciences, USA), they were perfused (Minipulse Peristaltic Pump, Gilson, France) for 45 min with Normal Locke at a flow rate of 50 l min−1 and the flow rate was increased to
47
100 l min−1 for 15 min. Collection of the perfusate over 5min periods started 60 min after loading the nerve terminals onto the filter. AVP content in each fraction was then determined by radioimmunoassay as already described [1,5] with necessary modifications [7]. The antisera were kindly given by Dr. John Bicknell, Babraham, UK. The AVP was raised in sheep. The final antibody dilution was 1:70,000. The crossreactivity of the AVP antiserum with OT was 0.001%. The sensitivity of the assay was 0.5 pg. The inter- and intra-assay coefficients of variation were 5–7%. The results represent the average AVP content of standard aliquots taken from the collected fractions arising from at least three separate groups of isolated nerve terminals. Internal control groups arising from the same pool of isolated nerve terminals were run for comparison to test groups. Differences in the total amount of AVP released between experiments resulted from differences in the amount of each preparation loaded onto the filters and from heterogeneity of secretary responsiveness between preparations. The mean evoked AVP release was calculated by subtracting the AVP release under basal conditions, i.e. prior to the stimulus from the peak amplitude observed after the application of the test substances. 2.5. Data analysis and statistics Results are expressed as mean ± S.E.M. Effects of test solutions on the basal [Ca2+ ]i level were expressed as percents of the mean [Ca2+ ]i value measured during 1 min just before
Fig. 1. Representative examples for fura-2 loaded isolated nerve terminals of rats and their response to 50 mM K+ . (A and B) Time courses of [Ca2+ ]i before, during and after 50 mM K+ challenge. The numbers (1–9) in the figures correspond to the number depicted in the images. The two round cells (see C) showed no response (B), whereas all nerve terminals showed robust increases in [Ca2+ ]i (A). (C) Images of fura-2 loaded nerve terminals taken in a phase contrast field (left most) and pseudo-color [Ca2+ ]i images before (a), during (b) and after (c) 50 mM K+ challenge (other three). The time when (a–c) were taken was shown with the arrows in (A) and (B). [Ca2+ ]i values are shown in the horizontal bar below.
48
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
the treatment, whereas the effects on the recovery phase time constant (τ) of [Ca2+ ]i after high K+ challenge were expressed as percents of the mean of responses measured prior and after the test in the same nerve terminal preparations. The Axograph software (version 4.0, Axon Instruments Inc., USA) was used for curve fitting. Statistical differences (P < 0.05) were determined by a Mann–Whitney’s U test when “n” was equal to or larger than 10, or by a Student’s t-test when “n” was smaller than 10.
3. Results 3.1. [Ca2+ ]i increases induced by high K+ Fig. 1 shows a typical recording of the [Ca2+ ]i increase induced by exposure to high K+ (50 mM) measured in an isolated nerve terminal. Increasing the K+ concentration in the perfusion solution from 5 to 50 mM for 30 s caused a fast and large increase in [Ca2+ ]i in all nerve terminals tested (Fig. 1A; mean K+ evoked [Ca2+ ]i increase, 902 ± 96 nM; resting [Ca2+ ]i , 59 ± 7 nM; n = 13). When viewed under
the microscope, healthy nerve terminals displayed irregular shapes and robust [Ca2+ ]i increases under high K+ exposure, therefore allowing easy identification (Fig. 1C). By contrast, the rare round cells observed in the preparation did not display any [Ca2+ ]i change under high K+ challenge (Fig. 1B). Upon returning the K+ concentration to 5 mM, [Ca2+ ]i in nerve terminals rapidly recovered (Fig. 1A and C). The amplitude of the [Ca2+ ]i increase (present study) or AVP release (see [4,5]) induced by exposure to a high K+ solution is concentration-dependent. As shown in Fig. 2A, exposing the nerve terminals to 30-s tests of increasing K+ concentration (10, 20, 30, 50 and 100 mM) evoked [Ca2+ ]i responses (Fig. 2B, respectively, 45 ± 6 nM; 119 ± 12 nM; 226 ± 26 nM; 649 ± 64 nM; 711 ± 70 nM; resting [Ca2+ ]i , 174 ± 17 nM; n = 25) displaying a single peak with progressively larger amplitude while the time course remained the same. Repetitive 30 s exposures to a constant high K+ concentration at intervals of 3–5 min caused comparable increases in [Ca2+ ]i (Fig. 2C; peak [Ca2+ ]i responses, respectively, 621 ± 47 nM; 630 ± 43 nM; 644 ± 50 nM; 650 ± 48 nM; resting [Ca2+ ]i , 68 ± 6 nM; n = 24) showing no significant inactivation.
Fig. 2. Increase in [Ca2+ ]i of isolated nerve terminals. (A) Traces represent changes in [Ca2+ ]i evoked by 30 s applications of increasing concentrations of K+ in a single neurohypophysial terminal. (B) The graph shows the dose–response relationship of evoked [Ca2+ ]i increase (mean ± S.E.M) as a function of K+ concentration. The resting [Ca2+ ]i concentration measured from each cell was subtracted from the peak amplitude evoked by each concentration of K+ . Data were obtained from 13 terminals, using the protocol described in (A). (C) Traces show changes in [Ca2+ ]i induced by 50 mM K+ exposures of increasing durations (30 s to 5 min) measured in a single neurohypophysial nerve terminal. Note the progressive change of [Ca2+ ]i increases from a single-phased response to a biphasic one.
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
The time course of the [Ca2+ ]i increase depends on the duration of high K+ exposure. Upon 50 mM K+ exposures of increasing duration (30 s to 5 min), the [Ca2+ ]i response progressively evolved from displaying an increase with a single peak to a biphasic increase that consisted of a rapid rising phase followed by a gradually declining plateau phase (Fig. 2C). Although the late phase declined in the presence of high K+ , we never observed a complete recovery of [Ca2+ ]i under high K+ , at least for the durations tested here. This result may result from inactivation of voltage-gated Ca2+ channels in the nerve terminals, that was already reported by others [7]. In the present study, we focus on the single-phased [Ca2+ ]i response, elicited by a 30 s-exposure to 50 mM K+ . 3.2. Effects of external Ca2+ removal The high K+ -induced [Ca2+ ]i increase or AVP release was abolished when external Ca2+ was reduced to 100 nM (see Section 2) from the bathing solution (Fig. 3), confirming that at least major part of the [Ca2+ ]i increase resulted from Ca2+ entry through voltage-gated Ca2+ channels in nerve ter-
49
minals. Reduction of external Ca2+ (from 2.2 to 100 nM) in the perfusion solution did not significantly modify the resting [Ca2+ ]i (Fig. 3A; 138 ± 16 nM versus 150 ± 16 nM; n = 17), indicating that there is no Ca2+ entry in resting conditions. The rate of [Ca2+ ]i recovery after K+ challenge was quantified by exponential fitting. Two exponentials were required to adequately fit the [Ca2+ ]i recovery phase (Fig. 3B). The first τ value ranged from 5.7 to 19.9 s and the second ranged from 313 to 2953 s. Since the latter showed much larger variation than the former, we focused on the time constant (τ) of the fast component because the first τ value in this study. Fig. 3C shows the AVP release responses in the presence of 2.2 or 100 nM external Ca2+ under basal or K+ -stimulated conditions. These results indicate that depolarizations-induced AVP release is dependent on the Ca2+ influx. 3.3. Effects of external Na+ replacement To examine the role of Na+ /Ca2+ exchange in intracellular clearance, Na+ in the perfusion solution was replaced by sucrose. As illustrated in Fig. 4A, substitution of Na+ had
Ca2+
Fig. 3. Effect of extracellular Ca2+ on K+ -induced [Ca2+ ]i and AVP release. (A) [Ca2+ ]i response to 50 mM K+ in the presence or absence of 2 mM external Ca2+ . The trace shows that 100 nM low Ca2+ -Locke buffer (indicated by open bar) totally suppressed the [Ca2+ ]i increase induced by a 30 s exposure to 50 mM K+ in neurohypophysial nerve terminals. The [Ca2+ ]i response rapidly recovered upon return to normal external Ca2+ . (B) An example of exponential curve fitting. The dotted line is exponential curve drawn with the time constant of 8.8 s. (C) The time course of AVP release from isolated nerve terminals. All data points are means (±S.E.) of five experiments.
50
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Fig. 4. Effect of external Na+ on basal and K+ -induced [Ca2+ ]i and AVP release. (A) Example of a recording showing that replacement of Na+ by sucrose in the bath had not effect neither on basal [Ca2+ ]i or on 50 mM K+ -induced [Ca2+ ]i increase. (B) Bar diagram represents the mean peak amplitude (percent of control) of the [Ca2+ ]i response observed in Na+ -free buffer (hatched column) and in control conditions after recovery (open column). All data points are means (±S.E.) of 10 experiments. (C) Fifty millimoles of K+ -induced AVP release profiles in the presence or absences of external Na+ . Data points are means (±S.E.) of five experiments.
no effect either on the basal [Ca2+ ]i (75 ± 11 nM versus 81 ± 13 nM; n = 10) or on the recovery phase after high K+ exposure (τ 10.32 ± 1.40 s versus τ 12. 07 ± 1.68 s; n = 10). The values of basal [Ca2+ ]i and τ in control versus Na+ free conditions obtained from 10 terminals are summarized in Fig. 4B. Similarly, the release of AVP under basal (34 ± 3 pg versus 40 ± 6 pg) or K+ -stimulated conditions (513 ± 33 pg versus 461 ± 43 pg; n = 4) was not affected by the removal of external Na+ . 3.4. Effects of drugs targeting Ca2+ stores To examine the role of Ca2+ stores, the response to high was observed in the presence of 1 M TG, or 30 M CPA, both of which are known to block Ca2+ pumps located on intracellular Ca2+ stores, or with a combination of 10 M ryanodine and 10 mM caffeine, which are known to deplete caffeine/ryanodine Ca2+ stores and thereby suppress subsequent Ca2+ release from such stores. Our data clearly show that neither TG (Fig. 5A; 151 ± 23 nM versus 142 ± 18 nM; n = 10), nor CPA (Fig. 5B; 303 ± 81 nM; n = 5) nor caffeine/ryanodine (298 ± 78 nM; n = 5) had a significant effect
on basal [Ca2+ ]i (297 ± 75 nM; n = 5) nor on the recovery phase after K+ challenge (Fig. 5A and B). The results of these experiments are summarized in Fig. 5C. In addition, basal AVP or K+ -evoked AVP was not affected in the absence or presence of TG or caffeine/ryanodine or CPA (Fig. 5D; basal AVP level, 63 ± 8 pg; K+ -evoked release, 285 ± 38 pg; K+ + TG, 287 ± 50 pg; K+ + caffeine/ryanodine, 291 ± 41 pg; K+ + CPA, 276 ± 30 pg; n = 4). We have also performed experiments using 10 mM caffeine alone in the presence or absence of external Ca2+ (2 mM). Two minutes continuous application of caffeine had no effect on the resting [Ca2+ ]i levels in all nerve terminals tested (n = 16). These terminals have responded only to high 50 mM K+ (data not shown).
K+
3.5. Effects of the mitochondrial uncoupler, CCCP To assess the role of mitochondrial Ca2+ uptake, the uncoupler of mitochondrial respiratory chain (CCCP) was used. CCCP (10 M) caused a small but significant increase in the resting [Ca2+ ]i (63 ± 10 nM versus 85 ± 13 nM; n = 10; P < 0.05) and significantly slowed down the recovery of [Ca2+ ]i (Fig. 6A; τ 9.07 ± 0.35 s versus τ 13.8 ± 1.12 s; n = 10; P<
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
51
Fig. 5. No effect of intracellular Ca2+ mobilizers on basal and K+ -induced [Ca2+ ]i increase or AVP release. (A) Trace showing a typical K+ (50 mM)-induced [Ca2+ ]i rise measured in the presence or absence of TG in an isolated nerve terminal. (B) In a different terminal, the effect of mixture of caffeine and ryanodine (Caf/Rya) or CPA was first on basal [Ca2+ ]i and then on K+ -induced [Ca2+ ]i increase. Note that none of the drugs did mobilize the Ca2+ from intracellular stores, as illustrated by the absence of any effect on basal [Ca2+ ]i . In addition, no effect of the drugs could be observed on the [Ca2+ ]i response to high K+ . (C) The bar graph represents the mean basal or evoked [Ca2+ ]i increase (expressed as percent of control) elicited by high K+ in absence (open column) and in presence (hatched column) of TG, Caf/Rya or CPA. Data are the means (±S.E.) of 5–10 experiments. (D) Basal or stimulated AVP release profiles of isolated nerve terminals in the presence or absence of above drugs. Data points are means (±S.E.) of five experiments.
0.05). Both these effects were rapidly reversible upon wash. The effects of CCCP on the basal [Ca2+ ]i and τ (n = 10) are summarized in Fig. 6B. Similarly, the resting AVP (47 ± 12 pg versus with CCCP, 164 ± 17 pg; n = 5; P < 0.05) or K+ evoked AVP release (K+ , 288 ± 26 pg versus K+ + CCCP, 802 ± 34 pg; n = 5; P < 0.05) was significantly altered by CCCP (Fig. 6C). 3.6. Effects of La3+ La3+ , a blocker of voltage-gated calcium channels [43], has also been reported to block the plasma membrane Ca2+ pump in some cell preparation [24,44]. To study further the effects of La3+ on nerve terminals, we first assessed its effects on basal and K+ -induced [Ca2+ ]i increase and AVP release responses. We observed that although La3+ (300 M) did not show significant effect on basal [Ca2+ ]i (125 ± 6 nM versus 121 ± 6 nM; n = 16) but it totally and reversibly blocked the high K+ -induced [Ca2+ ]i increase (Fig. 7A) and maintained the AVP release response due to increase in cy-
tosolic [Ca2+ ]I levels. These results strongly suggest that, as reported in other preparations, La3+ could inhibit voltagedependent Ca2+ channels in nerve terminals. Therefore, to assess the role of the plasma membrane Ca2+ pump without interfering with voltage-gated calcium channels, we only added La3+ (300 M) at the time high K+ was switched off from the bath (Fig. 7B). In all 10 terminals examined, La3+ nearly stopped the recovery of [Ca2+ ]i , preventing its return to the baseline for the duration of the exposure. This effect was reversible upon wash. The AVP release experiments also revealed that K+ -evoked response (432 ± 33 pg) was potentiated after La3+ application (673 ± 14 pg; n = 5; P < 0.05) as a consequence of sustained increase in [Ca2+ ]i (Fig. 7C).
4. Discussion In the present study we have investigated the cytosolic Ca2+ clearance mechanisms and Ca2+ regulation of AVP
52
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Fig. 6. Effect of CCCP on basal and K+ -induced [Ca2+ ]i and AVP responses. (A) The effect of the mitochondrial uncoupler, CCCP was assessed on the basal [Ca2+ ]i levels or the K+ -induced [Ca2+ ]i response. Profiles obtained from a single nerve terminal. (B) The bar graphs show the mean basal or stimulated [Ca2+ ]i response, expressed as percent of control, calculated from the peak amplitude responses (n = 10). Significant level: P < 0.05. (C) AVP release responses to CCCP. Data points are means (±S.E.) of five experiments.
release in freshly isolated rat neurohypophysial nerve terminals. The intracellular Ca2+ concentration ([Ca2+ ]i ) was monitored on single nerve terminal level. The [Ca2+ ]i displays complex fluctuations in response to many stimuli, and acts as signal for several neuronal functions [21]. The complex neuronal signaling involves activation of receptors, which leads to fluctuations in the [Ca2+ ]i that occurs different locations within the cell (see [45–47]). Increasing number of evidence has shown in many models that the Ca2+ signal is initiated by receptor- or voltage-gated opening of Ca2+ channels (see review [48]). It has been clear during recent years that the intracellular Ca2+ storage compartments such as mitochondria and the endoplasmic reticulum are also playing crucial roles in regulating Ca2+ signals and compartmentalized functions. These intracellular Ca2+ storage compartments are shown to initiate Ca2+ signals and also serve as a Ca2+ -buffering system. The Ca2+ -sensitive stores might act as a sink for increases in [Ca2+ ]i or might play an important role in neuronal aging [27].
4.1. The Ca2+ clearance mechanism of neurohypophysial nerve terminals The treatment that produced the largest effect on Ca2+ recovery phase after high K+ challenge was La3+ , the blocker of the plasma membrane Ca2+ pump. La3+ virtually arrested Ca2+ recovery. The result suggests that Ca2+ extrusion through the plasma membrane Ca2+ pump to the extracellular space may be the most critical mechanism for the Ca2+ clearance of the nerve terminal. La3+ is known to block other Ca2+ transport mechanisms such as Ca2+ channels and Na+ /Ca2+ exchange [49], however, the result that Na+ replacement with NMDG had little effect on the Ca2+ recovery phase suggest that the involvement of the Na+ /Ca2+ exchange is minor and that the arrest of Ca2+ recovery by La3+ was indeed due to the block of the Ca2+ pump. In contrast, in the same nerve terminals, it was claimed that Na+ /Ca2+ exchange is a major Ca2+ clearance mechanism [37] but our present results as well as the results from Stuenkel [33] on [Ca2+ ]i increase or AVP release have ruled out the above pos-
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
53
Fig. 7. Effect of La3+ on basal and K+ -induced [Ca2+ ]i and AVP responses. (A) The [Ca2+ ]i traces show the effect of La3+ under basal conditions followed by stimulation with high K+ . Note that the [Ca2+ ]i increase induced by high K+ was totally suppressed in the concomitant presence of La3+ in the bath. (B) To avoid voltage-gated calcium channels inhibition and allow proper assessment of the role of the plasma membrane Ca2+ -pump, La3+ was added to the bath at the end of high K+ stimulation. In these conditions, La3+ exposure almost prevented the high K+ -induced [Ca2+ ]i to recover. Similar results were obtained in 14 of nerve terminals. (C) AVP release profiles are shown under similar experimental conditions and (D) represents evoked AVP release. Data points are means (±S.E.) of five experiments.
sibility or any significant physiological role for Na+ /Ca2+ exchange. Another important and new finding in this study is concerning the role of mitochondria in the Ca2+ clearance mechanism. The mitochondrial uncoupler, CCCP, showed significant effect on the recovery phase as well as the basal [Ca2+ ]i and peptide release, suggesting only mitochondria plays an important role in the Ca2+ clearance of the nerve terminal both in basal (unstimulated) conditions. The significant effect of CCCP on the recovery phase suggests that mitochondria play a role also when Ca2+ overload occurred. In fact, electronmicrographs of isolated nerve terminal preparations as well as intact neurohypophysis have clearly showed the presence of mitochondria [2] (see also Fig. 8). CCCP, however, did not affect the peak [Ca2+ ]i level reached during the 50 mM K+ challenge. This suggests that mitochondria does not play a role as a Ca2+ sink during such rapid Ca2+ overload. It could be that Ca2+ uptake into mitochondria is not fast enough to counteract a massive [Ca2+ ]i increase caused by Ca2+ entry in response to 50 mM K+ . This interpretation may not be valid in conditions in which the fura-2 signal is saturated. Therefore, it should be confirmed with a lower affinity Ca2+ dye to make sure about the lack of differences in the amplitude of the [Ca2+ ]i signal dur-
ing the 50 mM K+ challenge in the presence or absence of CCCP. The lack of effect of the drugs acting on Ca2+ stores on the 2+ Ca -recovery phase as well as on basal [Ca2+ ]i and peptide release suggests that Ca2+ stores are absent or, if present, they play only minor physiological role in the nerve terminal. Surprisingly, using a signal mass approach to Ca2+ imaging in nerve terminals from mice, De Crescenzo et al. [39] has demonstrated an increase in [Ca2+ ]i due to caffeine in the absence of external Ca2+ , suggesting a possible presence of intracellular Ca2+ stores in the nerve terminals. However, such intracellular stores were not seen in the previous [33,35,50] and the present study from the neurohypophysial nerve terminals. One explanation for such contrasting results could be due to the use of high resolution camera to amplify the signal to noise ratio to observe small changes in the [Ca2+ ]i levels not only from the cytosol but also from the neurosecretory granules. Of interest, it should be noted that in a few terminals (about 10%) where the resting [Ca2+ ]i levels were very high (more than 400 nM), we observed an increase in [Ca2+ ]i due to caffeine (results not shown). In any case, our present results show that under our experimental conditions caffeine had no effect on AVP release and a previous report showed that caffeine had no effect on 45 Ca2+
54
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
Fig. 8. Schematic representations of possible Ca2+ clearance mechanisms in nerve terminals of the neurohypophysis. Arrows suggests that increased cytosolic Ca2+ can be extruded via plasma membrane (PM) Ca2+ pump or Na+ /Ca2+ exchange, or stored by mitochondrial or intracellular compartments (via store Ca2+ pump). We propose that only PM Ca2+ pump and mitochondria play a major role in Ca2+ clearance of the nerve terminals (indicated by thick solid line), whereas Na+ /Ca2+ exchange or intracellular Ca2+ stores do not (indicated by thin dotted line). The superimposed background picture is an electron micrograph of isolated nerve terminals, where large size mitochondria can be seen.
efflux [36]. These results suggest that the external Ca2+ independent increase in [Ca2+ ]i due to caffeine is not crucial to acute AVP secretion under physiological conditions. This shows clear contrast with Ca2+ signaling mechanisms of soma and dendritic part of SON neurons, where TG-sensitive Ca2+ stores play major roles in both Ca2+ signaling and peptide secretion [50,51]. The present results also suggest that, although previous immunohistochemical evidence indicated that Ca2+ -buffering proteins play important role in the soma and dendritic part of SON neurons [52] they may be absent or do not function in the terminal. 4.2. The correlation between [Ca2+ ]i and peptide secretion In most of the experiments in the present study, there was a good correlation between [Ca2+ ]i and AVP secretion. This is consistent with previous findings that Ca2+ plays a key role in the stimulus-secretion coupling in the neurohypophysial nerve terminals [2,9]. However, there was one exception. When La3+ was used to block the plasma membrane Ca2+ pump, the recovery of [Ca2+ ]i was arrested, whereas peptide secretion slowly decreased to the basal level. The difference can be explained by ‘fatigue’ of exocytotic machinery. It has been shown that the [Ca2+ ]i response to high K+ was well-sustained but the peptide secretion to high K+ was more
transient [5]. The [Ca2+ ]i response in the presence of La3+ was maintained stably high for more than 2 min, whilst the [Ca2+ ]i response to high K+ tends to be more transient. The response to high K+ seems to be because of inactivation of voltage-gated Ca2+ channels. Thus, the protocol of the La3+ experiment in this study may be a good tool to study the fatigue mechanism of exocytosis in the nerve terminal. 4.3. The physiological significance of the Ca2+ clearance mechanism It has been shown that phasic bursting activity of AVP neurons is directly related to AVP secretion and that the interval between phasic bursts as well as the frequency of bursts play an important role in the peptide secretion (see [1,53]). The importance of intervals between bursting and the frequency of the bursts suggests importance and necessity of quick Ca2+ clearance of a large capacity for peptide secretion. The La3+ -sensitive Ca2+ extrusion via the plasma membrane Ca2+ pump as well as respiration-coupled Ca2+ uptake into mitochondria revealed in the present study may serve as such large capacity Ca2+ clearance mechanisms. The present results together with the previous reports for Ca2+ stores and Ca2+ -binding proteins in the somato-dendritic region of SON neurons [50–52] also suggest that there are considerable differences in the Ca2+
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
signaling/clearance mechanisms between nerve terminals and somato-dendritic regions of SON neurons. Although the precise Ca2+ clearance mechanism in the somato-dendritic region of SON neurons remains unclear, our preliminary results indicate that ER could be a potential candidate to play a crucial role in Ca2+ clearance mechanisms [54]. In conclusion, the present study showed that the major Ca2+ clearance mechanism of nerve terminals is Ca2+ extrusion via the Ca2+ pump in the plasma membrane, that mitochondria also plays some role, and that other mechanisms such as Ca2+ uptake into Ca2+ stores and Ca2+ extrusion via Na+ /Ca2+ exchange certainly are not present in the nerve terminal of the neurohypophysis (Fig. 8). Acknowledgement We are grateful to the Grant-in-aid from the Ministry of Education and Science of Japan (15590216). References [1] M. Cazalis, G. Dayanithi, J.J. Nordmann, The role of patterned burst and interburst interval on the excitation-coupling mechanism in the isolated rat neural lobe, J. Physiol. 369 (1985) 45–60. [2] J.J. Nordmann, G. Dayanithi, Release of neuropeptides does not only occur at nerve terminals, Biosci. Rep. 8 (1988) 471–483. [3] A. Dutton, R.E.J. Dyball, Phasic firing enhances vasopressin release from the neurohypophysis, J. Physiol. 290 (1987) 433–440. [4] M. Cazalis, G. Dayanithi, J.J. Nordmann, Requirements for hormone release from permeabilized nerve endings isolated from the rat neurohypophysis, J. Physiol. 390 (1987) 71–91. [5] M. Cazalis, G. Dayanithi, J.J. Nordmann, Hormone release from isolated nerve endings of the rat neurohypophysis, J. Physiol. 390 (1987) 55–70. [6] G. Dayanithi, N. Martin-Moutot, S. Barlier, D.A. Colin, M. KretzZaepfel, F. Couraud, J.J. Nordmann, The calcium channel antagonist omega-conotoxin inhibits secretion from peptidergic nerve terminals, Biochem. Biophys. Res. Commun. 156 (1988) 255–262. [7] G. Wang, G. Dayanithi, S. Kim, D. Hom, L. Nadasdi, R. Kristipati, J. Ramachandran, E.L. Stuenkel, J.J. Nordmann, R. Newcomb, J.R. Lemos, Role of Q-type Ca2+ channels in vasopressin secretion from neurohypophysial terminals of the rat, J. Physiol. 502 (1997) 351–363. [8] J.L. Branchaw, S.F. Hsu, M.B. Jackson, Membrane excitability and secretion from peptidergic nerve terminals, Cell Mol. Neurobiol. 18 (1998) 45–63. [9] D. Brethes, G. Dayanithi, L. Letellier, J.J. Nordmann, Depolarization-induced Ca2+ increase in isolated neurosecretory nerve terminals measured with fura-2, Proc. Natl. Acad. Sci. U.S.A. 84 (1987) 1439–1443. [10] G.J. Augustine, M.P. Charlton, S.J. Smith, Calcium action in synaptic transmitter release, Annu. Rev. Neurosci. 10 (1987) 633–693. [11] J.R. Lemos, M.C. Nowycky, Two types of calcium channels coexist in peptide-releasing vertebrate nerve terminals, Neuron 2 (1989) 1419–1426. [12] X. Wang, S.N. Treistman, J.R. Lemos, Two types of high-threshold calcium currents inhibited by omega-conotoxin in nerve terminals of rat neurohypophysis, J. Physiol. 445 (1992) 181–199. [13] G. Wang, J.R. Lemos, Effects of funnel web spider toxin on Ca2+ currents in neurohypophysial terminals, Brain Res. 663 (1994) 215–222.
55
[14] T.E. Fisher, C.W. Bourque, Calcium-channel subtypes in the somata and axon terminals of magnocellular neurosecretory cells, Trends Neurosci. 19 (1996) 440–444. [15] G. Wang, G. Dayanithi, R. Newcomb, J.R. Lemos, An R-type Ca2+ current in neurohypophysial terminals preferentially regulates oxytocin secretion, J. Neurosci. 19 (1999) 9235–9241. [16] R. Newcomb, B. Szoke, A. Palma, G. Wand, G. Dayanithi, X.-H. Chen, R. Cong, L. Urge, K. Tarczy-Homoch, J. Loo, D.J. Dooley, L. Nadasdi, J.R. Lemos, G. Miljanich, Novel class E calcium channel antagonist from Tarantula venom, C.N.S. Drug Rev. 6 (2000) 153–173. [17] Y. Yin, G. Dayanithi, J.R. Lemos, Ca2+ -regulated, neurosecretory granule channel involved in release from neurohypophysial terminals, J. Physiol. 539 (2002) 409–418. [18] D.D. Friel, R.W. Tsien, A caffeine- and ryanodine-sensitive Ca2+ store in bullfrog sympathetic neurones modulates effects of Ca2+ entry on [Ca2+ ]i , J. Physiol. 450 (1992) 217–246. [19] E.C. Toescu, Activity of voltage-operated calcium channels in rat cerebellar granule neurons and neuronal survival, Neuroscience 94 (1999) 561–570. [20] P.B. Simpson, J.T. Russell, Mitochondrial Ca2+ uptake and release influence metabotropic and ionotropic cytosolic Ca2+ responses in rat oligodendrocyte progenitors, J. Physiol. 508 (1998) 413–426. [21] A. Verkhratsky, E.C. Toescu, Calcium and neuronal ageing, Trends Neurosci. 21 (1998) 2–7. [22] Z. Ahmed, J.A. Connor, Calcium regulation by and buffer capacity of molluscan neurons during calcium transients, Cell Calcium 9 (1988) 57–69. [23] S.A. Thayer, R.J. Miller, Regulation of the intracellular free calcium concentration in single rat dorsal root ganglion neurones in vitro, J. Physiol. 425 (1990) 85–115. [24] E. Neher, G.J. Augustine, Calcium gradients and buffers in bovine chromaffin cells, J. Physiol. 450 (1992) 273–301. [25] E.C. Toescu, Intraneuronal Ca2+ stores act mainly as a ’Ca2+ sink’ in cerebellar granule neurones, Neuroreport 9 (1998) 1227–1231. [26] E.C. Toescu, Mitochondria and Ca2+ signaling, J. Cell Mol. Med. 4 (2000) 164–175. [27] E.C. Toescu, A. Verkhratsky, Neuronal ageing from an intraneuronal perspective: roles of endoplasmic reticulum and mitochondria, Cell Calcium 34 (2003) 311–323. [28] A. Verkhratsky, The endoplasmic reticulum and neuronal calcium signalling, Cell Calcium 32 (2002) 393–404. [29] A. Verkhratsky, O.H. Petersen, Neuronal calcium stores, Cell Calcium 24 (1998) 333–343. [30] M.J. Berridge, The endoplasmic reticulum: a multifunctional signaling organelle, Cell Calcium 32 (2002) 235–249. [31] M.D. Bootman, O.H. Petersen, A. Verkhratsky, The endoplasmic reticulum is a focal point for co-ordination of cellular activity, Cell Calcium 32 (2002) 231–234. [32] J.T. Russell, N.A. Thorn, Isolation and purification of calciumbinding proteins from bovine neurohypophyses, Biochim. Biophys. Acta. 491 (1977) 398–408. [33] E.L. Stuenkel, Regulation of intracellular calcium and calcium buffering properties of rat isolated neurohypophysial nerve endings, J. Physiol. 481 (1994) 251–271. [34] E.C. Toescu, J.J. Nordmann, Effect of sodium and calcium on basal secretory activity of rat neurohypophysial peptidergic nerve terminals, J. Physiol. 433 (1991) 127–144. [35] E.L. Stuenkel, J.J. Nordmann, Intracellular calcium and vasopressin release of rat isolated neurohypophysial nerve endings, J. Physiol. 468 (1993) 335–355. [36] E.L. Stuenkel, J.J. Nordmann, Sodium-evoked, calcium-independent vasopressin release from rat isolated neurohypophysial nerve endings, J. Physiol. 468 (1993) 357–378. [37] S.H. Lee, M.H. Kim, K.H. Park, Y.E. Earm, W.K. Ho, K+ -dependent Na+ /Ca2+ exchange is a major Ca2+ clearance mechanism in axon terminals of rat neurohypophysis, J. Neurosci. 22 (2002) 6891–6899.
56
N. Sasaki et al. / Cell Calcium 37 (2005) 45–56
[38] N. Solovyova, N. Veselovsky, E.C. Toescu, A. Verkhratsky, Ca2+ dynamics in the lumen of the endoplasmic reticulum in sensory neurons: direct visualization of Ca2+ -induced Ca2+ release triggered by physiological Ca2+ entry, Embo. J. 21 (2002) 622–630. [39] V. de Crescenzo, R. ZhuGe, C. Velazquez-Maarrero, L.M. Lifshitz, E. Custer, J. Carmichael, A. Lai, R.A. Tuft, K.E. Fogarty, J.R. Lemos, J.V. Walsh Jr., Ca2+ syntillas, miniature Ca2+ release events in terminals of hypothalamic neurons, are increased in frequency by depolarization in the absence of Ca2+ influx, J. Neurosci. 24 (2004) 1226–1235. [40] K. Tanaka, I. Shibuya, Y. Uezono, Y. Ueta, Y. Toyohira, N. Yanagihara, F. Izumi, T. Kanno, H. Yamashita, Pituitary adenylate cyclaseactivating polypeptide causes Ca2+ release from ryanodine/caffeine stores through a novel pathway independent of both inositol trisphosphates and cyclic AMP in bovine adrenal medullary cells, J. Neurochem. 70 (1998) 1652–1661. [41] I. Shibuya, K. Tanaka, Y. Hattori, Y. Uezono, N. Harayama, J. Noguchi, Y. Ueta, F. Izumi, H. Yamashita, Evidence that multiple P2X purinoceptors are functionally expressed in rat supraoptic neurones, J. Physiol. 514 (1999) 351–367. [42] G. Grynkiewicz, M. Poenie, R.Y. Tsien, A new generation of Ca2+ indicators with greatly improved fluorescence properties, J. Biol. Chem. 260 (1985) 3440–3450. [43] N.C. McNaughton, D. Bleakman, A.D. Randall, Electrophysiological characterisation of the human N-type Ca2+ channel II: activation and inactivation by physiological patterns of activity, Neuropharmacology 37 (1998) 67–81. [44] Y.B. Park, J. Herrington, D.F. Babcock, B. Hille, Ca2+ clearance mechanisms in isolated rat adrenal chromaffin cells, J. Physiol. 492 (1996) 329–346.
[45] R.J. Miller, Calcium signalling in neurons, Trends Neurosci. 11 (1988) 415–419. [46] R.J. Miller, The control of neuronal Ca2+ homeostasis, Prog. Neurobiol. 37 (1991) 255–285. [47] R.J. Miller, Neuronal Ca2+ : getting it up and keeping it up, Trends Neurosci. 15 (1992) 317–319. [48] K. Dunlap, J.I. Luebke, T.J. Turner, Exocytotic Ca2+ channels in mammalian central neurons, Trends Neurosci. 18 (1995) 89– 98. [49] M. Wakamori, M. Strobeck, T. Niidome, T. Teramoto, K. Imoto, Y. Mori, Functional characterization of ion permeation pathway in the N-type Ca2+ channel, J. Neurophysiol. 79 (1998) 622– 634. [50] M. Ludwig, N. Sabatier, P.M. Bull, R. Landgraf, G. Dayanithi, G. Leng, Intracellular calcium stores regulate activity-dependent neuropeptide release from dendrites, Nature 418 (2002) 85– 89. [51] R.C. Lambert, G. Dayanithi, F.C. Moos, P. Richard, A rise in the intracellular Ca2+ concentration of isolated rat supraoptic cells in response to oxytocin, J. Physiol. 478 (1994) 275–287. [52] Z. Li, C. Decavel, G.I. Hatton, Calbindin-D28k: role in determining intrinsically generated firing patterns in rat supraoptic neurones, J. Physiol. 488 (1995) 601–608. [53] M. Muschol, P. Kosterin, M. Ichikawa, B.M. Salzberg, Activitydependent depression of excitability and calcium transients in the neurohypophysis suggests a model of “stuttering conduction”, J. Neurosci. 23 (2003) 11352–11362. [54] Y. Komori, I. Shibuya, M. Kuba, Y. Ueta, G. Dayanithi, Ca2+ Clearance and somatodendritic vasopressin release of rat supraoptic neurons, Jpn. J. Physiol. 52 (2002) S197.