spatial and temporal aspects of ca2+ oscillations in xenopus laevis melanotrope cells

Research Cell Caldum (1999) 19(3), 21S227 0 Pearson Pmfessional Ltd 19%

Spatial and temporal aspects of Ca*+ oscillations in X6nopus laevis melanotrope cells

Wim J.J.M. Scheenen, Bruce G. Jenks, Renier J.A.M. van Dinter, Eric W. Roubos Department of Animal Physiology, Nijmegen Institute for Neurosciences, Nijmegen, The Netherlands

Summary Spatio-temporal aspects of Ca2+ signaling in melanotrope cells of Xenopus laevis have been studied with confocal laser-scanning microscopy. In the whole-frame scanning mode, two major intracellular Caz+ compartments, the cytoplasm and the nucleus, were visualized. The basal [Ca2+] in the nucleus appeared to be lower than that in the cytoplasm and Ca2+ oscillations seemed to arise synchronously in both compartments. The N-type channel blocker o- conotoxin eliminated oscillations in both regions, indicating a strong coupling between the two compartments with respect to Ca2+ dynamics. Line-scanning mode, which gives higher time resolution, revealed that the rise phase of a Ca2+ oscillation is not a continuous process but consists of 3 or 4 discrete steps. Each step can be seen as a Ca2+-wave starting at the cell membrane and going through the cytoplasm at a speed of 33.3 f 4.3 pm/s. Before the Ca2+-wave enters the nucleus, a delay of 120.0 f 24.1 ms occurred. In the nucleus, the speed of a wave was 80.0 f 3.0 pm/s. Treatment with the Ca2+-ATPase inhibitor thapsigargin (1 PM) almost completely eliminated the apparent difference in the basal [Ca2+] in the cytoplasm and the nucleus, reduced the delay of a Ca2+-wave before entering the nucleus to 79.8 + 8.7 ms, and diminished the nuclear wave speed to 35.0 + 4.9 pm/s. These results indicate that a cytoplasmic thapsigargin-sensitive ATPase near the nuclear envelope is involved in buffering Ca2+ before the Ca2+ wave enters the nucleus. After sensitizing IP, receptors by thimerosal (10 PM) the speed of the cytoplasmic Ca2+-wave was increased to 70.3 + 3.6 pm/s, suggesting that IP, receptors may be involved in the propagation of the cytoplasmic Ca2+ wave. Our results indicate that in melanotropes the generation and propagation of Ca2+ oscillations is a complex event involving influx of Ca2+ through N-type Ca2+ channels, propagation of the cytoplasmic Ca2+ wave through mobilization of intracellular stores and a regulated Ca2+ entry into the nucleus. We propose that Ca2+-binding proteins may act as a Ca2+ store for propagation of the wave in the nucleus.

INTRODUCTION

In melanotrope cells of the amphibian Xen~pus Zuti, spontaneous Caz+ oscillations have been described [l-4].

Received 21 h/y 1995

Revised 7 December 1995 Accepted 20 December 1995

Correspondence to: Dr Bruce G. Jenks, Department of Animal Physiology, Nijmegen Institute for Neurosciences, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands

Present address: Dr Wim J.J.M. Scheenen, Department of Biomedical Sciences, via Trieste 75, 35121 Padova, Italy.

These cells secrete cl-melanophore-stimulating hormone (cl-MSH) causing darkening of the skin during adaptation of the animal to a dark background 151. Secretion of cx- MSH depends on Ca2+ influx through voltage-operated N-type Ca2+ channels [6]. It has also been shown that Caz+ oscillations in these cells depend on influx of extra- cellular Caz+ [4] through opening of N-type Ca*+ channels [ 11. Moreover, neural factors that inhibit ol-MSH secretion also inhibit Ca2+ oscillations, whereas secrete-stimulators induce these oscillations [2,3]. Therefore, it has been hypothesized that Caz+ oscillations are the driving force for hormone secretion [2,3,71. Although Ca*+ signaling in Xenopus melanotropes has been studied in some detail with respect to membrane channel opening, little is

219

220 W J J M Scheenen, B G Jenks, R J A M van Dinter, E W Roubos

known about the intracellular (cytoplasmic and nuclear) dynamics of these oscillations.

During the last few years the development of high-resolution Caz+ imaging techniques has been paralleled by an increased interest in spatiotemporal aspects of Caz+ signaling to see how Caz+ signals are propagated in a cell. For many cell types, it has been shown that changes in [Ca’+] (spikes and oscillations) occur in both the cytoplasm and the nucleus [&I-14]. Propagation of a Caz+ wave through the cytoplasm can occur through IP, and/or ryanodine-sensitive Ca2+-induced Ca*+-release (CICR) or through diffusion [ 151. However, little is known about the propagation of Ca2+ oscillations to the nucleus. For neuronal cell lines, it has been proposed that an increase in the nuclear [Ca2+] is a process driven by pas- sive diffusion [8], whereas in several smooth muscle cells an active component in nuclear Ca2+ oscillations has been proposed [9,10]. To gain further insight into spatiotemporal aspects of Ca2+ oscillations in Xenopus melanotropes, Ca dynamics were studied using confocal laser-scanning microscopy on isolated cells loaded with visual wavelength fluorescent Ca2+ indicators. Special attention has been paid to the dynamics of the Ca2+ oscillations in time (temporal aspect) and to the distribution of these oscillations in cytoplasm and nucleus (spatial aspect). The present study shows that Ca2+ oscillations in Xenopus melanotropes are complex signals which start with the opening of N-type Ca2+ channels at the plasma membrane. The oscillation rise phase consists of 3 or 4 discrete steps and the oscillation is propagated through the cytoplasm into the nucleus as a Ca2+ wave, in a process involving mobilization of Ca2+ from intracellular Ca2+ stores.

MATERIALS AND METHODS

Animals

Young-adult Xenopus la&s were taken from laboratory stock and kept on a black background for 3 weeks under continuous illumination, at 22°C. The animals were fed weekly with beef heart.

Preparation of single cells

Isolation of melanotrope cells was performed as described previously [ 1,2]. In short, after anesthetization in a MS222 solution (1 g/l, Sigma, St Louis, MO, USA), animals were perfused with Xenopu.r Ringer’s solution, con- taining 112 mM NaCl, 2 mM KCl, 2 mM CaCI, and 15 mM Ultra&HEPES (Calbiochem, La Jolla CA, USA; pH Z4), to remove blood cells. Then neurointermediate pituitary lobes were dissected and incubated for 45 min in Ringer’s solution without CaCl, to which 0.25% (w/v) trypsin

(Gibco, Renfrewshire, UK) had been added. Cells were subsequently dispersed in Leibovitz’s L 15 medium, con- taining 10% fetal calf serum (Gibco), by gentle trituration of the lobes with a siliconized Pasteur pipette. The medium had been adjusted to Xenopus blood osmolality (L15: ultrapure water = 2:l). After washing cells were plated on glass cover slips coated with poly-L-Lysine (MW > 300 kDa; Sigma) at a density of about 10,000 cells/slip, and cultured for 3 days at 22°C. The cells were readily identified on the basis of their characteristic round shape.

[Caz+] measurements

In order to determine possible differences in the relative Ca2+ concentrations in the cytoplasm and the nucleus, coverslips were placed in a Leiden chamber [ 161 and melanotropes were loaded with the two visible wavelength Ca2+-indicators Fluo-3 and Fura-red, 4 /.M and 6 @I of the AM ester forms, respectively, (Molecular Probes, Eugene, OR, USA) in the presence of 1 @vl Pluronic F-127 [17] (Molecular Probes) for 25 min at 20°C. For time-related measurements of changes in fluorescence intensity, cells were loaded with 6 @VI Fura-red in the presence of 1 @vl Pluronic F-127 for 25 min at 20°C. Then, cells were washed 3 times with medium and were allowed to equilibrate for 30 min. Preliminary experiments showed that the probe was released by 100 ,uM digitonin within 2 min, indicating that there is no sig- nificant compartmentalization of the probe. Prior to measurements, cells were placed on an Nikon Diaphot microscope (Nikon, Tokyo, Japan) equipped with a 60x oil-immersion objective (NA = 1.4). Confocal laser-scanning microscopy (CLSM) was performed on a BioRad MRC-600 (BioRad, Hemel Hempstead, UK). Excitation at 488 run was obtained by directing the light of a 15 mW Ar-laser through an excitation filter (488 DF 10). To reduce bleaching of the fluorescent probes, laser emission was attenuated to about 10% by introducing a neu- tral density filter in front of the laser. Fluorescent light was passed through a dichroic reflector (DR 5 10 LP) and a barrier filter (OG 515). Dual-emission measurements were performed via two photomultiplier tubes (PMT) by directing the fluorescence light through a dichroic reflector (DR 565 LP). Fluo-3 fluorescence was collected in a PMT, in front of which a bandpass filter (540 DF 30) was placed. Fura-red emission was collected simultaneously in the second PMT with a long-pass filter (EF 600 LP) in front. Optical sections of approximately 1 e were obtained by opening the confocal pinholes in front of the PMTs for 50%. On-line measurements of fluorescence intensity, utilizing the fast photon counting setup of the MRC-600, were obtained from preselected areas of interest using the time-course ratiometric software package

Cell Calcium (1996) 19(3), 2 19-227 Q Pearson Professional Ltd 1996

Spatial and temporal aspects of Ca2’ oscillations in Xenopus laevis melanotrope cells 227

12 -B \

.A 30s

Fig. 1 Loading properties of Flue-3 and Fura-red in Xenopus melanotropes (A) Whole-frame scanning of a melanotrope loaded with Flue-3 (F3) and Fura-red (Fr). After imaging the Ca*+ dynamics, this cell was fixed on stage and a nuclear staining with propidium iodine (PI) was performed. (B) Whole frame scanning of a spontaneously oscillating melanotrope with a time-interval of 2 s between frames. At the time of a Ca*+ oscillation, the Fluo-3 intensity increased and simultaneously the Fura-red intensity decreased.

(TCSM, BioRad). Changes in intracellular Ca2+ are expressed as changes in fluorescence intensity. During whole-frame s canning, the scan frame was sized at 192 x 128 pixels, allowing a scan time of 0.45 s/frame. In experiments where line-scanning was performed, the frame size was 384 x 256 pixels and the scan-time per line was 6 ms. The scanning line was set in such a way that it transsected the midline of a melanotrope cell. Four such lines were averaged to obtain a better signal-to-noise ratio, leading to a final time resolution of 24 ms in the x-t plots. Analyzing melanotropes in this way allowed scanning of one frame within 1.5 s. For presentation of results of the line scans, the average reciprocal of the signals for a given point in the cytosol or nucleus, 6 pixels wide, were calculated from the x-t plot of the line scan and plotted against time (reciprocal of signals were plotted in such traces because the data were collected using Fura-red, a probe which decreases fluorescence emission upon Ca2+ binding).

Nuclear staining

After Ca*+ experiments were performed, cells were fixed on stage in Bouins’ fmative for 15 min, after which they were washed 3 times with Xenopus Ringers’ solution and incubated with 10 pg/rnl propidium iodine (Sigma) for 10 min. After this staining, cells were washed 2 times with Xenopus Ringers’ solution, and studied with the CLSM (excitation at 5 14 mn; emission wavelength > 550 mu).

Calculations

III order to determine the time at which a Ca*+ wave started at each measure point, a fast-Fourier transforma-

8 Pearson Professional Ltd 1996

tion [ 181 was carried out on data obtained from line-scanning experiments. The transformed data points were sub- jected to a CUSUM-test [19]; the intersection of the CUSLJM-data with the X-axis was taken as the starting point of a Ca2+ change for that specific region.

RESULTS

In the present study, Caz+ oscillations in melanotrope cells were monitored using CLSM. Approximately 80% of the cells displayed spontaneous oscillations, consisting of individual Caz+ spikes arising from the basal [Caz+]. The increase period of a spike is referred to as the rise phase (see Figs l&2). The frequency of the oscillations proved to be remarkably constant for any given cell. The relative amplitude, as determined by comparing the maximum spike value with the basal fluorescence, was constant for each spike for any given area within a cell (see, for example, Fig. 2).

Spatial aspects of Ca2+ oscillations

Comparing the distribution of fluorescence intensity of both Caz+-sensitive indicators Fluo-3 and Fura-red, a clear compartmentalization of the Ca2+ signal was observed in all cells (Fig. lA, n = 6). The Fluo-3 fluorescence intensity was lower in the nucleus (Fn) than in the cytoplasm (Fc) (Fn/Fc = 0.8 + 0.1) whereas the Fura-red signal was highest in the nucleus (Fn/Fc = 1.2 + 0.1). In oscillating melanotropes, the changes in fluorescence intensity of the two probes in time had opposite directions (Fig. lB, 1z = 6). At the start of the experiment, the basal fluorescence intensities of Fluo-3 and Fura-red were similar but their bleaching ratios differed: at a sam-

Cell Calcium (1996) 19(3), 2 19-227


O.,

0.

0.:

8 a 0 % P 0. c

0 8 g 0.:

iz Ti H

0.

0.:

0.:

Fig. 2 Whole-frame scanning experiments of Xenopus melanotropes. (A) Whole frame scanning of a spontaneous oscillating melanotrope with a time interval of 0.45 s/frame. The reciprocal values of the Fura-red intensity for a cytoplasmic and nuclear region are presented because an increase in [Ca*+], leads to a decrease in Fura-red signal. Oscillations in cytosol and nucleus appeared to start at the same time point. (B,C) Several melanotropes were monitored at the same time with a time interval of 2 s. Shown are the reciprocal values of the Fura-red fluorescence intensity of a cytoplasmic and nuclear region of two cells. In 60% of the cells (22 of 36 cells), the changes in the nuclear fluorescence intensity were in harmony with those in the cytosol, leading to a stable ratio FnlFc (C). In the remaining 40 % of the cells the ratio FnlFc changed due to a stronger fluorescence change in the nucleus (B). (D) Both cytoplasmic and nuclear CaZ+ oscillations are sensitive to o-conotoxin GVIA. oconotoxin was added to the bathing solution by gentle pipetting, yielding a final concentration of 1kM.

D 1 pM o-Conotoxin ,

pling rate of 1 image every 2 s, Fluo-3 decreased in fluorescence intensity by B%/min and Fura-red by less than I%/& (data not shown). In order to follow the [Ca’+] dynamics in cytoplasm and nucleus in time, changes in fluorescence intensity of Fura-red were monitored. Performing whole frame scanning with a time resolution of 0.45 s, Ca2+ oscillations in cytoplasm and nucleus

Cell Calcium (1996) 19(3), 219-227

seemed to occur synchronously in all cells examined (Fig. ZA, n = 5). Recordings of 36 cells at a time resolution of 2 s showed that, in about 60% of the cells, the relative amplitude of the nuclear oscillations remained similar to that in the cytoplasm, leading to a stable fluorescence ratio (nucleus/cytoplasm) (Fig. 2C, 12 = 22), while in the remaining cells the relative amplitude of Ca*+ oscillations in the nucleus increased compared to the cytoplasm, leading to an increase in fluorescence ratio (Fig. 2B, n = 14). Oscillations in both cytoplasm and nucleus were fully inhibited by 1 p.M mconotoxin GVIA in all cells given this treatment (Fig. 2D, n = 13).

Observing Ca2+ oscillations with the line-scanning mode showed that increases in [Caz+] start at the cell membrane and form a Ca*+ wave that spreads in the direction of the nucleus. The average speed of the wave through the cytoplasm was 33.3 + 4.3 @s (n = 8). The speed variation between individual cells was between 25-40 &s, but the speed of a Ca2+ wave was very constant for any given cell. At the nuclear membrane, an apparent delay of 120.0 f 24.1 ms was obsenred before the Ca2+-wave entered the nucleus (Fig. 3). The nuclear speed of a wave was 80.0 f 3.0 ,umls. The rise phase of one wave was 300 ms, whereas the rise phase of a Ca2+ spike, as observed with whole-frame analysis, varied between 5-20 s (Fig. 2B,D). Consecutive frames of line scans showed that, when a Ca2+ spike was captured, each spike consisted of 3-4 distinct Ca2+ waves, seen as discrete steps of Ca2+ increase at each point in the cell. An example is shown in Figure 4A, where traces 2-5 show incremental Ca2+ increases. The 180 ms gaps between traces was the time required for the frame of the line scans to be stored to disk. These data have been filtered (moving average with window width 5 points) to obtain less noisy traces (Fig. 4B). For traces 2 and 3, the increase in Ca2+ was captured just at the start of the collection of the line scan frames, whereas for traces 4 and 5 the Ca2+ steps came within the frames. Sometimes, within a frame, there appeared to be an apparent decrease in Ca2+ follow- ing the discrete increase (e.g. trace 3 and to a lesser extent trace 4), possibly reflecting the functioning of endogenous Ca2+-ATPases. In contrast to the rise phase of the Ca2+ spike, the decrease in Ca2+ was relatively smooth (Fig. 4A,B traces 5-7).

A potential problem with collecting consecutive frames is that any photobleaching that occurred during the recording may recover by diffusion of unbleached probe into the bleached area during the period when the frame is being stored. However, in recordings where no Ca2+ changes were taking place, the first value of a new frame was always in the same range as the last value of the pre- ceding frame, thus indicating that this was not occurring. The data showed that each individual Ca2+ wave also entered the nucleus (e.g. Fig. 3).

0 Pearson Professional Ltd 7996

Spatial and temporal aspects of C$+ oscillations in Xenopus laevis melanotrope cells 223

7

0.14 -

0.12 -

i 0.1 -

\

0.08 ,‘r

CI

120 ms

Fig. 3 [Ca’+], dynamics in Xenopus melanotropes as revealed by data derived from line scans. (A) On-line analysis of two regions in the cell at the same distance from the cell membrane. Changes in Fura-red fluorescence are presented as reciprocal values. One region was situated in the cytosol (C), the other region in the nucleus (N). A delay of 120 ms in the start of the Ca*+ wave at the nuclear region compared to the cytoplasmic region was observed.

3 t 0.35

c

[ 0.30

s 6: 24

0.25

3 B 0 s 0.35

2 s 0.30

P

2 0.25 xi eu:

r

Fig. 4 (A) Consecutive frames of line scans were collected and analyzed for one region in the cell. The 180 ms gap between each trace is the time taken for each frame of line scans to be stored to disk. (B) Moving average of the data with window width of 5 points. For this cell one Ca*+ oscillation shows 4 consecutive waves in the rising phase, seen in traces 2, 3, 4 and 5 (shaded areas and indicated by arrows).

@Pearson Professional Lfd 1996 Cell Calcium (1996) 19(3), 219-227


Involvement of intracellular Ca2+ stores in Ca2+ oscillations

Thapsigargin (1 @!I) converged the Fura-red fluorescence intensity in cytoplasm and nucleus within 15 min

0.

0.

A 1 PM Thapsinargin

B: Control

C: Thapsigargin

300 ms

Fig. 5 Involvement of thapsigargin-sensitive Ca*+ stores. In all experiments thapsigargin was gently pipetted into the bathing solution, to a final concentration of 1 PM. (A) Whole-frame recording of a spontaneous oscillating melanotrope cell. Shown are the reciprocal values of the Fura-red fluorescence of a cytoplasmic (C) and nuclear (N) region. After thapsigargin treatment, the Ca2+ signal in both compartments increased and the fluorescence signals gradually converged. Indicated are the peak-heights of an oscillation during normal and thapsigargin-treatment (arrows). (B) Data derived from line-scanning experiment. Shown are reciprocals of Fura-red fluorescence of two regions at the same distance from the cell membrane, one region situated in the cytosol (C) and the other in the nucleus (N). The delay of the Ca*+ wave in entering the nucleus for this cell was 120 ms (dotted lines indicate where the Ca2+ starts to increase in the cytoplasm and nucleus). (C) Data derived from line-scanning of the same cell at the same position after 15 min thapsigargin treatment. The delay of the Cap+ wave in entering the nucleus was now reduced to 48 ms.

Cell Calcium (1996) 19(3), 2 19-227 8 Pearson Professional Ltd 1996

after application in all cells examined (Fig. 5A, n = 15). In control situations, the relative amplitude of the oscillations in the nucleus was always lower than in the cytoplasm (Fig. 5A). During thapsigargin treatment, the relative amplitude of the nuclear oscillations became higher than in the cytoplasm (Fig. 5A). After 15 min thapsigargin treatment, a reduction of the delay at the nuclear envelope from 120.0 + 24.1 ms to 79.8 f 8.7 ms was observed (Pig. 5B vs 5C, n = 5). Moreover, the speed of the Ca2+ wave in the nucleus was reduced from 80.0 Ifr 3.0 pm/s to 35.0 Ifr 4.9 &s. The speed of the Ca2+ wave in the cytosol remained unaffected.

Thimerosal (10 or 20 j&l), a thiol reagent that at low concentrations sensitizes IP, receptors [20], had no effect on the frequency (Fig. 6A, n = 20) nor on the relative amplitude of the Caz+ oscillations. Increasing the concentration to 30 j&l drastically decreased the frequency of the oscillations (Fig. 6A). Imaging melanotropes with the line-scanning mode showed that 10 m thimerosal increased the speed of the cytoplasmic Caz+ wave from 33.3 k 4.3 pm/s to 70.3 + 3.6 @s (Fig. 6B vs 6C, n = 4); thimerosal had no effect on the speed of the nuclear Caz+ wave nor on the delay observed at the nuclear envelope (data not shown).

DISCUSSION

In the present study, we investigated spat&temporal aspects of spontaneous Ca2+ oscillations in melanotrope cells of Xenapus kzevis using CLSM. It has previously been found that Caz+ oscillations in these cells arise from Ca2+ entry and that blocking voltage-operated Ca2+ channels during any time of the rise phase of a Caz+ spike stops the oscillations [4]. Subsequent studies have shown that the Ca2+ influx comes through opening of voltage- operated N-type Ca*+ channels [ 11. All studies in Xenv melanotropes have so far been concentrated on Caz+ entry mechanism and little is known about the intracellular aspects of these CaZ+ oscillations.

Loading melanotropes with both Fluo-3 and Fura-red showed that the fluorescence intensities of these probes in the nucleus were different from those in the cytoplasm: for Fluo-3 the fluorescence was less intense in the nucleus than in the cytoplasm, whereas Fura-red showed a reverse picture. Therefore, these results strongly suggest that the [Caz+] in the nucleus is lower than in the cytoplasm, although it can not be excluded that the differences in fluorescence intensities found are due to different loading capacities of the nucleus compared to cytoplasm for each probe. We show that changes in Fluo-3 and Fura-red fluorescence intensities during time occur in opposite directions, thus demon- strating that these changes reflect changes in [Caz+] because it is known that Fluo-3 has a higher fluores-

Spatial and temporal aspects of Ca2’ oscillations in Xenopus laevis melanotrope cells 225

0.4

0.3

0.2

0.18

0.14

0.1

A Thimerosal (PM) 10 20 30 40 I

I I I I I I Ii

I

N smin

B: Control 1 I

C: Thlmero$ , A

Fig. 6 Involvement of thimerosal-sensitive IP3 receptors in Ca*+ oscillations in Xenopus melanotropes. (A) Whole-frame scanning experiment. Per experiment, 7 melanotropes were monitored on- line. Thimerosal was gently pipetted in the bathing solution to reach the final concentrations indicated in the graph. Thimerosal (10 and 20 PM) had no effect on the frequency of spontaneous oscillating melanotropes. Increasing the concentration to 30 PM led to a drastic decrease in frequency. (B) Data derived from line-scanning experiment with an oscillating cell. Shown are reciprocals of Fura-red fluorescence of two regions; one situated in near the cell membrane (1) and a second region 7 pm deeper into the cytosol (2) close to the nucleus. The speed of the cytoplasmic Ca*+ wave for this cell was 38 pm/s. Dotted lines indicate the starting point of the Ca2+ increase for that point. (C) Line-scanning of the same cell at the same position after addition of 10 PM thimerosal. The speed of the cytoplasmic Caz+ wave was increased to 70 pm/s.

cence intensity when bound to Ca*+, whereas Fura-red has a lower fluorescence intensity when bound to Ca2+ [Zl-231. For myocytes, a calibration method for Ca*+ has been described based on a double loading with Fluo-3

Q Pearson Professional Ltd 1996 Cell Calcium (1996) 19(3), 219-227

and Fura-red. However, use of this calibration technique is impossible for Xenopus melanotropes because the bleaching properties of the two probes differ. Therefore, in experiments with a long time-course and with inten- sive laser-scanning, we consider the Fura-red signal to be the most reliable as it bleaches less than l%/min.

Two facts indicate that oscillations in the cytoplasm and nucleus arise from the same oscillator. First, even at the highest scanning frequency possible in whole-frame mode (0.45 s/frame) oscillations in cytoplasm and nucleus appeared at the same time and had the same duration. Second, both oscillations in cytoplasm and nucleus were inhibited by exonotoxin. The latter obser- vation indicates that the generation of the oscillations takes place at the plasma membrane through opening of c+conotoxin-sensitive N-type Ca2+ channels. The only difference between cytoplasmic and nuclear oscillations was seen when melanotropes were studied in the line- scanning mode. Increasing the time resolution to 24 ms showed that an oscillation concentrically moves inwards from the membrane into the cell as a Ca2+ wave. From comparing oscillations in cytoplasm& and nuclear regions at the same distance from the plasma membrane, it appeared that a cytoplasmic Ca2+ wave undergoes a delay of approximately 120 ms before entering the nucleus. Such a time delay was also found for RBL cells and it was hypothesized that changes in the nuclear [Caz+] passively follow that in the cytoplasm [13]. However, in Xenopxr melanotropes the fluorescence intensities in cytoplasm and nucleus converge after treatment with thapsigargin. Moreover, the delay in the Ca2+ signal entering the nucleus is substantially shortened by thapsigargin. These results suggest that a thapsigargin- sensitive Ca*+-ATPase activity buffers Caz+ when a Ca*+ wave reaches the vicinity of the nuclear envelope or is entering the nucleus. Possibly, Ca2+-ATPase activity delays the time required for Ca2+ to reach a threshold for the signal to enter the nucleus. Such a mechanism might provide a control point through which a cell can set the amplitude of the Ca2+ signal entering the nuclear com- partment. The location of this thapsigargin-sensitive Ca2+-ATPase activity, as well as its physiological sign& came remains to be determined. In smooth muscle cells, an active component sensitive to thapsigargin in trans- porting a Ca2+ signal to the nucleus has also been described [9, lo].

From analysis of the Ca2+ oscillations using whole- frame scanning, it can be calculated that the rise phase lasts about 5-20 s. This period is similar to that shown for Ca2+ oscillations in Xenopus melanotropes using digital imaging methods [ 1,4]. It was, therefore, surprising to see that the rise phase of a Ca2+ wave lasts only 300 ms. This apparent discrepancy was resolved when consecutive frames of line-scanning were taken, showing that each


Ca*+ oscillation consists of 3-4 consecutive Caz+ waves occurring over a period of 5-20 s. This is the first report of a stepwise increase in intracellular Ca*+ for secretory cells. For pancreatic acinar cells, the increase in [Caz+] during CaZ+ oscillations is a continuous process involving three intracellular regions with different sensitivities to ryanodine and IP, [ 15,24,25]. There are at least two possible mechanisms for the stepwise increase in Caz+ in Xenopus melanotropes. First, the stepwise increase in Caz+ might be a property of an intracellular Caz+ propaga- tor mechanism. Second, spontaneous membrane depolar- izations observed in Xenopus melanotropes [6] might occur quickly in succession accounting for successive openings of the N-type Ca2+ channel, and thus the stepwise increases in [Caz+]. Since inhibition of voltage-operated Caz+ channels during the rise phase of a Ca2+ spike blocks the formation of the spike [4], we feel that the stepwise increase during a Caz+ oscillation in Xenopus melanotropes most likely reflects plasma membrane events. Possibly, the number of consecutive Ca*+ waves in Xen~ melanotropes can be regulated in order to set the amplitude and duration of a CaZ+ spike.

The speed of a cytoplasmic Ca2+ wave ranged between 25-40 &s. The diffusion radius for Ca*+ is less than 1 c [26]. Therefore, it is unlikely that the propagation of the Ca2+ wave through the cytoplasm is through diffusion of intracellular Caz+ from the region of the membrane. According to Jaffe, a CICR mechanism would propagate a CaZ+ wave with a speed of 20-30 &s [271. It remains to be determined whether this propagation in Xenopus melanotropes is through a ryanodine-sensitive CICR or a Ca2+-sensitive IP,-receptor [ 15,281. A role for the Caz+-sensitive IP,-receptor in the propagation of a Ca*+ wave is suggested by the action of thimerosal. It has been demonstrated that this thiol reagent in low concentrations can increase the sensitivity of the IP, receptor for IP, and hence can increase oscillation frequency when such oscillations are sensitive to IP, [20]. While for Xenopus melanotropes we show no difference in frequency or amplitude of the Ca2+ oscillations after application of low concentrations of thimerosal (lo-20 @VI), the speed of the Ca*+ waves through the cytoplasm almost doubled with such treatment indicating involvement of IP,-sensitive stores in the propagation of the Ca2+ signal through the cytoplasm. Higher concentrations of thimerosal lead to a drastic reduction of the frequency, a result that likely reflects an inhibition of all cellular ATPase activity [20]. In general, the CaZ+-ATPase activity on the IP,-sensitive intracellular Caz+ pool is sensitive to thapsigargin 128,291. In Xenopus melanotropes, the thapsigargin-sensitive Ca2+ pool appears to be very small [l] and it is expected that additional intracellular Ca2+ stores will be involved in supporting propagation of the wave through the cytoplasm.

Cell Calcium (7 996) 19(3), 2 19-227

Another issue to be resolved is the source of Caz+ for the rapid nuclear wave. Consideration of the extreme speed of this wave precludes propagation through a sim- ple diffusion process from the cytosol. For hepatocytes, it has been postulated that an IP,-sensitive Ca2+ pool is present in the nucleus [30]. Involvement of such a pool in nuclear propagation of the Caz+ wave would seem unlikely in Xetzopus since low concentrations of thimerosal had no effect on the speed of the Ca2+ wave. While IP, receptors at the nuclear envelope could be involved in initiating the’wave in the nucleus, we suggest that propagation of the wave involves release of Ca2+ from nuclear Ca2+ binding proteins. It is well docu- mented that the nucleus contains a number of Ca2+ binding proteins including a calreticulin-like group which binds large quantities of Ca2+ with low affinity ([31] for review), characteristics expected for a Ca2+ storage protein.

In conclusion, our results indicate that Ca2+ oscillations in Xenopus melanotropes are generated by Ca2+ influx through voltage-sensitive N-type Ca2+ channels, are propagated through the cytoplasm to the nucleus by a CICR mechanism and we suggest that the wave may be carried through the nucleus by a mechanism involving release of Ca2+ from Ca2+ binding proteins. The Ca2+ wave may be of importance to a cell since it could function as a signal to coordinate a number of cellular processes occurring in different compartments of the cell, such as gene expression in the nucleus [ 11,32,33], intracellular transport of secretory products in the cytosol [33] and the exocytotic process at the membrane [fl.

ACKNOWLEDGEMENTS

We thank Mr P.MJ.M. Cruijsen for technical assistance. This study was supported by grants from The Nether- lands Organization for Scientific Research @IWO), from the European Community (I-KM grant ERRCHRXCT- 9200 17) and from INSERMNWOMW.

REFERENCES

Scheenen WJJ.M., Jerks B.G., Roubos E.W., Willems P.H.G.M. Spontaneous calcium 0sciBations in Xenopus kwk melanotrope cells are mediated by o-conotoxin sensitive calcium channels. Cell calcium 1994; 15: 36-44. Scheenen WJJ.M., Jerks B.G., WiIIems P.H.G.M., Roubos E.W. Action of stinndatory and inhibitory a-MSH secretagogues on spontaneous calcium oscillations in melanotrope cells of Xenops Levis. PjhZgen Arch 1994; 427: 244-25 1. Shibuya I., Douglas W.W. Spontaneous cytosok calcium pulsing detected in Xettopus melanotrophs: modulation by secrete-inhibitory and stimulant Iigands. Emiodnokgy 1993; 132: 2166-2175. Shibuya I., Douglas W.W. Spontaneous cytosolic calcium pulses in Xenqxf.s melanotrophs are due to calcium influx during

Q Pearson Professional Lfd 1996

Spatial and temporal aspects of CP oscillations in Xenopus laevis melanotrope cells 227

phasic increases in the calcium permeability of the cell membrane. Endocrinobgy 1993; 132: 2176-2183.

5. Jenks B.G., Leenders H J., Martens GJ.M., Roubos E.W. Adaptation physiology: the functioning of pituitary melanotrope cells during background adaptation of the amphibian Xenopus fae& Zoo2 Sn’ 1993; 10: l-l 1.

6. Scheenen WJ J.M., De Koning H.P., Jenks B.G., Vaudry H., Roubos E.W. The secretion of a-MSH from Xenopus melanotropes involves calcium it&x through wconotoxin- sensitive voltage-operated calcium channels. J Neurnerzdotitzol 1994; 6: 457-464.

7 Douglas W.W. Calcium stimulus-secretion coupling and exocytosis. Glancing back 30 years: and recent lessons on spontaneous secretion from melanotrophs revealing calcium- driven autonomous secretion, independent of action potential discharge, and its uncoupling by the secretacurbins: dopamine, GABA and neuropeptide Y. Biomed Res 1994; 14 (Suppl. 2): 9-19.

8. Birch B.D., Douglas L.E., Kocsis D. Intranuclear Ca*+ transients during neurite regeneration of an adult mammalian neuron. Proc Nat1 Acad Set USA 1992; 89: 7978-7982.

9. Himpens B., De Smedt H., Droogmans G., Casteels R. Differences in regulation between nuclear and cytoplasmic Caz+ in cultured smooth muscle cells. Am ]PkysioZ 1992; 263: c95-c105.

10. Himpens B., De Smedt H., Casteels R. Subcellular Caz+ gradients in A7r5 vascular smooth muscle. Cell Calcium 1994; 15: 55-65.

11. Negulescu P.A., Shastri N., Cahalan I.M. Intracellular calcium dependence of gene expression in single T lymphocytes. Pmc NatlAcadSci USA 1994; 91: 2873-2877

12. Przywara D.A., Bhave S.V., Bhave A., Wakade T.D., Wakade A.R. Stimulated rise in neuronal calcium is faster and greater in the nucleus than in the cytosol. FASEBJ 1991; 5: 217-222.

13. Allbritton N.L., Oancea E., Kuhn M.A., Meier T. Source of nuclear calcium signals. Proc Nat2 Acad Sci USA 1994; 91: 12458-12462.

14. Al-Mohamra EA., Caddy K.W.T., Bolsover S.R. The nucleus is insulated from large cytosolic calcium changes. Nature 1994; 367: 745-750.

15. Petersen O.H., Petersen C.C.H., Kasai H. Calcium and hormone action. Annu Rev Pkysioll994; 56: 297-3 19.

16. Ince C., Van Dissel J.T., Diesselhoff M.C. A teflon culture dish for high-magnification microscopy and measurements in single cells. @&TS Arch 1985; 403: 240-244.

1 Z Poenie M., Alderton J., Steinhardt R., Tsien R. Calcium rises abruptly and briefly throughout the cell at the onset of anaphase. Science 1986; 233: 886-889.

18 Lam R.B., Wiebolt R.C., Isenhour T.L. Practical computation

with Fourier transforms for data analysis. Anal C?tem 198 1; 53: 889A-900A.

19. Wohl H. The cusum plot: its utility in the analysis of clinical data. NE& JMed 1977; 296: 1044-1045.

20. Bootman M.D., Taylor C.W., Berridge MJ. The thiol reagent, thimerosal, evokes a Ca2+ spike in HeLa cells by sensitizing the inositol I,4,5&sphosphate receptor. .I Biol Ckem 1992; 267: 25113-25119.

2 1. Lipp P., Niggle E. Ratiometric Car+-measurements with visible wavelength indicators in isolated cardiac myocytes. Cell G&iu?n 1993; 14: 359-372.

22. Lipp P., Niggli E. Microscopic spiral waves reveal positive feedback in subcellular calcium signaling. Biopkyss] 1993; 65: 2272-2276.

23. Lipp P., Niggli E. Sodium current-induced calcium signals in isolated guinea-pig ventricular myocytes. JPkysioZ 1994; 474: 439-446.

24. Thorn P., Lawrie A.M., Smith P.M., Gallacher D.V., Petersen O.H. Local and global cytosolic Ca*+ oscillations in exocrine cells evoked by agonists and inositol t&phosphate. Cell 1993; 74: 661-668.

25. Kasai H., Li Y.X., Miyashita Y. Subcellular distribution of Caz+ release channels underlying CaZ+ waves and oscillations in exocrine pancreas. Cell 1993; 74: 669-672

26. Kasai H., Petersen O.H. Spatial dynamics of second messengers: IP, and CAMP as long-range and associative messengers. Trenak Neurosci 1994; 17: 95-101.

27 Jaffe L.F. Classes and mechanisms of calcium waves. Ceff Calcium 1993; 124: 736-745.

28. Berridge M J. Inositol ttisphosphate and calcium signalling. Nab-e 1993; 361: 3 15-325.

29. Foskett J.K., Wong D. Calcium oscillations in parotid acinar cells induced by microsomal Ca2+-ATPase inhibition. Am J Pkysiol 1992; 262: C656-C663.

30. Nicotera P., Orrenius S., Nilsson T., Berggren P.O. An inositol I,4,5-t&phosphate-sensitive Caz+ pool in liver nuclei. Proc Nat2

Acad Sci USA 1990; 87: 6858-6862. 3 1. Gilchrist J.S.C., Cxubryt M.P., Pierce G.N. Calcium and calcium-

binding proteins in the nucleus. Mel Cell Biockem 1994; 135: 79-88.

32. Loeiller J.P., Kley N., Louis J.C., Demeneix B.A. Caz+ regulates hormone secretion and proopiomelanocortin gene expression in melanotrope cells via the calmodulin and the protein kinase C pathway. JNeurachem 1989; 52: 1279-1283.

33. Hammers&lag R. Is the intersomal phase of fast axonal transport driven by oscillations of intracellular calcium? Neurockem Res 1994; 19: 1431-1437.

8 Pearson Professional L td 1996 Cell Calcium (1996) 19(3), 219-227

spatial and temporal aspects of ca2+ oscillations in xenopus laevis melanotrope cells

Documents