Spatiotemporal dynamics of Ca²⁺ signaling and its physiological roles

Abstract

Changes in the intracellular Ca²⁺ concentration regulate numerous cell functions and display diverse spatiotemporal dynamics, which underlie the versatility of Ca²⁺ in cell signaling. In many cell types, an increase in the intracellular Ca²⁺ concentration starts locally, propagates within the cell (Ca²⁺ wave) and makes oscillatory changes (Ca²⁺ oscillation). Studies of the intracellular Ca²⁺ release mechanism from the endoplasmic reticulum (ER) showed that the Ca²⁺ release mechanism has inherent regenerative properties, which is essential for the generation of Ca²⁺ waves and oscillations. Ca²⁺ may shuttle between the ER and mitochondria, and this appears to be important for pacemaking of Ca²⁺ oscillations. Importantly, Ca²⁺ oscillations are an efficient mechanism in regulating cell functions, having effects supra-proportional to the sum of duration of Ca²⁺ increase. Furthermore, Ca²⁺ signaling mechanism studies have led to the development of a method for specific inhibition of Ca²⁺ signaling, which has been used to identify hitherto unrecognized functions of Ca²⁺ signals.

Introduction

Intracellular Ca²⁺ concentration ([Ca²⁺]_i) is kept extremely low at resting conditions with approximately a 1:10,000 ratio compared to the extracellular Ca²⁺ concentration. Upon various stimuli to the cells, [Ca²⁺]_i increases due to Ca²⁺ influx via the plasma membrane and/or release of Ca²⁺ from the intracellular store to regulate various cell functions. Thus, [Ca²⁺]_i functions as a cellular switch. This notion was first established in skeletal muscle cells.^1),2) Subsequently, a long list of cell functions that are switched on by Ca²⁺ signals has been compiled, and the switching time and switching distance within the cell are extremely diverse (Fig. 1). The versatility of Ca²⁺ signals has fascinated many investigators and has been studied extensively. Furthermore, there should be many hitherto unrecognized functions that are regulated by Ca²⁺ signals. In this review, I would like to summarize how the studies on the switching mechanisms of Ca²⁺ signals evolved, referring to our work centering on the intracellular Ca²⁺ release mechanism. There have been extensive studies on Ca²⁺ signaling in striated muscle cells; that is, excitation-contraction coupling. Readers are referred to numerous excellent reviews on this subject.^3)–6) Here, I would like to concentrate on Ca²⁺ signaling in non-striated-muscle cells. The Ca²⁺ influx pathways are also important for the generation of Ca²⁺ signals, but a full account of the trans-plasmalemmal pathways is out of the scope of this review.

Fig. 1.

Switching time and distance of Ca²⁺ signals in various cell functions. Boxes, representing each cell function, indicate approximate period of time and distance regulated by Ca²⁺ signals.

1. Ca²⁺ signal and cell functions

Recent studies on Ca²⁺ signaling have been greatly influenced by imaging methods owing to the advent of fluorescent Ca²⁺ indicators.⁷⁾ Indeed, it can be said that the importance of spatiotemporal distribution of intracellular signaling molecules was first systematically recognized in Ca²⁺ signaling studies. Although fluorescence Ca²⁺ imaging was initially carried out in cultured cells or isolated single cells, this method was gradually applied to tissue preparations. We first imaged Ca²⁺ signals in vascular smooth muscle cells within the intact arterial tissue, which consists of endothelial, smooth muscle and sympathetic nerve layers from the inside to the outside of arterial walls.⁸⁾ Electrical stimulation of the perivascular sympathetic nerve network results in the contraction of smooth muscle cells, mimicking the vascular response to the sympathetic nerve activity.

When we imaged the Ca²⁺ signaling in Ca²⁺ indicator-loaded vascular smooth muscle cells receiving input from the perivascular sympathetic nerves, we found that the spatiotemporal characteristics of the Ca²⁺ response of vascular smooth muscle cells were extremely dynamic (Fig. 2A).⁸⁾ In the initial Ca²⁺ response shown in Fig. 2Ab, all the cells responded rather uniformly. This Ca²⁺ response was insensitive to α-adrenergic antagonists, and is due to purinergic P2X receptor stimulation by the sympathetic co-transmitter ATP released from the nerve endings. However, the purinergic response faded away rapidly due to receptor desensitization. In the following responses (Fig. 2Ac–f), wave-like increases in [Ca²⁺]_i (Ca²⁺ waves) traversed within the spindle-shaped smooth muscle cells along their longitudinal axis. Ca²⁺ waves occurred repeatedly during stimulation with intervals to generate oscillatory changes in [Ca²⁺]_i (Ca²⁺ oscillations, Fig. 2B). These Ca²⁺ waves and oscillations were blocked by α-adrenergic antagonists. Similar Ca²⁺ dynamics were also observed when noradrenaline, the sympathetic nerve transmitter, was directly applied to the arterial tissue. Interestingly, the frequency of Ca²⁺ oscillations, rather than their amplitude, increased with increasing concentration of noradrenaline (Fig. 2C). In other words, vascular smooth muscle contraction seems to be regulated by the frequency of Ca²⁺ oscillations (frequency modulation).

Fig. 2.

Spatiotemporal dynamics of Ca²⁺ signaling in vascular smooth muscle cells. (A) Confocal images of the Ca²⁺ concentration in the smooth muscle layer of rat tail artery before (a) and during (b–f) the electrical stimulation of sympathetic nerves. The fluorescence intensity of the Ca²⁺ indicator Fluo-3 was normalized by the initial fluorescence intensity (F/F₀) and pseudo-color-coded. Each frame was taken every 1 s and has a dimension of 115 × 115 μm. The frame numbers are shown above each panel. Electrical stimulation (5 Hz, 300 pulses) was applied between frames 5 and 64. (B) Fluorescence intensity changes of Fluo-3 (red, blue and green lines) in the three cells (white boxes in A) and the averaged value of all the cells in the field (thick black line). (C) Dependence of Ca²⁺ oscillation on noradrenaline concentration. Modified from M. Iino et al., Ref. 8.

It has been shown that many other cell functions are regulated by Ca²⁺ oscillations. Indeed, Ca²⁺ oscillation is one of the most ubiquitous forms of Ca²⁺ signaling.⁹⁾ Following from this, a couple of fundamental questions arise; primarily, how are Ca²⁺ oscillations generated? And is there any advantage to regulating cell functions through Ca²⁺ oscillations?

2. Potentiation of intracellular Ca²⁺ release by Ca²⁺

Agonist-induced breakdown of phosphatidyl inositol 4,5-bisphosphate generates inositol 1,4,5-trisphosphate (IP₃), which releases Ca²⁺ from the intracellular stores.¹⁰⁾ Today, IP₃-induced Ca²⁺ release is known to underlie Ca²⁺ oscillations in many cell types, including vascular smooth muscle cells. However, a simple signaling mechanism involving receptor activation, IP₃ generation, and Ca²⁺ release is not sufficient to cause oscillatory changes in [Ca²⁺]_i. There must be a complex mechanism to explain how IP₃-induced Ca²⁺ release becomes oscillatory. Studies on the basic properties of IP₃-induced Ca²⁺ release have provided clues to this question.

Inspired by the first report on IP₃-induced Ca²⁺ release in pancreatic acinar cells,¹¹⁾ I examined whether IP₃ had the same Ca²⁺ releasing activity in smooth muscle cells. The initial results were quite disappointing, and very little Ca²⁺ releasing activity of IP₃ was observed despite extensive efforts. I was about to conclude that IP₃-induced Ca²⁺ release plays no major role in smooth muscle cells, when I realized that I had always applied IP₃ in the virtual absence of Ca²⁺. A breakthrough was brought about when IP₃ was applied with a sub-micromolar concentration of Ca²⁺, which markedly enhanced IP₃-induced Ca²⁺ release (Fig. 3A).^12),13) As shown in Fig. 3B, the application of either IP₃ or Ca²⁺ alone resulted in almost no Ca²⁺ releasing activity. Only when both IP₃ and Ca²⁺ were applied simultaneously was there an activation of Ca²⁺ release.^12)–14) Following further increase in the Ca²⁺ concentration, there was an inhibition of Ca²⁺ release around 1 μM (Fig. 3C). These results made it clear that there is a bell-shaped dependence of IP₃-induced Ca²⁺ release on the cytoplasmic Ca²⁺ concentration.¹³⁾ This original finding in smooth muscle cells was then reproduced in various cell types.^15)–19)

Fig. 3.

Ca²⁺ dependence of IP₃-induced Ca²⁺ release. (A) Time courses of IP₃-induced Ca²⁺ release at different Ca²⁺ concentrations. The luminal Ca²⁺ concentration changes within the ER are shown. The experiments were carried out in the absence of ATP so that there was no uptake of Ca²⁺. (B) Dependence of the rate of IP₃-induced Ca²⁺ release on IP₃ and Ca²⁺ concentrations. (C) Bell-shaped dependence of IP₃-induced Ca²⁺ release on Ca²⁺ concentration. Modified from K. Hirose et al., Ref. 14 (A and B), and from M. Iino, Ref. 13 (C).

The bell-shaped Ca²⁺ dependence is the steady- state property of IP₃-induced Ca²⁺ release. It is also important to study the time-dependence of the effect of Ca²⁺. Thus, the kinetic property of the Ca²⁺ dependence of IP₃-induced Ca²⁺ release was studied using caged Ca²⁺, which can generate a step increase in Ca²⁺ concentration upon a brief ultraviolet light flash. The results indicated that Ca²⁺ had instantaneous potentiating and inhibitory effects on the IP₃-induced Ca²⁺ release depending on the magnitude of the step increase in Ca²⁺ concentration (potentiation <~0.3 μM, inhibition >~1 μM).²⁰⁾ These results indicate that the IP₃-induced Ca²⁺ release mechanism has the inherent property of becoming regenerative at submicromolar Ca²⁺ concentrations.

The Ca²⁺ sensitivity of IP₃-induced Ca²⁺ release bears a resemblance to the Ca²⁺-induced Ca²⁺ release (CICR) mechanism that was initially observed in skeletal muscle cells.^21),22) The CICR mechanism is now attributed to the ryanodine receptor, which binds ryanodine, a plant alkaloid, in an essentially irreversible manner.²³⁾ There are three subtypes of ryanodine receptors in mammals, and they function as the Ca²⁺ release channel as homotetramers. Each subunit of the RyR Ca²⁺ release channel consists of about 5,000 amino acids.^23)–25) On the other hand, the IP₃-induced Ca²⁺ release mechanism is mediated by the IP₃ receptor (IP₃R) with about 2,700 amino acid residues.²⁶⁾ There are three subtypes of IP₃R,^26)–28) which form Ca²⁺ release channels as homo- or heterotetramers. Interestingly, the primary structure of IP₃R has similarity to that of RyR.²⁶⁾ Hence, RyR and IP₃R are homologous proteins, and both function as a Ca²⁺-dependent Ca²⁺ release channel on the endoplasmic reticulum (ER) membrane. The major difference between the two types of Ca²⁺ release channels is that while RyR can be activated by Ca²⁺ alone, IP₃R requires both IP₃ and Ca²⁺ simultaneously for activation. There is another notable difference between IP₃R and RyR. Mg²⁺ is a strong inhibitor of the CICR mechanism of RyR.⁶⁾ However, Mg²⁺ has very little, if any, effect on the Ca²⁺ dependence of IP₃R activity.^13),29)

3. Regenerative Ca²⁺ release and Ca²⁺ wave/oscillation

The regenerativity of Ca²⁺ release via the Ca²⁺ release channels (Fig. 4A) may play an important role in the shaping of spatiotemporal patterns of Ca²⁺ signals. When Ca²⁺ is released locally in one part of the cell, this will further enhance Ca²⁺ release from the adjacent sites, and this will generate a Ca²⁺ wave like ‘toppling dominos’ (Fig. 4B). IP₃-dependent Ca²⁺ waves have been observed in many cell types including smooth muscle cells,³⁰⁾ oocytes,^31),32) neurons³³⁾ and exocrine cells.³⁴⁾ In most cases, the velocity of Ca²⁺ waves is about 10–40 μm s⁻¹.³⁵⁾ What may be the physiological significance of Ca²⁺ waves? Diffusion of Ca²⁺ within cells is limited due to the presence of high concentrations of Ca²⁺ binding proteins. Indeed, the diffusion coefficient of Ca²⁺ in cytoplasm is only ~1/30 of that in water and is even ~1/20 of the diffusion coefficient of IP₃ in cytoplasm.³⁶⁾ Thus, for rapid and uniform increase in the intracellular Ca²⁺ concentration, a mechanism to overcome the diffusion delay is required. Ca²⁺ waves can be regarded as a system for transmitting Ca²⁺ signals from one site of the cell to another, and is essentially the same mechanism as the action potentials for the propagation of information in nerve axons. The regenerative activation of sodium channels underlies the propagation of action potentials in neurons. Similarly, the regenerative activation of Ca²⁺ release channels underlies the propagation of Ca²⁺ waves in many types of cells. There have been many theoretical studies to simulate Ca²⁺ waves using mathematical models of the intracellular Ca²⁺ dynamics.^37),38) The regenerative Ca²⁺ release mechanism is the key factor of these models. An additional physiological significance has been attached to the Ca²⁺ wave in pancreatic acinar cells, in which Ca²⁺ waves always initiate from the apical side of the cell and spread toward the basolateral side. In this way, there is a time delay in the increase in [Ca²⁺]_i between the apical and basolateral membranes. The time difference in [Ca²⁺]_i is proposed to be important for the unilateral movement of solutes in the acinar cells for excretion.³⁴⁾

Fig. 4.

Regenerative Ca²⁺ release and Ca²⁺ waves. (A) Due to the Ca²⁺ dependence of IP₃R, IP₃-induced Ca²⁺ release has the inherent property of becoming regenerative. (B) The regenerative Ca²⁺ release may result in the generation of a Ca²⁺ wave (upper panel) like ‘toppling dominos’ (lower panel).

The importance of regenerative Ca²⁺ release in the generation of Ca²⁺ oscillations has been also postulated in theoretical studies of Ca²⁺ oscillations.^37)–40) This notion was subsequently tested experimentally. Based on the aforementioned structural similarity between IP₃R and RyR, the glutamate residue at position 2100 (E2100) was identified to be a critical amino acid residue of the type 1 IP₃R (IP₃R1) for the Ca²⁺ dependence of channel function (Fig. 5A).⁴¹⁾ When E2100 was replaced with aspartate (E2100D), there was a marked reduction in the Ca²⁺ sensitivity without significant change in the IP₃ dependence (Fig. 5C). This provided a unique opportunity to test the role of regenerative Ca²⁺ release in the generation of Ca²⁺ oscillations. Indeed, Ca²⁺ oscillation was suppressed in cells expressing E2100D IP₃R1 (Fig. 5B). These results established experimentally that the regenerative Ca²⁺ release is essential for the generation of Ca²⁺ oscillations.

Fig. 5.

Requirement of regenerative Ca²⁺ release in Ca²⁺ oscillation. (A) Schematic representation of the primary structure of IP₃R1. IP₃ binding region (green) and membrane spanning region (yellow) are shown. Position of the Ca²⁺ sensor amino acid (E2100) is also indicated. (B) Ca²⁺ responses of DT40 B cells expressing wild-type (left panel) and E2100D-mutant (right panel) IP₃R1 upon activation of B cell receptor (BCR). The time course of IP₃ concentration after BCR stimulation measured using IP₃ binding protein assay had two phases (Miyakawa, T. and Iino, M., unpublished observation). An initial transient increase after the start of BCR stimulation was followed by a decrease in the IP₃ concentration for 10–20 min. Then, a secondary prolonged increase in the IP₃ concentration was observed. The biphasic IP₃ concentration change seems to underlie the ~20-min silent period after the initial Ca²⁺ response in the left panel. (C) IP₃ and Ca²⁺ dependence of Ca²⁺ release via wild-type (open symbols) and E2100D (red symbols) IP₃R1. Modified from T. Miyakawa et al., Ref. 41.

The three subtypes of the IP₃R have functional differences^42)–44) and are differentially expressed in various tissues^45)–50) (Table 1). The order of IP₃ sensitivity is IP₃R2 > IP₃R1 > IP₃R3.^42),43) When multiple subtypes are expressed in the same cell, the IP₃ sensitivity becomes roughly the weighted average of the single subtypes.⁴²⁾ The Ca²⁺ sensitivity has a subtle difference between the three subtypes, and the Ca²⁺ concentration for half maximal activation (EC₅₀) is 30–100 nM.^42),44) These EC₅₀ values are similar to the resting intracellular Ca²⁺ concentrations, thus are ideal for the regenerative activation of Ca²⁺ release via the IP₃R. Another similarity between IP₃R and RyR is that both types of Ca²⁺ release channel are activated by ATP and other adenine nucleotides,^6),51) which enhance channel activity without altering the Ca²⁺ dependence. The ATP dependence is observed in IP₃R1 and IP₃R3, but is absent in IP₃R2.^42),43) Interestingly, when multiple IP₃R subtypes are coexpressed in the same cell, the property of IP₃R2 in terms of the ATP dependence becomes dominant.⁴²⁾ This suggests that heterotetramers containing IP₃R2 subtypes have no ATP dependence, indicating strong inter-subunit functional interaction. Together these functional differences are reflected in the patterns of Ca²⁺ oscillations, and IP₃R2 seems to be the most efficient subtype in generating Ca²⁺ oscillations.⁴²⁾ These findings provide a clear functional basis for the physiological significance of the differential expression of IP₃R subtypes in cell-type specific encoding of Ca²⁺ signaling.

Table 1.

Comparison of IP₃R subtypes

	Type 1	Type 2	Type 3
IP3 sensitivity	Medium	High	Low
permeabilized cell (chicken) (EC50)	4.7 μM	0.35 μM	18.6 μM
planar lipid bilayer (rat) (EC50)	0.27 μM	0.10 μM	0.40 μM
Ca2+ sensitivity (activation)*	Yes	Yes	Yes
permeabilized cell (chicken) (EC50)	50 nM	100 nM	100 nM
planar lipid bilayer (rat) (EC50)	30 nM	60 nM	60 nM
ATP sensitivity	Yes	No	Yes
permeabilized cell (chicken) [Fold increase (EC50)]	5.7-fold (0.4 mM)	1-fold	1.7-fold (n.a.)
planar lipid bilayer (rat) [Fold increase (EC50)]	5.6-fold (0.13 mM)	1-fold	5.9-fold (2 mM)
Tissue expression	neuron (enriched in cerebellum)		neuron (enriched in olfactory bulb)
	vascular smooth muscle
	pancreatic islet
	endothelium	endothelium	endothelium
	liver	liver
		parotid gland	parotid gland
		glia
		cardiac myocytes

^*Ca²⁺ sensitivity of the ascending limb of the bell-shaped Ca²⁺ dependence (see Fig. 3C). See text for references.

Mechanisms other than the regenerative activation of IP₃R may be involved in the generation of Ca²⁺ oscillations. There are positive and negative feedback mechanisms that regulate intracellular IP₃ production. First, the enzyme activity of phospholipase C (PLC) is positively regulated by the intra-cellular Ca²⁺ concentration at >100 nM concentrations.^52),53) Second, PLC activation induces protein kinase C activation, which in turn inhibits IP₃ generation via the inhibition of agonist receptor or PLC activity.⁵⁴⁾ These positive and negative feedback loops may cause oscillatory changes in the intracellular IP₃ concentration. Indeed, IP₃ oscillation has been observed using the IP₃-dependent translocation of GFP-tagged pleckstrin homology domain of PLCδ1 (GFP-PHD).^53),55),56) Such IP₃ oscillations may assist oscillatory release of Ca²⁺ from the ER.

4. Mitochondria as another key player in Ca²⁺ oscillation

Although regenerative Ca²⁺ release is essential for Ca²⁺ oscillations, it remains to be clarified how the regenerative Ca²⁺ release is initiated at the onset of each Ca²⁺ oscillation. Using an ER-targeted Ca²⁺ indicator, we observed intraluminal Ca²⁺ concentration during Ca²⁺ oscillations.⁵⁷⁾ Indeed, the ER Ca²⁺ concentration represented a mirror image of the cytoplasmic Ca²⁺ concentration: a decrease in the ER Ca²⁺ concentration was observed during an increase in the cytoplasmic Ca²⁺ concentration. However, there were important deviations from the perfect mirror image. In the first Ca²⁺ oscillation, the cytoplasmic Ca²⁺ concentration reached its peak even though the ER is still releasing Ca²⁺. This suggests that a considerable fraction of Ca²⁺ released from the ER enters non-ER Ca²⁺ stores during the first Ca²⁺ oscillation. In the second and subsequent Ca²⁺ oscillations, on the other hand, the Ca²⁺ release from the ER lagged behind the increase in the cytoplasmic Ca²⁺ concentration. These results indicate that there is a non-ER Ca²⁺ compartment that takes up Ca²⁺ in the first Ca²⁺ oscillation and supplies Ca²⁺ before the initiation of the second and subsequent ER Ca²⁺ release. Since mitochondria have been shown to function as Ca²⁺ buffering organelles,^58)–61) a mitochondria-targeted GFP-based Ca²⁺ indicator was used to image the intra-mitochondrial Ca²⁺ concentration.⁵⁷⁾ The results showed that there is a preferential loading of mitochondria with Ca²⁺ from the ER during the Ca²⁺ release phase in the first Ca²⁺ oscillation (Fig. 6A). After that, Ca²⁺ was released from the mitochondria, when the ER was taking up Ca²⁺. Thus, Ca²⁺ shuttles between the ER and mitochondria during each Ca²⁺ oscillation. As the ER Ca²⁺ loading increased, there was a decline in the ER Ca²⁺ uptake rate, which then allowed mitochondrial Ca²⁺ release to increase the cytoplasmic Ca²⁺ concentration. The increase in the cytoplasmic Ca²⁺ concentration was followed by regenerative Ca²⁺ release from the ER via the IP₃R to generate the second and subsequent Ca²⁺ oscillations (Fig. 6B). These findings suggest that mitochondria play a role in the pacemaking of Ca²⁺ oscillations, and that Ca²⁺ efflux from the mitochondria underlies the pacemaker Ca²⁺ increase before the regenerative Ca²⁺ release from the ER.⁵⁷⁾

Fig. 6.

Ca²⁺ regulation by endoplasmic reticulum and mitochondria during Ca²⁺ oscillations. (A) The first Ca²⁺ oscillation is generated by Ca²⁺ release from the ER via IP₃R. A considerable fraction of Ca²⁺ released from the ER enters mitochondria. (B) The second and subsequent Ca²⁺ oscillations are initiated by the Ca²⁺ release from mitochondria, which then triggers regenerative Ca²⁺ release via IP₃R from the ER. Mitochondrial Ca²⁺ is partially reloaded. The sequence is repeated until mitochondrial Ca²⁺ is depleted. From K. Ishii et al., Ref. 57.

5. Physiological significance of Ca²⁺ oscillations

Ca²⁺ oscillation regulates many important cell functions. One such function is T-cell activation. Introduction of foreign tissues to the host is detected by T cells, in which Ca²⁺ oscillations are generated.⁶²⁾ The Ca²⁺ oscillation then activates the Ca²⁺-calmodulin-dependent phosphatase calcineurin. Calcineurin then dephosphorylates nuclear factor of activated T cells (NFAT), which in turn translocates to the nucleus to enhance transcription of a set of genes.⁶³⁾ Because Ca²⁺ oscillation frequency regulates NFAT-dependent transcription,⁶⁴⁾ the dependence of nuclear translocation of NFAT on Ca²⁺ oscillations is a physiologically important mechanism.

We studied the relationship between GFP-tagged NFAT translocation to the nucleus and the frequency of Ca²⁺ oscillation.⁶⁵⁾ Indeed, there was a frequency-dependent translocation of NFAT to the nucleus. In Ca²⁺ oscillations, [Ca²⁺]_i increases only a fraction of time, i.e., the fractional Ca²⁺ increase time ([sum of the duration of Ca²⁺ increase] ÷ [total duration] × 100%) is less than 100%. When the extent of nuclear translocation of NFAT was plotted against the fractional Ca²⁺ increase time, it showed a convex upward dependence rather than a proportional dependence (Fig. 7). This means that the extent of NFAT translocation induced by Ca²⁺ oscillation is supra-proportional to the duration of Ca²⁺ increase. In other words, Ca²⁺ oscillation is an ‘energy-saving’ signaling mechanism. Since prolonged increase in [Ca²⁺]_i may have adverse effects on cell survival,⁹⁾ Ca²⁺ oscillation is an efficient system to regulate cell functions.

Fig. 7.

Nuclear translocation of GFP-tagged NFAT in response to Ca²⁺ oscillations. The steady state value of the NFAT in the nucleus during continuous or oscillatory changes in [Ca²⁺]_i were plotted against the fractional Ca²⁺ increase time ([sum of the duration of Ca²⁺ increase] ÷ [total duration] × 100%). The observed values lie above the proportional dependence (dashed line). Modified from T. Tomida et al., Ref. 65.

The major reason why NFAT nuclear translocation utilizes the Ca²⁺ increase efficiently in an energy-saving manner lies with the ability of NFAT to ‘memorize’ Ca²⁺ increase for a certain amount of time after termination of the Ca²⁺ signal. Dephosphorylation of NFAT by calcineurin is a rapid process and proceeds within a short period of time. However, rephosphorylation of NFAT is rather a slow process taking several minutes. Thus, dephosphorylated NFAT, which is ready for translocation to the nucleus, has a life time of ~7 min.⁶⁵⁾ Therefore, Ca²⁺ oscillations with an interval shorter than the life time of dephosphorylated NFAT are nearly as effective as a continuous Ca²⁺ increase. This property associated with Ca²⁺ oscillations can be generalized, and any molecule that can sustain the effect of Ca²⁺ for a certain length of time should be able to utilize Ca²⁺ oscillations efficiently. For example, phosphorylation of smooth muscle myosin by the myosin light chain kinase is a rapid process, whereas dephosphorylation by the myosin light chain phosphatase is a relatively slow process.⁶⁶⁾ Therefore, phosphorylated myosin can memorize transient increases in Ca²⁺ concentration, and Ca²⁺ oscillation is expected to be an efficient signal to regulate contractions of smooth muscle cells.

6. Search for the new functions of Ca²⁺ signaling

As shown in Fig. 1, Ca²⁺ has been shown to regulate numerous cell functions. It seems likely that there are still many unknown cell functions that are regulated by Ca²⁺ signals. Identification of such cell functions will not only further clarify the physiological significance of Ca²⁺ signals but also shed new light on various cell functions. For the search of hitherto unrecognized functions of Ca²⁺ signals, a new and specific method to inhibit Ca²⁺ signaling was developed.

IP₃ 5-phosphatase hydrolyzes IP₃ with a high specificity.⁶⁷⁾ Therefore, overexpression of the enzyme is expected to block Ca²⁺ signaling by preventing the increase in IP₃ concentration.^53),68) Indeed, when cerebellar Purkinje cells (PCs) were transduced with IP₃ 5-phosphatase by Sindbis viral infection, synaptic input-dependent IP₃ and Ca²⁺ signaling were effectively inhibited.⁶⁹⁾ PCs are the principal neurons in the cerebellar cortex and receive numerous inputs from the axons (parallel fibers, PFs) of granule cells to their dendrites. The PF-PC synapses are considered to play a major role in such functions as motor learning and motor coordination.⁷⁰⁾ This method clarified a new role of IP₃ signaling in the PF-PC synapse. In the spines of the PCs, that is, the postsynaptic side of the PF-PC synapse, both ionotropic and metabotropic glutamate receptors are expressed. When PFs are stimulated with burst stimulation consisting of several pulses at 50–100 Hz, which mimics the physiological stimulation pattern,⁷¹⁾ metabotropic glutamate receptor-dependent IP₃ signal is generated in the PCs.⁶⁹⁾ The IP₃ signal in turn induces a local and transient Ca²⁺ signal in the spines and dendrites of PCs receiving the PF inputs.^72),73) In other words, the postsynaptic IP₃-Ca²⁺ signaling functions as a detector of the presynaptic activity. When the IP₃ signaling was blocked in PCs by IP₃ 5-phosphatase, the synaptic strength of the PF-PC synapse was inhibited. Further analyses showed that postsynaptic IP₃ signaling drives a brain-derived neurotrophic factor (BDNF) signaling from the postsynaptic PCs to the presynaptic PF terminals and that the retrograde signaling maintains the presynaptic function.⁷⁴⁾ Therefore, the postsynaptic IP₃-Ca ²⁺ signaling mechanism plays an important role in the activity-dependent synaptic maintenance mechanism.

The IP₃ 5-phosphatase method clarified another mechanism that is regulated by IP₃-Ca²⁺ signaling. Astrocytes, the major glial cells in mammalian brain, generate spontaneous Ca²⁺ oscillations in vivo as well as in vitro.^75)–78) The astrocytic Ca²⁺ oscillations were blocked by IP₃ 5-phosphatase, and neurite growth over the Ca²⁺ signal-deficient astrocytes was inhibited. The analysis clarified that Ca²⁺ oscillations regulate the expression of N-cadherin on the surface of astrocytes, and that N-cadherin is important for the maintenance of neurite growth.⁷⁹⁾

7. Perspectives

Studies on the basic principle of Ca²⁺ signaling led to the clarification of the critical role of the regenerative nature of the IP₃-induced Ca ²⁺ release me chanism in the spatiotemporal generation pattern of Ca²⁺ signals, such as Ca²⁺ waves and oscillations. Ca²⁺ oscillations provide an efficient way to drive cell functions with a temporally-distributed short duration of Ca²⁺ increase, thus avoiding the adverse effects of continuous Ca²⁺ increases. In parallel with such studies, a new specific method to inhibit Ca²⁺ signaling was discovered and helped us to identify new cellular functions, i.e., an activity-dependent synaptic maintenance mechanism and a neuronglial cell interaction. Further clarification of the new roles of Ca²⁺ signaling is expected to shed new light on our understanding of many important biological functions.

Abbreviations

ER

endoplasmic reticulum

IP₃

inositol 1,4,5-trisphosphate

IP₃R

IP₃ receptor

CICR

Ca²⁺-induced Ca²⁺ release

PLC

phospholipase C

GFP-PHD

GFP-tagged Pleckstrin homology domain

NFAT

nuclear factor of activated T cells

PC

Purkinje cell

PF

parallel fiber

BDNF

brain-derived neurotrophic factor

Profile

Masamitsu Iino was born in 1950 in Yamagata, Japan. After receiving MD in 1976 from School of Medicine, Tohoku University, he started his research carrier in Department of Pharmacology headed by Professor Makoto Endo in the same university. After obtaining PhD from Graduate School of Medicine, Tohoku University in 1980, he carried out research in Department of Physiology, University College London as a postdoctoral fellow. After two years of study in the United Kingdom, he returned to Tohoku University in 1982. He then moved to The University of Tokyo in 1984. He was promoted to lecturer in 1991, and then became professor of Department of Pharmacology, Graduate School of Medicine, The University of Tokyo in 1995. Since then he led a research group strong in innovative generation of new imaging methods of signaling molecules. In the early years of his research, he studied Ca²⁺ signaling mechanism in muscle cells. He found a regenerative property of the intracellular Ca²⁺ release mechanism. His following research proved that the mechanism is physiologically essential to generate spatiotemporal patterns of Ca²⁺ signaling, which forms a basis for the versatility of Ca²⁺ as a cellular switch. He then changed his field to neuroscience, and is now extending his research to find novel functions of Ca²⁺ signaling in the brain to shed new light on the mystery of our brain. For his achievements he was awarded Uehara Prize in 2009. His laboratory has produced many talented young investigators. He is an editorial board member of The Journal of Physiology, and is serving as vice dean of the Graduate School of Medicine, The University of Tokyo since 2007.