Introductory Review Cellscience Reviews Vol 1 No.1 ISSN 1742-8130

Diversification of Function and Pharmacology in Intracellular Calcium Signalling

Abdelilah Arredouani

Introduction

A RISE in the intracellular free calcium concentration ([Ca²⁺]i) is a key signal in the initiation of a wide range of cellular events including fertilization, cell division, cell proliferation and differentiation, muscle contraction, hormone and neurotransmitter secretion, and ultimately cell death. The resting intracellular calcium concentration, or activity ([Ca²⁺]i), is maintained at around 100 nM, whereas the extracellular Ca²⁺ concentration is in the mM range. This high electrochemical gradient allows rapid rises in local [Ca²⁺]I, often referred to as events, which may take place over the millisecond timescale and which can easily reach µM levels. During evolution cells have developed many sophisticated and uniquely tailored systems to order to effect efficient and fast spatio-temporal changes in [Ca²⁺]i. Thus, both the outer cell plasma membrane and those of intracellular organelles are equipped with highly specialized proteins which allow [Ca²⁺]i changes through influx from the extracellular space and clearance by active extrusion, and those which allow mobilization from and uptake into intracellular stores such as the sarco-endoplasmic reticulum (SER), mitochondria, Golgi apparatus, nucleus and acidic granules. This review will only consider the contribution of intracellular Ca²⁺ stores to Ca²⁺ signaling and homeostasis. More detailed information about each of the cellular mechanisms mentioned may be found within the references cited.

I. Contribution of the sarco-endoplasmic reticulum to the control of [Ca²⁺]i changes

By far the sarco-endoplamic reticulum (SER) is the major intracellular Ca²⁺ store in most cell types. This organelle is equipped with sophisticated cellular systems allowing it to quickly buffer or “damp” [Ca²⁺]i rises (Arredouani et al., 2002c; Gilon et al., 1999), probably to prevent Ca²⁺ cytotoxicity, and to release Ca²⁺ in response to different extracellular stimuli. Ca²⁺ reuptake into the SER from the cell’s intracellular matrix, or cytoplasm, is mediated by a very efficient P-type Ca²⁺-ATPase, the Sarco-Endoplasmic Reticulum Ca²⁺-ATPase (SERCA). Signal induced Ca²⁺ mobilization into the cytoplasm on the other hand is mediated by intracellular Ca²⁺ channels located within the plasma membrane of the organelle: namely the IP3 and ryanodine receptors. The coordination between Ca²⁺ uptake and release is involved in the [Ca²⁺]i oscillatory pattern in pancreatic beta cells (Arredouani et al., 2002b), and essential for the for Ca²⁺ entry in non excitable cells via the capacitative calcium entry mediated by the store operated Ca²⁺ channels in the plasma membrane (Putney, Jr., 1990).

Figure 1. Schematic representation of major intracellular organelles involved in Calcium homeostasis including the Golgi apparatus, Nucleus, Endoplasmic Reticulum (ER) and Mitochondria (labeled). Routes of calcium entry into the cytoplasm via release and entry channels are given by white arrows, whilst Ca²⁺-ATPase extrusion mechanisms are depicted by blue arrows.

A. Role of the SER in the lowering of [Ca²⁺]i

a. Molecular structure of the SERCA pump

The SERCA pump is a protein with a hydrophobic region which is integrated into the lipidic bilayer of the endoplasmic reticulum (ER) and a hydrophilic region which protrudes into the cytosol (Herbette et al., 1977). The hydrophobic region is comprised of 10 transmembrane segments (M1-M10) organized into a helices (MacLennan et al., 1985). The two regions are linked by a stalk (Inesi & Asai, 1968) which is made of 5 segments (S1-S5) linked to 5 transmembrane ones. The C and N termini are both cytosolic (Matthews et al., 1989). Oxygen-rich residues localized in M4, M5, M6 and M8 segments form two high sites with a high affinity for Ca²⁺ (Toyoshima et al., 2000), forming a cavity responsible for the translocation of the calcium ion (MacLennan et al., 1997). The long loop between S4 and S5 of the stalk contains residues which are involved in the hydrolysis of ATP, and aspartate 351 to which the hydrolysed ATP ? phosphate is transferred (Lee & East, 2001). Moreover, the site of fixation for thapsigargin, a potent and specific inhibitor of the SERCA (Lytton et al., 1991) is composed from regions contributed by the M3/M4 and M7/M8 loops (Young et al., 2001). The basic alpha helical arrangement and two Ca²⁺ binding sites of the SERCA pump are represented in Figure 2.

Figure 2. Structure and predicted topology of the SERCA showing Ca²⁺ and Asp 351 binding sites.

b. Diversity and distribution of the SERCA

Three genes coding for 3 distinct proteins (SERCA1-SERCA3) have been cloned in higher vertebrates. In the human, these genes are located on chromosomes 16, 12 and 17 for SERCA1, 2 and 3 respectively (Dode et al., 1996; MacLennan et al., 1987). Alternative splicing leads to a total of 7 known isoforms (SERCA1a and b, SERCA2a and b and SERCA3a, b and c) which show greatest variability in sequence at their C terminal moieties (Andersen & Vilsen, 1998). SERCA1a is expressed exclusively in adult fast skeletal muscle fibers, whereas the SERCA1b is expressed in the same fibers but only in the fetus (Brandl et al., 1987). SERCA2a is expressed both in cardiac muscle and in the slow skeletal muscle fibers (Brandl et al., 1987), whilst SERCA2b is a housekeeping protein (Carafoli & Brini, 2000). The expression of SERCA3 is restricted to some tissues (Andersen & Vilsen, 1998), and is weakly expressed in muscle (Carafoli & Brini, 2000).

c. Characteristics of the different isoforms

At the functional level, expression of the different SERCA isoforms in vitro revealed no significant differences between SERCA1a and b (Dode et al., 1998). However, the turnover of SERCA1 is 2-fold faster than that of SERCA2a (Sumbilla et al., 1999), though their Ca²⁺ affinities appear similar (Lytton et al., 1992). The Ca²⁺ affinity of SERCA2b (Km ~ 0.17µM) is 2 fold higher than that of SERCA2a (Km ~ 0.31µM, (Verboomen et al., 1992). Functional studies of SERCA3 showed that this ATPase has properties distinct from the 2 other isoforms. It has a lower Ca²⁺ affinity (Km ~ 2µM) (Arredouani et al., 2002a), a high optimal pH (7.2-7.4 versus 6.8-7.0) and 10-fold higher sensitivity to inhibition by vanadate (Wuytack et al., 1995). The affinity for ATP is similar for all SERCA isoforms (0.02-0.05 µM). The particular biochemical characteristics and the restricted tissue distribution of SERCA3 might suggest a role in specialized signaling functions (Liu et al., 1997). Ablation of the expression of one SERCA isoform, e.g. SERCA3 ablation in pancreatic B-cells, often does not impair primary cellular function, in this case insulin secretion, suggesting distinct roles of different sarcoendoplasmic reticulum Ca²⁺ pumps for Ca²⁺ homeostasis (Arredouani et al., 2002).

d. Modulation of SERCA activity

Because of the prominent importance of SERCA for the control of [Ca²⁺]i changes in many cell types, any modulation of its activity should have a considerable effect upon Ca²⁺ homeostasis.

d.1. Physiological modulation of SERCA

1. By phospholamban

Phospholamban (PLN), a 52 AA peptide (6.1 kDa) with respectively cytosolic and transmembrane N- and C- termini (East, 2000), is expressed in cardiac, smooth and slow skeletal muscles. In the human, its gene is localized on chromosome 6 and no isoforms are presently known (Fujii et al., 1991). PLN is an integral part of the SER membrane and inhibits, in its non-phosphorylated state, the activity of SERCA. This inhibition is reflected by a diminution of the affinity of SERCA for Ca²⁺ (Kimura et al 1996), and thus a reduction of the amount of Ca²⁺ taken up into the SER. The inhibitory effect of PLN is removed when it’s phosphorylated by kinases or when [Ca²⁺]i rises (Odermatt et al., 1998). In vivo, the PLN might be phosphorylated by PKA and the CaM kinase independently of one another. In vitro, it could be phosphorylated by two other kinases, PKC and PKG (Movsesian et al., 1984). The PLN is abundantly expressed in cardiac muscle where it plays a major regulatory role in the inotropic response to ß-adrenergic stimulation (Odermatt et al., 1998). In vitro however, the PLN can inhibit the activity of both SERCA1 and SERCA2 (East, 2000), though the PLN is not expressed in the fast skeletal muscle fibers. SERCA3 is either not, or is only weakly regulated by PLN (Wuytack et al., 1995).

2. By sarcolipin

Sarcolipin (SLN) is a 31 AA polypeptide implicated in the regulation of SERCA1. Its expression profile resembles that of SERCA1 and its co-expression with SERCA1 in vitro reduces the Ca²⁺ affinity of the pump and augments its Vmax (Odermatt et al., 1998). It’s suggested that the modulation of SERCA1 by SLN is [Ca²⁺]i dependent: when [Ca²⁺]i is low the SLN would inhibit the Ca²⁺ uptake by reducing the Ca²⁺ affinity of the pump, whereas at high [Ca²⁺]i it stimulates Ca²⁺ reuptake into the SER by increasing the Vmax. The effect of SLN and PLN are additive, suggesting distinct mechanisms of modulation and binding to SERCA1 (Odermatt et al., 1998).

3. By CaM kinase

SERCA2 (a and b) activity can be modulated by direct phosphorylation by a CaM kinase associated with the SER by phosphorylating a Serine at position 38 (Toyofuku et al., 1994). This phosphorylation leads to a 50 to 70% increase of the Vmax of Ca²⁺ reuptake (Xu & Narayanan, 1999). SERCA1 and 3 are not however modulated by this pathway because Serine 38 is not conserved in these isoforms (Dode et al., 1996).

d.2. Pharmacological modulation of SERCA

The best known inhibitor of the SERCA Ca²⁺ pump is thapsigargin (TG), a potent and specific inhibitor of all SERCA isoforms. TG is extracted from the roots of Thapsia garganica, a plant of the umbelliferae family, and is widely used to deplete intracellular Ca²⁺ stores (Inesi & Sagara, 1994). It inhibits the SERCA by binding to the E2 conformation of the enzyme. This interaction is high affinity (Ki < 1nM) and the complete and irreversible inhibition is obtained with a 1:1 (TG:E2) stoichiometry (Young et al., 2001). There are however other known inhibitors of the SERCA family such as cyclopiazonic acid (CPA), 2,5-di(t-butyl) hydroquinone (DBHQ), and the thapsigargicin, but these are less potent than TG (reviewed in (Inesi & Sagara, 1994).

B. Role of the SER in [Ca²⁺]i rise

Beside its role in the lowering of [Ca²⁺]i exerted via the active reuptake of Ca²⁺ by the SERCA, the SER contributes to [Ca²⁺]i changes by rapidly releasing Ca²⁺. This is made possible owing to the presence within the organelle membrane of Ca²⁺ channel receptors which are activated in response to a wide range of external stimuli. Two type of the SER Ca²⁺ channels are widely expressed in many cell types and play primordial roles in some physiological functions: the IP3 receptors (IP3R), which are activated by IP3, and the ryanodine receptors (RyR), which are sensitive to ryanodine.

B.1. Ca²⁺ release via the IP3 receptors

This phenomenon is commonly termed “IP3-induced Ca²⁺ release (IP3-ICR)”. Receptors coupled to phospholipase C (PLC) via a G protein (Walters & Sepulveda, 1991) or tyrosine kinase (Walters et al., 1996) could activate this transduction pathway by producing IP3 from the hydrolysis of the phospholipid PI(4,5)P2 present within the plasma membrane. The IP3 released diffuses into the cytosol and binds to its receptor (IP3R) within the SER membrane. The ensuing activation of the Ca²⁺ channels allows Ca²⁺ to flow down its electrochemical gradient into the cytosol, thereby increasing the [Ca²⁺]i (Berridge, 1997).

a. Molecular structure and diversity of IP3 receptors

IP3R’s are the intracellular Ca²⁺ channels most commonly involved in allowing receptors in the plasma membrane to stimulate Ca²⁺ mobilization from intracellular stores. Molecular cloning has identified three mammalian subtypes of IP3R (I, II and III) showing 60-80% amino acid sequence homology (Berridge, 1993). This diversity is further increased by alternative splicing. The IP3R-I is alternatively spliced in 3 regions; S1, S2 and S3. Splicing at S2 gives rise to 5 receptor variants. Most of the structural and functional knowledge of IP3R’s are derived from studies on the type I isoforms, the first to be cloned and purified. The functional rat IP3R-I is a 2,749 amino acid tetramer (~313 kDa) (Furuichi et al., 1990), whilst the IP3R-II and III receptors are 2,710 (~307 kDa) and 2,670 (~304 kDa) amino acid residues in length respectively (Yoshida & Imai, 1997).

The overall structure of the IP3R is similar for the three isoforms. The IP3R-I comprises 3 domains: an N-terminal domain (~24%) containing the IP3 binding site, a large intermediary domain, which is involved in coupling and modulation (60%), and finally a C-terminal domain forming the Ca²⁺ channel itself (Marks, 1997). The binding site for IP3 is contributed by 2 basic residue-rich domains separated by a loop (Yoshikawa et al., 1999). The basic residues are likely to interact with phosphate groups of IP3. The N-terminal and the modulation domain form together the cytosolic moiety of the protein. The C-terminal domain, forming the channel, is composed of 6 transmembrane segments (S1- S6). A region between S5 and S6 is believed to form the pore of the channel (Michikawa et al., 1994). No function is known for the cytosolic C-terminal. The different IP3R isoforms can form homo- or hetero-tetramers (Nucifora, Jr. et al., 1996), but the functional significance of hetero-tetramer IP3R’s has not been established.

b. Electrophysiological analysis of the IP3R

The internal localization of the IP3R makes it difficult to study its electrophysiological properties in intact cells. Most of the studies were thus carried by incorporating recombinant IP3R in lipid bilayers (Bezprozvanny & Ehrlich, 1994). IP3R-I derived from canine cerebellum when incorporated into lipid bilayers has a unitary conductance, with Ca²⁺ as the current carrier, of about 50-70pS (Bezprozvanny & Ehrlich, 1994). Under the same conditions the unitary conductance of the rat cardiac IP3RII was estimated to be about 70pS (Perez et al., 1997). However it is difficult to study the electrophysiological properties and regulation of ion channels within organelles outside of their native environment and lipid microdomains.

c. Modulation of IP3R activity

The activity of IP3R is modulated by a variety a physiological and pharmacological agents such as IP3, Ca²⁺, ATP, phosphorylation, arachidonic acid, caffeine and heparin. The binding and the target sites of most of these modulators are dispersed along the coupling and modulation domain of the receptor (Patel et al., 1999).

c.1. Physiological modulation

c.1.1 By IP3

The IP3R is activated by I(1,4,5)P3 binding, the only known physiological activators being inositol triphosphate, with EC50 values of ~500nM, 58 nM and 3200 nM measured respectively for IP3R-I, II and III (Thrower et al., 2001). The IP3R channel doesn’t open unless IP3 is bound to its site, and it seems that IP3 must bind to all 4 subunits for the channel to open (Marchant & Taylor, 1997).

Figure 3. IP3 is produced by the receptor-stimulated hydrolysis of phosphatidylinositol 4,5 bisphosphate (PIP2) by phospholipase C (PLC). IP3 stimulates the release of Ca²⁺ from the endoplasmic reticulum (ER) via the IP3R. Ca²⁺ entry also occurs from the extracellular space via the activation of voltage and ligand-gated membrane Ca²⁺ channels.

c.1.2 By Ca²⁺

One of the critical properties of the IP3R is its regulation by Ca²⁺, but this mechanism of regulation is isoform-dependent. For IP3R-I, the regulation is biphasic and bell-shaped (Missiaen et al., 1999). Low [Ca²⁺]i (<300nM) potentiates the effect of IP3, whereas [Ca²⁺]i in the µM range inhibits the release of Ca²⁺ by IP3 (Yoshida & Imai, 1997). This bell-shaped regulation by Ca²⁺ would be mediated by calmodulin which would have an inhibitory effect (Thrower et al., 2001). The Ca²⁺ sensitivity of the two other isoforms (II and III) differs from that of type I. In fact the regulation of IP3R-II activity by Ca²⁺ is sigmoidal in nature (Thrower et al., 2001) . The absence of the Ca²⁺-dependent inhibition can be explained, at least partly, by the fact that high [Ca²⁺]i does not inhibit the binding of IP3 whereas it does inhibit IP3R-I activity. The regulation of IP3R-III activity by Ca²⁺ is controversial. Some studies suggested that it is biphasic (Mak & Foskett, 1997; Missiaen et al., 1998), whereas others showed that the open probability of the type IP3R-III increases monotonically with increasing cytoplasmic Ca²⁺ concentrations (Hagar et al., 1998). The reasons of theses discrepancies are not known. The bell-shaped Ca²⁺-dependence curve of type IP3R-I is likely to be ideal for supporting Ca²⁺ oscillations, whereas the properties of the type IP3R-III would be better suited to signal initiation.

Ca²⁺ binding sites were localized in the modulation domain of the IP3R-I (Sienaert et al., 1997). Thus Ca²⁺ acts as an allosteric regulator of the IP3R-I. It can’t open the channel in the absence of IP3, but the presence of Ca²⁺ augments the open probability of the IP3R (Ehrlich et al., 1994). The Ca²⁺ sensitivity of IP3R activation also suggests that the receptor may mediate the phenomenon of Ca²⁺-induced Ca²⁺ release, which also involves the Ryanodine receptor (RyR). This would allow the mobilized Ca²⁺ to stimulate the neighboring IP3R and thus triggering a regenerative Ca²⁺ wave (Berridge, 1997).

The Ca²⁺ concentration within the SER lumen ([Ca²⁺]SER) is also affects the sensitivity of the IP3R to IP3 (Horne & Meyer, 1995; Tanimura & Turner, 1996), but the mechanism by which it does so is still controversial. A Ca²⁺ binding site, with an affinity of approximately 0.3 µM, is localized on the intraluminal loop between transmembrane segments S5 and S6 (Sienaert et al., 1997). It’s however unlikely that the effect of the luminal Ca²⁺ is mediated by an interaction with this site. In fact, because of the high [Ca²⁺]SER, one can expect that this specific site is continuously saturated. Alternatively, the effect of luminal Ca²⁺ may likely arise from the cytosolic side following its diffusion into the cytosol (Taylor & Traynor, 1995), or else via Ca²⁺-binding proteins present within the lumen such as calsequestrin or calreticulin (Kawasaki & Kasai, 1994).

c.1.3. By ATP

Like Ca²⁺, ATP also exerts a biphasic effect on the IP3R. The activity of the receptor is increased by µM and decreased at mM concentrations of ATP (Patel et al., 1999). Two independent ATP binding sites have been characterized and localized to the modulation domain of IP3R-I with affinities of approximately 17 and 0.5 µM respectively (DeLisle et al., 1996; Maeda et al., 1991). The IP3R-III has only one ATP binding site with an affinity of ~177 µM (Maes et al., 2001). The regulation of IP3R-II by ATP has not been investigated. The regulation of IP3R by ATP is mediated by the binding of the nucleotide to one or more sites upon the modulation domain (Thrower et al., 2001). Because of the many charges of the ATP molecule, the nucleotide at high concentration is supposed to inhibit the IP3R by acting as a competitive antagonist for the binding site of IP3 (Nunn & Taylor, 1990). The unequal number of ATP binding sites and the different affinities of the type I and III IP3R would potentially explain the differences in ATP modulation in intact cells (Maes et al., 2001). Like Ca²⁺, ATP acts also as an allosteric regulator (Thrower et al., 2001). The effects of other nucleotides such as adenosine, AMP, ADP, GTP and UTP on IP3RI activity were investigated. Only AMP and GTP act to significantly increase channel activity (Mak et al., 1999).

Figure 4. High resolution imaging of IP3R structure showing two conformational states and with schematic representation of the influence of Ca²⁺ and the four bound IP3 molecules upon its three-dimensional structure.

c.1.4. By phosphorylation

The IP3R’s are the substrate of different protein kinases such as cAMP and cGMP-dependent protein kinases (PKA, PKG), and the Ca²⁺-calmodulin-dependent protein kinase II (CaMKII). The phosphorylation by PKA has been demonstrated both in vitro (Wojcikiewicz & Luo, 1998) and in vivo (Joseph & Ryan, 1993; Wojcikiewicz & Luo, 1998). The IP3R-I has two consensus sequences for PKA phosphorylation (Furuichi et al., 1989). The effect of phosphorylation on IP3R-I activity by PKA varies from tissue to tissue. Whereas in hepatocytes PKA strongly potentiates of Ca²⁺ release by IP3 (Hajnoczky et al., 1993), probably by increasing the sensitivity of the IP3R to cytosolic Ca²⁺ (Joseph & Ryan, 1993), whereas the capacity of IP3 to trigger Ca²⁺ release is diminished (Volpe & rson-Lang, 1990) or weakly potentiated by PKA (Nakade et al., 1994) in the cerebellum. This discrepancy could be explained by the presence of different variants of the IP3R in the two tissues. In fact, the 2 consensus sequences of phosphorylation are located on both sides of the alternative splicing site SII, which could generate isoforms with different phosphorylation sites (Thrower et al., 2001). IP3R-II and III isoforms are also phosphorylated by PKA, but on serine residues distinct from those on IP3R-I and with less efficacy (Wojcikiewicz & Luo, 1998).

The IP3R is also phosphorylated by PKG both in vivo and in vitro, and only one phosphorylation site is involved (Ser 1755) (Komalavilas & Lincoln, 1994; Komalavilas & Lincoln, 1996). The effects of PKA and PKG phosphorylation are however not additive, suggesting that both kinases target the same phosphorylation site. Like the PKA, the effect of PKG phosphorylation is tissue-dependent, with a potentiation of IP3-ICR in permeabilized hepatocytes (Guihard et al., 1996) and its inhibition in gastric smooth muscle (Murthy et al., 1993).

The phosphorylation of IP3R-I by PKC and CaMKII has not been well elucidated, although multiple phosphorylation sites are present in the sequence of this isoform. The IP3R activity may also be regulated by phosphorylation via a tyrosine kinase activity (Jayaraman et al., 1996) which increases the open probability of the Ca²⁺ channel. The effects of IP3R phosphorylation by different kinases are far from being fully elucidated. More investigations are needed to better identify their physiological importance.

c.1.5 By protein-protein interaction

It was demonstrated that purified IP3R-I from cerebellum binds to CaM (Maeda et al., 1991). The binding of CaM decreases the affinity of the receptor to IP3, thus reducing by about 10-fold the capacity of the IP3R to mobilize Ca²⁺ (Patel et al., 1997). This inhibitory effect of CaM is Ca²⁺-dependent (Hirota et al., 1999), and its physiological significance is still open. On the other hand, it has been reported that the FKBP12 is tightly associated with IP3R and that the breaking of this interaction by FK506, an immunosuppressor, strongly increases the capacity of IP3 to release Ca²⁺ from microsomes (Cameron et al., 1995b). Indeed a consensus sequence for the binding of FKBP12 has been localized on the 3 IP3R isoforms (Bultynck et al., 2001). In addition, an interaction between calcineurin, a Ca²⁺-dependent phosphatase, and the IP3R has been observed (Cameron et al., 1995a). This interaction is made via FKBP12 (Yoshida & Imai, 1997). It is possible that this phosphatase regulates the phosphorylation state of IP3R. When activated by Ca²⁺, calcineurin dephosphorylates the IP3R, thus diminishing the IP3-ICR (Thrower et al., 2001). The FK506 acts most probably by preventing the calcineurin-dependent dephosphorylation of IP3R, by dissociating the FKBP12 through which it binds to IP3R.

c.2. Pharmacological modulation

Only the effects of caffeine and heparin, the most used agents to study the regulation of IP3R, will be considered. For more details on other agents see (Michelangeli et al., 1995).

c.2.1. By caffeine

This xanthine inhibits IP3-ICR with an IC50 of ~20 mM (Brown et al., 1992). Although the caffeine augments the production of cAMP by inhibiting phosphodiesterase activity, its effect on IP3R activity is not due to an elevation of cAMP concentration, which would be expected to affect the receptor activity via PKA activation. Neither does it appear to be as a consequence of a displacement of IP3 binding (Brown et al., 1992), which suggests that caffeine inhibits the opening of the Ca²⁺ channel per se rather than the binding of IP3. The effect of the caffeine is counteracted by ATP, indicating that the inhibition results from a competitive interaction with an ATP binding site (Maes et al., 2001; Michelangeli et al., 1995).

c.2.2. By heparin

The most potent inhibitor of the IP3R is heparin. It has a high affinity for the IP3 binding site, thus acting as a competitive inhibitor (IC50 ~ 3 µM) (Michelangeli et al., 1995). However, the use of heparin as an antagonist has its limitations because of the reported non-specific effects such as activation of L-type voltage-dependent Ca²⁺ channels and the uncoupling of receptors from their G proteins (Michelangeli et al., 1995). It has also been reported that heparin may stimulate the release of Ca²⁺ through the RyR (Bezprozvanny et al., 1993). Morover, the use of this agent presents technical difficulties because it is not cell membrane permeant and needs to be injected.

B.2. Ca²⁺ Release via the ryanodine receptors

The second mechanism by which Ca²⁺ may be mobilized from the SER involves a Ca²⁺ channel of the organelle membrane commonly known as the ryanodine receptor (RyR), because of its high affinity for ryanodine, a plant alkaloid extracted from Ryana speciosa. Thus ryanodine is largely used as a ligand for the identification, purification, cloning and functional characterization of the RyR family (Masumiya et al., 2001).

a. Structure and molecular diversity of ryanodine receptors

Three RyR isoforms have been identified to date (RyR1, 2 and 3). They show ~ 70% sequence homology and hypervariability within their C-termini. The type 1 RyR presents two variants (ASI and ASII), the expression of which is developmentally regulated and tissue dependent (Shoshan-Barmatz & Ashley, 1998). The type 3 isoform undergoes extensive splicing, which would be expected to have a physiological functional significance. At least 3 variants of RyR3 are known (I, II and III) of which the expression is tissue-dependent (Miyatake et al., 1996).

The RyR1 is expressed mainly in skeletal muscle (Takeshima et al., 1989), but also in some smooth muscles (Neylon et al., 1995) and in some brain areas such as the cerebellar Purkinje cells (Furuichi et al., 1994). The RyR2 is highly expressed in cardiac muscle (Nakai et al., 1990) and the most distributed isoform in the brain but with low expression (Giannini et al., 1995). A weak expression of RyR2 in smooth muscle has also been reported (Neylon et al., 1995). As for RyR3, it is highly expressed within specific regions of the brain (hippocampus, thalamus and striatum) (Murayama & Ogawa, 1996) and in smooth muscle (Giannini et al., 1992), and also weakly in skeletal and cardiac muscle (Giannini et al., 1995), and in some non-excitable cells such as T-lymphocytes (Hakamata et al., 1994). Many cell types express more than one RyR isoform, but the physiological significance of this co-expression has not been established.

Figure 5. Three dimensional structure of the RyR1 receptor showing topology and calmodulin (CaM) and FKBP12 binding sites.

The RyR is an intracellular Ca²⁺ channel structurally resembling the IP3R (Grunwald & Meissner, 1995), but possesses distinct biophysical and pharmacological characteristics. The RyR1 was the first to be cloned from skeletal muscle and is the best characterized (Takeshima et al., 1989). It exists as a homo-tetrameric complex (in contrast to the IP3R which may exist as a hetero-tetrameric assembly) made of 4 subunits each 5,037 amino acids in length (~ 560 kDa). The primary domain of the receptor is present upon the cytosolic face of the membrane (the “foot” region). The transmembrane segments containing the channel pore are located within the C-terminal region, and both the N and C termini are cytosolic (Grunwald & Meissner, 1995).

b. Biophysical characteristics of ryanodine receptors

The cationic selectivity of the RyR for is low. The channel pore does not or only weakly discriminates between Ca²⁺ and Ba2+ (Shoshan-Barmatz & Ashley, 1998). All the RyR isoforms show multiple conductance states, the most frequent of which is 100 to 150pS. They are characterized by a high unity conductance, both for monovalent (~ 750 pS for K⁺) and for divalent cations (~ 150 pS for Ca²⁺) (Meissner, 1994). The permeability of the RyR to anions is negligible (Lindsay & Williams, 1991).

c. Modulation of RyR activity

The RyR activity is modulated by a variety of intracellular second messengers and many drugs. In this section only modulators of physiological importance are considered, along with the most widely used pharmacological agents. More detailed discussions about RyR pharmacology may be found in other sources (Franzini-Armstrong & Protasi, 1997; Shoshan-Barmatz & Ashley, 1998).

c.1. Physiological modulation

1. By Ca²⁺

The cytosolic Ca²⁺ has a major importance in the regulation of RyR activity and is considered to be its principal physiological activator, at least for RyR2 and RyR3, because the other ligands are unable to activate the channel in the absence of Ca²⁺, or else they require it to exert a maximal effect (Meissner, 1994). The three known isoforms have different sensitivities to Ca²⁺. Whereas the activation of RyR1 by Ca²⁺ is bell-shaped, that of RyR2 and RyR3 is not (Sarkozi et al., 2000). The maximal activation of RyR1 is obtained with ~ 5 µM, and decreases at [Ca²⁺]i levels in the mM range (Shomer et al., 1994). This biphasic effect of Ca²⁺ suggests the presence of 2 different binding sites for Ca²⁺: one with high affinity (Km ~ 1µM) which stimulates the opening of the channel, and another with low affinity (Km ~ 1mM) which inhibits the channel (Shoshan-Barmatz & Ashley, 1998). The RyR3 seems to be 10 times less sensitive to Ca²⁺ than the other two isoforms (Takeshima et al., 1995). The Ca²⁺ binding sites have not yet been identified though regions with EF hand motifs have been localized in RyR1 (Takeshima et al., 1989; Xiong et al., 1998).

The release of Ca²⁺ from the SER via the RyR may also be modulated by the luminal Ca²⁺ (Koizumi et al., 1999) but the underlying mechanism is controversial. Same hypothesizes as for IP3R have been suggested (see above) but they should be confirmed.

2. By cyclic adenosine 5’-diphosphoribose (cADPr)

cADPr is an endogenous compound present in many tissues from invertebrates, mammals and human (Lee, 2004). Its synthesis from nicotinamide-adenine dinucleotide (NAD+) is catalysed by ADPr cyclase, which also catalyses its hydrolysis to ADPr (Higashida et al., 2001). Two other cytoplasmic membrane proteins, CD38 ( a lymphocyte T anigen) and CD157 (a lymphocyte B antigen) may also produce cADPr from NAD⁺ (Lee, 2004), although the mechanism is still controversial. It has been recently reported that the basal cADPr concentrations are respectively ~ 3.9 and 0.63 fmol/µg protein from pancreatic acinar cells isolated from wild type and CD38 knock-out mice (Fukushi et al., 2001). Investigations are ongoing to better characterize the pathway (s) used in intact cell to produce cADPr.

cADPr-induced Ca²⁺ release (ICR) was first demonstrated in sea urchin eggs (Galione et al., 1991), and later in many cell types including neurons (Higashida et al., 2001), pancreatic beta cells (Takasawa et al., 1993), smooth, skeletal and cardiac cells (Zucchi & Ronca-Testoni, 1997), pancreatic acinar cells (Fukushi et al., 2001) and plant cells (Allen et al., 1995). This list is probably not exhaustive.

The mechanism by which the cADPr exerts its Ca²⁺ releasing effect is not yet fully elucidated. Pharmacological evidences suggest that the Ca²⁺ channel involved is the RyR. For instance, Ca²⁺ and caffeine-induced Ca²⁺ release are potentiated by cADPr (Lee, 1993), whereas ruthenuim red and high concentration of Mg²⁺ inhibit cADPr-ICR (Galione, 1993; Guse et al., 1996). The effect of cADPr is observed in cells expressing RyR2 (Cui et al., 1999) as well as in those expressing RyR3 (Guse, 1999). This effect is not clear in RyR1-expressing cells (Sitsapesan & Williams, 1995). The Ca²⁺ sensitivity of cADPr-ICR is controversial. An increase of the RyR2 open probability (Po) by cADPr was observed at sub-activating levels of [Ca²⁺]i (Meszaros et al., 1993), but another report showed that the Po can be increased only in the presence of activating [Ca²⁺]i (Sitsapesan & Williams, 1995). Moreover, the [cADPr] required to activate RyR2 (= 1µM), are largely higher than the physiological concentrations (Sitsapesan et al., 1995). More investigations are needed to unravel the physiological role of cADPr in Ca²⁺ signaling.

Until now, there is no experimental evidence of a direct interaction between cADPr and RyR. An indirect interaction can’t however be excluded. It’s possible that the effect of cADPr passes through proteins which interact with RyR. Pretreatment of RyR2 by FK506 reduces the cADPR-ICR, suggesting that cADPr and FK506 act at the same site on the FKBP12.6 (see bellow) (Noguchi et al., 1997).

3. By phosphorylation

The RyR is the substrate of many kinases (PKA, PKC, PKG and CaMKII), and potential phosphorylation sites by different kinases have been identified by analyzing the primary RyR structure (Shoshan-Barmatz & Ashley, 1998). The effect of phosphorylation is isoform-dependent. The phosphorylation of RyR2 by PKA for instance increases its Ca²⁺ sensitivity (Valdivia et al., 1995). The CaMKII effects are variable. It activates (Witcher et al., 1991) or inhibits the RyR2 (Lokuta et al., 1995). The phosphorylation of RyR1 is controversial. A serine at 2843 position has been reported to be phosphorylated by PKA, PKG and CaMKII (Suko et al., 1993). Other investigators have however observed only a very weak RyR1 phosphorylation by PKA or CaMKII (Strand et al., 1993).

On the other hand, it has been proposed that calcineurin, a Ca²⁺-dependent phosphatase, is associated with RyR2, and that this association would be the result of the activation of this phosphatase by the rise of [Ca²⁺]i by Ca²⁺ release from the SR (Bandyopadhyay et al., 2000). It’s tempting to believe that this association would dephosphorylate the RyR and terminate the Ca²⁺ release, but this remains to be confirmed.

Figure 6. Illustration of mechanism of Ca²⁺ Induced Ca²⁺ Release (CICR) by the activation of RyR’s by Ca²⁺ released from IP3 sensitive stores. Adapted from http://medweb.bham.ac.uk/research/toescu/Teaching/GIT/IP3.gif.

4. By ATP

During excitation-contraction coupling, ATP augments the release of Ca²⁺ from the SR by increasing the Ca²⁺ sensitivity of the RyR (Laver et al., 2001). The presence of at least 2 ATP binding sites on RyR1, and 2-4 sites on RyR2 and RyR3 is suggested by the identification of consensus sequences in the primary structure of these proteins (Zucchi & Ronca-Testoni, 1997). The mechanism underlying the potentiation of CICR by ATP remains unclear. One possibility would be a diminution of the free Mg²⁺ activity (inhibitory effect, see below) caused by the increase in [ATP]. Other nucleotides such as ADP, AMP, cAMP, adenine and adenosine are also known to potentiate CICR (Shoshan-Barmatz & Ashley, 1998), whereas CTP, GTP, ITP and UTP have only minor effects (Zarka & Shoshan-Barmatz, 1993).

5. By Mg²⁺

Experiments using ⁴⁵Ca²⁺ efflux from SR and measurements of lipid bilayers-incorporated RyR activity have showed an inhibitory effect of Mg²⁺ on CICR (Shoshan-Barmatz & Ashley, 1998). In skeletal muscle (RyR1) for example, the IC50 for Mg²⁺ is ~ 20 and 70-200 µM in the presence of 1 and 10 µM Ca²⁺ respectively. RyR2 is less sensitive to inhibition by Mg²⁺, as IC50 values higher than 300 µM were observed in the presence of 10 µM [Ca²⁺]i (Zucchi & Ronca-Testoni, 1997). The precise mechanisms of the Mg²⁺-RyR interaction remain to be determined.

The inhibitory effect of Mg²⁺ would result from a competitive displacement of Ca²⁺ from its high affinity stimulating site or from a binding to the low affinity inhibitory site (Shoshan-Barmatz & Ashley, 1998; Zucchi & Ronca-Testoni, 1997). The mechanism underlying the inhibitory effect of Mg²⁺ on RyR in vivo is however is unclear, as the physiological [Mg²⁺]I is higher than 1mM. It would be related to the phosphorylation state of the RyR. The Mg²⁺ would inhibit only the non-phosphorylated form of RyR, and this inhibition would be removed by a CaM-dependent mechanism (Shoshan-Barmatz & Ashley, 1998).

6. By protein-protein interaction

6.1. Calmodulin (CaM)

In cardiac muscle, CaM inhibits CICR and blocks the activation of RyR2 by caffeine, with an IC of ~ 0.1-0.2 µM (Zucchi & Ronca-Testoni, 1997). Theses effects take place only at a [Ca²⁺]i higher than 0.1 µM and do not require ATP, suggesting that they involve the Ca²⁺-CaM complex but not the Ca²⁺-CaM-dependent kinases. Many CaM binding sites have been identified on RyR. It has been proposed that at rest up to 16 CaM molecules may be bound to RyR. During EC coupling, when [Ca²⁺]i increases following the activation of RyR, the Ca²⁺ binds the CaM and it’s the Ca²⁺-CaM complex that maintains the channel in its inactivated form until most of the free Ca²⁺ is pumped back into the SR (Tripathy et al., 1995).

6.2. The FK506 binding protein (FKBP)

The FKBP is the cytosolic receptor of the immunosuppressor FK506. Two FKBP were identified, FKBP12 and FKBP12.6, which are co-purified with RyR1 and RyR2 respectively (Higashida et al., 2001). The dissociation of the FKBP/RyR1 complex induces an increase in the Ca²⁺ and caffeine sensitivities of RyR1, and the appearance of many subconducatance states (Ahern et al., 1997; Brillantes et al., 1994). It has also been suggested that the FKBP12 would act a physical link between the VDCC transverse tubules and RyR1during the EC coupling (Lamb & Stephenson, 1996).

The FK506 affects the association of calcineurin to RyR2 (Bandyopadhyay et al., 2000), suggesting that this association involves the FKBP12.6. As for the IP3R, it is possible that FK506, by taking down the FKBP12.6, prevents calcineurin from binding to RyR2. This later sees its activity increased because it can’t be dephosphorylated by the phosphatase any more (Bandyopadhyay et al., 2000).

6.3. Calsequestrin

Calsequestrin (CSQ) is a protein present in the lumen of the SR which plays an important role in the storage of Ca²⁺ because of its low affinity for the ion and its high binding capacity (Lytton et al., 1992). It has been suggested that CNQ interacts with RyR and, in particular, that RyR activation releases the CNQ-bound Ca²⁺ (McPherson & Campbell, 1993). It’s also known that the CNQ/RyR interaction involves the junction, another 26 kDa protein, which would play a role in the organization and/or the function of the Ca²⁺ release complex (Jones et al., 1995). Changes in [Ca²⁺]SR would induce conformational changes in CSQ, which in turn would modulate RyR activity (Hidalgo & Donoso, 1995).

c.2. Pharmacological modulation

Many pharmacological drugs are able to modulate RyR activity. Their actions are often complex and unclear. Only the effect of ryanodine, caffeine, and ruthenium red (the most commonly used drugs) will be considered here. The effects of many other substances are reviewed in detail elsewhere (Zucchi & Ronca-Testoni, 1997).

1. Ryanodine

The effects of this alkaloid on RyR activity are concentration-dependent. The modulation of this activity is biphasic (bell-shaped), suggesting the presence of low (Kd between 30 nM and 4 µM) and high (Kd between 2 and 200 nM) binding sites, probably on each of the monomers (Shoshan-Barmatz & Ashley, 1998). Low ryanodine concentrations (nM) open the Ca²⁺ channel (Smith et al., 1988), whereas high concentrations (>10µM) inhibit it irreversibly (Carroll et al., 1991). The binding of ryanodine to high affinity sites blocks the channel in a partially open state (40%), whereas the occupation of low affinity sites stabilizes it within a closed conformation. The exact localization of the ryanodine binding sites is not well known. It seems that both sites are in the C-terminal region near or even in transmembrane segment of the RyR (Callaway et al., 1994).

2. Caffeine

Caffeine was widely used in Ca²⁺ release studies from the SR. It increases the Ca²⁺ sensitivity of the release process, such that a mobilization of Ca²⁺ can take place even when [Ca²⁺]i is in the nM range (Zucchi & Ronca-Testoni, 1997). At relatively low concentrations (<2mM), the effect of caffeine depends on the presence of Ca²⁺, whereas at high concentrations (>5-10 mM) caffeine activates the RyR in the absence of the ion, indicating a more direct effect (Sitsapesan & Williams, 1990). The binding sites of caffeine however have not yet been identified.

3. Ruthenium red

This drug blocks Ca²⁺ release from SR in cardiac and muscles. The IC50 of its action is 19-90 nM for skeletal muscle (Zucchi & Ronca-Testoni, 1997). At the single-channel level, the drug decreases single-channel open probability (Buck et al., 1992). It is however technically difficult to use the drug because it is not cell permeant. Moreover it has been reported that it blocks many other processes such as Ca²⁺ uptake by mitochondria (Sparagna et al., 1995).

II. NAADP and Ca²⁺ signaling

The nicotinic acid dinucleotide phosphate (NAADP) is a relatively new Ca²⁺ mobilizing messenger which was first discovered in sea urchin eggs (Lee & Aarhus, 1995), and since the publication of its structure, cells responsive to this messenger have been found to be widespread and include plants, invertebrate eggs and neurons, amphibian neurons, as well as a variety of mammalian cells, including human (reviewed in (Lee, 2001)). The synthesis of NAADP results from a base-exchange reaction implicating NADP and nicotinic acid and which is catalyzed by the ADP-ribosyl cyclase (the same enzyme catalysing the synthesis of cADPr from NAD⁺) which replaces the nicotinamide group of NADP by nicotinic acid (Lee & Aarhus, 1995).

A. Which calcium store is the target for NAADP

Two main properties of NAADP-ICR indicate that the targeted Ca²⁺ store of this messenger is distinct from those of IP3 and cADPr: the insensitivity of NAADP-ICR to heparin and 8-NH2-cADPr, well known inhibitors of respectively IP3- and cADPr-ICR, and the lack of cross desensitization with these later processes (Chini et al., 1995; Lee & Aarhus, 1995). Moreover, it has been shown that the NAADP-ICR is insensitive to thapsigargin, suggesting that the involved Ca²⁺ store is not located in the endoplasmic reticulum (Genazzani & Galione, 1996). Recent studies from Galione’s group suggested that the NAADP-sensitive Ca²⁺ stores are reserve granules, the functional equivalent of lysosomes in sea urchin eggs (Churchill et al., 2002; Churchill et al., 2003; Yamasaki et al., 2004). These interesting findings will probably be consolidated by the identification of the NAADP receptor, which will greatly help to elucidate the mechanism of NAADP-ICR.

B. NAADP-ICR pharmacology

The NAADP-ICR is characterized particularly by its strong desensitization, and so far it’s the only specific means by which to block it (Aarhus et al., 1996). Sub-threshold concentrations (IC50 in the nM range) desensitize the NAADP receptor, such that subsequent challenge with even saturating concentrations of NAADP do not evoke Ca²⁺ release (Aarhus et al., 1996). The physiological interest of the inactivation is still not clear. It could be suitable for irreversible cellular events such as cell fertilization (Perez-Terzic et al., 1995), cell division and apoptosis (Genazzani & Galione, 1996). It has been reported that some L-type Ca²⁺ channel modulators and certain K + channel antagonists do inhibit NAADP-ICR (Genazzani et al., 1996; Genazzani et al., 1997). The underlying mechanisms of these blockers remains to be elucidated, and all of what is known is that their binding sites are distinct from those of NAADP (Genazzani et al., 1997).

III. Ca²⁺ Mobilization from other intracellular organelles

Besides the SER, other intracellular organelles have reported to behave as Ca²⁺ stores in many cell types. Most, if not all, of them express intracellular Ca²⁺ channels and Ca²⁺ pumps. The presence of multiple Ca²⁺ stores complicates the interpretation of [Ca²⁺]i changes in response to external stimuli, mainly when specific pharmacological tools are lacking.

A. The mitochondrion

The importance of mitochondria in the control of [Ca²⁺]i changes arises from the implication of this organelle in as major biological processes as apoptosis, oxidative stress and bioenergetics (Crompton, 1999; Lenaz et al., 2002; Montero et al., 2000). Ca²⁺ flux across the mitochondrial membrane are mediated through different transport systems. Ca²⁺ is sequestered by a Ca²⁺ uniporter (Montero et al., 2001) and as yet poorly characterized system, recently discovered, and called the rapid mode (RaM) (Buntinas et al., 2001). The Ca²⁺ uptake by the uniporter is slow, is coupled to the efflux of 2 H⁺ and depends mainly on the mitochondrial membrane potential and the [Ca²⁺]i in the vicinity of the organelle (the uniporter Kd for Ca²⁺ is ~ 1-5 µM) (Montero et al., 2000). RaM seems to be able to sequester large quantities of Ca²⁺, and is activated at low [Ca²⁺]i (Gunter et al., 2000). It would therefore likely intervene at the onset of high [Ca²⁺]i elevations in response to agonists (Buntinas et al., 2001). It’s however not known whether the RaM is a protein by itself, or whether it’s simply another conformation of the uniporter (Gunter et al., 2000). More studies are needed to better characterize the RaM phenomenon.

Measurements of intramitochondrial free Ca²⁺ ([Ca²⁺]mt) using directed aequorin showed intramitochondrial Ca²⁺ transients in response to cytosolic [Ca²⁺]i increases (Rizzuto et al., 1992). Further studies demonstrated a synchrony between [Ca²⁺]i and [Ca²⁺]mt oscillations (Hajnoczky et al., 1995). The functional significance of the Ca²⁺ uptake by mitochondria has long remained a much debated question. A substantial impulse was brought by the discovery of the activation by Ca²⁺ of key mitochondrial enzymes involved in the Krebs cycle: puruvate dehydrogenase, a-ketoglutarate-dehydrogenase and the NAD-isocitrate dehydrogenase (McCormack et al., 1990). The correlation between [Ca²⁺]mt changes and dehydrogenases activation was later demonstrated by simultaneously measuring the [Ca²⁺]mt changes and the production of NADH (Rizzuto et al., 1994).

The Ca²⁺ efflux from the mitochondria is mediated by a Na⁺/Ca²⁺ exchanger in cardiomyocytes and central nervous system, and via a H⁺/Ca²⁺ exchanger in hepatocytes (Montero et al., 2001). Another mechanism which would be implicated in Ca²⁺ efflux is the mitochondria permeability transition pore (PMT) (reviewed in (Crompton, 1999). This pore is about 2 nm in diameter (Duchen, 1999), is permeable to molecules < 1500 Da and is inhibited by cyclosporine A. Factors such as mitochondrial Ca²⁺, mitochondrial membrane depolarization, oxidized NAD and NADP, super-oxide anions (O^2-) all support pore opening, whereas opposite conditions stabilize the pore in a closed conformation. The opening of PMT triggers a rapid release of Ca²⁺ and mitochondrial metabolites, and the prolonged opening of the pore is associated with the release of apoptotic factors (such as the cytochrome c) (Montero et al., 2001). The PMT would also be involved in the development of a mitochondrial CICR. A rise of [Ca²⁺]mt associated with a small diminution of intramitochondrial pH would be enough to open the channel and release Ca²⁺ inducing a high [Ca²⁺]i microdomain which would in turn trigger a local CICR from the SER (Duchen, 1999). The modulation of the amplitude of [Ca²⁺]i changes by cyclosporine A, potent inhibitor of calcineurin, suggests that this mechanism could be involved in the physiological regulation of the Ca²⁺ signal, but this hypothesis awaits confirmation (Duchen, 1999).

Recent reports have demonstrated close contacts between mitochondria and ER (Rizzuto et al., 1998). This contact would permit to mitochondria to quickly sequester Ca²⁺ from high [Ca²⁺]i microdomains generated by Ca²⁺ release from ER. The mitochondrion would then have a dual role: 1) to modulate efficiently the mitochondrial metabolism in phase with cellular needs, 2) to prevent widespread non-specific activation of Ca²⁺-sensitive cellular processes.

B. Golgi apparatus

In the complex scenario of the regulation of [Ca²⁺]i, the role of the Golgi apparatus has just started to get attention. The free Golgi Ca²⁺ concentration ([Ca²⁺]Golgi) is estimated to be more than 0.1 mM (Pezzati et al., 1997; Pinton et al., 1998). The mechanisms by which a such high concentration is maintained have begun to be elucidated. Two Ca²⁺-ATPases were identified within the membrane of the organelle. The first is a pump similar to SERCA which is sensitive to thapsigargin (Pinton et al., 1998; Taylor et al., 1997). The second is insensitive to thapsigargin, is not a plasma membrane Ca²⁺-ATPase (PMCA- like), and its characteristics are reminiscent of the plasma membrane ATPase-related (PMR1), a P type Ca²⁺ ATPase from the Yeast Golgi apparatus which is essential for the normal secretory process in this unicellular organism (Pinton et al., 1998; Sorin et al., 1997).

The ability of agents such as the ionophore A23187 (Chandra et al., 1991) and histamine, an IP3 producing agent (Pinton et al., 1998), to release Ca²⁺ from the Golgi apparatus also supports a role of this organelle as a Ca²⁺ store, at least in certain cell types. The effect of IP3 suggests that the organelle expresses IP3R’s. Moreover, a type 2 RyR has been identified within the Golgi apparatus of neuronal cells (Cifuentes et al., 2001). All of these observations presuppose that the Golgi apparatus could be a central player in the regulation of [Ca²⁺]i changes, but this conclusion awaits further investigation.

C. The nucleus

The study of the nuclear Ca²⁺ ([Ca²⁺]nc) dynamics is arousing major interest mainly because of the involvement the Ca²⁺ in the regulation of many nuclear functions such as the cell cycle, the nucleo-cytosolic exchanges of diverse transcription factors, the modulation of gene expression, DNA synthesis and repair and of apoptosis (reviewed in (Gerasimenko & Gerasimenko, 2004)). By measuring simultaneously the [Ca²⁺]i and the Ca²⁺ concentraton within the nucleus ([Ca²⁺]NC), it has been demonstrated that [Ca²⁺] changes within these two compartments are synchronous, which suggests a passive Ca²⁺ diffusion from the cytosol into the nucleus via the nuclear pore complexes (NPC) (Nagai et al., 2001). This finding should however be confirmed because previous reports have suggested that this diffusion is not passive (Brini & Carafoli, 2000). On the other hand, different systems allowing the Ca²⁺ uptake and release were identified in the nuclear envelope (NE). The filling of the nucleoplasm with Ca²⁺ is mediated by a SERCA type Ca²⁺-ATPase (Nuclear Ca²⁺-ATPase or NCA), which is sensitive to thapsigargin and DBHQ (Lanini et al., 1992). This Ca²⁺ pump was located in the nuclear outer envelope and is stimulated by PKA (Rogue et al., 1998). It’s however not excluded that the NCA is also present in the inner envelope as the nuclear envelope is in continuity with the ER membrane an may than have the same properties (Santella & Carafoli, 1997).

For the nucleus to be a dynamic Ca²⁺ store, it should be able release Ca²⁺ taken up by the NCA. This is made possible by the presence of IP3 and cADPr receptors (Gerasimenko & Gerasimenko, 2004). The stimulation by IP3 induces a [Ca²⁺]NC rise, suggesting that the IP3R are located in the inner envelope (Humbert et al., 1996). This hypothesis is supported by the presence of the PLC, able to generate IP3, in the inner envelope (Brini & Carafoli, 2000). The CD38, producing cADPr, is also present in this envelope (Adebanjo et al., 1999). The activating mechanisms of CD38 and PLC are however not yet understood.

D.The secretory granules

The contribution of secretory granules in the regulation of [Ca²⁺]i was suggested because of the high intragranular Ca²⁺ concentration (~ 40 mM) (Winkler & Westhead, 1980), and due to the presence in these organelles of chromogranin A and B, two major proteins of this compartment able to bind Ca²⁺ with a high capacity and low affinity (Reiffen & Gratzl, 1986). It has also been proposed that IP3 releases Ca²⁺ from bovine adrenal medullary secretory vesicles (Yoo & Albanesi, 1990). These observations are however not universally accepted. Thus, the presence of IP3R in pancreatic beta cell secretory granules is controversial because of the non-specificity of the antibody used which presented a cross reaction with insulin (Blondel et al., 1994; Ravazzola et al., 1996). On the other hand, an IP3-ICR observed in pancreatic acinar cells was questioned because of possible contamination of the preparations by ER granules (Yule et al., 1997). The direct demonstration of the participation of the secretory vesicle in the control of [Ca²⁺]i was made by the simultaneous measurement of intravesicular Ca²⁺ ([Ca²⁺]v) and the [Ca²⁺]i proximal to the granules (Nguyen et al., 1998). This experiment showed that IP3 induces a rise of [Ca²⁺]i near the granules, whereas the Ca²⁺ uptake by the granules induces the opposite effect. The effects of IP3 are mediated by the IP3R located in the membrane of the vesicles (Quesada et al., 2001). On the other hand, a cADPr-ICR was recently reported in the insulin cell line MIN6, and the presence of RyR in vesicular membranes of these cells was detected (Mitchell et al., 2001). The presence of IP3 and ryanodine receptors in these vesicles suggests that a local IP3- or Ca²⁺-ICR from vesicles close to the plasmic membrane would contribute or constitute the triggering mechanism of granule exocytosis.

Whereas the Ca²⁺ release mechanisms from many organelles are increasingly well understood, those of Ca²⁺ sequestration in secretory granules are not. Ca²⁺ accumulation could be mediated by a Na⁺/Ca²⁺ (Troadec et al., 1998) or H⁺/Ca²⁺ (Goncalves et al., 1998) exchanger. However, the inhibition of these systems does not affect the Ca²⁺ uptake by secretory granules in MIN6 cells (Mitchell et al., 2001). The contribution of a P type Ca²⁺-ATPase distinct from SERCA, and recalling the PMR1 of Golgi apparatus was proposed (Mitchell et al., 2001), but it needs conformation in other cell types.

Conclusions

The recent advances in molecular biology and biophysical techniques have made it possible to identify and characterize intracellular Ca²⁺ channels within different cellular compartments and to monitor calcium changes directly in these stores. These advances have revealed not only a multitude of Ca²⁺ stores and Ca²⁺ messengers, but also many interactions between the different stores via CICR. This multiplicity of interactions and messengers provides cells with a flexible response by which to sense and distinguish a wide range of extracellular stimuli. More and more evidence suggests that these intracellular Ca²⁺ stores might be critical for some physiological processes, which in turn renders them potentially very interesting therapeutic targets. This exciting area of research will definitely attract more attention in the future; especially as clear differences in the mechanisms of Ca²⁺ signaling between different cell types begins to emerge.

REFERENCES

Aarhus R, Dickey DM, Graeff RM, Gee KR, Walseth TF, & Lee HC (1996). Activation and inactivation of Ca²⁺ release by NAADP+. J Biol Chem 271, 8513-8516.

Adebanjo OA, Anandatheerthavarada HK, Koval AP, Moonga BS, Biswas G, Sun L, Sodam BR, Bevis PJ, Huang CL, Epstein S, Lai FA, Avadhani NG, & Zaidi M (1999). A new function for CD38/ADP-ribosyl cyclase in nuclear Ca²⁺ homeostasis. Nat Cell Biol 1, 409-414.

Ahern GP, Junankar PR, & Dulhunty AF (1997). Subconductance states in single-channel activity of skeletal muscle ryanodine receptors after removal of FKBP12. Biophys J 72, 146-162.

Allen GJ, Muir SR, & Sanders D (1995). Release of Ca²⁺ from individual plant vacuoles by both InsP3 and cyclic ADP-ribose. Science 268, 735-737.

Andersen JP & Vilsen B (1998). Structure-function relationships of the calcium binding sites of the sarcoplasmic reticulum Ca²⁺-ATPase. Acta Physiol Scand Suppl 643, 45-54.

Arredouani A, Guiot Y, Jonas JC, Liu LH, Nenquin M, Pertusa JA, Rahier J, Rolland JF, Shull GE, Stevens M, Wuytack F, Henquin JC, & Gilon P (2002a). SERCA3 ablation does not impair insulin secretion but suggests distinct roles of different sarcoendoplasmic reticulum Ca²⁺ pumps for Ca²⁺ homeostasis in pancreatic beta-cells. Diabetes 51, 3245-3253.

Arredouani A, Henquin JC, & Gilon P (2002c). Contribution of the endoplasmic reticulum to the glucose-induced [Ca²⁺](c) response in mouse pancreatic islets. Am J Physiol Endocrinol Metab 282, E982-E991.

Arredouani A, Henquin JC, & Gilon P (2002b). Contribution of the endoplasmic reticulum to the glucose-induced [Ca²⁺](c) response in mouse pancreatic islets. Am J Physiol Endocrinol Metab 282, E982-E991.

Bandyopadhyay A, Shin DW, Ahn JO, & Kim DH (2000). Calcineurin regulates ryanodine receptor/Ca²⁺-release channels in rat heart. Biochem J 352 Pt 1, 61-70.

Berridge MJ (1997). Elementary and global aspects of calcium signalling. J Physiol 499 ( Pt 2), 291-306.

Berridge MJ (1993). Inositol trisphosphate and calcium signalling. Nature 361, 315-325.

Bezprozvanny I & Ehrlich BE (1994). Inositol (1,4,5)-trisphosphate (InsP3)-gated Ca channels from cerebellum: conduction properties for divalent cations and regulation by intraluminal calcium. J Gen Physiol 104, 821-856.

Bezprozvanny IB, Ondrias K, Kaftan E, Stoyanovsky DA, & Ehrlich BE (1993). Activation of the calcium release channel (ryanodine receptor) by heparin and other polyanions is calcium dependent. Mol Biol Cell 4, 347-352.

Blondel O, Moody MM, Depaoli AM, Sharp AH, Ross CA, Swift H, & Bell GI (1994). Localization of inositol trisphosphate receptor subtype 3 to insulin and somatostatin secretory granules and regulation of expression in islets and insulinoma cells. Proc Natl Acad Sci U S A 91, 7777-7781.

Brandl CJ, deLeon S, Martin DR, & MacLennan DH (1987). Adult forms of the Ca²⁺ATPase of sarcoplasmic reticulum. Expression in developing skeletal muscle. J Biol Chem 262, 3768-3774.

Brillantes AB, Ondrias K, Scott A, Kobrinsky E, Ondriasova E, Moschella MC, Jayaraman T, Landers M, Ehrlich BE, & Marks AR (1994). Stabilization of calcium release channel (ryanodine receptor) function by FK506-binding protein. Cell 77, 513-523.

Brini M & Carafoli E (2000). Calcium signalling: a historical account, recent developments and future perspectives. Cell Mol Life Sci 57, 354-370.

Brown GR, Sayers LG, Kirk CJ, Michell RH, & Michelangeli F (1992). The opening of the inositol 1,4,5-trisphosphate-sensitive Ca²⁺ channel in rat cerebellum is inhibited by caffeine. Biochem J 282 ( Pt 2), 309-312.

Buck E, Zimanyi I, Abramson JJ, & Pessah IN (1992). Ryanodine stabilizes multiple conformational states of the skeletal muscle calcium release channel. J Biol Chem 267, 23560-23567.

Bultynck G, De SP, Rossi D, Callewaert G, Missiaen L, Sorrentino V, De SH, & Parys JB (2001). Characterization and mapping of the 12 kDa FK506-binding protein (FKBP12)-binding site on different isoforms of the ryanodine receptor and of the inositol 1,4,5-trisphosphate receptor. Biochem J 354, 413-422.

Buntinas L, Gunter KK, Sparagna GC, & Gunter TE (2001). The rapid mode of calcium uptake into heart mitochondria (RaM): comparison to RaM in liver mitochondria. Biochim Biophys Acta 1504, 248-261.

Callaway C, Seryshev A, Wang JP, Slavik KJ, Needleman DH, Cantu C, III, Wu Y, Jayaraman T, Marks AR, & Hamilton SL (1994). Localization of the high and low affinity [3H]ryanodine binding sites on the skeletal muscle Ca²⁺ release channel. J Biol Chem 269, 15876-15884.

Cameron AM, Steiner JP, Roskams AJ, Ali SM, Ronnett GV, & Snyder SH (1995a). Calcineurin associated with the inositol 1,4,5-trisphosphate receptor-FKBP12 complex modulates Ca²⁺ flux. Cell 83, 463-472.

Cameron AM, Steiner JP, Sabatini DM, Kaplin AI, Walensky LD, & Snyder SH (1995b). Immunophilin FK506 binding protein associated with inositol 1,4,5-trisphosphate receptor modulates calcium flux. Proc Natl Acad Sci U S A 92, 1784-1788.

Carafoli E & Brini M (2000). Calcium pumps: structural basis for and mechanism of calcium transmembrane transport. Curr Opin Chem Biol 4, 152-161.

Carroll S, Skarmeta JG, Yu X, Collins KD, & Inesi G (1991). Interdependence of ryanodine binding, oligomeric receptor interactions, and Ca²⁺ release regulation in junctional sarcoplasmic reticulum. Arch Biochem Biophys 290, 239-247.

Chandra S, Kable EP, Morrison GH, & Webb WW (1991). Calcium sequestration in the Golgi apparatus of cultured mammalian cells revealed by laser scanning confocal microscopy and ion microscopy. J Cell Sci 100 ( Pt 4), 747-752.

Chini EN, Beers KW, & Dousa TP (1995). Nicotinate adenine dinucleotide phosphate (NAADP) triggers a specific calcium release system in sea urchin eggs. J Biol Chem 270, 3216-3223.

Churchill GC, O'Neill JS, Masgrau R, Patel S, Thomas JM, Genazzani AA, & Galione A (2003). Sperm deliver a new second messenger: NAADP. Curr Biol 13, 125-128.

Churchill GC, Okada Y, Thomas JM, Genazzani AA, Patel S, & Galione A (2002). NAADP mobilizes Ca²⁺ from reserve granules, lysosome-related organelles, in sea urchin eggs. Cell 111, 703-708.

Cifuentes F, Gonzalez CE, Fiordelisio T, Guerrero G, Lai FA, & Hernandez-Cruz A (2001). A ryanodine fluorescent derivative reveals the presence of high-affinity ryanodine binding sites in the Golgi complex of rat sympathetic neurons, with possible functional roles in intracellular Ca²⁺ signaling. Cell Signal 13, 353-362.

Crompton M (1999). The mitochondrial permeability transition pore and its role in cell death. Biochem J 341 ( Pt 2), 233-249.

Cui Y, Galione A, & Terrar DA (1999). Effects of photoreleased cADP-ribose on calcium transients and calcium sparks in myocytes isolated from guinea-pig and rat ventricle. Biochem J 342 ( Pt 2), 269-273.

DeLisle S, Blondel O, Longo FJ, Schnabel WE, Bell GI, & Welsh MJ (1996). Expression of inositol 1,4,5-trisphosphate receptors changes the Ca²⁺ signal of Xenopus oocytes. Am J Physiol 270, C1255-C1261.

Dode L, De GC, Mountian I, Attard M, Town MM, Casteels R, & Wuytack F (1998). Structure of the human sarco/endoplasmic reticulum Ca²⁺-ATPase 3 gene. Promoter analysis and alternative splicing of the SERCA3 pre-mRNA. J Biol Chem 273, 13982-13994.

Dode L, Wuytack F, Kools PF, Baba-Aissa F, Raeymaekers L, Brike F, van d, V, Casteels R, & Brik F (1996). cDNA cloning, expression and chromosomal localization of the human sarco/endoplasmic reticulum Ca²⁺-ATPase 3 gene. Biochem J 318 ( Pt 2), 689-699.

Duchen MR (1999). Contributions of mitochondria to animal physiology: from homeostatic sensor to calcium signalling and cell death. J Physiol 516 ( Pt 1), 1-17.

East JM (2000). Sarco(endo)plasmic reticulum calcium pumps: recent advances in our understanding of structure/function and biology (review). Mol Membr Biol 17, 189-200.

Ehrlich BE, Kaftan E, Bezprozvannaya S, & Bezprozvanny I (1994). The pharmacology of intracellular Ca²⁺-release channels. Trends Pharmacol Sci 15, 145-149.

Franzini-Armstrong C & Protasi F (1997). Ryanodine receptors of striated muscles: a complex channel capable of multiple interactions. Physiol Rev 77, 699-729.

Fujii J, Zarain-Herzberg A, Willard HF, Tada M, & MacLennan DH (1991). Structure of the rabbit phospholamban gene, cloning of the human cDNA, and assignment of the gene to human chromosome 6. J Biol Chem 266, 11669-11675.

Fukushi Y, Kato I, Takasawa S, Sasaki T, Ong BH, Sato M, Ohsaga A, Sato K, Shirato K, Okamoto H, & Maruyama Y (2001). Identification of cyclic ADP-ribose-dependent mechanisms in pancreatic muscarinic Ca²⁺ signaling using CD38 knockout mice. J Biol Chem 276, 649-655.

Furuichi T, Furutama D, Hakamata Y, Nakai J, Takeshima H, & Mikoshiba K (1994). Multiple types of ryanodine receptor/Ca²⁺ release channels are differentially expressed in rabbit brain. J Neurosci 14, 4794-4805.

Furuichi T, Shiota C, & Mikoshiba K (1990). Distribution of inositol 1,4,5-trisphosphate receptor mRNA in mouse tissues. FEBS Lett 267, 85-88.

Furuichi T, Yoshikawa S, Miyawaki A, Wada K, Maeda N, & Mikoshiba K (1989). Primary structure and functional expression of the inositol 1,4,5-trisphosphate-binding protein P400. Nature 342, 32-38.

Galione A (1993). Cyclic ADP-ribose: a new way to control calcium. Science 259, 325-326.

Galione A, Lee HC, & Busa WB (1991). Ca²⁺-induced Ca²⁺ release in sea urchin egg homogenates: modulation by cyclic ADP-ribose. Science 253, 1143-1146.

Genazzani AA, Empson RM, & Galione A (1996). Unique inactivation properties of NAADP-sensitive Ca²⁺ release. J Biol Chem 271, 11599-11602.

Genazzani AA & Galione A (1996). Nicotinic acid-adenine dinucleotide phosphate mobilizes Ca²⁺ from a thapsigargin-insensitive pool. Biochem J 315 ( Pt 3), 721-725.

Genazzani AA, Mezna M, Dickey DM, Michelangeli F, Walseth TF, & Galione A (1997). Pharmacological properties of the Ca²⁺-release mechanism sensitive to NAADP in the sea urchin egg. Br J Pharmacol 121, 1489-1495.

Gerasimenko O & Gerasimenko J (2004). New aspects of nuclear calcium signalling. J Cell Sci 117, 3087-3094.

Giannini G, Clementi E, Ceci R, Marziali G, & Sorrentino V (1992). Expression of a ryanodine receptor-Ca²⁺ channel that is regulated by TGF-beta. Science 257, 91-94.

Giannini G, Conti A, Mammarella S, Scrobogna M, & Sorrentino V (1995). The ryanodine receptor/calcium channel genes are widely and differentially expressed in murine brain and peripheral tissues. J Cell Biol 128, 893-904.

Gilon P, Arredouani A, Gailly P, Gromada J, & Henquin JC (1999). Uptake and release of Ca²⁺ by the endoplasmic reticulum contribute to the oscillations of the cytosolic Ca²⁺ concentration triggered by Ca²⁺ influx in the electrically excitable pancreatic B-cell. J Biol Chem 274, 20197-20205.

Goncalves PP, Meireles SM, Gravato C, & Vale MG (1998). Ca²⁺-H⁺ antiport activity in synaptic vesicles isolated from sheep brain cortex. Neurosci Lett 247, 87-90.

Grunwald R & Meissner G (1995). Lumenal sites and C terminus accessibility of the skeletal muscle calcium release channel (ryanodine receptor). J Biol Chem 270, 11338-11347.

Guihard G, Combettes L, & Capiod T (1996). 3':5'-cyclic guanosine monophosphate (cGMP) potentiates the inositol 1,4,5-trisphosphate-evoked Ca²⁺ release in guinea-pig hepatocytes. Biochem J 318 ( Pt 3), 849-855.

Gunter TE, Buntinas L, Sparagna G, Eliseev R, & Gunter K (2000). Mitochondrial calcium transport: mechanisms and functions. Cell Calcium 28, 285-296.

Guse AH (1999). Cyclic ADP-ribose: a novel Ca²⁺-mobilising second messenger. Cell Signal 11, 309-316.

Guse AH, Silva CP, Weber K, Ashamu GA, Potter BV, & Mayr GW (1996). Regulation of cADP-ribose-induced Ca²⁺ release by Mg²⁺ and inorganic phosphate. J Biol Chem 271, 23946-23953.

Hagar RE, Burgstahler AD, Nathanson MH, & Ehrlich BE (1998). Type III InsP3 receptor channel stays open in the presence of increased calcium. Nature 396, 81-84.

Hajnoczky G, Gao E, Nomura T, Hoek JB, & Thomas AP (1993). Multiple mechanisms by which protein kinase A potentiates inositol 1,4,5-trisphosphate-induced Ca²⁺ mobilization in permeabilized hepatocytes. Biochem J 293 ( Pt 2), 413-422.

Hajnoczky G, Robb-Gaspers LD, Seitz MB, & Thomas AP (1995). Decoding of cytosolic calcium oscillations in the mitochondria. Cell 82, 415-424.

Hakamata Y, Nishimura S, Nakai J, Nakashima Y, Kita T, & Imoto K (1994). Involvement of the brain type of ryanodine receptor in T-cell proliferation. FEBS Lett 352, 206-210.

Herbette L, Marquardt J, Scarpa A, & Blasie JK (1977). A direct analysis of lamellar x-ray diffraction from hydrated oriented multilayers of fully functional sarcoplasmic reticulum. Biophys J 20, 245-272.

Hidalgo C & Donoso P (1995). Luminal calcium regulation of calcium release from sarcoplasmic reticulum. Biosci Rep 15, 387-397.

Higashida H, Hashii M, Yokoyama S, Hoshi N, Chen XL, Egorova A, Noda M, & Zhang JS (2001). Cyclic ADP-ribose as a second messenger revisited from a new aspect of signal transduction from receptors to ADP-ribosyl cyclase. Pharmacol Ther 90, 283-296.

Hirota J, Michikawa T, Natsume T, Furuichi T, & Mikoshiba K (1999). Calmodulin inhibits inositol 1,4,5-trisphosphate-induced calcium release through the purified and reconstituted inositol 1,4,5-trisphosphate receptor type 1. FEBS Lett 456, 322-326.

Horne JH & Meyer T (1995). Luminal calcium regulates the inositol trisphosphate receptor of rat basophilic leukemia cells at a cytosolic site. Biochemistry 34, 12738-12746.

Humbert JP, Matter N, Artault JC, Koppler P, & Malviya AN (1996). Inositol 1,4,5-trisphosphate receptor is located to the inner nuclear membrane vindicating regulation of nuclear calcium signaling by inositol 1,4,5-trisphosphate. Discrete distribution of inositol phosphate receptors to inner and outer nuclear membranes. J Biol Chem 271, 478-485.

Inesi G & Asai H (1968). Trypsin digestion of fragmented sarcoplasmic reticulum. Arch Biochem Biophys 126, 469-477.

Inesi G & Sagara Y (1994). Specific inhibitors of intracellular Ca²⁺ transport ATPases. J Membr Biol 141, 1-6.

Jayaraman T, Ondrias K, Ondriasova E, & Marks AR (1996). Regulation of the inositol 1,4,5-trisphosphate receptor by tyrosine phosphorylation. Science 272, 1492-1494.

Jones LR, Zhang L, Sanborn K, Jorgensen AO, & Kelley J (1995). Purification, primary structure, and immunological characterization of the 26-kDa calsequestrin binding protein (junctin) from cardiac junctional sarcoplasmic reticulum. J Biol Chem 270, 30787-30796.

Joseph SK & Ryan SV (1993). Phosphorylation of the inositol trisphosphate receptor in isolated rat hepatocytes. J Biol Chem 268, 23059-23065.

Kawasaki T & Kasai M (1994). Regulation of calcium channel in sarcoplasmic reticulum by calsequestrin. Biochem Biophys Res Commun 199, 1120-1127.

Koizumi S, Lipp P, Berridge MJ, & Bootman MD (1999). Regulation of ryanodine receptor opening by lumenal Ca²⁺ underlies quantal Ca²⁺ release in PC12 cells. J Biol Chem 274, 33327-33333.

Komalavilas P & Lincoln TM (1994). Phosphorylation of the inositol 1,4,5-trisphosphate receptor by cyclic GMP-dependent protein kinase. J Biol Chem 269, 8701-8707.

Komalavilas P & Lincoln TM (1996). Phosphorylation of the inositol 1,4,5-trisphosphate receptor. Cyclic GMP-dependent protein kinase mediates cAMP and cGMP dependent phosphorylation in the intact rat aorta. J Biol Chem 271, 21933-21938.

Lamb GD & Stephenson DG (1996). Effects of FK506 and rapamycin on excitation-contraction coupling in skeletal muscle fibres of the rat. J Physiol 494 ( Pt 2), 569-576.

Lanini L, Bachs O, & Carafoli E (1992). The calcium pump of the liver nuclear membrane is identical to that of endoplasmic reticulum. J Biol Chem 267, 11548-11552.

Laver DR, Lenz GK, & Lamb GD (2001). Regulation of the calcium release channel from rabbit skeletal muscle by the nucleotides ATP, AMP, IMP and adenosine. J Physiol 537, 763-778.

Lee AG & East JM (2001). What the structure of a calcium pump tells us about its mechanism. Biochem J 356, 665-683.

Lee HC (2001). Physiological functions of cyclic ADP-ribose and NAADP as calcium messengers. Annu Rev Pharmacol Toxicol 41, 317-345.

Lee HC (1993). Potentiation of calcium- and caffeine-induced calcium release by cyclic ADP-ribose. J Biol Chem 268, 293-299.

Lee HC (2004). Multiplicity of Ca²⁺ messengers and Ca