|Home | About | Journals | Submit | Contact Us | Français|
In this review we describe the background and implications of our recent discovery that Two pore channels (TPCs) comprise a novel class of calcium release channels gated by the intracellular messenger nicotinic acid adenine dinucleotide phosphate (NAADP). Their localization to the endolysosomal system highlights a new function for these organelles as targets for NAADP-mediated Ca2+ mobilization. In addition, we describe how TPCs may also trigger further Ca2+ release by coupling to the endoplasmic reticular stores through activation of IP3 receptors and ryanodine receptors.
NAADP is the most potent Ca2+ releasing molecule described, discharging Ca2+ from Ca2+ storage organelles to elicit cytoplasmic Ca2+ signals at picomolar to low nanomalor concentrations in sea urchin egg preparations (34). In some mammalian cells, Ca2+ responses are evoked by as low as 10 nM NAADP (4). Measurement of cell and tissue levels have shown that NAADP is synthesised in response to activation of various cell surface receptors and other stimuli, confirming NAADP's role as an intracellular messenger (18, 53). In contrast to the other established Ca2+ mobilizing messengers, inositol 1,4,5-trisphosphate (IP3) and cyclic ADP ribose (cADPR), that release Ca2+ from the endoplasmic reticulum (ER) by activating IP3 receptors (IP3Rs) and ryanodine receptors (RyRs), respectively, NAADP is enigmatic since in many cells it appears to act via a distinct Ca2+ release mechanism, which resides not on the ER but on acidic organelles (14).
NAADP was discovered as a trace contaminant of commercially available NADP (35). This may not be too surprising since the structure of NAADP is very close to that of NADP, differing only by the substitution of nicotinic acid as the base instead of nicotinamide. This small change, however, has enormous biological effects since whilst NADP is inactive, NAADP is very potent at generating Ca2+ signals. This discovery was made by Lee and colleagues whilst studying calcium release mechanisms in homogenates prepared from sea urchin eggs (15). This preparation has done much to advance our understanding of novel Ca2+ signalling mechanisms including the first demonstration that cADPR mobilizes Ca2+ through the activation of RyRs (20, 44), and much of the ground work in studying the mechanisms of NAADP action has been gleaned from this system before being widened to mammalian preparations.
The initial report on the Ca2+ mobilizing action of the then unidentified NADP-related molecule implicated a different organellar target to the ER since it released Ca2+ from a partially separate subcellular fraction of sea urchin egg homogenate (15). Later, NAADP evoked Ca2+ release was found to persist in homogenates and intact eggs even when the ER Ca2+ pool had been abrogated by the SERCA pump inhibitor, thapsigargin. Another feature of sea urchin eggs is that when centrifuged they stratify, that is they change from spherical to a more oblong shape with different organelles migrating to different poles. Under these conditions, IP3 and cADPR mobilize Ca2+ from the nuclear pole where the ER is, but NAADP releases Ca2+ predominantly from the opposite pole (36). Taken together, these studies favoured a target for NAADP in organelles distinct from the ER, the traditional site of action of Ca2+ mobilizing messengers.
A major new lead in the physiology of NAADP came from the finding that in sea urchin eggs the major target for NAADP was acidic organelles such as reserve granules, the urchin egg's equivalent of lysosomes (14). Stratification studies revealed that acidic stores stained with lysotracker localised to the non-ER pole exactly where NAADP was found to act as described above. Further fractionation of sea urchin egg homogenates showed that binding sites for [32P]NAADP co-localized with lysosomal markers. The isolated NAADP-sensitive stores showed ATP-dependent Ca2+ uptake that was insensitive to thapsigargin, but abrogated by protonophores and bafilomycin A1, a vacuolar proton pump inhibitor, suggesting a role for a proton-dependent Ca2+ uptake mechanism. The lysosomotropic agent glycyl-L-phenylalanine 2-naphthylamide (GPN) selectively abolished NAADP-induced Ca2+ release in both homogenates and intact eggs whilst leaving IP3- and cADPR-sensitive mechanisms unaffected.
NAADP-evoked Ca2+ release has now been shown to be widespread in mammalian cells, and even in plants (21). In most cases where investigated, including neurones, cardiac myocytes, smooth muscle, exocrine cells, and platelets, NAADP also appears to release Ca2+ from acidic stores rather than from the ER. The study of NAADP thus has highlighted acidic stores as important Ca2+ storage organelles that can be discharged by their own intracellular messenger. The role of Ca2+ storage in the endolysosomal system may be pivotal to endomembrane fusion and trafficking and for other physiological functions of these organelles (47, 48). Recently, it has been proposed that reduced lysosomal Ca2+ storage is a primary defect in Niemann-Pick type C1 disease, a lysosomal storage disease, and this is associated with abolition of NAADP-induced Ca2+ release (37).
Given the smaller size, and perhaps the lower Ca2+ content, Ca2+ release from acidic stores is often smaller than that evoked from the larger reticular ER network. However, even the first report of NAADP-evoked Ca2+ release in a mammalian cell type proposed that NAADP-induced Ca2+ signals also trigger a subsequent, larger Ca2+ release from the ER through the Ca2+-induced Ca2+ release (CICR) mechanisms of IP Rs and RyRs (9). That Ca2+ release in response to NAADP in turn triggers CICR through both IP3Rs and RyRs has been extensively studied in sea urchin eggs (11, 12) and demonstrated in a wide range of mammalian systems (2, 4, 7, 9, 22, 39). Indeed, tight junctions, likely less than 100 nm across, between the RyR-containing sarcoplasmic reticulum and a subpopulation of lysosomes have been observed in vascular smooth muscle and these may constitute a “trigger zone” for the initiation of propagating Ca2+ signals by NAADP (32, 33). Thus, bafilomycin completely abolishes NAADP responses whilst thapsigargin and IP3R and RyR blockage may only reduce the amplitude of the responses (7, 22, 32, 39, 41, 52). NAADP thus not only has a role in regulating endolysosomal luminal Ca2+ content and cytosolic Ca2+ signals arising from these organelles, but also a second role in coordinating ER-mediated Ca2+ release. Such interactions have often confounded experiments designed to elucidate the identity of the NAADP receptor, with some proposing that RyRs are a direct target (16). As we have stated previously, the identification of NAADP binding proteins are required to resolve such controversies (21).
The initial descriptions of NAADP-evoked Ca2+ release from sea urchin egg homogenates showed that NAADP likely activated a novel Ca2+ release mechanism distinct from IP3Rs and RyRs. This was based on the individual ability of IP3, cADPR and NAADP to induce homologous desensitization of their target mechanisms without affecting the responsiveness of the other two and regardless of the order of the additions of each messenger (35). Log10 concentration-response curves for NAADP-induced Ca2+ release are quite unlike the familiar sigmoid curves seen for either IP3 or cADPR. In sea urchin eggs, subthreshold concentrations of NAADP for Ca2+ release can nevertheless fully desensitize NAADP receptors (24), whilst in mammalian cells the concentration-response curves are bell shaped, with micromolar concentrations of NAADP inducing desensitization in the absence of apparent Ca2+ release. This latter phenomenon has been exploited to identify receptor mechanisms coupled to NAADP-mediated Ca2+ signalling (9, 18).
In another approach, pharmacological antagonism of IP3Rs by heparin or RyRs by ryanodine, or the selective cADPR antagonist analogue 8-NH2-cADPR, did not affect the ability of NAADP to evoke a Ca2+ release response (19). Whilst IP3Rs and RyRs are also Ca2+-activated channels and can propagate Ca2+ release as regenerative waves through CICR, this property is apparently not shared by the intrinsic NAADP-sensitive Ca2+ release mechanism either in egg homogenates (10, 23) or in intact cells (11). As outlined above, rather NAADP evokes local Ca2+ signals that may only be propagated and globalized by subsequent recruitment of IP3Rs and RyRs. In addition, NAADP-induced Ca2+ release is markedly less pH sensitive than IP3Rs and RyRs (11) allowing the messenger to efficiently act at acidic stores, where large pH variation is inevitable because protons are co-released with Ca2+ (43).
Some pharmacological agents have been reported to selectively block NAADP-induced Ca2+ release without affecting either IP3Rs or RyRs. These include voltage-gated calcium channel (VGCC) blockers such as dihydropyridines (25) and the store-operated channel and TRPC channel blocker SKF96365 (42). Recently, a potent and selective membrane permeant NAADP antagonist, NED19, has been described, which will be invaluable in helping dissect the contribution of NAADP during Ca2+ signalling (45). Taken together, these data support the hypothesis that NAADP activates a unique Ca2+ release mechanism involving channels that are perhaps more similar to VGCC or TRP families than the two established ER Ca2+ release channel families, the IP3Rs and RyRs.
The search for the molecular identity of NAADP receptors had been a bumpy ride to say the least. Initially, multi-step fractionations of sea urchin egg homogenates were attempted and a protein target that bound to [32P]NAADP, but was substantially smaller than IP3Rs and RyRs, was demonstrated (3). However, the NAADP-binding proteins were not identified at the molecular level. Since the demonstration that NAADP elicited Ca2+ release from lysosome-related acidic stores, lysosomal localized cation channels have been considered as candidates of NAADP receptors. For example, mucolipin 1, or TRPML1, was tested and failed to show [32P]NAADP binding or an NAADP-evoked Ca2+ response in heterologous systems (49). Despite these findings, another group showed that a TRPML1 antibody blocked NAADP-evoked currents in planar lipid bilayers incorporated with lysosomal membranes prepared from rat liver and coronary arterial myocytes (55, 56) and siRNA silencing of TRPML1 expression attenuated lysosome-associated Ca2+ release in the myocytes (55). However, other experimental data suggest that TRPML1 functions either as a proton channel or an iron release channel at the lysosomes (17, 50).
In a recent study, we demonstrated that two-pore channels (TPCs or TPCNs for gene names) function as NAADP receptors (8). Based on primary sequences, TPCs are novel members of the superfamily of voltage-gated cation channels. Hydropathy analyses revealed that the full-length protein contains 12 putative transmembrane (TM) segments that are segregated into two homologous domains of 6 TM segments each with an intervening hydrophobic loop between the fifth and the sixth segments (Fig. 1). This 6-TM organization resembles that of all 6-TM voltage-gated cation channel family members including the four repeated domains found in the pore-forming subunits of voltage-gated Ca2+ and Na+ channels (54). The sequence homology of the TPC TM domains is also closer to voltage-gated Ca2+ and Na+ channels than to K+ channels. In fact among the single domain 6-TM channels, Catspers and some TRP channels are more closely related to TPCs than Kv channels are, but nevertheless their homologies to TPCs are slightly lower than that of Cav and Nav channels. This has led to the hypothesis that TPCs may represent evolutional intermediates of the two round duplications of single 6-TM domain channels that gave rise to the four 6-TM domain Cav and Nav channels (1). The fact that related TPCs are widely expressed in plants suggests that this is an ancient ion channel family that had been in existence before the divergence of plants and animals. On the other hand, TPCN genes are completely missing in some model animal species, such as flies and nematodes, indicating that these channels may not be essential for survival.
With the report of the first mammalian TPC (TPC1) sequence some years ago (28), so was the name given but this should not be confused with two-pore potassium channels, which have two repeats of a 2-TM architecture. However, functional data had been completely lacking, at least in animal cells, for this unique group of ion channels that apparently have an important position in the evolution of voltage-gated cation channels. By contrast, active research on plant cells revealed that plant TPC1, which is not equivalent to animal TPC1 but rather an equally distant homologue of all animal TPCs, forms a Ca2+ release channel in intracellular vacuoles (46). Indeed, even before the work on the plant TPC1 was reported, we had noticed that human TPC1 and TPC2 expressed in HEK293 cells were not expressed on the plasma membrane but rather on membranes of endosomes and lysosomes. Given that NAADP acts at acidic endolysosomal stores (14) and its action is inhibited by Ca2+ channel antagonists (25), the specific subcellular locations and the homology to Cav and TRP channels prompted us to examine whether TPCs form NAADP receptors.
Because of its prominent localization on lysosomal membranes, we focused initially on TPC2 and obtained several lines of evidence that TPC2 is involved in NAADP binding and NAADP-mediated Ca2+ release from acidic stores (8). First, NAADP binds to membranes from cells over-expressing TPC2 with a high affinity of 5-10 nM. Second, NAADP, at concentrations as low as 10 nM, elicited intracellular Ca2+ concentration increases in HEK293 cells that overexpress TPC2. Interestingly, the rise in intracellular Ca2+ exhibited two phases, an initial pacemaker phase with low to moderate Ca2+ transients akin to a “Ca2+ burst” (4), followed by a secondary rapid and much larger Ca2+ increase. The initial phase represents lysosomal Ca2+ release evoked by NAADP, as it was unaffected by blocking ER Ca2+ release with thapsigargin, heparin, or ryanodine, or all of these together, but completely abolished after treating cells with bafilomycin. The secondary phase, however, was eliminated by thapsigargin or heparin, or both of these together with ryanodine, and therefore represents globalization of the Ca2+ signal through CICR, resulting, originally, from the NAADP-induced Ca2+ release from acidic stores. This biphasic phenomenon and the dependence of the secondary phase on the ER stores are entirely consistent with NAADP-elicited Ca2+ responses in native cells as described above. Interestingly, the wild type HEK293 cells lacked the biphasic response and exhibited only small and short-lived Ca2+ transients upon stimulation by higher concentrations of NAADP, suggesting that the relatively low endogenous level of TPC expression in this cell type does not favour globalization of the NAADP-evoked Ca2+ signal and / or that lysosomes form “loose junctions” with the ER in HEK293 cells and thereby fail to adequately mediate, via endogenous TPC2, sufficient Ca2+ release to breach the threshold for CICR via IP3Rs. Third, shRNA specifically targeted against TPC2 abolished NAADP-evoked Ca2+ signals in HEK293 cells over-expressing TPC2 and the native NAADP response in a human hepatoblastoma cell line. Finally, NAADP-evoked Ca2+ signals activated non-selective cation currents in pancreatic b cells prepared from wide type mice, but no such current was detected in cells prepared from Tpc2 knockout mice. Taken together, these data strongly support a role for TPC2 in NAADP-evoked Ca2+ release in mammalian cells.
We further noticed that although overexpression of TPC1 did not result in globalization of the Ca2+ signal in response to 10 nM NAADP, the localized Ca2+ transients did become more pronounced at this low ligand concentration than in wild type cells. This, perhaps, reflects the more discrete intracellular distribution of TPC1 than TPC2, the lower Ca2+ content in the TPC1-expressing endosomes than lysosomes and or the fact that endosomal TPC1 does not form true junctions with the ER in HEK293 cells and therefore fails to elicit CICR via IP3Rs. Nevertheless this finding does suggest that TPC1 may also act as an NAADP receptor. Therefore, the TPCs represent a family of endolysosomal cation channels targeted by NAADP for regulated Ca2+ release from acidic stores. Ca2+ release from this dynamic and mobile source contributes to the diversity and versatility of Ca2+ signalling not only because of its unique origin, but also because of its ability to couple to the ER Ca2+ release under certain but not all conditions (Fig. 2).
There is still much to learn about TPC-mediated Ca2+ signalling. The NAADP-induced Ca2+ responses in sea urchin and starfish eggs are pivotal events at fertilization (13, 42). In mammalian systems, this ligand appears to play key roles in digestive enzyme and fluid secretion of pancreatic acinar cells (52), glucose-induced insulin secretion in β cells (40), smooth muscle contraction in the vasculature (4, 32), T lymphocyte activation (2), neurotransmitter release and neurite outgrowth (5, 6), and platelet activation (38). In some cases NAADP evokes specific Ca2+ responses distinct from either IP3 or cADPR such as the cortical flash at marine invertebrate egg fertilization (13,42), and differentiation of PC12 cells (5). In addition, NAADP modulates luminal pH of its target organelles suggesting another distinct signalling event evoked by this messenger (43). Notwithstanding, the study of NAADP signalling had been greatly hindered by the lack of knowledge about its molecular target(s) and lack of specific blockers. Now two recent reports have revealed not only a potent and specific NAADP antagonist (45), but also the molecular identity of the NAADP receptor (8). It is expected that these findings will greatly accelerate functional characterization of NAADP-induced Ca2+ signalling in many different cell types and biological systems (Fig. 2).
Human genetic studies have revealed some physiological and pathophysiological functions that may be attributed to TPC2. The polymorphic variants of the TPCN2 gene at two locations are associated with hair colour in Northern Europeans (51), indicative of an important role for this channel in pigmentation. This would be consistent with the lysosomal function of TPC2 in melanosomes, lysosomal-related organelles of the melanocytes, which produce and release melanin to keratinocytes for pigmentation. The TPCN2 locus is also within the chromosomal region that is linked to autosomal recessive nonsyndromic hearing impairment in at least five families (29, 31), and for one of them, the TPCN2 gene has been sequenced to show five exonic and four intronic variants (29). Given that progressive hearing loss has been linked to several lysosomal storage diseases (30), it is possible that dysfunction of the lysosomal Ca2+ release channel, TPC2, contributes to this disease. Furthermore, TPCN2 is one of three genes within the 11q13 amplicon core found in many human oral cancers that are frequently overexpressed in nonamplified primary tumors, indicating that it may contribute to growth or metastatic potential of cancer cells (27). Recent studies have linked alteration of Ca2+ signalling in apoptosis and autophagy of cancer cells (26). Autophagy is a physiological process invoked by cellular metabolic stress that is known to involve lysosome-mediated proteolysis. However, there has been very limited study of the role of lysosomal-mediated Ca2+ signalling in regulation of autophagy and apoptosis. Understanding the functional relationship between NAADP and TPC2-mediated Ca2+ release could provide an attractive therapeutic strategy for modulating the synergistic relationship between apoptosis and autophagy in cancer cells.
The discovery of TPCs as NAADP-gated Ca2+ release channels in the endolysosomal membranes opens many new directions for investigating the role of Ca2+ signalling in human health and diseases. Continued investigation in this area should offer great insights into the roles of the local NAADP signalling through endolysosomes and its coupling to global Ca2+ events through communication with the ER Ca2+ store, as well as to other systems, such as plasma membrane channels and Ca2+-dependent enzymes (Fig. 2). Therefore, this diverse and versatile signalling capability likely contributes to many physiological processes. Moreover, although our current study demonstrates unequivocally that TPCs function as the NAADP-regulated Ca2+ release channels from acidic organelles, we cannot rule out the possibility that other proteins tightly associate with these channels and modulate their activities. It may also be too early to rule out the direct involvement of other candidate channels, e.g. RyRs and TRPML1, in NAADP-evoked Ca2+ mobilization because this will require additional experimentation under specific conditions in appropriate cell types.