Since its discovery in 1987 (Clapper et al., 1987) NAADP has been shown to mobilise Ca2+ in a wide variety of organisms including invertebrates, vertebrates and plants (Albrieux et al., 1998; Berg et al., 2000; Cancela et al., 1999; Chini et al., 1995a; Chini et al., 1995b; Mojzisova et al., 2001; Navazio et al., 2000; Santella et al., 2000; Zhang et al., 2006a) and has been described as the most potent Ca2+- mobilising molecule identified to date (Galione et al., 2009). Given the broad range of species that exhibit NAADP-dependent Ca2+ signalling, it is clear that this Ca2+-
mobilising second messenger plays an important role in cellular Ca2+ homeostasis.
Yet despite this, the molecular identity of the NAADP receptor has, until now remained elusive.
Recently the transient receptor potential channel mucolipin 1 (TRPML1) which is a lysosome-bound channel, was proposed to function as a NAADP receptor (Zhang
et al., 2008). However, in normal rat kidney cells (NRK) over-expression of TRPML1 failed to increase NAADP binding (Pryor et al., 2006). Other groups have also suggested that the primary function of TRPML1 is to act as either a Fe2+ (Dong
et al., 2008) or a H+ release channel (Soyombo et al., 2006). Further evidence contrary to the view that TRPML1 is an NAADP receptor comes from the finding that the antibody used by Zhang et al., to label / block TRPML1 would appear to label elements within TRPML1 null cells suggesting that the antibody is non-selective (Prof. Antony Galione, personal correspondence). NAADP has also been proposed to act directly on ryanodine receptors (RyRs) to release Ca2+ (Dammermann et al., 2005; Hohenegger et al., 2002). However, RyR antagonists such as ryanodine, procaine and ruthenium red failed to abolish NAADP-dependent Ca2+ signalling in both mammalian and non-mammalian cells (Boittin et al., 2002; Chini et al., 1995a; Lee et al., 1995). Moreover, RyR1 and RyR2 reconstituted into lipid bilayers did not appear to be activated by NAADP (Copello et al., 2001). Therefore, there is significant evidence to suggest that NAADP does not directly activate either TRPML1 or RyRs but acts on a discrete receptor. In 2000 the two-pore channel subtype 1 (TPC1) was first cloned and identified as a novel two-domain channel with an unknown function (Ishibashi et al., 2000). In this thesis I have provided evidence in support of the view that TPCs represent a family of NAADP receptors
6.1 Two-pore channels in relation to other cation channels
Two-pore channels (TPCs) are novel members of the voltage-gated cation channel superfamily. Examination of their primary sequence revealed a predicted structure consisting of 12 transmembrane (TM) regions that are divided into 2 homologous domains of 6 TM regions each (S1-S6). Importantly, and characteristic of members of this superfamily, TPCs contain a re-entrant P-loop between S5 and S6 of each domain. This basic 6-TM unit is representative of all members within the voltage-gated ion channel superfamily, from the single domain channels such as transient receptor potential channels (TRPs) and the cation channel of spermatozoa (CatSpers), to the 4 repeat domain of the pore-forming subunit of voltage-gated Ca2+ channels (Cav) (Yu et al., 2005). Of the voltage-gated cation channel superfamily
members, TPCs share the greatest homology with Cav and voltage-gated Na+
channels. TPCs are also closely related to CatSpers and some TRPs albeit to a lesser extent than either Nav or Cav. This is particularly intriguing given that single domain
channels such as TRPs form tetramers whereas only 1 of the 4-repeat domain containing α-subunit of Cav is required for the formation of a functional pore. This
has led others to propose that TPCs channels function as dimers with the 4 P-loops (2 from each monomer) coming together in order to form the central pore (Zong et al., 2009). Given the similarity to both TRPs and Cav and the conserved 6-TM
organisation found in each protein, TPCs are purported to be evolutionary intermediates resulting from 2 rounds of duplication from a common 6 TM, single domain protein to the 4-fold symmetry of 24 TM channels such as Cav and Nav
(Anderson et al., 2001; Ishibashi et al., 2000).
Further support for the view that TPCs may be intermediary steps is taken from the finding that the 3 animal TPC subtypes (TPC1, TPC2 and TPC3) and plant TPC1 are expressed in a wide variety of species in both plant and animal kingdoms. This suggests that they represent an ancient gene family. However, it is important to note that TPC3 appears to be absent in primates and rodents such as mice and rats. Furthermore, species such as honeybees and silkworms contain only TPC1 encoding genes (TPCN1) whereas C. elegans and D.melanogaster do not appear to contain any sequence encoding TPCs. Conversely, sequences for TPCN1 and TPCN3 but not
TPCs suggests that although the genes encoding TPCs appear to be from an ancient gene family, the absence of 1 or more TPC subtypes (including a complete absence of TPCs) from a number of species would indicate that TPCs represent an important ion channel family that may not be essential for cell survival.
6.2 NAADP mobilises Ca
2+via two-pore channels that are targeted
to lysosome-related organelles
Examination of the cellular distribution of TPCs revealed that TPC2 is targeted to the membranes of lysosomes but not those of early or late endosomes, Golgi apparatus, mitochondria or the ER. Conversely, TPC1 and TPC3 showed little or no co-localisation with either lysosomal membrane markers and TPC2, but appeared to be primarily targeted to the membranes of endosomes in addition to other unidentified vesicular organelles. Clearly, all mammalian TPCs are expressed on the membranes of endolysosomes, which is consistent with the targeting of plant TPC1 to the membranes of vacuoles (Peiter et al., 2005). Given that NAADP releases Ca2+ from acidic lysosome-related organelles (Churchill et al., 2002; Kinnear et al., 2004; Yamasaki et al., 2004; Zhang et al., 2004) and that NAADP-mediated Ca2+ release is inhibited by both L-type Ca2+ channel agonists and antagonists (phenylalkylamines and dihydropyridines) (Genazzani et al., 1996a; Genazzani et al., 1997) the structure and localisation of TPCs and TPC2 in particular, naturally led me to investigate whether or not TPCs represent NAADP receptors.
Consistent with a role for TPCs in NAADP-dependent Ca2+ signalling, HEK293 cells that stably over-expressed TPC2 were used to provide, via cell lysates, membranes enriched with TPC2 upon which NAADP was found to bind at two sites, a high and low affinity site with Kd values of ~5 nM and ~7 µM, respectively. This is
consistent with the view that NAADP receptors exhibit high and low affinity binding sites that are proposed to underpin the bell-shaped concentration-response curve to NAADP. The high affinity site has been suggested to mediate activation, while the low affinity binding site may confer homologous desensitisation of the receptor (Berg
In HEK293 cells stably over-expressing TPC2 NAADP elicited a biphasic global Ca2+ transient which was abolished by depletion of lysosomal Ca2+ stores with the vacuolar H+-ATPase inhibitor bafilomycin A1 or when cells were preincubated with shRNA against TPC2. Moreover, the NAADP antagonist Ned-19, which was shown to fluorescently label TPC2 on the lysosomal membrane, selectively blocked NAADP-mediated Ca2+ signalling in HEK293 cells. Together these findings suggest that TPC2 underpins NAADP-mediated Ca2+ release from lysosomes. Consistent with the two site binding model, I further demonstrated that NAADP-dependent Ca2+ signalling via TPC2 exhibited an approximate activation threshold of 10 nM NAADP whereas no response was observed when the concentration of NAADP was increased to 1 mM suggesting significant desensitisation of the channel. This characteristic bell-shaped concentration-response curve has been observed in other mammalian preparations (Berg et al., 2000; Cancela et al., 1999; Masgrau et al., 2003) and homologous desensitisation is also observed in sea urchin eggs (Aarhus et al., 1996; Dickinson et al., 2003; Genazzani et al., 1996a). However, in sea urchin egg a distinction with respect to the desensitisation mechanism is apparent where low sub- threshold concentrations of NAADP desensitise / inactivate the receptor to subsequent activation by maximal NAADP concentrations (Aarhus et al., 1996; Dickinson et al., 2003; Genazzani et al., 1996a). Given that sea urchin eggs express all 3 TPC subtypes whereas only TPC1 and TPC2 are present in human cells, the difference in the mechanisms of homologous desensitisation / inactivation between the cell types may be due to the variation in the individual properties of each TPC subtype which, for TPC1 and TPC3 remain to be determined.
As mentioned above, TPC1 appears to be primarily targeted to the membranes of endosomes which likely represent a smaller Ca2+ store. This is consistent with the spatially restricted Ca2+ release in response to NAADP in HEK293 cells stably over- expressing TPC1. It would appear, therefore that NAADP mobilises Ca2+ via TPC2 and TPC1 from lysosomes and most likely endosomes respectively. It is therefore probable that TPCs may represent a family of NAADP receptors each with a unique cellular distribution conferred by the organelles to which they are targeted, and this raises the possibility that each TPC may confer a degree of versatility to NAADP- mediated Ca2+ signalling. Yet more complexity is possible when one considers the fact that both lysosomes and endosomes are mobile within the cytoplasm (Herman et al., 1984; Prekeris et al., 1999; Taunton, 2001) and may utilise actin filaments and