Diversity of two-pore channels and the accessory NAADP receptors in intracellular Ca²⁺ signaling

Abstract

Intracellular Ca²⁺ signaling via changes or oscillation in cytosolic Ca²⁺ concentration controls almost every aspect of cellular function and physiological processes, such as gene transcription, cell motility and proliferation, muscle contraction, and learning and memory. Two-pore channels (TPCs) are a class of eukaryotic cation channels involved in intracellular Ca²⁺ signaling, likely present in a multitude of organisms from unicellular organisms to mammals. Accumulated evidence indicates that TPCs play a critical role in Ca²⁺ mobilization from intracellular stores mediated by the second messenger molecule, nicotinic acid adenine dinucleotide phosphate (NAADP). In recent years, significant progress has been made regarding our understanding of the structures and function of TPCs, including Cryo–EM structure determination of mammalian TPCs and characterization of a plastid TPC in a single–celled parasite.. The recent identification of Lsm12 and JPT2 as NAADP-binding proteins provides a new molecular basis for understanding NAADP-evoked Ca²⁺ signaling. In this review, we summarize basic structural and functional aspects of TPCs and highlight the most recent studies on the newly discovered TPC in a parasitic protozoan and the NAADP-binding proteins LSM12 and JPT2 as new key players in NAADP signaling.

1. Introduction

Ca²⁺ mobilization from intracellular stores mediated by second messenger molecules is an important source of cytosolic Ca²⁺. There are three known Ca²⁺-mobilizing second messengers in mammalian cells: nicotinic acid adenine dinucleotide phosphate (NAADP), inositol 1,4,5–trisphosphate (IP₃), and cyclic ADP–ribose (cADPR). Among them, NAADP, which differs from NADP by an OH group replacing the –NH₂ group, is the most potent (i.e., effective in the low nanomolar range) [1–3]. Distinct from IP₃ and cADPR, which induce the release of Ca²⁺ from the endoplasmic reticulum (ER) by activating the IP₃ and ryanodine receptors, NAADP uniquely mobilizes Ca²⁺ from acidic organelles (e.g., endosomes and lysosomes, collectively called endolysosomes), as first observed in sea urchin egg homogenates [4,5] and then in numerous mammalian cell types [6,7]. Two-pore channels (TPCs) are considered as the key endolysosomal cation channels responsive to NAADP stimulation for intracellular Ca²⁺ release. TPCs are mainly found in endolysosomes in animal cells and also in plant vacuoles. The ion homeostasis carried out by endolysosomal channels, including TPCs in particular, regulates the membrane dynamics (fusion/fission), the acid nature, and enzymatic activity of the endolysosomal organelles [7–9]. The endolysosomal system plays a vital role in cellular functions from endocytosis and secretory events to degradation of proteins and organelles. NAADP signaling is involved in a broad range of cellular and physiologic processes such as neurotransmitter release and neurosecretion, membrane excitability, autophagy, exocytosis, fertilization, contraction of cardiac and smooth muscles, cell differentiation (neurons and skeletal myocytes), insulin secretion, and glucose uptake [7,10,11]. Defects in NAADP signaling have been implicated in many diseases, including lysosomal storage diseases, diabetes, autism, and cardiovascular, blood, and muscle diseases [6,7].

2. Diversity of TPCs

TPCs are ligand- and/or voltage-gated cation channels. Deuterostome animals such as vertebrates, sea stars, and crinoids have three TPC paralogues (TPC1, TPC2, and TPC3, encoded by TPCN1, TPCN2, and TPCN3, respectively), but the gene encoding TPC3 has undergone lineage-specific loss, e.g., in primates and rodents (Fig. 1A, Table S1) [12]. Humans and mice have two functional TPC isoforms: TPC1, which is broadly expressed in different stages of endosomes and lysosomes, and TPC2, which is expressed mainly in late-stage endosomes and lysosomes [13,14]. Activation of TPC1 requires both the presence of endolysosome-enriched lipid phosphatidylinositol 3,5–bisphosphate (PI (3,5)P₂) and the occurrence of membrane depolarization, whereas TPC2 is insensitive to voltage and can be potently activated by PI(3,5)P₂ alone [9,15–17]. They are both Na⁺-selective channels when activated by PI (3,5)P₂ [9,15–17], inhibited by ATP via mTORC1 [18], and slightly blockaded by cytoplasmic and luminal Mg²⁺ [16]. TPC1 is potentiated by cytoplasmic Ca²⁺ but suppressed by luminal Ca²⁺ [19], whereas TPC2 is not sensitive to Ca²⁺ [16]. The endolysosomal TPCs regulate endomembrane dynamics [8,9]. With mutant mouse models, TPCs were reported to be involved in muscle contraction, hormone secretion, fertilization, cell differentiation, and pathological disorders such as Parkinson disease, Ebola virus infection, and fatty liver disease [21,22]. TPC2 is also found in the melanosome, an intracellular organelle responsible for the synthesis of the pigment melanin in mammalian pigment cells [23,24]. TPC2 is implicated in regulation of human hair color [25] and pigmentation in Xenopus oocytes [26]. TPC2 activity in melanosomes regulates the organelle’s size, pH, membrane potential, and pigment synthesis [23,24].

Fig. 1. Diversity and molecular basis of TPCs.

(A) Phylogeny tree of TPCs from different species. Amino acid seuqneces of TPC paralogues and orthologues from same and different species were retrived from NCBI protein database and then aligned and analyzed with methods of ClustalW and Minimum-Evolution Tree using program MEGA (v. 11) [66]. (B) Structure of open-state human TPC2 in complex with PI(3,5)P₂ (pdb: 6NQ0) [37]. (C) Amino acid sequence alignment of selected TPCs of different species. Most amino acids are shaded collectively in colors according to their physicochemical properties (hydrophobic in yellow, polar in green, positively charged in blue and negatively charged in red) and special amino acids (glycine, histidine, tyrosine, and cysteine) are each shaded in other different colors. Residues discussed to be important for channel gating are shown in stick mode (sidechain only) in (B) and labeled according to human TPC2 sequence on top in (C).

Similar to TPC1, the TPC3 channel is also voltage–activated Na⁺- selective channels [27]. Zebrafish (D. rerio) and frog (e.g., X. laevis and X. tropicalis) TPC3 channels were expressed on the plasma membrane when heterologously expressed in HEK293 cells [27]. TPC3 provides slow-activating and non-inactivating Na⁺ currents that could underlie ultra-long action potentials observed in some oocytes and eggs although the biological functions of ultra-long action potentials and TPC3 are unclear [27]. TPC3 in sea urchins appeared to be localized in acidic organelles when heterologously expressed in X. laevis oocytes and SKBR3 cells [28]. TPC3 was initially reported to be insensitive to PI(3,5) P₂ [27]. However, a later study of a similar frog (X. tropicalis) TPC3 channel showed that the channel was sensitive to both PI(3,4)P₂ and PI (3,5)P₂ and that the slow process of channel-activation by voltage is due to depolarization-induced slow production of PI(3,4)P₂ in Xenopus oocytes [29]. All plants express at least a TPC1 channel localized to the vacuolar membrane [30]. The plant TPC1 from A. thaliana (AtTPC1) is a voltage-gated, Ca²⁺-regulated, nonselective (Ca²⁺ permeable) cation channel, whose biological function might be related to stress adaption but remains largely enigmatic [30].

Genes encoding TPCs also exist in invertebrates and eukaryotic unicellular organisms (Fig. 1A, Table S1), but not in prokaryotes [12, 30]. Despite the absence of the TPC gene in the model organisms D. melanogaster and C. elegans, some insects such as bees (e.g., A. mellifera), wasps, and silkworms have TPCN1, arachnids such as ticks (e.g., I. ricinus) may have both TPCN1 and TPCN3, and some flatworms, e.g., S. mansoni, have TPC2-like channels [12,31]. Sponges (e.g., A. queenslandica), the oldest multicellular living animals, have six TPC homologs (3 TPC1-like, 1 TPC2, and 2 TPC2-like), a higher number compared to other animals [32]. Genes encoding TPC-like proteins have also been found in unicellular eukaryotes such as choanoflagellates (e.g, M. brevicollis and S. rosetta), amoebae (e.g., D. disoideum), and algae (e. g., V. brassicaformis) [12,30] and lately reported in Apicomplexa [33] as discussed below.

3. General biochemical and biophysical properties of TPCs

TPCs are homodimeric channels with each subunit containing two pore-forming 6-transmembrane (TM) domains (6-TM domain I and II) of the basic structural unit (six TM helixes and a pore loop) of a voltage–gated ion channel (Fig. 1B). TPCs are believed to represent an evolutionary intermediate between homotetrameric voltage-gated K⁺ (K_V) channels and the single subunit (four 6-TM domains) voltage-gated Ca²⁺ and Na⁺ (Ca_V and Na_V) channels [34]. The X–ray structure of plant A. thaliana TPC1 [35,36] and Cryo-EM structures of mouse TPC1 [20] and human TPC2 [37] provide insights for understanding diversity in TPC structure and function. As seen in voltage-gated K⁺ channels, the transmembrane region in TPCs is domain-swapped and the homodimer forms rectangle-shaped functional channels. Multiple cytosolic components within each TPC subunit—including the N-terminus, i.e., the linker between the two 6-TM domains and the C-terminal region—are assembled into a tightly packed cytosolic domain. The plant AtTPC1 contain a Ca²⁺ binding EF-hand domain in the linker of the two 6-TM domains, whereas both mouse TPC1 and human TPC2 lack essential Ca²⁺-chelating acidic residues although their linker regions also adopt structures similar to the EF-hand domain. Compared to the voltage-activated PI(3,5)P₂-insenstive non-selective cation channel of the A. thaliana TPC1, the voltage- and/or PI(3,5)P₂-activated mammalian TPC1 and TPC2 channels are highly Na⁺-selective, e.g., P_Na:P_Ca: P_K≈10:1:0.3 from whole-lysosome recordings of human TPC2 [9] or similarly P_Na:P_Ca:P_K≈17:1:0.8 for inside-out recording of plasma membrane–targeted human TPC2^L11A/L12A mutant channels in excised membrane patches [15]. The structures of selectivity filter regions among plant and human TPCs are highly conserved. Structure-guided mutations showed that the ion selectivity is slightly affected by alteration in residue 272 (human TPC2) and mainly determined by variation in residues 652 and 653 (human TPC2) [15]. More than 10-fold changes in ion permeability of Ca²⁺ and K⁺ relative to Na⁺ were observed for both plant TPC1 and human TPC2 upon swapping their amino acid residues at only 3 positions (A272S, V652M, and N653G for human TPC2 and S265A, M629V, and G630N for plant TPC1) [15]. The human TPC2 dominant negative mutation L265P [38] likely results in loss of ion conduction through affecting the neighboring filter residues (T271, A272). Except in plant TPC1 and some choanoflagellate (e.g., S. rosetta) TPC-like channels, the Na⁺ selectivity appears to be a highly conserved feature in TPCs of different species according to the amino acid sequences (Fig. 1C, Table S1).

The human TPC2 ion conduction pore adopts both closed and open conformations in the ligand-bound state [37]. In the closed conformation, four pairs of residues—T308 and Y312 from the 6th TM helix of the 6-TM domain I (IS6) and L690 and L694 from the 6th TM helix of the 6-TM domain II (IIS6) (other such helixes are abbreviated similarly hereafter)—are proposed to form the constriction points on the cytosolic side, i.e., the cytosolic gate [37]. In both mouse TPC1 and human TPC2 structures, the ligand PI(3,5)P₂ is situated on the cytosolic side of the membrane at the junction formed by the IS3 and IS4 and the linker of the IS4 and IS5 (Fig. 1B) [20,37]. The inositol 1,3,5-trisphosphate head group of PI(3,5)P₂ interacts with multiple basic residues (K203, K204, K207, R210, R329 in human TPC2) and defines most of the ligand-protein interactions. Pore opening and closing is triggered by a conformational change at the IS6 in that the IS6 breaks into two halves at G317 (human TPC2) just below the cytosolic gate in the closed state, whereas IS6 becomes a long continuous helix in the open state (Fig. 1B). This IS6 conformational change is likely initiated by the interaction between PI(3,5)P₂ and residues on IS6, particularly Ser322 and the basic Arg329 in human TPC2 or Lys331 (equal to S322 of human TPC2 in position) in mouse TPC1 whose mutation to Ala abolished the channels’ responses to PI(3,5)P₂ [20,37]. The Gly residue at the 317 position (human) is fully conserved in TPC2 across different species whereas its presence is lost in TPC1 and TPC3 of most species (Fig. 1C), suggesting an unique and essential role of this Gly hinge point in PI(3,5)P₂-induced activation of the voltage-insensitive TPC2.

Both voltage sensor domains (VSD1 and VSD2) of each subunit are structurally similar between the voltage-sensitive mouse TPC1 and the insensitive human TPC2. Although several conserved features of a canonical VSD are absent in VSD1 and most of them are preserved in VSD2 of human TPC2, the presence or absence of an arginine at the R3 position of VSD2 underlies the drastic difference in the voltage sensitivity of mammalian TPC1 and TPC2 channels [20,37]. An arginine at the R3 position of VSD2 is fully conserved in TPC1, TPC3, and TPC-like channels in all examined different species while it is replaced by an Ile or Val residue in TPC2 (Fig. 1C). Replacing the R3 arginine (Arg540) with Gln or Ile in mouse TPC1 led to the loss of the channel’s voltage sensitivity [20,39], while introduction of an arginine to the R3 position by mutation of I551R in human TPC2 converted the channel into a voltage-dependent channel in a manner similar to the activation of TPC1 [37]. The opening of TPC1 requires interplay between the PI(3,5) P₂-induced conformational changes of IS6 and depolarization-induced facilitation of IIS6 movement [20].

4. An ancestral TPC with novel features localized in apicoplast of Toxoplasma gondii

A recent study by Li et al. [33] reported a novel TPC (TgTPC) identified in a single-celled parasite, T. gondii, the causative agent of toxoplasmosis. Compared to TPCs found in vertebrate animals and plants, TgTPC is larger (160 kDa) because of its much longer N-terminus and the linker between the channel’s two transmembrane domains. The linker lacks EF-hand domains found in other TPCs. Phylogenetic analysis showed TgTPC together with other Apicomplexa TPCs formed a distinct clade that is most closely related to those in chromerids (e.g., V. brassicaformis) which are Apicomplexa’s photosynthetic ancestors. TgTPC is more closely related to unicellular TPC-related proteins and plant TPC1 than to deuterostome animal TPC1, TPC2, and TPC3. Interestingly, TgTPC was absent on the lysososome-like organelle named plant-like vacuole but localized on the peripheral membranes of apicoplast, a non-photosynthetic plastid found in most apicomplexan parasites. The apicoplast originated from algae via secondary endosymbiosis and has four layers of membranes, of which the outermost layer is derived from the host’s endosomal system [40]. Using conditional knockout, complementary expression, and pore-region mutation of TgTPC, researchers found that silencing TgTPC resulted in greatly reduced growth and the lytic cycle of T. gondii, owing to a loss of apicoplast integrity and function that likely caused a reduction in apicoplast-dependent production of essential metabolites for cell growth [33]. With Ca²⁺-imaging analysis of genetically encoded Ca²⁺ indicator targeted to the apicoplast and electron microscopy, it was found that deletion of the TgTPC gene deletion reduced apicoplast Ca²⁺ uptake presumably from ER via membrane contact site between the apicoplast and the ER. Importantly, TgTPC has significant sequence similarity to other TPCs in the pore regions and the substitution of conserved Leu residues (equivalent to Leu265 and Leu646) by Pro nullified the rescuing effect of the exogenous channel expression in TgTPC-silencing cells, supporting a role of the TgTPC ion conducting activity in the observed function of the channel in T. gondii. Overall, this well-designed study revealed a new TPC that is localized in the apicoplast, a non-acidic organelle important for the growth and the lytic cycle of the parasitic protozoan, possibly by direct involvement of organelle communication and intracellular Ca²⁺ signaling. The biophysical properties of TgTPC remain unknown. Future electrophysiological studies will be needed to clearly demonstrate the channel activities, identify the stimuli (voltage and/or endogenous ligands) of channel activation, and determine the ion selectivity.

5. TPCs in NAADP-evoked Ca²⁺ mobilization from acidic organelles

NAADP-evoked Ca²⁺ signaling is believed to occur via an initial NAADP-evoked local Ca²⁺ release from the small endolysosomal organelles, and this release subsequently triggers a global and large-scale Ca²⁺ release from the large ER Ca²⁺ store via a Ca²⁺-induced Ca²⁺ release (CICR) mechanism through the Ca²⁺-sensitive IP₃ receptors and/or ryanodine receptors (Fig. 2) [3, 5,41–43]. Despite the significance and uniqueness of NAADP-evoked Ca²⁺ signaling, the biochemical process of NAADP-evoked Ca²⁺ release from endolysosomes is poorly understood. Since 2009, accumulating evidence supports that endolysosomal TPCs are critical to NAADP-evoked Ca²⁺ release: (i) manipulation of TPC expression in cell lines resulted in changes in NAADP-evoked Ca²⁺ release [12,13,38,44,45]; (ii) cells’ response to NAADP was eliminated in TPC knockout mice [13,46,47]; and (iii) cells’ response to NAADP was rescued by re-expression of TPCs in TPC1/2 knockout mice [47]. NAADP-responsive Ca²⁺ currents were observed with planar or traditional patch-clamp recording of whole lysosomes [16,38,48–50]. Recently, a small molecule agonist namely TPC2-A1-N was reported to activate TPC2 and produce similar Ca²⁺ currents [51]. Naturally, there has been a long-held belief that TPCs are the NAADP-responsive Ca²⁺-release channels underlying the Ca²⁺ mobilization from acidic stores. However, such a notion is at odds with the high Na⁺-selectivity of mammalian TPCs when activated by voltage and/or PI(3,5)P₂ and with the observation that the channels’ response to NAADP was completely undetectable in some other studies [9,15,17, 19]. Thus, controversies or doubts arose, and to some extent remain, as to how TPCs are involved in NAADP-evoked Ca²⁺ release from the endolysosomal system. Compared to the PI(3,5)P₂-activated robust and large TPC currents, the currently reported NAADP-activated currents were mostly small in amplitude. Given the presence of numerous other endogenous channels on endolysosomes [52] which could potentially contaminate TPC currents, more thorough studies are needed to establish whether NAADP-activated TPCs are more Ca²⁺ permeable than the voltage- and PI(3,5)P₂-induced TPCs. From another viewpoint, the limited TPC Ca²⁺ permeability together with the large Ca²⁺ gradient across the lysosome membranes might be able to produce enough Ca²⁺ release from acidic stores to result in the much larger Ca²⁺-induced Ca²⁺ release from the ER. Alternatively, the TPC-mediated Na⁺ currents could contribute to endolysosomal Ca²⁺ release by membrane depolarization that could activate some voltage-gated Ca²⁺ channels on the endolysosomal membrane. Regardless of the uncertainty on biophysical properties of NAADP-activated TPC currents, it was later generally agreed that TPCs are not the direct receptor of NAADP, i.e., some other protein, such as an accessory NAADP-binding protein, is needed for the channel’s response to NAADP [54–56]. This concept originated from the absence of photocrosslinking labeling of TPCs with a photoactive NAADP analogue [55] and is consistent with the labile nature of the channels’ response to NAADP, which cannot be reproduced by different groups or under different conditions. The recent identification and characterization of NAADP binding proteins, as discussed below, demonstrates the necessity of an NAADP-binding protein for the NAADP-evoked TPC activation and Ca²⁺ mobilization from intracellular stores.

Fig. 2. A hypothetic model of NAADP-induced Ca²⁺ mobilization signaling pathways.

Lsm12 as an NAADP receptor interactes with TPCs and mediates NAADP-evoked and TPC-dependent Ca²⁺ release from the endolysosome that subsequently triggers a larger scale Ca²⁺ release from endoplasmic reticulum (ER) through ryanodine receptors and/or IP3 receptors via a mechanism of CICR. The possibility of limited Ca²⁺ permeability through TPCs is indicated with a gray arrow. JPT2 as another NAADP receptor interacts with ryanodine receptor 1 and likely mediates Ca²⁺ release direclty from endoplasmic reticulum without the involvement of endolysosome system. The uncertainty on the role of JTP2 in TPC-dependent NAADP signaling is indicated by a question mark.

6. Lsm12 as an NAADP-binding protein and TPC regulatory protein in NAADP-evoked Ca²⁺ signaling

To identify the elusive NAADP receptor protein, our group performed affinity purification and a quantitative proteomic analysis of the interacting proteins of immobilized NAADP (crosslinked to agarose beads) and human TPCs [57]. We have identified a Sm-like (Lsm) protein, Lsm12, complexed with NAADP, TPC1, and TPC2. Lsm12 possesses an N-terminal Lsm domain and a putative C-terminal anticodon-binding domain. Lsm12 directly binds to NAADP via its Lsm domain, colocalizes with TPC2, and mediates the apparent association of NAADP to isolated TPC2 or TPC2-containing membranes. Thus, our findings of Lsm12’s relationship to TPCs and NAADP clearly consolidate the notion that TPCs are not direct receptors of NAADP. We observed that Lsm12 is essential to and immediately participates in NAADP-evoked TPC activation and Ca²⁺ mobilization. With a combination of whole-cell patch-clamp recording and intracellular application of NAADP via microinjection, we were able to record Lsm12-dependent NAADP-induced plasma membrane–targeted TPC2 and TPC1 mutant channel currents. The physical association between Lsm12 and TPC2 likely plays a role in NAADP signaling because the Lsm12^Δ45−50 mutant, which has a weakened association with TPC2 but not NAADP, was deficient in NAADP-evoked TPC2 activation and intracellular Ca²⁺ elevation. These findings led us to conclude that Lsm12 is an NAADP receptor and a TPC regulatory protein [57]. According to The Human Protein Atlas (www.proteinatlas.org), Lsm12 at the RNA level is ubiquitous and abundantly expressed in various tissues and cell lines. Given that Lsm12 was functionally critical for NAADP-evoked Ca²⁺ release in all 3 tested cell types (HEK293, SK-BR-3, and MEFs), we anticipate that Lsm12 as an NAADP receptor plays a critical role in NAADP signaling in various cells.

Although an Lsm protein could potentially regulate expression of some other proteins via its RNA-binding function, our data (including the following) strongly support that Lsm12 directly participates in NAADP-evoked Ca²⁺ release: (i) Lsm12 is a highly selective and high-affinity receptor for NAADP; (ii) Lsm12 mediates the apparent association of NAADP to TPCs and the NAADP-induced activation of TPCs; and (iii) importantly, microinjection of the purified Lsm12 protein immediately restored the NAADP-evoked Ca²⁺ release and TPC2 activation in Lsm12-null cells, suggesting that the loss of NAADP response in the Lsm12-null cells is unlikely caused by a defect in cellular structure or content whose restoration process is usually not immediate, e.g. ~ 5 minutes for lysosomal Ca²⁺ refilling if depleted [58]. The function of Lsm12 in NAADP binding and its association with TPC2 are both mainly mediated by its Lsm domain. The Lsm domain is predicted to form a highly conserved tertiary structure of an Lsm fold that consists of a 5-stranded anti-parallel β-sheet. As seen in many other Lsm proteins, the Lsm domain may oligomerize into a hexameric or heptameric Lsm ring that allows binding to an oligonucleotide (adenine or uracil) [59]. NAADP is essentially a type of dinucleotide. Thus, Lsm12 appears to be well positioned to serve as an NAADP receptor via its Lsm domain. Overall, our findings establish Lsm12 as a receptor protein with a high affinity for NAADP and as a TPC regulatory protein necessary for NAADP-evoked TPC activation and TPC-dependent Ca²⁺ mobilization from intracellular stores [57].

Lsm12 belongs to the Lsm protein family, which includes smaller Lsm proteins (Lsm1-10) that contain only an Lsm domain and larger Lsm proteins (Lsm11, 12, 14A, 14B, and 16 in human) that contain an extra C-terminal non-Lsm domain [60]. Given that Lsm12 binds to NAADP mainly via its Lsm domain, it is reasonable that some other Lsm proteins may potentially function as NAADP binding proteins and may play a role in NAADP signaling when conditions allow. However, given that Lsm12 knockout or knockdown caused a full or nearly full loss of the cells’ response to NAADP, Lsm12 must be the dominant NAADP receptor in the examined cells of HEK293, SK-BR-3, and MEFs [57]. Consistently, knockdown of Lsm5 and Lsm11 produced only minor effect on NAADP-evoked Ca²⁺ elevation in HEK293 cells [57].

7. JPT2 as an NAADP-binding protein in NAADP signaling

A distinct protein, JPT2 (also known as HN1L), was also recently identified as an NAADP-binding protein by two research groups via a crosslinking approach with a photoactive NAADP analogue in erythrocytes [62] and human Jurkat T cells [63], suggesting the presence of different NAADP-binding proteins. Both reports of this protein used the same all-in-one-clickable alkyne-NAADP photoprobe followed a previous study [55] that an NAADP analogue (5-azido-NAADP) was photo-crosslinked to some low-molecular-weight protein(s) (22/23-kDa doublet) but not TPCs in mammalian cells. NAADP binds to recombinant JPT2 [62, 63]. Knockdown of JPT2 partially reduced the photolabeling of the low-molecular-weight protein in both HEK293 and U2OS cells and largely abolished the U2OS cells’ response (intracellular Ca²⁺ elevation) to NAADP [62]. Co-immunoprecipitation of JPT2 was observed with TPC1 but not TPC2 [62]. No experimental data have been presented yet to determine whether JPT2 can mediate TPC-dependent NAADP signaling and NAADP-evoked TPC activation. Interestingly, in human Jurkat and primary rat T cells, the knockout of JPT2 affected the early (presumably NAADP-mediated) phase of artificial antigen-stimulated Ca²⁺ microdomain signals in a ryanodine receptor 1, but not in a TPC-dependent manner [63]. JPT2 was co-immunoprecipitated and partly co-localized with ryanodine receptor 1 at ER and plasma membrane junctions in Jurkat T cells [63]. Overall, the two reports establish JPT2 as an NAADP-binding protein involved in NAADP signaling in certain cells. However, given that NAADP signaling could be TPC-independent in some cells, it remains to be examined whether JPT2 is involved in TPC-dependent NAADP signaling.

8. Discussion

In recent years, significant progress has been made in expanding our knowledge of the structures and function of TPCs and in the identification of NAADP-binding proteins involved in NAADP signaling. The newly characterized plastid TPC in T. gondii revealed the importance of an ancestral TPC in the unicellular life of a human parasite [33]. The identification of Lsm12 as an NAADP receptor mediating the TPC-dependent NAADP signaling fills a molecular signaling process gap between NAADP and TPCs [57]. The report of the JPT2’s dependence on ryanodine receptor 1 but not TPCs in its effect on NAADP signaling in Jurkat T cells [63] differs from the canonical view of the essential role of TPCs in NAADP signaling. This finding necessitates the need to revisit some early idea that distinct ion channels, including ryanodine receptors [64] and TRPML1 [65] channels, might also underlie different cells’ response to NAADP in Ca²⁺ signaling. The relationship between Lsm12 and JPT2 in NAADP signaling remains to be closely examined. The currently reported studies [62,63] suggest that JPT2 likely functions differently than Lsm12 [57] in NAADP signaling. In agreement with the previous report of ryanodine receptors as NAADP-responsive channels [64], Lsm12 and JPT2 might underlie different NAADP signaling processes mediated by different ion channel (TPCs vs. ryanodine receptors) (Fig. 2). Without doubt, the recent advance in the field will expedite our understanding of the structure and function of TPCs and the elusive biochemical process of NAADP-evoked Ca²⁺ release from acidic stores.

Abbreviations:

NAADP

nicotinic acid adenine dinucleotide phosphate

TPC

two-pore channel

IP₃

inositol 1,4,5–trisphosphate

TM

transmembrane

VSD

voltage sensor domains

CICR

Ca²⁺-induced Ca²⁺ release

ER

endoplasmic reticulum