Upon activation of the TCR/CD3 complex, formation of NAADP rapidly increases within 10–20 s in Jurkat T cells. Following a subsequent decrease within the first minute, a continuously elevated [NAADP] remains for 5–20 min (10). It has been proposed that NAADP may act as an early triggering messenger, mediating initial localized Ca²⁺ events which are subsequently amplified to a global signal, e.g., by recruitment of further channels, other second messengers like cADPR, IP₃, and/or Ca²⁺ induced Ca²⁺ release (CICR). In T-lymphocytes a bell-shaped concentration-response curve following NAADP microinjection is observed. Compared to other second messengers such as IP₃, already low concentrations in the nanomolar range (30–100 nM) induce Ca²⁺ signaling in T-lymphocytes (18).

The mechanism and very early kinetics of receptor-mediated formation of NAADP in vivo remains unclear and has been discussed previously [e.g., (19)]. In brief, NAADP is formed in vitro by a base-exchange of NADP in presence of nicotinic acid and a pH of 5 [Figure 1; (20)]. Further, at pH 5, but also at pH 7.4, 2′-phospho-cADPR may be converted to NAADP (21). Both reactions are catalyzed by the membrane bound, multifunctional enzyme CD38 and require presence of up to millimolar concentrations of nicotinic acid in vitro (20, 21). Furthermore, influx of extracellular NAADP may also induce Ca²⁺ signals as shown in a rat basophilic cell line (22). Interestingly, gene silencing of CD38 in Jurkat T-lymphocytes did not result in decreased NAADP levels. Rather, in thymus and spleen of CD38 knock-out mice increased NAADP levels were observed, thus indicating that CD38 may particularly drive degradation of NAADP (23, 24). Accordingly, in T cells to date CD38 may be primarily understood as degrading enzyme while its role in NAADP synthesis in vivo remains to be elucidated. In T-lymphocytes and other CD38⁺ cells, NAADP is degraded to 2′-phospho-ADPR at neutral and acidic pH by CD38, but degradation may also occur non-specifically via nucleotide pyrophosphatases (25).

The targeted receptor(s) and hence target organelle(s) of NAADP are still under debate (19, 26–29). In general, RyR have been implicated as NAADP targets in different cell types, e.g., skeletal muscle cell (30), or pancreatic acinar cells (31, 32). Nonetheless, in the majority of mammalian cells as well as in sea urchin eggs there is evidence that NAADP may primarily target acidic stores and that the RyR located on the ER may rather play a central role in the amplification of the Ca²⁺ signal (33–35).

However focusing particularly on data obtained in T-lymphocytes, NAADP Ca²⁺ signaling strongly depends on RyR activity in the ER (36–40). Following either knock-down or inhibition of RyR by ryanodine in Jurkat T cells, subcellular and global Ca²⁺ signals by NAADP microinjection were inhibited or almost completely abolished (36, 38). In primary effector T-lymphocytes the NAADP antagonist BZ194 inhibits Ca²⁺ signaling, e.g., during formation of the immunological synapse. Furthermore, BZ194 has been shown to selectively inhibit NAADP dependent binding of [³H]ryanodine to RyR1 (39). Interestingly, in primary effector T cells of an animal model of multiple sclerosis, experimental autoimmune encephalomyelitis (EAE), BZ194 leads to a decrease in cell motility and invasive capacity as well as a decrease in cytokine expression, all of which indicate the central role of NAADP-mediated Ca²⁺ signaling in T cells possibly via RyR (40). In contrast to these results obtained in CD4⁺ T cells, in cytotoxic T cells NAADP appears to target two-pore channels (TPC) on cytolytic granules (41). In general, overexpression or inhibition of the endolysosomal TPC1 and TPC2 suggest that NAADP initiates Ca²⁺ events via TPC [e.g., (42–44)]. Recently, the N-terminus of TPC1 has been identified as functional region for NAADP-mediated Ca²⁺ signaling (26). In contrast, it was shown that NAADP-mediated Ca²⁺ signaling in TPC1^−/−/TPC2^−/− mice does not differ from wild-type mice (27). Thus, whether NAADP primarily targets TPCs is controversial and particularly the effect of NAADP on TPCs in T-lymphocytes is not yet clear.

Further, in T-lymphocytes and neutrophils TRPM2 is activated by micromolar concentrations of NAADP per se, but particularly in synergism with cADPR (45, 46). The effect of cADPR on TRPM2 however, could not be confirmed in HEK293 cells overexpressing TRPM2 and contamination of commercial cADPR preparations with ADPR have been discussed (47, 48). NAADP has been shown to target the unspecific Ca²⁺ channel TRPML1 in smooth muscle myocytes (49). Whether TRPML1 and TRPM2 are of functional relevance within NAADP – mediated Ca²⁺ signaling in T-lymphocytes, remains to be elucidated.

Despite the questions which organelles are targeted by NAADP and which specific downstream mechanisms may underlie the initiated Ca²⁺ events, also the identity of the NAADP receptor remains unclear. Photoaffinity labeling in mammalian cells using a probe specific for NAADP binding proteins was not altered upon overexpression or knock-out of TPC1 or TPC2, but suggests that a yet not identified 22/23 kDa protein binds NAADP and may hence couple NAADP to its respective Ca²⁺ channels (50, 51), a mechanism recently introduced as unifying hypothesis of NAADP action (29).

Cyclic adenosine diphosphoribose (cADPR) was the first Ca²⁺ mobilizing second messenger discovered as derivative of an adenine dinucleotide (52, 53). Though first described in sea urchin egg homogenates, the Ca²⁺ mobilizing activity of cADPR was soon detected in many cells types. In 1995 we published the first report demonstrating specific Ca²⁺ release in human Jurkat T cells (54). Central aspects of the role of cADPR in T cell Ca²⁺ signaling were subsequently published by our laboratory: (i) formation of cADPR upon TCR/CD3 ligation (14, 55), and (ii) mode of action of cADPR by activation of Ca²⁺ release via RyR, as shown by gene silencing of RyR (56). Further, we demonstrated tyrosine phosphorylation of RyR upon TCR/CD3 ligation; in permeabilized T cells enhancement of cADPR evoked Ca²⁺ release by tyrosine kinase p59fyn was observed (57). Importantly, we demonstrated amplification and propagation of pacemaker Ca²⁺ signals by cADPR (58). A connection of cADPR signaling to Ca²⁺ entry was also observed: microinjection of cADPR in the absence of extracellular Ca²⁺ or in the presence of Ca²⁺ channel blockers resulted in much reduced Ca²⁺ signals (59). Finally, using a specific cADPR antagonist it was shown that downstream activation parameters of primary human T cells, such as activation antigen expression or proliferation, were concentration-dependently inhibited (14) suggesting a pivotal role of cADPR in T cell biology.

A detailed structure-activity analysis of cADPR in T cells has been conducted over the past couple of years. The main results from these studies were recently reviewed (60) and are (i) critical dependence of agonist vs. antagonist properties on the substituent at the C-atom 6 of the purine base, (ii) maintenance of biological activity, albeit at a lower level, when both southern and northern ribose were replaced by carbocyclic moieties or simplified ether/alkane bridges, and (iii) the possibility of radical simplification of the purine structure, e.g., the 1,2,3-triazole-4-amide mimic of adenine within cADPR retains biological activity.

A relatively new addition to the realm of adenine based Ca²⁺ mobilizing second messengers in T-lymphocytes is ADPR. Presence of ADPR in eukaryotic cells has been known for quite a while (61), but since ADPR is rather dangerous for the cell – its reactive ribose can non-enzymatically form Schiff-bases with amino groups of cellular proteins (62, 63) – it was mostly considered a toxic cellular waste product. This casually explained the presence of efficient mechanisms for the degradation of ADPR in form of cytosolic (64, 65) and mitochondrial ADPR pyrophosphatases (66, 67). These enzymes hydrolyze the pyrophosphate bridge of ADPR yielding AMP and ribose-5′-phosphate that are fed back into metabolism.

Two discoveries suggested that there might be more to ADPR: in 2001 it was reported that TRPM2 (formerly termed TRPC7 or LTRPC2), a Ca²⁺-permeable cation channel of the melastatin subfamily of TRP channels, can be activated by binding of ADPR to a cytoplasmic domain homologous to the mitochondrial ADPR pyrophosphatase NUDT9 (68, 69). This channel shows expression in a variety of tissues with highest levels being found in brain and cells of the immune system. A year later Bastide et al. showed that ADPR is also able to activate type I RyR isolated from rat skeletal muscle in the presence of micromolar concentrations of Ca²⁺ (70). Since there has been little news on the action of ADPR on RyR, we will focus on the role of ADPR for TRPM2 activation in T-lymphocytes.

Most of the work on ADPR and TRPM2 in T cells so far has been done in Jurkat cells that express TRPM2 on transcript and protein level and respond with a typical TRPM2 current to ADPR infusion (11, 45, 69). Microinjection of ADPR (11) and uncaging of photoactivatable ADPR (71) in these cells results in Ca²⁺ entry-dependent Ca²⁺ signals. By HPLC analysis the cellular ADPR concentration of roughly 40 μmol/L in resting Jurkat cells (72) was shown to nearly double after stimulation with high concentrations of concanavalin A (11).

There are different conceivable ways how this ADPR might be generated. CD38 expressed in Jurkat as well as primary T cells (73) can metabolize β-NAD⁺ and cADPR to yield ADPR (74). The topological paradox initially described for cADPR also holds true for ADPR (1). This paradox might be resolved by specific uptake mechanisms for ADPR as have been reported for erythrocytes (75, 76). Another possibility is the presence of CD38 in a type III orientation (2). While the contribution of CD38 to the increase in ADPR after stimulation is still unclear, the basal ADPR seems to be independent of CD38 as the murine T-lymphoma line BW5147 that lacks transcripts for CD38 (77) has even higher basal ADPR levels [73 μmol/L (72)] than Jurkat cells.

Another way that has been discussed for the production of ADPR is the consecutive action of poly-ADPR polymerase (PARP) and poly-ADPR glycohydrolase (PARG) (78). While PARP activity and poly-ADPR levels are quite low in non-stimulated cells, there is a constant turn-over due to the low K_M of PARG [reviewed in (79)] that might contribute to the basal ADPR detected in Jurkat cells. Under DNA damaging conditions like strong oxidative stress the activity of PARP increases to such levels that a large part of cellular β-NAD⁺ can be metabolized as has been shown for DT-40 cells (80). Data for a range of cells suggest activation of TRPM2 by oxidative stress results in cell death by apoptosis (78, 81), most likely due to mitochondrial calcium overload and downstream activation of caspases [reviewed in (82)]. While murine CD4⁺ T cells also die after exposure to hydrogen peroxide, this apparently does not involve TRPM2 (83).

Interestingly there have been reports that in T cells PARP-1 activation can occur after TCR stimulation in a way independent of oxidative stress or DNA damage resulting in poly-ADP ribosylation of NFAT [(84, 85); reviewed in (86)]. It might be speculated that this increased pADPR turn-over will result in increased cellular ADPR and TRPM2 activation hinting to a possible role for ADPR/TRPM2 in TCR signaling. In accordance with this, naïve CD4⁺ T cells from the wild-type mice upregulated TRPM2 after stimulation with α-CD3/α-CD28-beads and CD4⁺ T-lymphocytes from TRPM2^−/− mice showed not only reduced proliferation, but also reduced production of pro-inflammatory cytokines upon activation (83).

Most work on the role of TRPM2 in the immune response has been done using the TRPM2^−/− mouse (87). In a model for ulcerative colitis the inflammation was suppressed, but this was shown to be due to a reduced production of the chemokine CXCL2 in monocytes whereas the infiltration of T cells in the colon was not affected by the knock-out of TRPM2 (87). Recent work has shown that TRPM2 knock-out does not affect airway inflammation either induced by oxidative stressors (88) or as a result of exposure to ovalbumin in a mouse model for acute asthma (89). On the other hand TRPM2^−/− noticeably reduced inflammation and spinal cord lesions in EAE induced by a peptide from myelin oligodendrocyte glycoprotein (83). Since the effect of TRPM2^−/− on EAE might also involve reduced neuronal cell death or microglia activation, it will be interesting to see whether proliferation and effector functions of T cells from such animals are affected.

Taken together, adenine derived Ca²⁺ mobilizing second messengers play essential roles in T cell Ca²⁺ signaling, both during activation/re-activation or apoptosis. While NAADP is important as rapid Ca²⁺ trigger, particularly in effector T cells, cADPR apparently holds a central role in maintenance of long-lasting Ca²⁺ signaling. Interestingly, also apoptosis induction via TRPM2 involves an adenine derived second messenger, the dinucleotide ADPR. The documented involvement of these adenine derived Ca²⁺ mobilizing second messengers in central aspects of immune regulation make the pathways described in this review suitable targets for therapeutic intervention. In fact, we have recently shown that the NAADP antagonist BZ194 ameliorated the clinical course of transfer EAE, an animal model of multiple sclerosis (40).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.