nicotinic acid adenine dinucleotide phosphate (NAADP) is one of the latest and most potent second messengers to be implicated in the process of Ca2+ release from intracellular stores. First reported to release Ca2+ from stores in sea urchin eggs in 1995 (10, 26), it may be involved in cholecystokinin and endothelin-1 Ca2+ signaling in mammalian cells (8, 23). The article by Soares et al. (Ref. 32; see p. C227 in this issue) now also implicates NAADP in responses to histamine. The receptor that mediates NAADP-induced Ca2+ release has not yet been isolated or identified, but it appears to demonstrate unusual properties of inactivation/desensitization that are different in sea urchin eggs and mammalian cells (17).
The Ca2+ released initially in response to NAADP is generally not from inositol trisphosphate (IP3)-sensitive or ryanodine-sensitive Ca2+ stores (23). The prevalent opinion has been that it releases Ca2+ from lysosomal or other acidic stores (33, 34). In certain systems, however, it does release Ca2+ from stores that involve ryanodine receptors (7, 13, 18). In the case of pancreatic acinar cells, this may involve direct action of NAADP on ryanodine receptors (18), but in other cases, the Ca2+ released in response to NAADP may load up other NAADP-insensitive stores (12) or bring about secondary Ca2+-induced Ca2+ release involving ryanodine-sensitive stores (7, 23).
NAADP can be produced by the same enzyme that produces another Ca2+ releasing second messenger, cADP ribose (cADPR). This ADP ribosyl cyclase was originally isolated from Aplysia (25). Although it is a soluble enzyme, its closest mammalian counterpart, the lymphocyte surface antigen CD38, a less-active ADP ribosyl cyclase, has clearly been implicated in the synthesis of cADPR extracellularly (21), its vectorial transport into the cell (16), and in NAADP production (1). At first glance, it seems a lot to ask of any enzyme for it to produce two distinct second messengers extracellularly and then transport at least cADPR back into the cell. However, ADP-ribosyl cyclases like CD38 are truly remarkable enzymes. As seen in reactions indicated by red and green arrows in Fig. 1, not only have they been implicated in the three functions above (cADPR synthesis via i in Fig. 1; NAADP synthesis via ii in Fig. 1), but also in the generation of dimeric ADP ribose (ADPR2; Ref. 14; iii in Fig. 1), several adenine dinucleotides, adenosine 5′-pyrophosphate-5′-adenosine (Ap2a), and its N1-isoadenosine and N3 isoadenosine derivatives (2; iv in Fig. 1), and generation of ADPR (31; v and vi in Fig. 1). The presence of so many distinct functions for the same enzyme is extremely unusual.
NAADP produced extracellularly by CD38 would still have to be transported into the cell by some transporter to release Ca2+ from lysosomes, but such a transport mechanism has been recently described (5). NAADP might also be produced intracellularly, since soluble (19) and mitochondrial (27) forms of ADP-ribosyl cyclase capable of producing NAADP have also been described. As with the more active Aplysia enzyme, researchers including Chini concluded that the ADP ribosyl cyclase CD38 was responsible for most mammalian NAADP production as well (11).
Soares et al. (32) now challenge this perception. Chini et al. (11) had previously examined NAADP synthesis by one particular reaction in CD38−/− knockout mice to support the claim that most NAADP was produced by CD38, but they had not examined NAADP content of any tissues in the knockout mice. Chini et al. (11) and Ceni et al. (9) showed that some cADPR was still produced in the brains of the CD38−/− knockout mice, indicating that CD38 was not required to generate cADPR in mouse brain. With the use of the same mice, Soares et al. (32) measured NAADP content in the uterus and several other tissues lacking CD38, and demonstrate it is not reduced, even if the tissues are devoid of the one CD38-mediated enzymatic reaction known to make it. They show that the Ca2+ release induced by NAADP from myometrial cells and microsomes exhibits the same principal characteristics observed in other systems, namely the ability to release Ca2+ from stores distinct from those released by IP3 or cADPR/Ca2+. Thus, the system they are studying still responds to the NAADP that is produced, and consequently the mammalian enzyme responsible for the production of most NAADP is called into question. The CD38 requirement for NAADP generation is not challenged just in the uterus, but also in the heart, liver, kidney, and spleen. However, while these CD38-deficient tissues produced NAADP, their cADPR levels were undetectable, and thus none produced significant cADPR in the absence of CD38 (32).
Thus, the enzyme responsible for NAADP generation in mammalian cells is unknown, and this represent a challenge of equal weight in the field to the identification of the receptor for NAADP action.
The hunt for an alternative enzyme for generating NAADP should now begin in earnest. Candidates include other ADP ribosyl cyclases, including CD157 (15, 29). CD157, formerly known as BST-1 (20, 22), is very closely situated to CD38 in the 4p15 region of human chromosome 4 and is likely a product of gene duplication (15). Since CD157−/− mice are also available, breeding double knockouts with CD38−/− mice (29) should readily resolve this issue. The aforementioned soluble and mitochondrial ADP ribosyl cyclases (19, 27) might represent other family members capable of doing the job. However, the observation that cADPR is not produced in CD38−/− tissues that still generate NAADP (32) suggests that the alternative NAADP-producing enzyme is less likely to be a cADPR-generating ADP ribosyl cyclase. CD157 is much less active at cyclizing cADPR than CD38 (22), but its ability to generate NAADP has not been reported.
There may be other reasons why an ADP ribosyl cyclase is perhaps an unlikely candidate. In terms of mechanism, ADP ribosyl cyclases such as CD38 or CD157 generate NAADP by what is known as the base-exchange reaction, removing nicotinamide from nicotinamide adenine dinucleotide phosphate (NADP) and substituting nicotinic acid at an acidic pH optimum (Refs. 1, 4; ii in Fig. 1, with a similar base exchange reaction leading to NAAD shown in vii of Fig. 1). The base-exchange reaction leading to NAADP is the enzymatic reaction that Chini and colleagues (11, 32) had measured and found deficient in the CD38−/− mice. Although not much NAADP needs to be generated to release Ca2+ (1), the base-exchange reaction is weak at neutral pH (1) and may suffer due to the absence of nicotinamide to nicotinic acid conversion by nicotinamidase in mammals (30). It is not clear whether exogenous niacin in the diet contributes sufficient nicotinic acid to support the base-exchange reaction which utilizes it to produce NAADP (4). Perhaps not surprisingly, then, since CD38−/− tissues also produced neglible NAADP by way of in vitro base-exchange reaction relative to wild-type tissues, the results of Soares et al. (32) also indicate that most mammalian NAADP is now unlikely to be produced by the base-exchange reaction.
Other pathways converging on NAADP are shown in Fig. 1, bottom left. The nicotinic acid requirement reduces the likelihood of one recently suggested alternative, in which NAADP could be produced by linking nicotinic acid to 2′ phospho cADP ribose (cADPRP; see viii in Fig. 1) by some ADP-ribosyl cyclase (28), perhaps CD157 (29) or some other as yet uncharacterized cyclase which generates very little cADPR relative to NAADP.
Rather, Soares et al. suggest that NAADP might be produced: 1) from NADP by amide to acid conversion instead (Refs. 3, 11, 32; ix in Fig. 1) or 2) by phosphorylation of NAAD by some NAD kinase (Refs. 3, 11, 32; x in Fig. 1).
However, there may be other possibilities. For example, Billington et al. (6) have reported that NAADP can be reduced to NAADPH, which is inactive at inducing Ca2+ release. If so, NAADP could be produced from NAADPH by an NAD(P)H oxidase (Ref. 6; xi in Fig. 1).
Let the hunt begin!
- Copyright © 2007 the American Physiological Society