Biochemical Pharmacology 77 (2009) 1612–1620

Contents lists available at ScienceDirect

Biochemical Pharmacology
journal homepage: www.elsevier.com/locate/biochempharm

Detection and pharmacological modulation of nicotinamide mononucleotide

(NMN) in vitro and in vivo
Laura Formentini, Flavio Moroni, Alberto Chiarugi *
Department of Preclinical and Clinical Pharmacology, University of Florence, Italy

A R T I C L E I N F O A B S T R A C T

Article history: The emerging key role of NAD-consuming enzymes in cell biology has renewed the interest in NAD
Received 24 December 2008 resynthesis through the rescue pathways. The first step of the nicotinamide-dependent NAD-rescue
Accepted 19 February 2009 pathway is operated by nicotinamide phosphoribosyl transferase (NaPRT) forming nicotinamide
mononucleotide (NMN). Because of the difficulties in measuring NMN, numerous open questions exist
Keywords: about the pathophysiological relevance of NaPRT and NMN itself. Here, we describe a new method of
NAD fluorimetric NMN detection upon derivatization of its alkylpyridinium group with acetophenone. By
PARP
adopting this method, we analyzed the kinetics of nicotinamide-dependent NAD recycling in HeLa and
Mitochondria
U937 cells. Measurement of NMN contents in subcellular fractions revealed that the nucleotide is highly
FK866
NMNAT enriched in mitochondria, suggesting intramitochondrial NAD synthesis. NMN increases in cells
NaPRT undergoing hyperactivation of the NAD-consuming enzyme poly(ADP-ribose) polymerase (PARP)-1, or
exposed to gallotannin, a putative inhibitor of NMN-adenylyl transferases. Evidence that the inhibitor of
NAD resynthesis FK866 selectively inhibits NaPRT having no effect on NMNAT activity is also provided.
Importantly, NMN reduces NAD and ATP depletion in cells undergoing PARP-1 hyperactivation,
significantly delaying cell death. Finally, we show that a single injection of FK866 in the mouse induces
long-lasting (up to 16 h) but mild (20%) reduction of NMN contents in different organs, suggesting slow
rate of basal NAD consumption in vivo. Data provide new information on the biochemistry and
pharmacology of NAD biosynthesis, allowing a better understanding of pyridine nucleotide metabolism.
ß 2009 Elsevier Inc. All rights reserved.

1. Introduction NAD [1,4,5]. Notably, a recently identified metabolic route leading to

NAD formation from nicotinamide riboside also exists [6,7].
Pyridine nucleotides have long been considered hydride ion Conversely, vertebrates are unable to convert nicotinamide into
carriers exclusively involved in oxidoreduction reactions. However, nicotinic acid [5].
the recent identification of different enzyme families having A central metabolite in NAD resynthesis is nicotinamide
nicotinamide adenine dinucleotide (NAD) as substrate has sig- mononucleotide (NMN, Fig. 1). It is produced by nicotinamide
nificantly renewed the interest in the biochemistry and pharmacol- phosphoribosyl transferase (NaPRT) and by nicotinamide riboside
ogy of pyridine nucleotides [1–3]. In particular, being these enzyme kinase (NRK). NMN is then converted into NAD by three isoforms of
families responsible for the irreversible transformation of NAD into NMN-adenylyl transferase (NMNAT). Recently, NaPRT received
different products, attention has been focused on the mechanisms of much attention by the scientific community because of its
NAD resynthesis into eukaryotic cells. Classically, two metabolic pleiotypic functions. Specifically, besides being an intracellular
pathways regulate the formation of NAD in mammalian cells; i.e. the enzyme involved in NAD resynthesis, the protein has been also
de novo (also called the ‘‘kynurenine’’) pathway, leading to NAD from identified as a released factor behaving as an adipokine and as an
tryptophan, and the rescue pathway using the nicotinamide moiety inflammatory cytokine [8]. Adipokines are fat tissue-derived
produced as a by-product by the NAD-consuming enzymes. The hormones with central roles in metabolism and disease pathogen-
Preiss–Handler route also contributes to pyridine nucleotide esis. A protein with the same amino acid structure of NaPRT has
formation, transforming nicotinic acid absorbed from the gut into been identified as a 52-kDa release product of visceral fat and
therefore called visfatin [9,10]. Controversies exist as for the
endocrine properties of visfatin/NaPRT. Although originally
* Corresponding author at: Dept. Pharmacology, University of Florence, Viale
identified as an insulin receptor-interacting protein able to
Pieraccini 6, 50139 Firenze, Italy. Tel.: +39 055 4271230. improve glucose-stimulated insulin secretion, the protein is now
E-mail address: alberto.chiarugi@unifi.it (A. Chiarugi). thought to be devoid of insulin receptor-interacting properties but

0006-2952/$ – see front matter ß 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.bcp.2009.02.017
 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620 1613

Fig. 1. Derivatization of NMN. NMN is converted into a fluorescent compound upon reaction with acetophenone and formic acid.

still able to regulate glucose homeostasis by promoting pancreatic Bethesda, USA. Cell viability was evaluated by measuring lactate
islet insulin secretion. Of note, these properties are entirely dehydrogenase (LDH) release in the incubating media or reduc-
dependent on NMN production by secreted visfatin/NaPRT [11]. tion of methylthiazolyl tetrazolium (MTT) and described [17]. An
The secreted protein can also behave as a cytokine called Pre-B cell inverted Nikon TE-2000U microscope equipped with a CDD
colony-enhancing factor (PBEF). NaPRT/visfatin/PBEF is released camera was used for cell visualization.
by various immune cells, promotes TNFa, IL1b and IL-6 produc-
tion, and is increased in the sera of patients affected by sepsis, 2.2. Cell fractionation
acute lung injury, myocardial infarction, rheumatoid arthritis and
inflammatory bowel disease. The protein is also able to inhibit As previously described [17], cells were disrupted using a glass/
neutrophil apoptosis, thereby promoting the immune response glass homogenizer in 500 ml of buffer A (Tris HCl 50 mM, pH 7.4,
[12]. Whether NMN production contributes to the immunoregu- mannitol 225 mM, saccarose 75 mM, 1 mM PMSF, 10 ml of
latory functions of NaPRT/visfatin/PBEF is still unknown but a protease inhibitor cocktail), and centrifuged at 600  g to obtain
specific receptor has not been identified. Also, the permeability of the nuclear pellet. Supernatants were centrifuged at 12,000  g to
NMN through the plasma membrane is not clear and, again, a obtain the mitochondrial pellet and the cytosolic fraction. Cell
possible NMN-interacting receptor waits to be identified. fractions were then processed for NMN determination as described
Given the potential relevance of NaPRT/visfatin/PBEF to above.
pathophysiology, several ELISA kits able to detect the protein
have been recently made commercially available. Yet, these kits 2.3. NMN, NAD and ATP measurement
have different sensitivities and are heterogeneous in nature. These
inconsistencies are probably responsible for the apparent con- Ultra pure NMN standards (Sigma, Milan, Italy) were dissolved
trasting properties of secreted NaPRT/visfatin/PBEF present in the in water and solutions analyzed by HPLC with UV or fluorimetric
literature. Also, since the functional properties of NaPRT/visfatin/ detection. For UV detection, cells grown in a 48-well plate were
PBEF may in part depend on its enzymatic activity [11,13], scraped with 100 ml of HCl 0.6. The cell extract was centrifuged at
evaluation of intra as well as extracellular NMN production can 14,000  g/5 min and 25 ml of the supernatant injected in an HPLC
certainly help understanding the pathophysiological role of the system consisting in a mobile phase of 0.1 M buffer phosphate pH
protein. Unfortunately, analytical determination of NMN is 6.5, 1% acetonitrile, 10 mM tetra butyl-ammonium bromide
difficult because of its physicochemical features. Indeed, very (TBAB), a Supelco 25 cm column (5 mm) and an UV detector
few studies have measured NaPRT activity and/or NMN concen- (PerkinElmer) set at 260 nm. For fluorimetric detection, we
trations in biological fluids or tissue extracts using radioactive or modified a method previously described [18]. Briefly, cells grown
complex HPLC/MS techniques [11,13,14]. In the present study we in a 48-well plate were lysed with 100 ml of HClO4 1N, whereas
provide new information on NMN metabolism and pharmacolo- mouse organs were sonicated in HCLO4 1N (1:4, w/v). Then, 100 ml
gical modulation by using an original method of NMN measure- of the extract were neutralized with KOH 1N and, after 5 min,
ment as well as the newly identified inhibitor of NaPRT (E)-N-[4- additional 100 ml of 0.1 M bicine pH 7.4 were added. The cell
(1-benzoylpiperidin-4-yl) butyl]-3-(pyridin-3-yl) acrilamide extract was centrifuged at 14,000  g/5 min and 10 ml of the
(FK866) [15,16]. supernatant were mixed with 100 ml of KOH 1N and 50 ml of
acetophenone. The solution was incubated for 15 min at 4 8C, then
2. Materials and methods 100 ml formic acid were added and the solution incubated 5 min at
100 8C. By means of this derivatization procedure, NMN is
2.1. Cell culture and treatment converted into a highly fluorescent compound as shown in
Fig. 1. Excitation and emission spectra of the fluorescent
Unless otherwise stated, all chemicals and cell culture compound were determined by means of a spectro fluorophot-
products were from Sigma (Milan, Italy). HeLa or U937 cells were ometer RF5000 (Shimadzu, Milan, Italy). Samples were injected
cultured in Dulbecco’s modified Eagle’s medium (DMEM) into the HPLC system consisting in a mobile phase of 0.1 M buffer
supplemented with 2 mM glutamine, 10% fetal bovine serum phosphate pH 6.5, 10% acetonitrile, a Supelco 25 cm column
and antibiotics. Cultures were brought to 50–70% confluence and (5 mm) and a fluorimetric detector (PerkinElmer) with excitation
exposed to N-methyl-N0 -nitro-N-nitrosoguanidine (MNNG, and emission wavelength of 332 and 454 nm, respectively.
100 mM), nicotinamide 10–1000 mM, NMN 10–1000 mM, NAD NAD contents were quantified by means of an enzymatic
0.1–1 mM. (E)-N-[4-(1-benzoylpiperidin-4-yl) butyl]-3-(pyridin- cycling procedure as described [19]. ATP was measured by the
3-yl) acrilamide (FK866, 1-100 mM) was obtained from NIH, ATPlight kit (PerkinElmer, Milan).
1614 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620

Fig. 2. Optimization on HPLC analysis of NMN. (A) NAD biosynthetic pathways in mammals. NAD is synthesized de novo through the kynurenine pathway originating from
tryptophan. Intracellular NAD degrading enzymes such as PARPs, mono(ADP-ribosyl)transferases (MARTs) and sirtuins hydrolyze NAD forming different products and
nicotinamide (NAM). The later is utilized to re-synthesize NAD through the NAD-rescue pathway including NaPRT and NMNAT1-3. Nicotinic acid (NA) originating from food is
transformed into NAD through the Preiss–Handler pathway composed by NA phosphoribosyl transferase (NAPRT), NMNAT1-3 and NAD deamidase (NADASE). Nicotinamide
riboside kinase (NRK) converts nicotinamide riboside (NR) into NMN. The site of action of the NaPRT inhibitor FK866 is shown. (B), Chromatogram of NMN (250 pmol)
obtained with an HPLC apparatus coupled to an UV detector (see Section 2). (C), Chromatogram of a whole cell extract (U937 cells) injected into the same apparatus described
in (B). (D) Absorption and emission spectra of NMN derivatized as described in Section 2. (E), Chromatogram of derivatized NMN (250 pmol) obtained with an HPLC apparatus
coupled to a fluorimetric detector (see Section 2). (F) Chromatogram of a whole cell extract (U937 cells) derivatized as described in Section 2 and injected into the same
 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620 1615

2.4. In vivo experiments by NMNAT. Reportedly, NaPRT is the rate-limiting enzyme of the
mammalian NAD-rescue pathway [21]. When NMN was added to the
CD1 male albino mice (20–25 g) (n = 4 per group) were injected culture medium for 1 h, its intracellular contents increased in both
i.p. with FK866 (100 mg/kg) and sacrificed 4, 8 and 16 h later. cells types (Fig. 3C), indicating that the nucleotide readily permeates
Organs were rapidly collected and processed as described above the plasma membrane. Accordingly, NAD contents increased in both
for NMN content determination. Procedures involving animals and HeLa and U937 cells exposed to NMN (Fig. 3D). We also measured the
their care were conducted in compliance with the Italian guide- subcellular distribution of NMN in these cell lines. As shown in Fig. 2E,
lines for animal care (DL 116/92) in application of the European NMN was significantly more concentrated in the mitochondrial
Communities Council Directive (86/609/EEC) and was formally fraction than in the cytosolic or nuclear ones. However, when
approved by the Animal Care Committee of the Department of subcellular NMN levels were expressed as % of total cellular content,
Pharmacology of the University of Florence. the nucleotide amount appeared higher in the nuclear fraction
(Fig. 3F).
3. Results
3.3. Pharmacological modulation of intracellular NMN contents
3.1. Spectrofluorometric HPLC analysis of NMN
It has been repeatedly reported that, in condition of massive
Isocratic HPLC analysis of NMN with UV absorbance detection is DNA damage, hyperactivation of the NAD-consuming enzyme
virtually impossible because of its extremely short elution time. poly(ADP-ribose) polymerase (PARP)-1 leads to depletion of the
Indeed, under our experimental conditions (see Section 2), standard nucleotide pool and cell death [22]. The cell’s capability to face
NMN had a retention time of 1 min 54 s (Fig. 2B). Addition of tetra PARP-1-dependent NAD depletion by activating the NAD-rescue
butyl-ammonium bromide (up to 10 mM) to the mobile phase to pathway is therefore central to cell survival. Yet, how NMN
increase NMN retention only minimally influenced this parameter concentrations vary when PARP-1 hyperactivates is still
(not shown). We were unable to identify an unequivocal peak with a unknown. As shown in Fig. 4A and B, exposure of U937 cells
retention time corresponding to NMN in a whole cellular extract to the prototypical PARP-1 activator MNNG [23–25] only
from different cell lines (Fig. 2C and not shown). marginally increased NMN contents despite massive NAD
It is well known that N-alkylpyridinium compounds are depletion. The increase of NMN and depletion of NAD were
transformed into fluorescent compounds through reactions with prevented by the PARP-1 inhibitor phenanthridinone (PHE).
ketone moieties followed by heating in acidic environment. These findings suggested that the rate of PARP-1-dependent NAD
Accordingly, methyl-nicotinamide and NAD have been quantified utilization exceeded that of NMNAT-mediated NAD resynthesis.
spectrofluorometrically for different purposes [18,20]. We there- However, the possibility of a compartmentalized pool of NMN
fore attempted to adapt this method to NMN measurement (see not readily convertible into NAD must also be considered.
Section 2). Upon derivatization, NMN became fluorescent with Additionally, massive ATP depletion that typically follows PARP-
excitation and emission maxima at 332 and 454 nm, respectively 1 activation could differently affect the activity of the ATP-
(Fig. 2D). Of note, elution time of derivatized NMN standard was dependent enzymes NaPRT and NMNAT. Gallotannin has been
increased up to 4 min 18 s (Fig. 2E), and a single peak with the reported to inhibit NMNAT [26]. In keeping with this finding, we
identical elution time was present in a whole cellular extract found that NMN contents increased in cells exposed to
derivatized as described above (Fig. 2F). The peak area linearly gallotannin (Fig. 4C). Evaluating the effect of the latter in cells
increased by spiking the extract with different amounts on NMN undergoing PARP-1 hyperactivation indicated that the gallotan-
(not shown). Optimization of the derivatization procedure showed nin- and PARP-1 activation-dependent increases of NMN were
that NMN fluorescence increased linearly by augmenting the additive (Fig. 4C). FK866 is a recently identified inhibitor of
temperature of step 2 up to 100 8C (Fig. 2G). At this temperature, NaPRT [15,16]. It is unknown, however, whether FK866 also
incubations longer than 5 min reduced fluorescence (Fig. 2H). inhibits NMNAT. We report here that FK866 similarly reduced
Maximal efficiency of step 1 was reached with derivatization times both NAD and NMN contents in U937 cells (Fig. 4D and E),
of 15 min or longer. NMN fluorescence increased linearly over the confirming its ability to inhibit NaPRT activity. However,
range of 8–2500 pmol with a threshold sensitivity of 750 fmol increases in NAD contents prompted by the addition of NMN
(Fig. 2I). to the culture medium were proportionally similar in the
presence or absence of FK866 (Fig. 4F), indicating that the drug
3.2. Modulation of NMN content in HeLa and U937 cells does not affect NMNAT activity.

Basal contents of NMN in U937 cells were 3.7  0.1 nmol/mg 3.4. Effect of NMN or NAD on PARP-1-dependent cell death
prot, and about 10-fold lower in HeLa cells (0.48  0.02 nmol/mg
prot). These findings suggested low basal activity of NaPRT in HeLa It has been reported that strategies aimed at increasing
cells. We therefore attempted to evaluate in intact cells the intracellular NAD contents provide cytoprotection under different
contribution of NaPRT to NMN contents by adding nicotinamide to pathological conditions (see Refs. [2,3,21,27] for reviews). Hence,
the cell medium. As shown in Fig. 3A, nicotinamide increased NMN to evaluate the potential cytoprotective properties of NMN, we
contents in a concentration-dependent manner in U937 but not in analyzed various cell death parameters of HeLa and U937 cells
HeLa cells, again suggesting low expression levels/activity of NaPRT in undergoing NAD and ATP depletion because of PARP-1 activation.
this cell type. In contrast with this interpretation, however, addition As shown in Fig. 5A and B, both NMN and NAD were able to
of nicotinamide to the culture medium similarly increased NAD counteract the reduction in NAD and ATP contents in HeLa and U937
contents in both cell types. Taken together, these findings suggested cells exposed 1 h to the PARP-1-activating agent MNNG. Twenty-
that NaPRT is active in both U937 and HeLa cells, and that NMN does three hours later, MNNG-challenged cells appeared shrunken and
not accumulate in HeLa because of its rapid transformation into NAD joined in clusters of dead cells. Conversely, those exposed to MNNG

apparatus described in (E). (G) Effect of the temperature of the step 2 of the derivatization procedure on efficiency of NMN derivatization. (H) Effect of duration of step 1 or 2 of
the derivatization procedure on efficiency of NMN derivatization. (I) NMN calibration curve. Fluorescence of NMN is linear up to 2500 pmol injected. In G-I, bars/points
represent the mean  S.E.M. of three experiments conducted in duplicate.
1616 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620

Fig. 3. NMN and NAD contents in HeLa and U937 cells. NMN and NAD contents were measured in HeLa and U937 cells under control conditions or after exposure for 1 h to
different concentrations of nicotinamide (Nam) (A and B) or NMN (C and D). (E and F) Contents of NMN in cytosol, nuclei and mitochondria of HeLa or U937 cells. *p < 0.05,
**p < 0.01 vs. Cytosol (ANOVA + Tukey’s post hoc test). In A–E, each point/bar represents the mean  S.E.M. of three experiments conducted in duplicate.

plus NMN or NAD in part conserved their healthy morphology high in liver, intermediate in heart, and under detectable levels in
(Fig. 5C). Accordingly, MTT reduction and LDH release assays brain [11]. Accordingly, we found that NMN contents were higher
showed that both NMN and NAD partially counteracted cell death in the liver than in the heart (6.7  0.3 and 4.4  0.2 nmol/mg tissue,
24 h after PARP-1 hyperactivation (Fig. 5D and E). NMN- or NAD- respectively). However, the brain contents of NMN (3.86  0.1 nmol/
treated cells died 36 h after MNNG exposure (not shown), indicating mg tissue) were similar to those of the heart. Importantly, injection of
that nucleotide cytoprotection was only transient. FK866 reduced NMN contents in the three organs after 4 and 8 h. At
16 h from the injection, NMN contents in the organs were still lower
3.5. NMN content in mouse organs and effect of FK866 than control with a tendency to return to basal values (Fig. 6).

To our knowledge, the impact of FK866 on NMN contents in vivo 4. Discussion

has not been reported. We therefore injected the drug in mice
(100 mg/kg, i.p.) and measured NMN contents in various organs at In the last several years we have witnessed a renewed interest
different times after injection. Reportedly, mouse NaPRT activity is in NAD biology and basic biochemistry of pyridine nucleotide
 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620 1617

Fig. 4. Effect of PARP-1 activation, gallotannin or FK866 on NMN or NAD contents in U937 cells. The NMN (A) or NAD (B) contents were measured in cells under control
conditions and at different times after exposure to the PARP-1-activating agent MNNG (100 mM) in the presence or absence of the PARP-1 inhibitor phenanthridinone (PHE,
30 mM). (C) The contents of NMN were measured in cells exposed for 1 h to different concentrations of gallotannin (GLT) in the presence or absence of MNNG 100 mM/1 h.
Effect of a 4-h exposure to FK866 on intracellular NMN (D) or NAD (E) contents. (F) NAD contents in cells exposed 1 h to different concentrations of NMN and preincubated 4 h
with FK866 (10 mM). Each bar represents the mean  S.E.M. of at least three experiments conducted in duplicate. *p < 0.05, **p < 0.01 vs. Crl (ANOVA + Tukey’s post hoc test). In
(F) §p < 0.05 vs. Crl in the presence of FK866 (ANOVA + Tukey’s post hoc test).

metabolism [1]. Among the biochemical routes leading to NAD limiting step of the rescue pathway [29,30]. Also, evidence that
formation, the rescue pathway operated by NaPRT and NMNATs is intracellular NMN contents promptly increase when the nucleo-
of pivotal importance to NAD homeostasis in mammals. Indeed, tide is added to the culture media indicates that plasma membrane
this pathway rescues nicotinamide which is a by-product of is permeable to this nucleotide. These findings suggest that the
several enzymes such as poly and mono-ADP-ribose transferases, pharmacologic effects of exogenous NMN in cultured cells and
sirtuins as well as ADP-ribosyl cyclase, which are responsible for mice [11,13] are due to cellular uptake and changes in NAD
substantial consumption of NAD under constitutive and patholo- contents. Whether NMN permeates the plasma membrane
gical conditions. However, the intracellular contents as well as through the putative NAD uptake mechanisms [31] remains to
subcellular compartmentalization of NMN, the first metabolite of be clarified. We also report that NMN was more concentrated in
the nicotinamide-dependent NAD-rescue pathway, are in large the mitochondrial fraction than in the nuclear or cytosolic ones.
part unknown. Furthermore, whether NMN formation mediates Although we adopted a crude cell fractionation technique with
the signalling properties of extracellular NaPRT/visfatin/PBEF possible subcellular fraction contamination, data suggest high
waits to be clearly understood. Current lack of knowledge about mitochondrial NMN contents. This finding is of particular
the biological relevance of NMN can be ascribed, at least in part, to relevance if we consider that it is still debated whether NaPRT
difficulties in nucleotide measurements in biological samples. activity is present in the mitochondria. NaPRT can be detected by
Previous studies have quantified NMN by means of complex HPLC Western blotting in mitochondria extracted from cultured cells or
systems coupled to UV detection and subsequent mass spectro- mouse liver, and the NaPRT inhibitor FK866 reduces mitochondrial
metry analysis [11,13,28]. Indirect methods or radioactive NAD content when added to a pure organelle preparation [28].
precursors have been also adopted to follow NMN formation Overall, these findings suggest that NMN is synthesized in
and degradation [6]. In the present study, we report a new method mitochondria by NaPRT. Although the presence of a mitochondrial
to quantify NMN in biological samples, based on the ability of its N- NaPRT is conceivable if we consider that mitochondria contain a
alkylpyridinium moiety to be transformed into fluorescent specific NMNAT isoform [26], how NaPRT enters the organelles is
compounds. unknown given that a mitochondrial-targeting sequence has not
By monitoring the effect of nicotinamide addition to the been reported [16]. In light of the recent identification of
incubating media of HeLa or U937 cells on their NMN and NAD mitochondrial NAD carriers [28,32], the possibility exists that
content, we confirm that, at least in HeLa cells, NaPRT is the rate- part of the mitochondrial NMN pool is uptaken from the
1618 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620

Fig. 5. Cytoprotection form PARP-1-dependent cell death by NMN or NAD. Effect of NMN or NAD added to the incubating media of HeLa and U937 cells (1 h preincubation) on
reduction of NAD (A) or ATP (B) contents prompted by 1 h exposure to 100 mM MNNG. (C) Phase contrast visualization of the two cell types under control conditions or 11 h
after 1 h exposure to MNNG in the presence or absence of NMN or NAD, both at 1 mM. The two nucleotides were present during and after MNNG exposure. Effect of NMN or
NAD added to the incubating media of HeLa and U937 cells on MTT reduction (D) or LDH release (E) evaluated 11 h after 1 h exposure to 100 mM MNNG. The two nucleotides
were present during and after MNNG exposure. Each bar represents the mean  S.E.M. of at least three experiments conducted in duplicate. In (C) representative images of three
experiments are shown. *p < 0.05, **p < 0.01 vs. MNNG (ANOVA + Tukey’s post hoc test). In (C), bar = 40 or 20 nm for HeLa and U937, respectively.

cytoplasm. Regardless, evidence that mitochondria contain sig- occurred upon PARP-1 activation (Fig. 4A). This finding suggests
nificant amounts of NMN (Fig. 3E), as well as NMNAT3 [26], clearly that NMNAT becomes limiting the rate of the NAD-rescue pathway
indicates that at least the last step of the NAD-rescue pathway when the dinucleotide is massively consumed by PARP-1. Also, it
occurs in these organelles. This makes sense if one consider that indicates that strategies aimed at circumventing the ‘‘NMNAT
NAD-consuming enzymes such as sirtuins and mono-ADP-ribose bottleneck’’ could be of cytoprotective relevance in conditions of
transferases are present in mitochondria [21,33]. Conversely, the hyperactivation of PARP-1. In good agreement with this hypothesis
presence of a mitochondrial PARP-1 is still debated [17,34–36]. and previous findings [37], we report that exogenous NAD reduced
A large body of evidence identifies PARP-1 as a key player of cell early PARP-1-dependent bioenergetic derangement and cell death.
death. It has been repeatedly proposed that the molecular pathway On the one hand this result further corroborates the hypotheses
responsible for PARP-1-dependent cell death stems from an that NAD depletion is causative in this type of cell demise [38], and
excessive consumption and ensuing depletion of NAD pools on the other that NAD permeates through the plasma membrane
[22]. It is unknown, however, whether/how PARP-1 hyperactiva- [31]. Of note, NMN also reduced PARP-1-dependent energy failure
tion affects the kinetics of the NAD-rescue pathway. We show here and cell death, despite to a lower extent when compared to NAD.
that the massive depletion of intracellular NAD that follows PARP- This is in keeping with the NMNAT rate-limiting activity for NAD
1 activation is not accompanied by a concomitant depletion of resynthesis during overactivation of PARP-1. We report, however,
NMN. Rather, a small but significant increase in NMN content that protection afforded by NMN or NAD on cells undergoing PARP-
 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620 1619

Acknowledgments

This study was supported by grants from the University of

Florence, the Italian Ministry of University and Scientific and
Technological Research, Associazione Italiana Sclerosi Multipla
and Ente Cassa di Risparmio di Firenze.

References

[1] Berger F, Ramirez-Hernandez MH, Ziegler M. The new life of a centenarian:

signalling functions of NAD(P). Trends Biochem Sci 2004;29:111–8.
[2] Belenky P, Bogan KL, Brenner C. NAD+ metabolism in health and disease.
Trends Biochem Sci 2007;32:12–9.
[3] Magni G, Orsomando G, Raffelli N, Ruggieri S. Enzymology of mammalian NAD
metabolism in health and disease. Front Biosci 2008;13:6135–54.
[4] Magni G, Amici A, Emanuelli M, Orsomando G, Raffaelli N, Ruggieri S. Enzy-
Fig. 6. Effect of FK866 on NMN contents in different mouse organs. NMN was mology of NAD+ homeostasis in man. Cell Mol Life Sci 2004;61:19–34.
measured in different mouse organs under control conditions (C) and at different [5] Brenner C. Evolution of NAD biosynthetic enzymes. Structure 2005;13:1239–
times after i.p. injection of FK866 (100 mg/kg). Each bar represents the mean of two 40.
experiments (n = 4 per group). *p < 0.05 vs. C (ANOVA + Tukey’s post hoc test). [6] Bieganowski P, Brenner C. Discoveries of nicotinamide riboside as a nutrient
and conserved NRK genes establish a Preiss–Handler independent route to
NAD+ in fungi and humans. Cell 2004;117:495–502.
1 hyperactivation is only transient. This indicates that the partial [7] Bogan KL, Brenner C. Nicotinic acid, nicotinamide, and nicotinamide riboside: a
prevention of NAD/ATP depletion is not compatible with cell molecular evaluation of NAD+ precursor vitamins in human nutrition. Annu
survival, and/or that delayed events in addition to nucleotide Rev Nutr 2008;28:115–30.
[8] Yang H, Lavu S, Sinclair DA. Nampt/PBEF/Visfatin: a regulator of mammalian
depletion [i.e. poly(ADP-ribose)-dependent mitochondrial AIF health and longevity? Exp Gerontol 2006;41:718–26.
release [39]] contribute to cell demise prompted by PARP-1. [9] Fukuhara A, Matsuda M, Nishizawa M, Segawa K, Tanaka M, Kishimoto K, et al.
Indeed, NMN can improve energy dynamics working as a precursor Visfatin: a protein secreted by visceral fat that mimics the effects of insulin.
Science 2005;307:426–30.
of NAD but also favour poly(ADP-ribose) formation by the
[10] Fukuhara A, Matsuda M, Nishizawa M, Segawa K, Tanaka M, Kishimoto K, et al.
detrimental NMNAT–PARP-1 complex [40]. Retraction. Science 2007;318:565.
Despite the emerging complexity and pathophysiological [11] Revollo JR, Korner A, Mills KF, Satoh A, Wang T, Garten A, et al. Nampt/PBEF/
visfatin regulates insulin secretion in beta cells as a systemic NAD biosynthetic
relevance of NAD metabolism, pharmacology of NAD resynthesis
enzyme. Cell Metab 2007;6:363–75.
is extremely poor, with FK866 being the most studied chemical. [12] Luk T, Malam Z, Marshall JC. Pre-B cell colony-enhancing factor (PBEF)/
Evidence that gallotannin increases intracellular NMN contents visfatin: a novel mediator of innate immunity. J Leukoc Biol 2008;83:804–
(Fig. 4C) is in keeping with prior work showing that gallotannin 16.
[13] Wang P, Xu TY, Guan YF, Su DF, Fan GR, Miao CY. Perivascular adipose tissue-
inhibits NMNAT activity [26]. Although gallotannin is a complex derived visfatin is a vascular smooth muscle cell growth factor: role of
mixture of tannins extracted from oak gall [41], these findings nicotinamide mononucleotide. Cardiovasc Res 2008.
could be of relevance for the identification of the pharmaco- [14] Rongvaux A, Shea RJ, Mulks MH, Gigot D, Urbain J, Leo O, et al. Pre-B-cell
colony-enhancing factor, whose expression is up-regulated in activated lym-
phore of more potent and specific NMNAT inhibitors. Further, phocytes, is a nicotinamide phosphoribosyltransferase, a cytosolic enzyme
the demonstration that FK866 reduced basal NMN and NAD involved in NAD biosynthesis. Eur J Immunol 2002;32:3225–34.
contents in cultured cells, but did not prevent the increase of [15] Hasmann M, Schemainda I. FK866, a highly specific noncompetitive inhibitor
of nicotinamide phosphoribosyltransferase, represents a novel mechanism for
NAD in cells exposed to NMN (Fig. 4D–F), indicates for the first induction of tumor cell apoptosis. Cancer Res 2003;63:7436–42.
time that the drug selectively inhibits NaPRT, having no effects [16] Khan JA, Tao X, Tong L. Molecular basis for the inhibition of human NMPRTase,
on NMNAT activity. This is of significance if one considers that a novel target for anticancer agents. Nat Struct Mol Biol 2006;13:582–8.
[17] Cipriani G, Rapizzi E, Vannacci A, Rizzuto R, Moroni F, Chiarugi A. Nuclear
the chemical is currently in clinical trial for cancer treatment poly(ADP-ribose) polymerase-1 rapidly triggers mitochondrial dysfunction. J
[42]. Biol Chem 2005;280:17227–34.
In keeping with data on tissue distribution of NaPRT expression [18] Putt KS, Hergenrother PJ. An enzymatic assay for poly(ADP-ribose) polymer-
ase-1 (PARP-1) via the chemical quantitation of NAD(+): application to the
[11] and activity [43], we report that NMN is more concentrated in
high-throughput screening of small molecules as potential inhibitors. Anal
the mouse liver than in the heart or brain. Notably, NMN contents Biochem 2004;326:78–86.
in the brain are comparable to those of the heart, despite evidence [19] Shah GM, Poirier D, Duchaine G, Desnoyers S, Brochu G, Lageaux J, et al.
that NaPRT expression is almost undetectable in the brain [11]. Methods for biochemical study of poly(ADP-ribose) metabolism in vitro and in
vivo. Anal Biochem 1995;227:1–13.
Although we do not know the cell type(s) in which NMN is [20] Musfeld C, Biollaz J, Belaz N, Kesselring UW, Decosterd LA. Validation of an
contained, data suggest active NMN uptake from neural cells. HPLC method for the determination of urinary and plasma levels of N1-
Treatment of mice with FK866 induced long-lasting but mild methylnicotinamide, an endogenous marker of renal cationic transport and
plasma flow. J Pharm Biomed Anal 2001;24:391–404.
reduction of NAD contents in liver, heart and brain. Obviously, mild [21] Sauve AA. NAD+ and vitamin B3: from metabolism to therapies. J Pharmacol
reduction can be primarily ascribed to the pharmacodynamic and Exp Ther 2008;324:883–93.
pharmacokinetic features of FK866. Indeed, metabolism of FK866 [22] Chiarugi A. PARP-1: killer or conspirator? The suicide hypothesis revisited.
Trends Pharmacol Sci 2002;23:122–9.
is still to be clearly understood. Data, however, hint that under [23] D’Amours D, Desnoyers S, Poirier GG. Poly(ADP-ribosyl)ation reactions in the
constitutive conditions NAD consumption and/or recycling regulation of nuclear functions. Biochem J 1999;342:249–68.
through nicotinamide occurs at a slow rate. This hypothesis is [24] Ha HC, Snyder SH. Poly(ADP-ribose) polymerase is a mediator of necrotic cell
death by ATP depletion. Proc Natl Acad Sci 1999;69:13978–82.
consistent with knowledge that NAD half-life in rat liver is about [25] Yu SW, Poitras MF, Coombs C, Bowers WJ, Federoff HJ, Poirier GG, et al.
10 h [44]. Mediation of poly(ADP-ribose) polymerase-1-dependent cell death by apop-
In conclusion, the development of a new method for NMM tosis-inducing factor. Science 2002;297:259–63.
[26] Berger F, Lau C, Dahlmann M, Ziegler M. Subcellular compartmentation and
measurement allowed us to gather new information on the
differential catalytic properties of the three human nicotinamide mononu-
biochemistry and pharmacology of NAD biosynthesis. Given that cleotide adenylyltransferase isoforms. J Biol Chem 2005;280:36334–41.
the pathophysiological relevance of NaPRT and NMN is still in large [27] Porcu M, Chiarugi A. The emerging therapeutic potential of sirtuin-interacting
part obscure, the present study helps addressing some of the drugs: from cell death to lifespan extension. Trends Pharmacol Sci 2005;26:
94–103.
numerous unanswered questions existing in the field of pyridine [28] Yang H, Yang T, Baur JA, Perez E, Matsui T, Carmona JJ, et al. Nutrient-sensitive
nucleotide metabolism and NAD signalling. mitochondrial NAD+ levels dictate cell survival. Cell 2007;130:1095–107.
1620 L. Formentini et al. / Biochemical Pharmacology 77 (2009) 1612–1620

[29] Revollo JR, Grimm AA, Imai SI. The NAD biosynthesis pathway mediated by [37] Ying W, Garnier P, Swanson RA. NAD+ repletion prevents PARP-1-induced
nicotinamide phosphoribosyltransferase regulates Sir2 activity in mammalian glycolytic blockade and cell death in cultured mouse astrocytes. Biochem
cells. J Biol Chem 2004. Biophys Res Commun 2003;308:809–13.
[30] Pillai JB, Isbatan A, Imai S, Gupta MP. Poly(ADP-ribose) polymerase-1-depen- [38] Berger NA. Poly (ADP-ribose) in the cellular response to DNA damage. Radiat
dent cardiac myocyte cell death during heart failure is mediated by NAD+ Res 1985;101:4–15.
depletion and reduced Sir2alpha deacetylase activity. J Biol Chem 2005;280: [39] Yu SW, Andrabi SA, Wang H, Kim NS, Poirier GG, Dawson TM, et al. Apoptosis-
43121–30. inducing factor mediates poly(ADP-ribose) (PAR) polymer-induced cell death.
[31] Billington RA, Travelli C, Ercolano E, Galli U, Roman CB, Grolla AA, et al. Proc Natl Acad Sci USA 2006;103:18314–9.
Characterization of NAD uptake in mammalian cells. J Biol Chem 2008;283: [40] Berger F, Lau C, Ziegler M. Regulation of poly(ADP-ribose) polymerase 1
6367–74. activity by the phosphorylation state of the nuclear NAD biosynthetic enzyme
[32] Todisco S, Agrimi G, Castegna A, Palmieri F. Identification of the mitochondrial NMN adenylyl transferase 1. Proc Natl Acad Sci 2007;104. 3765-3760.
NAD+ transporter in Saccharomyces cerevisiae. J Biol Chem 2006;281:1524– [41] Formentini L, Arapistas P, Pittelli M, Jacomelli M, Pitozzi V, Menichetti S, et al.
31. Mono-galloyl glucose derivatives are potent poly(ADP-ribose) glycohydrolase
[33] Richter C, Winterhalter KH, Baumhuter S, Lotscher HR, Moser B. ADP-ribosyla- (PARG) inhibitors and partially reduce PARP-1-dependent cell death. Br J
tion in inner membrane of rat liver mitochondria. Proc Natl Acad Sci USA Pharmacol 2008.
1983;80:3188–92. [42] Holen K, Saltz LB, Hollywood E, Burk K, Hanauske AR. The pharmacokinetics,
[34] Du L, Zhang X, Han YY, Burke NA, Kochanek PM, Watkins SC, et al. Intra- toxicities, and biologic effects of FK866, a nicotinamide adenine dinucleotide
mitochondrial poly-ADP-ribosylation contributes to NAD+ depletion and cell biosynthesis inhibitor. Invest New Drugs 2008;26:45–51.
death induced by oxidative stress. J Biol Chem 2003;278:18426–33. [43] Dietrich LS, Fuller L, Yero IL, Martinez L. Nicotinamide mononucleotide
[35] Scovassi AI. Mitochondrial poly(ADP-ribosylation): from old data to new pyrophosphorylase activity in animal tissues. J Biol Chem 1966;241:188–91.
perspectives. FASEB J 2004;18:1487–8. [44] Ijichi H, Ichiyama A, Hayaishi O. Studies on the biosynthesis of nicotinamide
[36] Niere M, Kernstock S, Koch-Nolte F, Ziegler M. Functional localization of two adenine dinucleotide. 3. Comparative in vivo studies on nicotinic acid, nico-
poly(ADP-ribose)-degrading enzymes to the mitochondrial matrix. Mol Cell tinamide, and quinolinic acid as precursors of nicotinamide adenine dinucleo-
Biol 2008;28:814–24. tide. J Biol Chem 1966;241:3701–7.