Cell Metabolism

Article

The NAD+ Precursor Nicotinamide Riboside

Enhances Oxidative Metabolism
and Protects against High-Fat Diet-Induced Obesity
Carles Cantó,1,6,7 Riekelt H. Houtkooper,1,6,8 Eija Pirinen,1,2 Dou Y. Youn,3 Maaike H. Oosterveer,1 Yana Cen,3
Pablo J. Fernandez-Marcos,1 Hiroyasu Yamamoto,1 Pénélope A. Andreux,1 Philippe Cettour-Rose,1 Karl Gademann,4
Chris Rinsch,5 Kristina Schoonjans,1 Anthony A. Sauve,3 and Johan Auwerx1,*
1ÉcolePolytechnique Fédérale de Lausanne, CH-1015 Lausanne, Switzerland
2Biotechnology and Molecular Medicine, A.I. Virtanen Institute for Molecular Sciences, Biocenter Kuopio, University of Eastern Finland,
70210 Kuopio, Finland
3Department of Pharmacology, Weill Cornell Medical College, New York, NY 10065, USA
4Department of Chemistry, University of Basel, CH-4056 Basel, Switzerland
5Amazentis, Quartier de L’innovation EPFL, CH-1015 Lausanne, Switzerland
6These authors contributed equally to this work
7Present address: Nestlé Institute of Health Sciences, CH-1015 Lausanne, Switzerland
8Present address: Laboratory Genetic Metabolic Diseases, Academic Medical Center, 1105 AZ Amsterdam, The Netherlands

*Correspondence: admin.auwerx@epfl.ch
DOI 10.1016/j.cmet.2012.04.022

SUMMARY the GPR109A receptor (Benyo et al., 2005). We hence became

interested in the possible therapeutic use of alternative NAD+
As NAD+ is a rate-limiting cosubstrate for the sirtuin precursors that do not activate GPR109A.
enzymes, its modulation is emerging as a valuable NR was recently identified as a NAD+ precursor, with
tool to regulate sirtuin function and, consequently, conserved metabolism from yeast to mammals (Bieganowski
oxidative metabolism. In line with this premise, and Brenner, 2004). Importantly, NR is found in milk (Bieganow-
decreased activity of PARP-1 or CD38—both NAD+ ski and Brenner, 2004), constituting a dietary source for NAD+
production. Once it enters the cell, NR is metabolized into nico-
consumers—increases NAD+ bioavailability, result-
tinamide mononucleotide (NMN) by a phosphorylation step cata-
ing in SIRT1 activation and protection against meta-
lyzed by the nicotinamide riboside kinases (NRKs) (Bieganowski
bolic disease. Here we evaluated whether similar and Brenner, 2004). In contrast to NR, NMN has not yet been
effects could be achieved by increasing the supply found in dietary constituents, and its presence in serum is
of nicotinamide riboside (NR), a recently described a matter of debate (Hara et al., 2011; Revollo et al., 2007). This
natural NAD+ precursor with the ability to increase highlights how NR might be an important vehicular form of
NAD+ levels, Sir2-dependent gene silencing, and a NAD+ precursor whose levels could be modulated through
replicative life span in yeast. We show that NR nutrition.
supplementation in mammalian cells and mouse The sirtuins are a family of enzymes that use NAD+ as
tissues increases NAD+ levels and activates SIRT1 a cosubstrate to catalyze the deacetylation and/or mono-ADP-
and SIRT3, culminating in enhanced oxidative ribosylation of target proteins. One of their major particularities
is that their Km for NAD+ is relatively high, making NAD+ a rate-
metabolism and protection against high-fat diet-
limiting substrate for their reaction (Canto and Auwerx, 2012).
induced metabolic abnormalities. Consequently,
Initial work by yeast biologists indicated that the activity of Sir2
our results indicate that the natural vitamin NR could (the yeast SIRT1 homolog) as a NAD+-coupled enzyme could
be used as a nutritional supplement to ameliorate provide a link between metabolism and gene silencing (Imai
metabolic and age-related disorders characterized et al., 2000a; Imai et al., 2000b). In this way, Sir2 was proposed
by defective mitochondrial function. to mediate metabolic transcriptional adaptations linked to situa-
tions of nutrient scarcity, which are generally coupled to
increased NAD+ levels (for review, see Houtkooper et al.,
INTRODUCTION 2010). During the last decade, an overwhelming body of
evidence indicates that the activity of mammalian sirtuins,
The administration of NAD+ precursors, mostly in the form of most notably SIRT1 and SIRT3, have the ability to enhance fat
nicotinic acid (NA), has long been known to promote beneficial oxidation and prevent metabolic disease (Hirschey et al., 2011;
effects on blood lipid and cholesterol profiles and even to induce Lagouge et al., 2006; Pfluger et al., 2008). Therefore, strategies
short-term improvement of type 2 diabetes (Karpe and Frayn, aimed to increase intracellular NAD+ levels have gained interest
2004). Unfortunately, NA treatment often leads to severe flush- in order to activate sirtuins and battle metabolic damage. Valida-
ing, resulting in poor patient compliance (Bogan and Brenner, tion of this concept was achieved recently, by demonstrating
2008). These side effects are mediated by the binding of NA to that pharmacological and genetic approaches aimed to reduce

838 Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc.
Cell Metabolism
NR Increases NAD+ and Prevents Metabolic Disorders

NAD+ content NAD+ content

A NAD+ content B GPR109A-induced C (mmol / kg of prot) D (mmol / kg of prot)
(mmol / kg of prot) calcium mobilization
1,2 4000 (RLU – increase over basal) 1,5 Veh NR 0,75
NMN NA
3000
0,9 1 * * 0,5 *
NR * * * * **
2000 * * * *
NMN
0,6 0,5 0,25 *
C2C12 1000 NA
Hepa1.6
HEK293T
0,3 0 0 0
0 200 400 600 800 1000 10-9 10-8 10-7 10-6 10-5 10-4 10-3 C2C12 Hepa1.6 HEK293T Liver Muscle
Nicotinamide Riboside (µM) Concentration (M)

E F G H

Nampt mRNA levels

(relative to Veh)
1,5
1
3 2 *

PAR
liver mitochondria

0,5
NAD+ content in

(mmol / kg prot)
(mmol / kg prot)

*
Mitochondrial

1,5
NAD+ levels

2 0
1
Nampt
1
0,5
Tubulin PARP-1
0 0
Veh NR Veh NR

Figure 1. Nicotinamide Riboside Supplementation Increases NAD+ Content and Sirtuin Activity in Cultured Mammalian Cells
(A) C2C12 myotubes, Hepa1.6, and HEK293T cells were treated with nicotinamide riboside (NR) for 24 hr, and acidic extracts were obtained to measure total
NAD+ intracellular content.
(B) GPR109A-expressing Chem-4 cells were loaded with 3 mM Fura-2 acetoxymethyl ester derivative (Fura-2/AM) for 30 min at 37
C. Then cells were washed
three times with Hank’s balanced salt solution. Finally, calcium fluxes in response to increasing concentrations of nicotinic acid (NA; as positive control), NR, or
nicotinamide mononucleotide (NMN) were determined as indicated in the Experimental Procedures.
(C) C2C12 myotubes, Hepa1.6, and HEK293T cells were treated with either PBS (as vehicle) or 0.5 mM of NR, NMN, or NA for 24 hr. Then total NAD+ intracellular
content was determined as in (A).
(D) C57Bl/6J mice were fed with chow containing vehicle (water) or either NR, NMN,or NA at 400 mg/kg/day (n = 8 mice per group). After 1 week, NAD+ content
was determined in liver and quadriceps muscle.
(E) HEK293T cells were treated with NR (0.5 mM, black bars) or vehicle (white bars) for 4 hr. Then cells were harvested, and mitochondria were isolated for NAD+
measurement.
(F) C57Bl/6J mice were fed with chow containing vehicle (water) or NR at 400 mg/kg/day (n = 8 mice per group). After 1 week, mitochondria were isolated from
their livers to measure NAD+ content.
(G) HEK293T cells were treated with either PBS (as vehicle) or 0.5 mM of NR for 24 hr. Then mRNA and protein was extracted to measure Nampt levels by
RT-qPCR and western blot, respectively.
(H) HEK293T cells were treated with either PBS (as vehicle) or 0.5 mM of NR for 24 hr. Then protein homogenates were obtained to test global PARylation and
PARP-1 levels. Throughout the figure, all values are presented as mean ±SD. Asterisk indicates statistical significant difference versus respective vehicle group at
p < 0.05. Unless otherwise stated, the vehicle groups are represented by white bars, and NR groups are represented by black bars.

the activity of major NAD+-consuming activities in the cell, such effects at concentrations between 0.5 and 1 mM. In C2C12 my-
as PARP-1 (Bai et al., 2011b) and CD38 (Barbosa et al., 2007), otubes, the Km for NR uptake was 172.3 ± 17.6 mM, with a Vmax of
prompted an increase in NAD+ bioavailability and enhanced 204.2 ± 20.5 pmol/mg of protein/min. Unlike NA, both NR and
SIRT1 activity, ultimately leading to effective protection against another well-described NAD+ precursor, NMN (Revollo et al.,
metabolic disease. In this work we aimed to test whether similar 2007), did not activate GPR109A (Figure 1B), hence constituting
effects could be achieved through dietary supplementation with valuable candidates to increase NAD+ levels without activating
a natural NAD+ precursor, such as NR. GPR109A. Strikingly, the ability of NR to increase intracellular
NAD+ in mammalian cells was, at least, similar to that of these
RESULTS other precursors (Figure 1C). We next evaluated the efficacy of
NR, NMN, and NA to increase NAD+ in vivo by supplementing
NR Increases Intracellular and Mitochondrial NAD+ mouse chow with NR, NMN, or NA at 400 mg/kg/day for
Content in Mammalian Cells and Tissues 1 week. All compounds increased NAD+ levels in liver, but only
NR treatment dose-dependently increased intracellular NAD+ NR and NA significantly enhanced muscle NAD+ content. (Fig-
levels in murine and human cell lines (Figure 1A), with maximal ure 1D). These results illustrate how NR administration is a valid

Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc. 839
 Cell Metabolism
+
NR Increases NAD and Prevents Metabolic Disorders

tool to boost NAD+ levels in mammalian cells and tissues without SOD2 activity (Figure 2E). To ensure that NR-induced SOD2 de-
activating GPR109A. acetylation was consequent to SIRT3 activation, we used mouse
Given the existence of different cellular NAD+ pools and the embryonic fibroblasts (MEFs) established from SIRT3 KO mice.
relevance of mitochondrial NAD+ content for mitochondrial and The absence of SIRT3 was reflected by the higher basal acetyla-
cellular function (Yang et al., 2007a), we also analyzed whether tion of SOD2 (Figure 2F). Importantly, NR was unable to
NR treatment would affect mitochondrial NAD+ levels. In decrease the acetylation status of SOD2 in SIRT3/ MEFs (Fig-
contrast to what has been observed with other strategies aimed ure 2F), even though NAD+ levels increased to similar levels as in
to increase NAD+ bioavailability, such as PARP inhibition (Bai SIRT3+/+ MEFs (Figure 2G). These results clearly indicate that NR
et al., 2011a), we found that mitochondrial NAD+ levels were triggers SIRT3 activity, probably by increasing mitochondrial
enhanced in cultured cells (Figure 1E) and mouse liver (Figure 1F) NAD+ levels, inducing the concomitant deacetylation of its mito-
after NR supplementation. chondrial targets. Strikingly, not all sirtuins were affected by NR,
To further solidify our data, we also wondered whether the as the acetylation of tubulin, a target of the cytoplasmic SIRT2
enhanced NAD+ levels upon NR treatment could derive from (North et al., 2003), was not altered (data not shown).
alterations in the NAD+ salvage pathway or PARP activity.
However, we could not see any change in Nampt mRNA or NR Supplementation Enhances Energy Expenditure
protein content in response to NR treatment (Figure 1G). Simi- Given the promising role of sirtuins to protect against metabolic
larly, PARP activity and PARP-1 content were not affected by disease, we next evaluated the effects of long-term NR adminis-
NR (Figure 1H). Altogether, these results suggest that NR tration in vivo. We fed 10-week-old male C57Bl/6J mice with
increases NAD+ by direct NAD+ biosynthesis rather than by indi- either chow diet (CD) or high-fat diet (HFD), supplemented or
rectly affecting the major NAD+ salvage (Nampt) or consumption not with NR at 400 mg/kg/day. While NR had no effect on the
(PARPs) pathways. Importantly, this increase in NAD+ was not body weight (BW) on CD, HFD-induced BW gain was signifi-
linked to changes in cellular glycolytic rates or ATP levels (data cantly attenuated by NR (Figure 3A), due to reduced fat mass
not shown), which would be expected if NAD+/NADH ratios (Figure 3B). This was visibly translated into a significant lower
had been altered to the point of compromising basic cellular weight of the epididymal depot in NR-fed mice (Figure S2A).
functions. Importantly, this was not due to redistribution of lipids to other
tissues (Figure S2A), most notably to liver, which actually con-
NR Treatment Enhances SIRT1 and SIRT3 Activity tained 40% less triglycerides (Figure S2B).
The ability of NR to increase intracellular NAD+ levels both in vivo The reduced BW gain of NR-fed mice upon HFD was not due
and in vitro prompted us to test whether it could activate sirtuin to reduced food intake, as NR-fed mice actually had a tendency
enzymes. Confirming this hypothesis, NR dose-dependently to eat more, especially on HFD conditions (Figure 3C). Similarly,
decreased the acetylation of FOXO1 (Brunet et al., 2004) in NR did not affect the activity pattern of mice (Figure 3D), indi-
a SIRT1-dependent manner (Figure 2A). This deacetylation of cating that the lower BW on HFD was not consequent to different
FOXO1 by SIRT1 upon NR treatment resulted in its transcrip- physical activity. Rather, the phenotype was due to enhanced
tional activation, leading to higher expression of target genes, energy expenditure (EE). Mice on CD had a marked tendency
such as Gadd45, Catalase, Sod1, and Sod2 (see Figure S1 on- to display higher O2 consumption rates when fed with NR, and
line) (Calnan and Brunet, 2008). The lack of changes in SIRT1 this tendency became clearly significant under HFD conditions
protein levels upon NR treatment (Figure 2A) suggests that NR (Figure 3E). Of note, NR-fed mice became more flexible in their
increases SIRT1 activity by enhancing NAD+ bioavailability. use of energy substrates, as reflected in the higher amplitude
The higher SIRT1 activity in NR-treated cells was supported by of the changes in RER between feeding and fasting periods (Fig-
mRNA expression analysis. Consistent with SIRT1 being a nega- ure S2C) in CD conditions. Altogether, these results indicate that
tive regulator of Ucp2 expression (Bordone et al., 2006), NR NR lowers HFD-induced BW gain by enhancing EE.
decreased Ucp2 mRNA levels (Figure 2B). Importantly, knocking From a metabolic perspective, NR- and vehicle-fed mice had
down Sirt1 prevented the action of NR on Ucp2 expression (Fig- similar fasting blood glucose levels in either CD or HFD condi-
ure 2B). Similarly, the higher expression of a FOXO1 target gene, tions (Figure 3F). However, fasting insulin levels were much lower
Sod2, upon NR treatment was also prevented by the knockdown in NR-supplemented mice (Figure 3G). This lower insulin/glucose
of either Foxo1 or Sirt1 (Figure 2B). This suggested that NR leads ratio is indicative of insulin sensitization after NR administration.
to a higher Sod2 expression through the activation of SIRT1, This speculation was further supported by glucose tolerance
which then deacetylates and activates FOXO1. Importantly, the tests. NR promoted a slight, albeit not significant, improvement
knockdown of Sirt1 did not compromise the ability of NR to in glucose tolerance (Figure 3H) in mice fed a HFD, accompanied
increase intracellular NAD+ content, indicating that NR uptake by a robust reduction in insulin secretion (Figure 3I). Therefore,
and metabolism into NAD+ are not affected by SIRT1 deficiency NR-fed mice on HFD display a better glucose disposal with lower
(Figure 2C). insulin levels. In order to conclusively establish whether NR
In line with the increase in mitochondrial NAD+ levels (Figures fed mice were more insulin sensitive, we performed insulin toler-
1E and 1F) and the potential consequent activation of mitochon- ance tests and hyperinsulinemic-euglycemic clamps on CD and
drial sirtuins, NR also reduced the acetylation status of Ndufa9 CD-NR mice. We chose not to perform this analysis on the HFD
immunoprecipitates and SOD2 (Figures 2D and 2E, respec- groups in order to avoid the possible influence of differential BW.
tively), both targets for SIRT3 (Ahn et al., 2008; Qiu et al., Glucose disposal upon insulin delivery was largely enhanced in
2010). SOD2 deacetylation has been linked to a higher intrinsic NR-fed mice (Figure 3J). In agreement, mice supplemented
activity. In line with these observations, NR treatment enhanced with NR required an almost 2-fold higher glucose infusion rate

840 Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc.
Cell Metabolism
NR Increases NAD+ and Prevents Metabolic Disorders

A IP: FOXO1 B C
3 Vehicle
IB: Ac-Lys NAD+ content
NR (mmol / kg prot)

(arbitrary units)
*

mRNA levels
IB: FOXO1 1
2
* *
IB (SN):SIRT1
1 0,5
*
IB (SN): Tubulin *
NR (mM) 0 0.2 0.5 0 0.2 0.5 0 0
Scramble SIRT1 FOXO1 siRNA - - + + - - - - ++ - - - - ++ - - - - ++ - -
SIRT1 siRNA - - - - + + - - - - + + - - - - + + - - - - + +
siRNA siRNA
FOXO1 SIRT1 UCP2 SOD2

D E F G
IP: SOD2 IP: SOD2
IP: Ndufa9 NAD+ content
IB: Ac-Lys IB: Ac-Lys
(mmol / kg prot)

IB: SOD2 IB: SOD2 1

IB: Ac-Lys
Veh NR * *
(relative to Veh)

SIRT3
SOD2 activity

1,5 * 0,5
IB: Ndufa9 1 Porin
Veh NR 0,5
0
0 Veh NR Veh NR

MEF SIRT3 +/+ MEF SIRT3 -/-

Figure 2. Nicotinamide Riboside Supplementation Increases Sirtuin Activity in Cultured Mammalian Cells
(A) HEK293T cells were transfected with a pool of either scramble siRNAs or SIRT1 siRNAs. After 24 hr, cells were treated with vehicle (PBS) or NR at the
concentrations indicated, and, after an additional 24 hr, total protein extracts were obtained. FOXO1 acetylation was tested after FOXO1 immunoprecipitation (IP)
from 500 mg of protein, while tubulin and SIRT1 levels were evaluated in the supernatant of the IP.
(B) HEK293T cells were transfected with a pool of either scramble siRNAs, FOXO1 siRNAs, or SIRT1 siRNAs. After 24 hr, cells were treated with NR (0.5 mM; black
bars) or vehicle (PBS; white bars) for an additional 24 hr. Then total mRNA was extracted and the mRNA expression levels of the markers indicated were evaluated
by qRT-PCR.
(C) HEK293T cells were transfected with a pool of either scramble siRNAs, FOXO1 siRNAs, or SIRT1 siRNAs. After 24 hr, cells were treated with NR (0.5 mM; black
bars) or vehicle (PBS; white bars) for an additional 24 hr. Then acidic extracts were obtained to measure intracellular NAD+ levels.
(D and E) HEK293T cells were treated with NR (0.5 mM) or vehicle (PBS) for 24 hr and total protein extracts were obtained to measure (D) Ndufa9 or (E) SOD2
acetylation after IP. The extracts were also used to measure SOD2 activity (E, bottom panel).
(F and G) SIRT3+/+ and SIRT3/ mouse embryonic fibroblasts (MEFs) were treated with NR (0.5 mM) or vehicle (PBS) for 24 hr, and either (F) total extracts to test
SOD2 acetylation were obtained or (G) acidic extracts were used to measure intracellular NAD+ content. Throughout the figure, all values are presented as
mean ±SD. Asterisk indicates statistical significant difference versus respective vehicle group at p < 0.05. Unless otherwise stated, the vehicle groups are
represented by white bars, and NR groups are represented by black bars. This figure is complemented by Figure S1.

to maintain euglycemia in hyperinsulinemic-euglycemic clamps was significantly accentuated upon HFD (Figure 4A), suggesting
(Figure 3K). Together, these observations unequivocally demon- an enhanced muscle oxidative performance. Similarly, NR-fed
strate that NR-fed mice are more insulin sensitive. Furthermore, mice, both on CD and HFD, showed enhanced thermogenic
NR partially prevented the increase in total (Figure 3K) and LDL capacity, as manifested in the ability to maintain body tempera-
cholesterol levels (Figure S2D) induced by HFD, even though ture during cold exposure (Figure S3B and Figure 4B). The latter
HDL-cholesterol levels were unaffected (Figure S2E). The observation hints toward an improvement in brown adipose
amelioration of cholesterol profiles is fully in line with previous tissue (BAT) oxidative performance. To gain further insight into
observations from the use of other NAD+ precursors, such as the ability of BAT and muscle to enhance their oxidative perfor-
NA (Houtkooper et al., 2010). mance, we performed some histological analysis. Gastrocne-
mius muscles from NR mice displayed a more intense SDH
NR Enhances the Oxidative Performance of Skeletal staining than gastrocnemius muscles from their vehicle-fed
Muscle and Brown Adipose Tissue littermates, indicating a higher oxidative profile (data not shown).
NR-fed mice had a clear tendency to display a better endurance Electron microscopy revealed that mitochondria in BAT of
performance than vehicle-fed mice (Figure S3A). This tendency NR-fed mice, despite not being significantly larger, had more

Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc. 841
 Cell Metabolism
+
NR Increases NAD and Prevents Metabolic Disorders

A B C D

E F G H

I J K L

Figure 3. NR Supplementation Prevents Diet-Induced Obesity by Enhancing Energy Expenditure and Reduces Cholesterol Levels
Ten-week-old C57Bl/6J mice were fed with either chow diet (CD) or high-fat diet (HFD) mixed with either water (as vehicle) or NR (400 mg/kg/day) (n = 10 mice
per group).
(A) Body weight evolution was monitored for 12 weeks.
(B) Body composition was evaluated after 8 weeks of diet through Echo-MRI.
(C–E) Food intake, activity, and VO2 were evaluated using indirect calorimetry.
(F and G) Blood glucose and insulin levels were measured in animals fed with their respective diets for 16 weeks after a 6 hr fast.
(H and I) After 10 weeks on their respective diets (CD, squares; HFD, circles), an intraperitoneal glucose tolerance test was performed in mice that were fasted
overnight. At the indicated times, blood samples were obtained to evaluate either (H) glucose or (I) insulin levels. Areas under the curve are shown at the top right
of the respective panels.
(J) Insulin tolerance tests were performed on either CD or CD-NR mice (4 weeks of treatment). At the indicated times, blood samples were obtained to evaluate
blood glucose levels. The area above the curve is shown at the top right of the panel.
(K) Hyperinsulinemic-euglycemic clamps were performed on either CD or CD-NR mice (4 weeks of treatment). Glucose infusion rates (GIR) were calculated after
the test.
(L) Serum levels of total cholesterol were measured in animals fed with their respective diets for 16 weeks, after a 6 hr fast. Throughout the figure, white represents
the vehicle group, and black represents the NR-supplemented mice. All values are presented as mean ±SD. Asterisk indicates statistical significant difference
versus respective vehicle-treated group. This figure is complemented by Figure S2.

abundant cristae (Figure 4C), which has been linked to ingly, NAD+ was also higher in the BAT of NR-fed mice, but
increased respiratory capacity (Mannella, 2006). Altogether, only on HFD (Figure 5B and Figure S4A). These differences could
the above results suggest that NR-supplemented mice display be due to the differential expression of NRKs in tissues. NRKs
a higher oxidative capacity due to enhanced mitochondrial initiate NR metabolism into NAD+ (Houtkooper et al., 2010).
function. There are two mammalian NRKs: NRK1 and NRK2 (Bieganowski
and Brenner, 2004). While we found Nrk1 expressed ubiquitously
Chronic NR Feeding Increases NAD+ In Vivo in a Tissue- (Figure S4B), Nrk2 was mainly present in cardiac and skeletal
Specific Manner muscle tissues, as previously described (Li et al., 1999), but
We next wondered how chronic NR feeding would affect NAD+ also detectable in BAT and liver (Figure S4C), in line with the
metabolism in mice. Chronic NR supplementation increased better ability of these tissues to respond to NR.
NAD+ levels in both CD (Figure S4A) and HFD (Figure 5A) condi- We also tested whether the increase in NAD+ would be
tions in some tissues, including liver and muscle, but not in concomitant to changes in other NAD+ metabolites. Strikingly,
others, such as brain or white adipose tissue (WAT). Interest- NADH and nicotinamide (NAM) levels were largely diminished

842 Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc.
Cell Metabolism
NR Increases NAD+ and Prevents Metabolic Disorders

that NAD+ salvage pathways do not explain the differences in

AOC (a.u.)
A B 20
1000 38
15 NAD+ levels. We furthermore could not detect differences in
* *
Body temperature ºC
10
800 37 * mRNA expression of the different NMN adenylyltransferase
Distance run (m)

5
0 (NMNAT) enzymes (Figure 5D). Altogether, these results rein-
600 36
force the notion that the higher NAD+ levels observed in tissues
400 35 * from NR-fed mice are consequent to an increase in direct NAD+
synthesis from NR.
200 34

0 33 NR Enhances Sirtuin Activity In Vivo

0 1 2 3 4 5 Higher NAD+ levels were also accompanied by higher sirtuin
Time at 4ºC (hours) activity in vivo. A prominent deacetylation of SIRT1 and
SIRT3 targets (FOXO1, Brunet et al., 2004; and SOD2 Qiu
C Veh
(relative to vehicle group)
2 et al., 2010, respectively) was observed in the skeletal muscle,
liver, and BAT, where NAD+ content was induced by NR, but
Mitochondrial size

1,5 not in brain and WAT, where NAD+ levels were unaffected by
NR supplementation (Figure 6A and Figure S5A). We also eval-
1 uated PGC-1a acetylation as a second readout of SIRT1
activity (Rodgers et al., 2005). We were unable to detect
0,5 PGC-1a in total lysates from WAT or brain (Figure S5B), but
in muscle, liver, and BAT, PGC-1a was deacetylated upon
0
NR treatment (Figure S5C). These observations highlight how
NR can only induce sirtuin activity in tissues where NAD+ accu-
NR 7
(relative to vehicle group)

* mulates. Like in cultured cells, we could not detect changes in

6 the acetylation status of the SIRT2 target tubulin (data not
Cristae content

5 shown), suggesting either that increasing NAD+ might not affect

4 the activity of all sirtuins equally, that the increase is only
3 compartment specific, or that additional regulatory elements,
2 like class I and II HDACs, also contribute to tubulin acetylation
1 status.
0 In line with the changes in acetylation levels of PGC-1a, a
key transcriptional regulator of mitochondrial biogenesis, we
could observe either an elevated expression or a strong
Figure 4. NR Enhances Skeletal Muscle and BAT Oxidative Function tendency toward an increase (p < 0.1) of nuclear genes encoding
Ten-week-old C57Bl/6J mice were fed a high-fat diet (HFD) mixed with either
transcriptional regulators of oxidative metabolism (Sirt1,
water (as vehicle; white bars and circles) or NR (400 mg/kg/day; black bars and
circles) (n = 10 mice per group).
Ppargc1a, mitochondrial transcription factor A [Tfam]) and mito-
(A) An endurance exercise test was performed using a treadmill in mice fed chondrial proteins (Mitofusin 2 [Mfn2], cytochrome c [Cycs],
with either HFD or HFD-NR for 12 weeks. medium-chain acyl-coA dehydrogenase [Acadm], carnitine
(B) A cold test was performed in mice fed with either HFD or HFD-NR for palmitoyltransferase-1b [Cpt1b], citrate synthase [Cs], or ATP
9 weeks. The area over the curve (AOC) is shown on the top right of the graph. synthase lipid binding protein [Atp5g1]) in quadriceps muscles
(C) Electron microscopy of the BAT was used to analyze mitochondrial content
from NR-fed mice (Figure 6B). Conversely, in brain, where
and morphology. The size and cristae content of mitochondria were quantified
as specified in the Experimental Procedures. Throughout the figure, all values
NAD+ and sirtuin activity levels were not affected by NR feeding,
are shown as mean ±SD. Asterisk indicates statistical significant difference the expression of these genes was not altered (Figure 6B).
versus vehicle-supplemented group at p < 0.05. This figure is complemented Consistently also with enhanced mitochondrial biogenesis,
by Figure S3. mitochondrial DNA content was increased in muscle, but not in
brain from NR-fed mice (Figure 6C). Finally, mitochondrial
protein content also confirmed that mitochondrial function was
in muscles from NR-fed mice (Figure 5B), indicating that NR only enhanced in tissues in which NAD+ content was increased
specifically increases NAD+, but not necessarily other byprod- (Figure 6D and Figure S5D). This way, while muscle, liver, and
ucts of NAD+ metabolism. We analyzed in vivo whether the BAT showed a prominent increase in mitochondrial proteins
activity of major NAD+-degrading enzymes or the levels of (complex V—ATP synthase subunit a and porin), such change
Nampt could also contribute to the increase in NAD+ after was not observed in brain or WAT. Altogether, these results
chronic NR supplementation. As previously observed in suggest that NR feeding increases mitochondrial biogenesis in
HEK293T cells (Figures 1G and 1H), PARP-1 levels and global a tissue-specific manner, consistent with the tissue-specific
PARylation were similar in muscle (Figure 5C) and livers (Fig- nature of the increases in NAD+ and sirtuin activity observed in
ure S4D) from NR- and vehicle-fed mice, indicating that the NR-fed mice. The higher number of mitochondria, together
enhanced NAD+ content cannot be explained by differential with the different morphological mitochondrial profiles found in
NAD+ consumption through PARP activity. Nampt mRNA (Fig- NR-fed mice (Figure 4C), would contribute to explain the higher
ure 5D) and protein (Figure 5C, Figure S4E, and data not shown) oxidative profile, EE, and protection against metabolic damage
levels were also similar in NR- and vehicle-fed mice, suggesting observed upon NR feeding.

Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc. 843
 Cell Metabolism
+
NR Increases NAD and Prevents Metabolic Disorders

A 0,5 B 0,3 Figure 5. Chronic NR Supplementation Increases

Plasma and Intracellular NAD+ Content in
*
(mmol / kg tissue)

0,4

(mmol / kg tissue)
a Tissue-Specific Manner
NAD+ content

0,2 Tissues from C57Bl/6J mice were collected after 16 weeks

0,3
* of HFD supplemented with either water (as vehicle; white
0,2
bars) or NR (400 mg/kg/day; black bars).
* 0,1
* (A) NAD+ levels were measured in acidic extracts obtained
0,1
* from different tissues.
0 0
(B) NADH and NAM levels were measured in gastrocne-
Brain Muscle Liver BAT WAT NADH NAM
mius muscle.
(C) Quadriceps muscle protein homogenates were ob-
tained to test global PARylation, PARP-1, and Nampt
C D protein levels.
2
PAR (D) Total mRNA was isolated from quadriceps muscles,
(relative to Vehicle)

1,5 and the mRNA levels of the markers indicated were

mRNA levels

measured by RT-qPCR. Throughout the figure, all values

1 are expressed as mean ±SD. Asterisk indicates statistical
significant difference versus respective vehicle-treated
0,5
PARP-1 group.

0
Nampt Nampt NMNAT-1 NMNAT-2 NMNAT-3

Vehicle NR

DISCUSSION elevating NAD+ levels, activating SIRT1 and its downstream

targets. However, future experiments will be required to verify
While increased NAD+ levels in response to NR supplementation this hypothesis. The activation of SIRT1 and SIRT3, impacting
were already reported in yeast (Belenky et al., 2007; Bieganow- on targets such as PGC-1a, FOXOs, and SOD2, also provides
ski and Brenner, 2004) and cultured human cells (Yang et al., a likely explanation for the mitochondrial fitness and metabolic
2007b), we extend these observations to a wide range of flexibility of NR-supplemented mice. In this sense, it is remark-
mammalian cell lines and demonstrate that NR supplementation able that while SIRT1 transgenic mice have some shared pheno-
also enhances NAD+ bioavailability in mammalian tissues. Also, types with NR-treatment mice, such as protection against
our work provides evidence that the increase in NAD+ after NR metabolic damage, there are also a number of discrepancies.
administration stimulates the activity of mammalian sirtuins. For example, SIRT1 transgenics show similar BW gain upon
This further supports the role of sirtuins as a family of proteins HFD to wild-type littermates (Pfluger et al., 2008). The differ-
whose basal activity can be largely modulated by NAD+ avail- ences between NR-supplemented mice and SIRT1 trangenics
ability (Houtkooper et al., 2010). can be explained by numerous reasons. First, NR affects
The fact that the activity of both SIRT1 and SIRT3 is positively different sirtuin activities; second, sirtuin expression does not
regulated by NR both in vitro and in vivo favors the hypothesis necessarily correspond with sirtuin activity; third, NR actions
that the increase in NAD+ promoted by NR affects, at least, the are tissue specific. Importantly, we should also consider that
mitochondrial and the nuclear compartments. This is a key differ- NR might trigger other actions unrelated to sirtuin activity that
ence compared to other strategies that boost intracellular NAD+ might also contribute to the beneficial effects of NR (Figure 7).
levels. For example, Parp-1 knockout mice show a 2-fold In conclusion, our work shows that NR is a powerful supple-
increase in intracellular NAD+, yet SIRT3 activity is not affected ment to boost NAD+ levels, activate sirtuin signaling, and improve
(Bai et al., 2011b). Given that PARP-1 is a nuclear enzyme, it is mitochondrial function, suggesting that this vitamin could be
tempting to hypothesize that, in the Parp-1 knockout mice, the used to prevent and treat the mitochondrial decline that is a hall-
increase in NAD+ is mainly confined to the nucleus, while the mark of many diseases associated with aging. Very recent work,
NR-induced increase in NAD+ levels, despite being more discrete showing that intraperitoneal administration of NMN could
in amplitude, reaches different subcellular compartments. improve the metabolic damage induced by high-fat feeding
The inability of NR to activate GPR109A illustrates how NR (Yoshino et al., 2011), further supports this concept. To date,
provides an alternative way to increase NAD+ and ameliorate however, only NR has been identified as a naturally occurring
metabolic homeostasis in the absence of the undesirable effects component of the human diet (Bieganowski and Brenner, 2004).
seen with NA. Since NR does not activate GPR109A, we specu- Furthermore, NR protects against metabolic dysfunction at lower
late that NR reduces cholesterol levels through the activation of concentrations than those reported for NMN, and we proved that
SIRT1 in liver, which influences the activity of a number of tran- it is effective after oral administration when mixed with food, in
scription factors and cofactors linked to cholesterol homeostasis contrast to NMN, which is injected intraperitoneally (Yoshino
(Kemper et al., 2009; Li et al., 2007; Shin et al., 2003; Walker et al., 2011). So, how might NR and NMN biology intertwine and
et al., 2010). Interestingly, the mechanism of action of NA on impact on NAD+ synthesis in mammalian physiology? Recent
VLDL/LDL and hepatic metabolism is not clear, as GPR109A is observations suggest that, at least for some cell types, NMN
not expressed in the liver (Tunaru et al., 2003). Our data raise needs to be converted into NR through dephosphorylation
the possibility that NA, like NR, achieves such effects by performed by extracellular 50 -nucleotidases in order to penetrate

844 Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc.
Cell Metabolism
NR Increases NAD+ and Prevents Metabolic Disorders

A Muscle Brain Figure 6. NR Stimulates Sirtuin Activity

C
8000 In Vivo and Enhances Mitochondrial Gene

IP: FOXO1

mtDNA copy number

IB: Ac-K Expression
6000 * Tissues from C57Bl/6J mice were collected after
IB: FOXO1 16 weeks of HFD supplemented with either water
4000 (as vehicle; white bars) or NR (400 mg/kg/day;
black bars).

IP: SOD2
IB: Ac-K 2000 (A) Total protein extracts were obtained from
quadriceps muscle and brain as indicated to
IB: SOD2 0 evaluate the acetylation levels of FOXO1 and
Muscle Brain SOD2 through immunoprecipitation assays, using
1 and 0.5 mg of protein, respectively.
B D (B) Total mRNA from quadriceps muscle and brain
4 Muscle was extracted to measure the abundance of the
*
(relative to vehicle)

Muscle markers indicated by RT-qPCR.

mRNA levels

3
Complex V (C) Mitochondrial DNA content was measured in
2 * * * DNA extracted from quadriceps muscle and brain.
Porin The results are expressed as mitochondrial copy
1 number relative to genomic DNA.
Tubulin (D) The abundance of mitochondrial marker
0 proteins in 20 mg of protein from total quadriceps
muscle and brain lysates. Throughout the figure,
Vehicle NR
all values are shown as mean ±SD. Asterisk indi-
cates statistical significant difference versus
Brain
vehicle-supplemented group at p < 0.05. This
2 Brain
figure is complemented by Figure S4.
(relative to vehicle)

Complex V
mRNA levels

1,5
Porin
1

0,5 Tubulin
Our data, combined with the evidence
Vehicle NR
0 that other NAD+ precursors can improve
d 1a 1a rt
1 age-related insulin resistance (Yoshino
ca c- pt Si
M Pg C et al., 2011) and that NR increases yeast
replicative life span (Belenky et al.,
2007), therefore warrant future investiga-
into the cell (Nikiforov et al., 2011). Given that the presence of tions to see if boosting NAD+ levels by NR supplementation
NMN in plasma is debated (Hara et al., 2011), the most likely might also improve the health and life span of humans.
scenario is that NR, rather than NMN, is the entity directly taken
up by the cell, to then become metabolized in the cytosolic EXPERIMENTAL PROCEDURES
compartment to NMN. The lack of NRK1 and NRK2 in the mito-
Materials
chondrial compartment (Nikiforov et al., 2011) supports the notion
All chemicals and reagents were purchased from Sigma-Aldrich unless stated
that NMN might act as the precursor entering the mitochondria or otherwise. NR was custom synthesized as previously described (Yang et al.,
other compartments for further metabolism into NAD+ by the 2007b) and was kindly provided by Amazentis and Merck Research
various NMNAT enzymes (see Figure 7 for a scheme). Laboratories.

Figure 7. Schematic Representation of the

Different Actions of NR in Metabolic Homeostasis
The scheme summarizes the hypothesis by which NR
supplementation would increase NAD+ content in key
metabolic tissues, leading to SIRT1 and SIRT3 activation
and the deacetylation and modulation of the activity of key
metabolic regulators. This model does not rule out the
participation of additional mechanisms of action for NR to
achieve its beneficial effects. Abbreviations can be found
in the text, and enzymes are indicated in italics.

Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc. 845
 Cell Metabolism
+
NR Increases NAD and Prevents Metabolic Disorders

Animal Experiments NAD+, NADH, and NAM Determination

Eight-week-old male C57Bl/6J mice were purchased from Charles River, and NAD+ levels were determined using a commercial kit (Enzychrom, BioAssays
powder chow (D12450B) and high-fat (D12492) diets were from Research Diets Systems, CA), and tissue samples were also measured by HPLC (Ramsey
Inc. (New Brunswick, NJ, USA). CD pellets were made by adding 80 ml of water et al., 2009). Other NAD metabolites were determined as previously described
per kg of powder CD. HFD pellets were made by adding 40 ml of water per kg of (Yang and Sauve, 2006).
powder HFD. For NR-, NMN-, and NA-supplemented diets, the appropriate
amount of these compounds was added to the water used to create the pellets, GPR109A—Calcium Mobilization Assay
taking into account the differences in the daily intake of each diet. Mice were Ready-to-Assay GRP109A nicotinic acid receptor cells were used to measure
housed separately, had ad libitum access to water and food, and were kept calcium mobilization as specified by the manufacturer (Millipore). Calcium flux
under a 12 hr dark-light cycle. Mice were fed with homemade pellets from was determined using excitation at 340 and 380 nm in a fluorescence spectro-
10 weeks of age. To make the pellets, the powder food was mixed with water photometer (VictorX4, Perkin Elmer) in a 180 s time course, adding the ligand at
(vehicle) or with NR. All animal experiments were carried according to national 60 s. Internal validation was made using 0.1% Triton X-100 for total fluoro-
Swiss and EU ethical guidelines and approved by the local animal experimen- phore release and 10 mM EGTA to chelate free calcium. Similarly, GPR109A
tation committee of the Canton de Vaud under license #2279. specificity was internally validated using control cells devoid of GRP109A
overexpression.
Animal Phenotyping
Mice were weighed and food consumption was measured each week on the
Liver Triglyceride Measurement
same day. Most clinical tests were carried out according to standard opera-
Liver triglycerides were measured from 20 mg of liver tissue using a variation of
tional procedures (SOPs) established and validated within the Eumorphia
the Folch method, as described (Bai et al., 2011a).
program (Champy et al., 2008). Body composition was determined by Echo-
MRI (Echo Medical Systems, Houston, TX, USA), and oxygen consumption
(VO2), respiratory exchange ratios (RER), food intake, and activity levels Statistics
were monitored by indirect calorimetry using the comprehensive laboratory For the statistical analysis of the animal studies, all data were verified for
animal monitoring system (Columbus Instruments, Columbus, OH, USA). normal distribution. To assess significance, we performed Student’s t test
Treadmill, cold tests, and hyperinsulinemic-euglycemic clamps were per- for independent samples. Values are expressed as mean ±SD unless other-
formed as described (Lagouge et al., 2006). Glucose and insulin tolerance wise specified.
was analyzed by measuring blood glucose and insulin following intraperitoneal
injection of 2 g/kg glucose or 0.25 U insulin/kg, respectively, after an overnight SUPPLEMENTAL INFORMATION
fast. Plasma insulin was determined in heparinized plasma samples using
specific ELISA kits (Mercodia). All animals were sacrificed at 13:00, after Supplemental Information includes five figures and can be found with this
a 6 hr fast, using CO2 inhalation. Blood samples were collected in heparinized article online at doi:10.1016/j.cmet.2012.04.022.
tubes, and plasma was isolated after centrifugation. All plasma parameters
were measured using a Cobas c111 (Roche Diagnostics). Tissues were ACKNOWLEDGMENTS
collected upon sacrifice and flash frozen in liquid nitrogen.
This work was supported by grants of the École Polytechnique Fédérale de
Histology and Microscopy Lausanne, Swiss National Science Foundation, the European Research
Succinate dehydrogenase staining and transmission electron microscopy Council Ideas program (Sirtuins; ERC-2008-AdG231-118), and the Velux foun-
(TEM) were performed as described (Bai et al., 2011b). Mitochondrial size and dation. R.H.H. has been supported by a Rubicon fellowship of the Netherlands
cristae content were analyzed as previously described (St-Pierre et al., 2003). Organization for Scientific Research and by an AMC postdoc fellowship. E.P.
is funded by the Academy of Finland. J.A. is the Nestle chair in energy metab-
Cell Culture and Transfection olism. A.A.S. received grants to support this research from the Ellison Medical
Murine C2C12 myoblasts were grown and differentiated into myotubes as Foundation New Scholar Award 2007 and contract C023832 from NY State
described (Canto et al., 2009). Murine Hepa1.6 and human HEK293T cells Spinal Cord Injury Board. The authors thank Robert Myers, Peter Meinke,
were cultured in DMEM (4.5 g/l glucose, 10% FCS). Human FOXO1 and and Thomas Vogt at Merck Research Laboratories, Rahway, and Charles
SIRT1 siRNAs were obtained from Dharmacon (Thermo Scientific) and used Thomas at Amazentis, Lausanne, for the kind gift of NR; all the members of
following the instructions from the manufacturer. SIRT3 MEFs were estab- the Auwerx lab for inspiring discussions; and Graham Knott of the BioEM
lished according to standard techniques (Picard et al., 2002) from conditional facility at EPFL for EM imaging.
SIRT3/ mice. Deletion of the SIRT3 gene was induced via infection with
adenovirus encoding for the Cre recombinase. Received: October 6, 2011
Revised: March 2, 2012
Gene Expression and Mitochondrial DNA Abundance Accepted: April 25, 2012
RNA was extracted and quantified by qRT-PCR as described (Lagouge et al., Published online: June 5, 2012
2006). The qPCR primers used have been previously described. Mitochondrial
DNA abundance (mtDNA) was quantified as described (Lagouge et al., 2006) REFERENCES
after isolating total DNA using a standard phenol extraction.
Ahn, B.H., Kim, H.S., Song, S., Lee, I.H., Liu, J., Vassilopoulos, A., Deng, C.X.,
Immunoprecipitation, SDS-PAGE, Western Blotting and Finkel, T. (2008). A role for the mitochondrial deacetylase Sirt3 in regu-
Cells were lysed in lysis buffer (50 mM Tris, 150 mM KCl, EDTA 1 mM, NP40 lating energy homeostasis. Proc. Natl. Acad. Sci. USA 105, 14447–14452.
1%, nicotinamide 5 mM, Na-butyrate 1 mM, protease inhibitors [pH 7.4]).
Bai, P., Canto, C., Brunyanszki, A., Huber, A., Szanto, M., Cen, Y., Yamamoto,
Immunoprecipitation from cultured cells or tissue samples was performed
H., Houten, S.M., Kiss, B., Oudart, H., et al. (2011a). PARP-2 regulates SIRT1
exactly as described (Canto et al., 2009). For western blotting, proteins were
expression and whole-body energy expenditure. Cell Metab. 13, 450–460.
separated by SDS-PAGE and transferred onto nitrocellulose membranes.
Anti-acetyl-lysine antibodies were from Cell Signaling; FOXO1 and SOD2 anti- Bai, P., Canto, C., Oudart, H., Brunyanszki, A., Cen, Y., Thomas, C.,
bodies were from Santa Cruz Biotechnology; SIRT1 and PGC-1a antibodies Yamamoto, H., Huber, A., Kiss, B., Houtkooper, R.H., et al. (2011b). PARP-1
were from Millipore; tubulin and acetyl-tubulin antibodies were from Sigma inhibition increases mitochondrial metabolism through SIRT1 activation. Cell
Inc. Antibodies for mitochondrial markers were purchased from Mitosciences. Metab. 13, 461–468.
Antibody detection reactions were developed by enhanced chemilumines- Barbosa, M.T., Soares, S.M., Novak, C.M., Sinclair, D., Levine, J.A., Aksoy, P.,
cence (Amersham, Little Chalfont, UK). and Chini, E.N. (2007). The enzyme CD38 (a NAD glycohydrolase, EC 3.2.2.5)

846 Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc.
Cell Metabolism
NR Increases NAD+ and Prevents Metabolic Disorders

is necessary for the development of diet-induced obesity. FASEB J. 21, Li, J., Mayne, R., and Wu, C. (1999). A novel muscle-specific beta 1 integrin
3629–3639. binding protein (MIBP) that modulates myogenic differentiation. J. Cell Biol.
Belenky, P., Racette, F.G., Bogan, K.L., McClure, J.M., Smith, J.S., and 147, 1391–1398.
Brenner, C. (2007). Nicotinamide riboside promotes Sir2 silencing and extends Li, X., Zhang, S., Blander, G., Tse, J.G., Krieger, M., and Guarente, L. (2007).
lifespan via Nrk and Urh1/Pnp1/Meu1 pathways to NAD+. Cell 129, 473–484. SIRT1 deacetylates and positively regulates the nuclear receptor LXR. Mol.
Benyo, Z., Gille, A., Kero, J., Csiky, M., Suchankova, M.C., Nusing, R.M., Cell 28, 91–106.
Moers, A., Pfeffer, K., and Offermanns, S. (2005). GPR109A (PUMA- Mannella, C.A. (2006). Structure and dynamics of the mitochondrial inner
G/HM74A) mediates nicotinic acid-induced flushing. J. Clin. Invest. 115, membrane cristae. Biochim. Biophys. Acta 1763, 542–548.
3634–3640.
Nikiforov, A., Dolle, C., Niere, M., and Ziegler, M. (2011). Pathways and subcel-
Bieganowski, P., and Brenner, C. (2004). Discoveries of nicotinamide riboside lular compartmentation of NAD biosynthesis in human cells: from entry of
as a nutrient and conserved NRK genes establish a Preiss-Handler indepen- extracellular precursors to mitochondrial NAD generation. J. Biol. Chem.
dent route to NAD+ in fungi and humans. Cell 117, 495–502. 286, 21767–21778.
Bogan, K.L., and Brenner, C. (2008). Nicotinic acid, nicotinamide, and nicotin- North, B.J., Marshall, B.L., Borra, M.T., Denu, J.M., and Verdin, E. (2003). The
amide riboside: a molecular evaluation of NAD+ precursor vitamins in human human Sir2 ortholog, SIRT2, is an NAD+-dependent tubulin deacetylase. Mol.
nutrition. Annu. Rev. Nutr. 28, 115–130. Cell 11, 437–444.
Bordone, L., Motta, M.C., Picard, F., Robinson, A., Jhala, U.S., Apfeld, J., Pfluger, P.T., Herranz, D., Velasco-Miguel, S., Serrano, M., and Tschop, M.H.
McDonagh, T., Lemieux, M., McBurney, M., Szilvasi, A., et al. (2006). Sirt1 (2008). Sirt1 protects against high-fat diet-induced metabolic damage. Proc.
regulates insulin secretion by repressing UCP2 in pancreatic beta cells. Natl. Acad. Sci. USA 105, 9793–9798.
PLoS Biol. 4, e31. 10.1371/journal.pbio.0040031.
Picard, F., Gehin, M., Annicotte, J., Rocchi, S., Champy, M.F., O’Malley, B.W.,
Brunet, A., Sweeney, L.B., Sturgill, J.F., Chua, K.F., Greer, P.L., Lin, Y., Tran, Chambon, P., and Auwerx, J. (2002). SRC-1 and TIF2 control energy balance
H., Ross, S.E., Mostoslavsky, R., Cohen, H.Y., et al. (2004). Stress-dependent between white and brown adipose tissues. Cell 111, 931–941.
regulation of FOXO transcription factors by the SIRT1 deacetylase. Science
Qiu, X., Brown, K., Hirschey, M.D., Verdin, E., and Chen, D. (2010). Calorie
303, 2011–2015.
restriction reduces oxidative stress by SIRT3-mediated SOD2 activation.
Calnan, D.R., and Brunet, A. (2008). The FoxO code. Oncogene 27, 2276–
Cell Metab. 12, 662–667.
2288.
Ramsey, K.M., Yoshino, J., Brace, C.S., Abrassart, D., Kobayashi, Y.,
Canto, C., and Auwerx, J. (2012). Targeting sirtuin 1 to improve metabolism: all
Marcheva, B., Hong, H.K., Chong, J.L., Buhr, E.D., Lee, C., et al. (2009).
you need is NAD+? Pharmacol. Rev. 64, 166–187.
Circadian clock feedback cycle through NAMPT-mediated NAD+ biosyn-
Canto, C., Gerhart-Hines, Z., Feige, J.N., Lagouge, M., Noriega, L., Milne, J.C., thesis. Science 324, 651–654.
Elliott, P.J., Puigserver, P., and Auwerx, J. (2009). AMPK regulates energy
Revollo, J.R., Korner, A., Mills, K.F., Satoh, A., Wang, T., Garten, A., Dasgupta,
expenditure by modulating NAD+ metabolism and SIRT1 activity. Nature
B., Sasaki, Y., Wolberger, C., Townsend, R.R., et al. (2007). Nampt/PBEF/
458, 1056–1060.
Visfatin regulates insulin secretion in beta cells as a systemic NAD biosynthetic
Champy, M.F., Selloum, M., Zeitler, V., Caradec, C., Jung, B., Rousseau, S., enzyme. Cell Metab. 6, 363–375.
Pouilly, L., Sorg, T., and Auwerx, J. (2008). Genetic background determines
Rodgers, J.T., Lerin, C., Haas, W., Gygi, S.P., Spiegelman, B.M., and
metabolic phenotypes in the mouse. Mamm. Genome 19, 318–331.
Puigserver, P. (2005). Nutrient control of glucose homeostasis through
Hara, N., Yamada, K., Shibata, T., Osago, H., and Tsuchiya, M. (2011). a complex of PGC-1alpha and SIRT1. Nature 434, 113–118.
Nicotinamide phosphoribosyltransferase/visfatin does not catalyze nicotin-
Shin, D.J., Campos, J.A., Gil, G., and Osborne, T.F. (2003). PGC-1alpha acti-
amide mononucleotide formation in blood plasma. PLoS ONE 6, e22781. 10.
vates CYP7A1 and bile acid biosynthesis. J. Biol. Chem. 278, 50047–50052.
1371/journal.pone.0022781.
Hirschey, M.D., Shimazu, T., Jing, E., Grueter, C.A., Collins, A.M., Aouizerat, St-Pierre, J., Lin, J., Krauss, S., Tarr, P.T., Yang, R., Newgard, C.B., and
B., Stancakova, A., Goetzman, E., Lam, M.M., Schwer, B., et al. (2011). Spiegelman, B.M. (2003). Bioenergetic analysis of peroxisome proliferator-
SIRT3 deficiency and mitochondrial protein hyperacetylation accelerate the activated receptor gamma coactivators 1alpha and 1beta (PGC-1alpha and
development of the metabolic syndrome. Mol. Cell 44, 177–190. PGC-1beta) in muscle cells. J. Biol. Chem. 278, 26597–26603.

Houtkooper, R.H., Canto, C., Wanders, R.J., and Auwerx, J. (2010). The secret Tunaru, S., Kero, J., Schaub, A., Wufka, C., Blaukat, A., Pfeffer, K., and
life of NAD+: an old metabolite controlling new metabolic signaling pathways. Offermanns, S. (2003). PUMA-G and HM74 are receptors for nicotinic acid
Endocr. Rev. 31, 194–223. and mediate its anti-lipolytic effect. Nat. Med. 9, 352–355.

Imai, S., Armstrong, C.M., Kaeberlein, M., and Guarente, L. (2000a). Walker, A.K., Yang, F., Jiang, K., Ji, J.Y., Watts, J.L., Purushotham, A., Boss,
Transcriptional silencing and longevity protein Sir2 is an NAD-dependent O., Hirsch, M.L., Ribich, S., Smith, J.J., et al. (2010). Conserved role of SIRT1
histone deacetylase. Nature 403, 795–800. orthologs in fasting-dependent inhibition of the lipid/cholesterol regulator
SREBP. Genes Dev. 24, 1403–1417.
Imai, S., Johnson, F.B., Marciniak, R.A., McVey, M., Park, P.U., and Guarente,
L. (2000b). Sir2: an NAD-dependent histone deacetylase that connects chro- Yang, T., and Sauve, A.A. (2006). NAD metabolism and sirtuins: metabolic
matin silencing, metabolism, and aging. Cold Spring Harb. Symp. Quant. regulation of protein deacetylation in stress and toxicity. AAPS J. 8, E632–
Biol. 65, 297–302. E643.
Karpe, F., and Frayn, K.N. (2004). The nicotinic acid receptor—a new mecha- Yang, H., Yang, T., Baur, J.A., Perez, E., Matsui, T., Carmona, J.J., Lamming,
nism for an old drug. Lancet 363, 1892–1894. D.W., Souza-Pinto, N.C., Bohr, V.A., Rosenzweig, A., et al. (2007a). Nutrient-
Kemper, J.K., Xiao, Z., Ponugoti, B., Miao, J., Fang, S., Kanamaluru, D., Tsang, sensitive mitochondrial NAD+ levels dictate cell survival. Cell 130, 1095–1107.
S., Wu, S.Y., Chiang, C.M., and Veenstra, T.D. (2009). FXR acetylation is nor- Yang, T., Chan, N.Y., and Sauve, A.A. (2007b). Syntheses of nicotinamide
mally dynamically regulated by p300 and SIRT1 but constitutively elevated in riboside and derivatives: effective agents for increasing nicotinamide adenine
metabolic disease states. Cell Metab. 10, 392–404. dinucleotide concentrations in mammalian cells. J. Med. Chem. 50, 6458–
Lagouge, M., Argmann, C., Gerhart-Hines, Z., Meziane, H., Lerin, C., Daussin, 6461.
F., Messadeq, N., Milne, J., Lambert, P., Elliott, P., et al. (2006). Resveratrol Yoshino, J., Mills, K.F., Yoon, M.J., and Imai, S. (2011). Nicotinamide mononu-
improves mitochondrial function and protects against metabolic disease by cleotide, a key NAD(+) intermediate, treats the pathophysiology of diet- and
activating SIRT1 and PGC-1alpha. Cell 127, 1109–1122. age-induced diabetes in mice. Cell Metab. 14, 528–536.

Cell Metabolism 15, 838–847, June 6, 2012 ª2012 Elsevier Inc. 847