Lithium increases PGC-1a expression and mitochondrial
biogenesis in primary bovine aortic endothelial cells
Ian T. Struewing, Corey D. Barnett, Tao Tang and Catherine D. Mao
Graduate Center for Nutritional Sciences, University of Kentucky, Lexington, USA
Lithium is commonly used as a therapeutic agent in
the treatment of bipolar disorder (or maniac depres-
sion) [1], and as a mimic of Wnt signaling both in vivo
and in vitro [2]. The beneficial effects of lithium in
the treatment of bipolar disorder are thought to be
due to a combination of activation of the Wnt ⁄ b-catenin
signaling pathway, via inhibition of the glycogen syn-
thase kinase-3b (GSK3b) [3], and depletion of the
intracellular inositol pool via the inhibition of various
enzymes in the phosphoinositide pathways, for example,
the rate-limiting enzyme inositol monophospha-
tase 1 (IMPase-1) [4,5]. In addition, evidence suggests
that lithium is neuroprotective and is beneficial in the
treatment of ischemia–reperfusion injuries and neuro-
degenerative diseases, including Alzheimer’s, Parkin-
son’s and Huntington’s disease [6]. For Huntington’s
disease, it has been proposed that lithium increases
the degradation of aggregated Hungtintin-mutated
Keywords
cell signaling; CREB; FOXO; gene
expression; mitochondria
Correspondence
C. D. Mao, Graduate Center for Nutritional
Sciences, University of Kentucky, 900
Limestone Street, Lexington, KY 40536,
USA
Fax: +1 859 257 3646

BAEC, bovine aortic endothelial cell; COX, cytochrome c oxidase; CREB, c-AMP responsive element binding; DCF, 2¢-7¢-dichlorofluorescein;
FOXO, Forkhead box class O; GSK3, glycogen synthase kinase; IMP, inositol monophosphatase; LPS, lipopolysaccharide; MTP, mitochondria
transmembrane potential; NRF, nuclear respiratory factor; OXPHOS, oxidative phosphorylation; PGC, peroxisome proliferators-activated
receptor-gamma coactivator; PRC, PGC-1a-related coactivator; TFAM, transcription factor A mitochondria; TFB, transcription factor B; UCP,
uncoupling protein.
FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS 2749
proteins via autophagy in an IMPase-1-dependent manner
[7]. By contrast, the preconditioning and protective
effects of lithium in brain and heart ischemia–reper-
fusion injury models appear to depend upon the inhi-
bition of GSK3b [8,9]. Although much attention has
been paid to the inhibitory effects of lithium on
GSK3b and IMPase-1 activity, lithium acts as a com-
petitive inhibitor of numerous Mg
2+
-dependent fac-
tors, transporters and enzymes, including a key
glycolytic enzyme, phosphoglucomutase [10]. Such a
wide spectrum of potential targets is consistent with
the narrow range of lithium doses that can be used
therapeutically in the absence of toxicity and with lim-
ited side effects [1].
GSK3b plays a pivotal role in the canonical Wnt ⁄
b-catenin signaling pathway by phosphorylating and
targeting b-catenin to the proteasomal degradation
pathway in the absence of Wnt signals. In the presence
of Wnt signals, GSK3b becomes phosphorylated on
Ser9 and is inactivated, allowing the cytosolic stabilization
and nuclear translocation of b-catenin. In the nucleus,
b-catenin interacts with the TCF ⁄ LEF transcription

chondria of cerebellar cells [19] and enrichment of act-
ive GSK3b in mitochondria and nuclei has also been
observed in neuronal cells [20]. In the immortalized
neuronal cell line SH-SY5Y, mitochondrial and active
GSK3b was shown to interact with p53 and thereby
promote the pro-apoptotic activities of p53 [21].
By contrast, activation ⁄ inhibition of GSK3b was
shown to control glycolysis–oxidative phosphorylation
(OXPHOS) coupling and apoptosis in HeLa tumor
cells via the release ⁄ binding of hexokinase II from the
mitochondria outer membrane [22]. However, these
effects appear to be cell-type dependent, because lithium
had the opposite effect in B16 melanoma cells in
association with a cell-cycle arrest [17]. Recently, a
novel p53 target has been identified, synthesis of cyto-
chrome oxidase 2 (SCO2), which is responsible for bio-
genesis of the cytochrome oxidase complex in the inner
mitochondrial membrane and thus links the tumor
suppressor p53 with the control of energy metabolism
and glycolysis switching in tumor cells [23]. Although
lithium has been shown to affect glycolysis via direct
inhibition of phosphoglucomutase [10], its effects on
mitochondrial energy metabolism and biogenesis have
not been addressed.
Mitochondrial biogenesis is highly orchestrated, and
involves signal cross-talk between the nucleus and
mitochondria leading to the coordinated regulation of
gene expression [24,25]. The mitochondrial genome
encodes only 13 OXPHOS components; the other
OXPHOX components and all the factors required for

2750 FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS
mitochondrial dysfunction by interfering with cAMP-
responsive element binding (CREB) transcription fac-
tor-dependent regulation of PGC-1a expression [30].
Inversely, increasing the expression of PGC-1a was
shown to be neuroprotective in the mutated-hungtintin
transgenic mouse model of Huntington disease [30].
Interestingly, the preconditioning effects of various
substances and factors in brain, heart and vessel
ischemic–reperfusion injuries also seem to be mediated
in part by an increase in mitochondrial biogenesis and
preservation of mitochondrial function [31,32]. The
protective effects of lithium, both during precondition-
ing in heart and brain ischemic–reperfusion injury
models [8,9], and in neurodegenerative disease models
[6], have been studied only in relation to the apoptotic
function of mitochondria, and not their energy homeo-
static function.
In this study, we show that lithium increases mitoch-
ondrial biogenesis in BAEC leading to an increase in
ATP production. Unexpectedly, this novel effect of
lithium was independent of the inhibition of GSK3b
and of inositol depletion. Moreover, our results reveal
that lithium treatment affects two cascades known to
converge to the upregulation of PGC-1a expression by
CREB and Forkhead box class O (FOXO1) trans-
cription factors and hence to increase mitochondrial
biogenesis.
Results
Lithium increases mitochondrial mass in BAEC

1.5
2
10 mM
Na
5 m
M
Li
10 m
M
Li
1 m
M
VPA
airdnohcotiMehtfoslevelevitaleR
laitnetoPenarbmemsnarT
B
)UA(ssaMa
ird
nohcot
iMevitaleR
AND-uNsusrevAN
D-tiM
12S/TFAM Cytb/ATPbeta
CR/RRS1
0
0.25
0.5
0.75
1
1.25

em535
. The
MTP levels were corrected for the variation in cell number and the
results are expressed as relative levels with the control NaCl-trea-
ted cells equal to 1. The graph represents mean ± SEM obtained in
eight independent experiments preformed in triplicate. (C) Lithium
increases mitochondrial mass in BAEC. BAEC were treated with
10 m
M NaCl (Na), 10 mM LiCl (Li) or 1 mM valproate (VPA) for 36 h
prior to isolation of total DNA. Levels of mitochondrial DNA and lev-
els of nuclear DNA were quantified by real-time PCR with specific
bovine primers for the mitochondrial encoded genes, cytochrome b
(cyt-b) and 12S rRNA, as well as for the mitochondria genome
control region (CR) and for the nuclear encoded genes: TFAM,
ribosome biogenesis regulator-1 (RRS1) and ATP synthase-b
(ATP-beta). The ratio of mitochondrial DNA to nuclear DNA was
determined for each treatment and for each of the pair of
MitDNA ⁄ NuDNA: 12S ⁄ TFAM, Cytb ⁄ ATPb and CR ⁄ RRS1. Results
are expressed in relative levels with the control NaCl equal to 1.
Mean ± SEM obtained from 3–4 independent experiments are
reported in the graph. In all cases, after Student’s t-test analysis,
the results were considered significant at P < 0.05 (*).
I. T. Struewing et al. Lithium increases mitochondrial biogenesis
FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS 2751
treatment (P<0.05) (Fig. 1A). Because lithium was
shown to inhibit various enzymes of the glycolytic and
tricarboxylic acid pathways and to decrease ATP pro-
duction via glycolysis [17,34], we first tested whether
lithium increases ATP production via changes in mito-
chondrial activity and ⁄ or mass in BAEC. The mito-

three encoded by the nuclear genome to avoid bias of
differential efficiency of amplification between primer
sets. As shown in Fig. 1C, lithium treatment increased
significantly the relative levels of mitochondrial DNA,
about 1.35 ± 0.14, 1.4 ± 0.07 and 1.49 ± 0.07-fold
for 12S ⁄ TFAM, cytochrome b ⁄ ATP synthase-b and
CR ⁄ RRS1 MitDNA ⁄ NuDNA pairs, respectively, which
indicated an increase of the mitochondrial mass,
whereas VPA had no significant effect. Taken together,
these results show that lithium treatment in BAEC
increases mitochondrial mass significantly, leading to
an increase in ATP production without changes in mit-
ochondrial efficiency. This lithium-induced increase in
mitochondrial mass was not accompanied by any signi-
ficant changes in mitochondrial morphology or distri-
bution, as shown by immunofluorescence microscopy
of BAEC stained with Mitotracker-Deep Red (Fig. 2).
Mitochondria in lithium-treated BAEC were found
around the nucleus and protrusion, as in control cells,
repeated treatments with 50 lm of the NO donor,
DETA-NO, a known inducer of mitochondrial biogen-
esis [36], also increased the mitochondrial mass but the
mitochondria were mainly perinuclear (Fig. 2).
Lithium increases mitochondrial biogenesis
markers
To determine whether the increase in mitochondrial
mass was due to an increase in mitochondrial bio-
genesis, the mRNA levels of various OXPHOS compo-
nents encoded by either the nuclear genome or the
mitochondrial genome were assessed using real-time

out reaching statistical significance (Fig. 3). Levels of
uncoupling protein 2 (UCP2) mRNA were also signifi-
cantly increased after 36 h lithium treatment
 1.4 ± 0.1 fold, which was similar to the increase
observed for the mitochondrial biogenesis markers.
This coordinated increase in mRNA levels for the mit-
ochondrial biogenesis markers and UCP2 is in agree-
ment with the unaffected OXPHOS efficiency observed
after lithium treatment (Fig. 1). To confirm that the
increase in mRNA was accompanied by an increase in
protein levels, we assessed the expression of the ATP
synthase-b protein by immunoblotting. A significant,
1.7 ± 0.3-fold, increase in ATP synthase-b was
observed after lithium treatment, although valproate
had no effect (Fig. 3C). Our results show that lithium
increased the expression of mitochondrial biogenesis
markers in a coordinated fashion, although the effects
of lithium on mRNA levels for mitochondrial-encoded
genes are stronger than the effects on the expression of
nuclear-encoded genes. This may be due to an increase
in mitochondrial mass (Fig. 1) and ⁄ or a greater
increase in the expression of mitochondrial transcrip-
tion factors.
Lithium increases mRNA levels for transcription
factors involved in mitochondrial biogenesis
Expression of OXPHOS genes encoded by the mito-
chondrial genome is under the control of specific tran-
scription factors: TFAM, TFB1 and TFB2, whose
expression, as well as expression of the nuclear-enco-
ded OXPHOS genes, is mainly under the control of

were considered significant at P < 0.05 (*) (Student’s t-test).
A
0
1
2
3
4
5
**
Fold increase of mRNA levels
of nuclear enclosed genes
Fold increase of mRNA levels of
mitochondrial enclosed genes
*
Na Li VPA
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
*
Na Li VPA
120.96
Ratio
α
-Tubulin

synthase
−
β
COX-VIc COX-VIa cyt-c
Mit-DNA
polymerase
0
1
2
3
4
5
24h 36h
*
**
*
*
*
I. T. Struewing et al. Lithium increases mitochondrial biogenesis
FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS 2753
c-Myc, a b-catenin target gene [37], has also been
shown to increase the levels and activity of TFAM
[38]. Therefore, we tested the mRNA levels of all these
transcription factors in response to lithium. A signifi-
cant increase in mRNA levels for TFAM and TFB2
was seen after 24 h lithium treatment,  2.2 ± 0.5-
and 1.7 ± 0.2-fold, respectively, and these effects
were even more pronounced at 36 h with a 4.4 ± 1.1-
and 4.1 ± 1-fold increase, respectively (Fig. 4A).
Among the nuclear transcription factors, only NRF2b

Treatment of BAEC with 10 mm lithium for between
30 min and 2 h had no significant effect on intra-
cellular H
2
O
2
levels, whereas treatment with 1 lm LPS
resulted in a small but significant increase after 2 h
treatment ( 1.36 ± 0.1-fold; Fig. 4B). Longer lithium
treatments resulted in a decrease in peroxide produc-
tion (not shown). Therefore, lithium-induced mito-
chondrial biogenesis was not due to a compensatory
mechanism following mitochondrial damage induced
by an increase in oxidative stress.
Lithium effects on mitochondrial biogenesis are
partially dependent on inositol depletion
Lithium is a competitive inhibitor of various enzymes
of the inositol pathway including the limiting enzyme
IMPase-1, resulting in a marked depletion of the intra-
cellular inositol pool that can be restored by the addi-
tion of myo-inositol [5]. We tested whether lithium was
able to increase the expression of mitochondrial bio-
genesis markers after pretreatment with 1 mm myo-
inositol. As shown in Fig. 5A, addition of 1 mm
myo-inositol attenuated the effects of lithium with a
20–25% decrease in mRNA levels for TFAM, cyto-
chrome b and ATP synthase-6, although it did not sig-
nificantly affect these levels in NaCl-treated cells.
However, the changes between LiCl + myo-inositol-
treated cells and LiCl-treated cells were not statistically

0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
30 min 2h
NaCl LiCl LPS
leveledixorepnegordyhfoeg
n
a
h
cd
l
oF
*
*
B
Fig. 4. Lithium increases the mRNA levels of transcription factors
involved in the control of mitochondrial biogenesis in the absence
of oxidative stress. (A) Lithium increases the mRNA levels of
mitochondrial biogenesis transcription factors. BAEC were treated
for the indicated times with either 10 m
M NaCl as a control or
10 m
M LiCl prior to RNA extraction, and the levels of target mRNAs
were quantified using real-time RT-PCR. The graph represents
mean ± SEM results obtained from five independent experiments.

comparing the effects of two other unrelated inhibitors
of GSK3b, valproate and indirubin-3¢-monoxime. Val-
proate has been shown to indirectly inhibit GSK3b
[39], although its activation of the Wnt ⁄ b-catenin sign-
aling pathway in various cells appears to depend
mainly on histone deacetylases 2 inhibition [40].
Among these various inhibitors, only lithium treatment
led to a significant increase in the mRNA l evels of mito-
chondrial-encoded genes, ATP synthase-6 and cyto-
chrome b, nuclear-encoded genes, ATP synthase-b,
cytochrome oxidase VIc and TFAM (Fig. 6A). By
contrast, indirubin had no effect, whereas valproate
increased,  1.5-fold, mRNA levels for the nuclear-
encoded genes cytochrome oxidase VIc and TFAM
(Fig. 6A). These results were in agreement with a weak
or lack of effect of valproate on ATP production and
mitochondrial mass, respectively (Fig. 1). To further
rule out a role for GSK3b in the lithium-dependent
increase in mitochondrial biogenesis markers, BAEC
were transfected with wild-type GSK3b, constitutive
active S9A-GSK3b and the inactive kinase-dead
K85A-GSK3b for 36 h prior to the analysis of mito-
chondrial and nuclear gene expression. If the effects of
lithium on mitochondrial biogenesis were dependent
on GSK3b inhibition, expression of the catalytic
inactive K86R-GSK3b form should also increase
expression of the mitochondrial biogenesis markers.
However, mRNA levels of the nuclear genes ATP syn-
thase-b, cytochrome oxidase VIc and TFAM, and of
the mitochondrial genes ATP synthase-6 and cyto-

0
1
2
3
4
5
6
Na Na + myo-inositol Li Li + myo-inositol
*
*
*
*
*
*
*
*
A
0
0.25
0.5
0.75
1
1.25
1.5
1.75
lairdnohcotim fo slevel evitaleR
AND raelcunsv AND
NaCl LiCl
LiCl +
myo-inositol

activating Akt-S473 phosphorylation significantly,
about twofold, as early as 30 min into treatment
(Fig. 7B), and this decrease persisted over 24 (Fig. 7B)
and 36 h (not shown). This decrease in active Akt was
consequently associated with a decrease at 6 h in the
transcription factor FOXO1 phosphorylation on
Thr24, an Akt substrate site in vivo [45] (Fig. 7B). Val-
proate, like lithium, did not induce an increase but
rather a decrease in Akt-S473 phosphorylation,
although the latter occurred later after 24 h treatment
(Fig. 7B). Similarly, although to a lesser extent than
lithium, valproate increased inhibitory GSK3b-S9
phosphorylation. By contrast, indirubin induced rapid
disappearance of both the activating Akt-S473 and the
inhibitory GSK3b-S9 phosphorylations at treatment
periods between 5 min and 24 h, compared with con-
trol-treated cells (Fig. 7B). Therefore, valproate and
lithium appeared to induce similar changes in the
Akt ⁄ FOXO1 signaling cascade in primary BAEC
except for stronger and faster effects with lithium.
Lithium increases the expression of PGC-1a
in BAEC
The finding that the Akt ⁄ FOXO1 cascade is affected
by lithium in BAEC prompted us to investigate the
effects of lithium on the expression of PGC-1a as it
has previously been shown that activation of Akt led
to downregulation of PGC-1a expression via nuclear
exclusion of FOXO1 in skeletal muscle cells [46]. Lev-
els of PGC1-a mRNA, as well as those of the related
coactivator PRC, were determined using real-time

in treated Vs control NaCl treated BAEC
Li Val Ind
*
*
*
*
*
*
A
Fold increase of target mRNA levels
in GSK3
β
versus control transfected BAEC
B
ATP
synthase–
β
COX-VIc
cyt-b TFAM
ATP
synthase-6
0
1
2
3
4
5
6
Control WT S9A K85R
IL-8

mechanisms: activation of FOXO1 and CREB.
Discussion
Lithium is commonly used to treat bipolar disorder [1]
and recent evidence suggests that it might also be
beneficial in the treatment of neurodegenerative dis-
eases [6]. However, the mechanisms involved in both
the beneficial effects and side effects of lithium are not
fully identified. We report a novel effect of lithium at
doses commonly used to inhibit GSK3b activity and
mimic Wnt signaling [33]. In primary endothelial cells,
lithium treatment triggered an increase in mitochond-
rial mass and ATP production without changing
mitochondrial efficiency. This in crease in mitochondrial
biogenesis correlated with the upregulation of key
master controllers of mitochondrial biogenesis: tran-
scription factors NRF1 and NRF2b and coactivator
PGC-1a [25,26]. In addition, we showed that two
different signaling cascades known to regulate PGC-1a
expression, inactivation of Akt [46] and activation of
CREB [26], were triggered by lithium treatment.
An increase in mitochondrial biogenesis has been
described in numerous physiological conditions as an
adaptive mechanism during muscle exercise, calorie
restriction, hormone treatment and cell differenti-
A
Akt
β
β
-actin
phosphoS473-Akt

VPA
PhosphoT24-FOXO1/FOXO1
0
0.5
1
1.5
2
2.5
3
Li
VPA
C
Na Li VPA IndNa Li VPA IndNa Li VPA IndNa Li VPA Ind
24 h
NaCl LiCl VPA Indirubin
Fig. 7. Lithium increases BAEC cell size and affects Akt ⁄ FOXO1 signaling cascade. (A) Lithium increases the spreading and size of BAEC.
BAEC were plated for 12 h prior to being treated with 10 m
M NaCl, 10 mM LiCl, 1 mM VPA or 5 lM indirubin for 36 h. Phase-contrast micro-
scope images were taken and representative images are shown for each treatment. (B) Lithium increases the inhibitory phosphorylation of
GSK3b on Ser9 in absence of Akt activation. BAEC were treated as indicated in (A) for 5 min, 30 min, 2 h and 24 h prior to cell harvesting
and analysis of GSK3b and Akt phosphorylations using immunoblotting with specific phospho-S9-GSK3b and phospho-S473-Akt antibodies.
Akt-dependent phosphorylation of FOXO1 on Thr24 was also studied in parallel. Total levels of GSK3b, Akt and FOXO1 were used to normal-
ize for changes in expression and b-actin was used as loading control. A representative experiment is shown and the fold changes obtained
after lithium treatment from four independent experiments are reported in the graphs.
I. T. Struewing et al. Lithium increases mitochondrial biogenesis
FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS 2757
ation, as well as in various pathological situations to
compensate for mitochondrial dysfunction or damage
[24,25]. It is possible that lithium as a potential com-
petitive inhibitor of some Mg

ding BAEC [13,18]. The microtubule polymerizing
agent, taxol, has been shown to induce both an increase
in mitochondrial biogenesis and G
2
⁄ M cell-cycle arrest
in the human 143B osteosarcoma cell line, but unlike
lithium, these changes were associated with an abnor-
mal distribution of mitochondria around the nucleus
[52]. By contrast, mitochondrial DNA replication starts
at the G
1
⁄ S phase transition, whereas mitochondrial
biogenesis peaks in the G
2
⁄ M phase, allowing equal
distribution of mitochondria between the two daughter
cells during cytokinesis [49,53]. Thus, the lithium-
induced cell-cycle arrest in G
2
⁄ M might be sufficient to
explain lithium-induced mitochondrial biogenesis.
Our results also showed that this lithium-dependent
increase in mitochondrial biogenesis in BAEC was
associated with an increase in mRNA levels for coacti-
vator PGC-1a but not coactivator PRC (Fig. 8A). This
is consistent with the cell-cycle arrest induced by lith-
ium. Indeed, regulation of PRC expression is mainly
dependent on cell-proliferation status, i.e. increased
in the presence of growth factors and decreased in
contact-inhibited cells [26]. However, regulation of

Na Li VPA Ind
8 h
B
Treatments
phosphoS133-CREB
phosphoS133-ATF1
CREB
β
-actin
0
0.5
1
1.5
2
2.5
Li
VPA
Fold changes versus NaCl
PhosphoS133-CREB/CREB
8 h
Fig. 8. Lithium increases the expression of the coactivator PGC-1a.
(A) Lithium increases the levels of PGC-1a mRNA. BAEC were trea-
ted for 36 h with 10 m
M NaCl, 10 mM LiCl, 1 mM VPA or 5 lM
indirubin prior to RNA extraction and the levels of PGC-1a and PRC
mRNAs were quantified using real-time RT-PCR. The graph repre-
sents mean ± SEM results obtained from five independent experi-
ments. (B) Lithium increases the levels of phospho-S133-CREB.
BAEC were treated for 8 h with 10 m
M NaCl, 10 mM LiCl, 1 mM

rial biogenesis is not secondary to the establishment of
cell senescence.
However, regulation of PGC-1a is also triggered by
metabolic and environmental stresses [25,26]. Our
results showed that lithium induced a decrease in Akt
phosphorylation on Ser473 and Akt activity, because
phosphorylation of FOXO1 on Thr24, an Akt sub-
strate site, also decreased (Fig. 7). Although lithium
has been shown to increase Akt phosphorylation in
various cell lines, it is important to note that, in all
these studies, cells were serum-starved prior to the
addition of lithium, which was not the case in this
study. The observed decrease in Akt and FOXO1 phos-
phorylation was consistent with changes in the pattern
of gene expression observed after lithium treatment
(unpublished), which includes the increase in PGC-1a
expression (Fig. 8). Indeed, levels of PGC-1a expres-
sion have been shown to depend on the Akt ⁄ FOXO1
cascade in skeletal muscle [46]. Both the decrease in
activating Akt phosphorylation and increase in PGC-
1a expression are reminiscent of the induction of a
stress pathway, which remains to be identified as lith-
ium does not induce oxidative stress in BAEC (Fig. 4).
Interestingly, mild stresses triggered, for example, by
physical exercise or preconditioning agents such as the
K-ATP-dependent channel opener diazoxide, increase
mitochondrial biogenesis allowing the preservation of
mitochondrial functions during stronger stress [59,60].
The preconditioning effects of lithium have been des-
cribed both in brain and heart ischemia–reperfusion

biogenesis that appears highly relevant for the beneficial
effects of lithium treatment in bipolar disorder and
neurodegenerative diseases.
Experimental procedures
Materials
The chemicals lithium chloride, sodium valproate, d-
myo-inositol, DETA-NO (2,2¢-(hydroxynitrosohydrazono)
bis-ethanimine), indirubin-3¢-monoxime and the mouse
anti-(tubulin-a) mAb were from Sigma-Aldrich (St Louis,
MO). Mitochondria probes, Mitotracker-CMXRos and
Mitotracker-633-Deep Red, and the intracellular H
2
O
2
probe, 5- (and 6)-chloromethyl-2¢-7¢-dichlorodihydrofluo-
rescein diacetate (CM-H
2
DCFDA) and the rabbit poly-
clonal ATP synthase-b were from Molecular Probes
(Eugene, OR). Mouse anti-GSK3b mAb were from Transduc-
tion Laboratories (Lexington, KY) and the rabbit polyclonal
anti-(b-actin), anti-(phospho-S9-GSK3b), anti-(phospho-
S473-Akt), anti-(Akt), anti-(phosphoT24 ⁄ T32-FOXO1 ⁄ 3a),
anti-(FOXO1, antiphosphoS133-CREB) and anti-CREB
sera, and secondary horseradish-peroxidase-conjugated anti-
bodies were from Cell Signaling (Beverly, MA).
I. T. Struewing et al. Lithium increases mitochondrial biogenesis
FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS 2759
Cell culture and transfection
Primary BAEC were purchased from Cambrex (Walkers-

BAEC were treated with either 10 mm NaCl or 10 mm LiCl
for 48 h, whereas treatment with 50 lm DETA-NO was
performed for 4 days with repeated treatments every 24 h
as described previously [36], prior to being stained with
200 nm Mitotracker-Deep Red for 45 min accordingly to
the manufacturer’s recommendation (Molecular Probes).
Slides were then mounted in presence of the antifading
agent vectashield containing the nuclear stain DAPI (Vec-
tor Laboratories, Burlingame, CA). Serial images were
taken using Leica immunofluorescence confocal micro-
scope.
Whole-cell extracts and western blot analysis
After washing with cold NaCl ⁄ P
i
, cells were lysed in
50 mm Hepes pH 7.4, 0.1% Chaps, 5 mm dithiothreitol
and 2 mm EDTA supplemented with protease and phos-
phatase cocktail inhibitors (Sigma-Aldrich). Equal amounts
of the proteins were denaturated by boiling in Laemmli
buffer, fractionated on SDS–PAGE and transferred onto
Immobilon P membrane (Millipore, Billerica, MA). After
blocking, membranes were incubated with the various pri-
mary antibodies as indicated and subsequently with the
appropriate secondary antibodies conjugated to horseradish
peroxidase (Cell Signaling). Immunoreactive proteins were
detected using SuperSignalÒ chemiluminescence (Pierce
Chemical Co, Rockford, IL) and the intensity of the result-
ing bands was determined by densitometry using scion
software. To control for variations in loadings, the same
membrane was stripped, washed and blocked prior to

and centrifuged at 500 g for 5 min.
The cell pellet was then resuspended in 200 lL of NaCl ⁄ P
i
with 2 lgÆmL
)1
RNAse A prior to being subjected to
DNA extraction using the DNeasy kit (Qiagen, Valencia,
CA). The amounts of mitochondrial and nuclear DNA
were determined by real-time PCR using the SyBr green
Core reagent kit (Applied Biosystems). Two different
genes for each genome were studied in parallel to minim-
ize variation in primer pair efficiency. We used specific
primer pairs for cytochrome b, mitochondria control
region and 12S ribosomal RNA as markers of the mitoch-
ondrial genome and ATP synthase subunit-b, ribosome
biogenesis regulator-1 and mitochondrial transcription fac-
tor A as markers of the nuclear genome. The sequences
of the various primers are provided in Table 1. For each
primer set, quantification was performed in duplicate with
Lithium increases mitochondrial biogenesis I. T. Struewing et al.
2760 FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS
10 ng total DNA per reaction. The differences in the
MitDNA-Ct versus the NuDNA-Ct (DMit–NuDNA-Ct)
were calculated for each condition and for each MitDNA–
NuDNA pair: 12S ⁄ TFAM, cytochrome b ⁄ ATP-synthase-b
and CR ⁄ RRS1. The fold variations of the MitDNA levels
versus NuDNA levels following treatments were deter-
mined using the equation 2DMit–NuDNA-Ct treated )
DMit–NuDNA-Ct control, where control cells represent
NaCl-treated cells. Three to four independent experiments

Cytochrome c1 Nuclear Fwd: CCAGGTAGCCAAGGATGTGT
Rev: GACCCTGAAGCTCAGGACAG
COX VIa Nuclear Fwd: GGAAGGCCCTCACCTACTTC
Rev: CGGGTTCACATGAGGGTTAT
COX VIc Nuclear Fwd: GCTTTGGCAAAACCTCAGAT
Rev: ACCAGCCTTCCTCATCTCCT
ATP synthase subunit b Nuclear Fwd: ACAGGACCCTATGTGCTTGG
Rev: ATCAGCAAATTCCCCAACAG
Uncoupling protein-2 Nuclear Fwd: ATGACAGACGACCTCCCTTG
Rev: GGCATGAACCCTTTGTAGAAG
Mit-transcription factor A
(TFAM)
Nuclear Fwd: GGGAGGAACAAATGATGGAA
Rev: CCATGGGCTACAGAAAAGGA
Mit-transcription factor B2
(TFB2)
Nuclear Fwd: GTACAAGTCCCGTTCCGAGAC
Rev: CACTCTGGCACCACTTTCAAG
NRF1 Nuclear Fwd: ACCGCCGAATAATTCACTTG
Rev: CACAAACACAGGCCACAACC
NRF2b ⁄ GABP-b Nuclear Fwd: CATTGTGACCATGCCAGATG
Rev: GTAGGCCTCTGCTTCCTGTTC
c-myc Nuclear Fwd: CTCCTCACAGCCCGTTAGTC
Rev: CGCCTCTTGTCATTCTCCTC
PGC1-a Nuclear Fwd: CCGAGAATTCATGGAGCAAT
Rev: GATTGTGTGTGGGCCTTCTT
PRC Nuclear Fwd:
GCTGAGAATGTGGCTGTTGA
Rev: TCACTGATGAAAGCCTGCAC
rpL30 Nuclear Fwd: CTCAACGAGAACAAGCTATC

in NaCl ⁄ P
i
with Ca
2+
⁄ Mg
2+
. After three washes in NaCl ⁄ P
i
with Ca
2+
⁄ Mg
2+
, the fluorescence intensities associated with
active mitochondria were measured at k
exc550
⁄ k
em590
using a
Fusion
TM
plate reader (Perkin–Elmer, Waltham, MA). Cell
number was determined using CyQuant staining accordingly
to the manufacturer’s instructions (Molecular Probe) and
measurement of the fluorescence intensity at k
exc485
⁄ k
em535
.
The mitochondria membrane potential per cell was calcula-
ted as the ratio of the CMXRos fluorescent intensity versus

as described above, the
results were normalized for variation in cell numbers using
CyQuant staining and quantification, thought in this case
separate wells were used. The results obtained from five inde-
pendent experiments performed in triplicate are presented as
the relative levels of hydrogen peroxide production with the
control NaCl-treated cells being equal to 1.
Statistical analysis
After verification of the normal distribution of the values
obtained in at least three independent experiments,
Student’s t-test and one-way anova were performed using
graphpad prism v. 4.0 (GraphPad Software, San Diego,
CA). Results were considered significant at P < 0.05.
Acknowledgements
We thank I. Dominguez for the kind gift of wt- and
K85R-GSK3b constructs. This work was supported by
NIH grant HL68698 to CDM.
References
1 Pilcher HR (2003) The ups and downs of lithium.
Nature 425, 118–120.
2 Hedgepeth C, Conrad LJ, Zhang J, Huang HC, Lee
VM & Klein PS (1997) Activation of the Wnt signaling
pathway: a molecular mechanism for lithium action.
Dev Biol 185, 82–91.
3 Klein PS & Melton DA (1996) A molecular mechanism
for the effect of lithium on development. Proc Natl
Acad Sci USA 93, 8455–8459.
4 Belmaker RH, Livne A, Agam G, Moscovich DG,
Grisaru N, Schreiber G, Avissar S, Danon A & Kofman
O (1990) Role of inositol-1-phosphatase inhibition in

Lithium increases mitochondrial biogenesis I. T. Struewing et al.
2762 FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS
12 Williams RS & Harwood AJ (2000) Lithium therapy and
signal transduction. Trends Pharmacol Sci 21, 61–64.
13 Mao CD, Hoang P & DiCorleto PE (2001) Lithium
inhibits cell cycle progression and induces stabilization
of p53 in bovine aortic endothelial cells. J Biol Chem
276, 26180–26188.
14 Rossig L, Badorff C, Holzmann Y, Zeiher AM &
Dimmeler S (2002) Glycogen synthase kinase-3 couples
AKT-dependent signaling to the regulation of p21Cip1
degradation. J Biol Chem 277, 9684–9689.
15 Cheng CW, Smith SK & Charnock-Jones DS (2003)
Wnt-1 signaling inhibits human umbilical vein endothe-
lial cell proliferation and alters cell morphology. Exp
Cell Res 291, 415–425.
16 Stump RJ, Lovicu FJ, Ang SL, Pandey SK & McAvoy
JW (2006) Lithium stabilizes the polarized lens epithelial
phenotype and inhibits proliferation, migration, and
epithelial mesenchymal transition. J Pathol 210, 249–
257.
17 Penso J & Beitner R (2003) Lithium detaches hexoki-
nase from mitochondria and inhibits proliferation of
B16 melanoma. Mol Genet Met 78, 74–78.
18 Smits VA, Essers MA, Loomans DS, Klompmaker R,
Rijksen G & Medema RH (1999) Inhibition of cell pro-
liferation by lithium is associated with interference in
cdc2 activation. FEBS Lett 457, 23–27.
19 Hoshi M, Sato M, Kondo S, Takashima A, Noguchi K,
Takahashi M, Ishiguro K & Imahori K (1995) Different

78–90.
27 Suliman HB, Carraway MS, Welty-Wolf KE, Whorton
AR & Piantadosi CA (2003) Lipopolysaccharide stimu-
lates mitochondrial biogenesis via activation of nuclear
respiratory factor-1. J Biol Chem
278, 41510–41518.
28 Kwong JQ, Beal MF & Manfredi G (2006) The role of
mitochondria in inherited neurodegenerative diseases.
J Neurochem 97, 1659–1675.
29 Sun X, Wang J-F, Tseng M & Young LT (2006) Down-
regulation in components of the mitochondrial electron
transport chain in the post mortem frontal cortex of
subjects with bipolar disorder. J Psychiat Neurosci 31,
189–196.
30 Cui L, Jeong H, Borovecki F, Parkhurst CN, Tanese N
& Krainc D (2006) Transcriptional repression of PGC-
1alpha by mutant huntingtin leads to mitochondrial
dysfunction and neurodegeneration. Cell 127, 59–69.
31 McLeod CJ, Jeyabalan AP, Minners JO, Clevenger R,
Hoyt RFJ & Sack MN (2004) Delayed ischemic precon-
ditioning activates nuclear-encoded electron-transfer-
chain gene expression in parallel with enhanced posta-
noxic mitochondrial respiratory recovery. Circulation
110, 534–539.
32 Gidday JM (2006) Cerebral preconditioning and ischae-
mic tolerance. Nat Rev Neurosci 7, 437–448.
33 Williams RSB, Cheng L, Mudge AW & Harwood AJ
(2002) A common mechanism of action for three mood-
stabilizing drugs. Nature 417, 292–295.
34 Fonseca CP, Jones JG, Carvalho RA, Jeffrey FM,

The mood-stabilizing agent valproate inhibits the activ-
ity of glycogen synthase kinase-3. J Neurochem 72,
1327–1330.
40 Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar
MA & Klein PS (2001) Histone deacetylase is a
direct target of valproic acid, a potent anticonvulsant,
mood stabilizer, and teratogen. J Biol Chem 276,
36734–36741.
41 Nemeth ZH, Deitch EA, Fekete CSZ, Hauser CJ &
Hasko G (2002) Lithium induces NF-jB activation and
interleukin-8 production in human intestinal epithelial
cells. J Biol Chem 277, 7713–7719.
42 Vyas DR, Spangenburg EE, Abraha TW, Childs TE &
Booth FW (2002) GSK-3beta negatively regulates skele-
tal myotube hypertrophy. Am J Physiol Cell Physiol
283, C545–C551.
43 Haq S, Choukroun G, Kang ZB, Ranu H, Matsui T,
Rosenzweig A, Molkentin JD, Alessandrini A,
Woodgett J, Hajjar R et al. (2000) Glycogen synthase
kinase-3beta is a negative regulator of cardiomyocyte
hypertrophy. J Cell Biol 151, 117–130.
44 Zhang F, Phiel CJ, Spece L, Gurvich N & Klein PS
(2003) Inhibitory phosphorylation of glycogen synthase
kinase-3 (GSK-3) in response to lithium. Evidence for
autoregulation of GSK-3. J Biol Chem 278,
33067–33077.
45 Biggs WHR, Meisenhelder J, Hunter T, Cavenee WK &
Arden KC (1999) Protein kinase B ⁄ Akt-mediated phos-
phorylation promotes nuclear exclusion of the winged
helix transcription factor FKHR1. Proc Natl Acad Sci

mitochondria in mammalian cells. J Cell Sci 114,
281–291.
53 Martinez-Diez M, Santamaria G, Ortega AD & Cuezva
JM (2006) Biogenesis and dynamics of mitochondria
during the cell cycle: significance of 3¢-UTRs. Public
Library of Science One 1, e107–118.
54 Duguez S, Feasson L, Denis C & Freyssenet D (2002)
Mitochondrial biogenesis during skeletal muscle
regeneration. Am J Physiol Endocrin Metab 282,
E802–E809.
55 van der Velden JL, Langen RC, Kelders MC, Wouters
EF, Janssen-Heininger YM & Schols AM (2006) Inhibi-
tion of glycogen synthase kinase-3beta activity is suffi-
cient to stimulate myogenic differentiation. Am J
Physiol Cell Physiol 290, C453–C462.
56 Polesskaya A, Seale P & Rudnicki MA (2003) Wnt sig-
naling induces the myogenic specification of resident
CD45+ adult stem cells during muscle regeneration.
Cell 113, 841–852.
57 Kang S, Bajnok L, Longo KA, Petersen RK, Hansen
JB, Kristiansen K & MacDougald OA (2005) Effects of
Wnt signaling on brown adipocyte differentiation and
metabolism mediated by PGC-1alpha. Mol Cell Biol 25,
1272–1282.
58 Hutter E, Renner K, Pfister G, Stockl P, Jansen-Durr P
& Gnaiger E (2004) Senescence-associated changes
in respiration and oxidative phosphorylation in
primary human fibroblasts. Biochem J 380,
919–928.
59 Kevelaitis E, Oubenaissa A, Peynet J, Mouas C &

dorsoventral axis formation in Xenopus embryos. Proc
Natl Acad Sci USA 92, 8498–8502.
I. T. Struewing et al. Lithium increases mitochondrial biogenesis
FEBS Journal 274 (2007) 2749–2765 ª 2007 The Authors Journal compilation ª 2007 FEBS 2765