ORIGINAL
E n d o c r i n e
ARTICLE
R e s e a r c h
Activation of Peroxisome Proliferator-Activated
Receptor Pathway Stimulates the Mitochondrial
Respiratory Chain and Can Correct Deficiencies in
Patients’ Cells Lacking Its Components
Jean Bastin, Flore Aubey, Agnès Rötig, Arnold Munnich, and Fatima Djouadi
Université Paris Descartes (J.B., F.A., F.D.), Centre National de la Recherche Scientifique Unité Propre de Recherche 9078, Faculté NeckerEnfants Malades, 75015 Paris, France; and Institut National de la Santé et de la Recherche Médicale U781 (A.R., A.M.), Hôpital NeckerEnfants Malades, 75015 Paris, France
Context: The mitochondrial respiratory chain (RC) disorders are the largest group of inborn errors
of metabolism and still remain without treatment in most cases.
Objective: We tested whether bezafibrate, a drug acting as a peroxisome proliferator-activated
receptor (PPAR) agonist, could stimulate RC capacities.
Design: Fibroblasts or myoblasts from controls or patients deficient in complex I (CI), complex III
(CIII), or complex IV (CIV) were cultured with or without bezafibrate.
Main Outcome Measures: Enzyme activities, mRNA and protein expression, and respiration rates
were measured.
Results: In control cells, bezafibrate increased the CI, CIII, and CIV enzyme activities (⫹42 to ⫹52%),
as well as RC mRNAs (⫹40 to ⫹120%) and RC protein levels (⫹50 to ⫹150%). Nine of 14 patient cell
lines tested exhibited a significant increase in the activity of the deficient RC complex after bezafibrate treatment (⫹46 to ⫹133%), and full pharmacological correction could be achieved in seven
cell lines. Similar effects were obtained using a PPAR␦ agonist. These changes were related to a
drug-induced increase in the mutated mRNAs and RC protein levels. Finally, the molecular mechanisms by which the PPAR pathway could induce the expression of genes encoding structural
subunits or ancillary proteins of the RC apparatus, leading to stimulate the activity and protein
levels of RC complex, likely involved the PPAR␥ coactivator-1␣.
Conclusions: This study suggests a rationale for a possible correction of moderate RC disorders due to
mutations in nuclear genes, using existing drugs, and brings new insights into the role of PPAR in the
regulation of the mitochondrial RC in human cells. (J Clin Endocrinol Metab 93: 1433–1441, 2008)
M
itochondrial disorders are among the most common
inborn metabolic diseases, affecting at least one in
8500 individuals (1). They are characterized by deficient activity of one, or more, of the mitochondrial respiratory chain
(RC) complexes [complex I (CI) to V] involved in oxidative
phosphorylation (OXPHOS). This group of orphan diseases
exhibits an extraordinary diversity of clinical presentations
affecting muscle and many different tissues, with almost any
age of onset, and course ranging from early death in the neonatal period to adult mild forms (2). The last 10 yr have seen
significant advances in the diagnosis and molecular analysis of
mitochondrial disorders, whereas little progress was made in
their treatment over the same period. Various pharmacological agents have been tested, including vitamins, cofactors,
0021-972X/08/$15.00/0
Abbreviations: CI, Complex I; CIII, complex III; CIV, complex IV; COX, cytochrome c oxidase;
KCN, potassium cyanide; NADH, reduced nicotinamide adenine dinucleotide; NQR, NADHdecylubiquinone oxidoreductase; NRF, nuclear respiratory factor; OXPHOS, oxidative phosphorylation; PGC-1␣, peroxisome proliferator activated receptor ␥ coactivator-1␣; PPAR, peroxisome proliferator-activated receptor; PPRE, PPAR response element; QCCR, decylubiquinol
cytochrome c reductase; RC, respiratory chain; RT-QPCR, real-time quantitative PCR; Tfam,
mitochondrial transcription factor A; TMPD, N,N,N⬘,N⬘-tetramethyl-p-phenylenediamine.
Printed in U.S.A.
Copyright © 2008 by The Endocrine Society
doi: 10.1210/jc.2007-1701 Received July 31, 2007. Accepted January 10, 2008.
First Published Online January 22, 2008
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
jcem.endojournals.org
1433
1434
Bastin et al.
Mitochondrial Disorders and PPAR Agonists
quinone derivatives, artificial electron acceptors, or free radical scavengers, but despite clinical improvements observed in
isolated cases, there is no hard evidence in support of these
treatments (3). Furthermore, most of them focus on symptoms
rather than on the cause of the disease, and in particular do not
target to the RC enzyme deficiency. In the present study, we
explored a different approach based on the hypothesis that
compounds capable of stimulating the expression of RC complexes might allow to improve or correct partial RC deficiencies, which are often encountered in patients with mitochondrial disorders (4).
One possible strategy to increase residual enzyme activity is to
act at the level of gene expression to stimulate the transcription.
Regulation of RC genes involves several specific transcription
factors, including the nuclear respiratory factors (NRFs) 1 and 2,
which bind and activate the promoters of various nuclear genes
encoding RC subunits or RC-associated proteins, and also regulate gene expression of the mitochondrial transcription factor A
(Tfam) (5, 6). Among recently characterized regulatory factors of
RC genes is the peroxisome proliferator activated receptor
(PPAR)␥ coactivator-1␣ (PGC-1␣). PGC-1␣ does not bind directly to DNA but rather docks on and coactivates the transcription factors already bound to the promoter region of target genes,
like NRF1 and NRF2 (5, 6). Accordingly, PGC-1␣ potentially
up-regulates not only the expression of nuclear genes encoding
RC components, but also the expression of Tfam, and, therefore,
is now considered to be pivotal in the coordinate regulation of
nuclear and mitochondrial genomes. This is supported by gainof-function and loss-of-function experiments, which underscore
the importance of PGC-1␣ in OXPHOS regulation in the mouse
skeletal muscle and heart (7).
PPARs are ligand-activated nuclear receptors involved in
the regulation of energy metabolism by controlling the expression levels of numerous genes, in particular in the mitochondrial fatty acid ␤-oxidation pathway (8, 9). Accordingly,
the PPARs have focused much attention as potential therapeutic targets for treatment of diabetes, obesity, or cardiovascular diseases. Interestingly, recent data clearly suggest
that PPAR-signaling pathway could also impact the RC. Thus,
targeted overexpression of PPAR␦ in the mouse muscle results
in a marked increase in the proportion of fibers with high
mitochondrial content, as revealed by succinate dehydrogenase (10) and cytochrome c oxidase (COX) expression studies
(11). More importantly, it has recently been shown that PPAR
agonists such as fibrates, drugs that have long been known to
ameliorate dyslipidemia, stimulate the expression of PGC-1␣
gene in mice skeletal muscle, both in vitro and in vivo (12).
Altogether, these data provide a rationale by which PPAR
agonists might regulate the RC.
In the present study, we therefore tested whether bezafibrate
or high-affinity PPAR agonists might up-regulate the expression
of the mitochondrial RC complexes, and if this could lead to
stimulate the expression of deficient RC enzyme activity or associated proteins in fibroblasts from patients with mitochondrial
disorders due to CI, CIII, or CIV deficiencies.
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
Patients and Methods
Patients
Patients 1– 8, reported previously, harbored nuclear mutations in RC
enzyme subunits or assembly factors (Table 1) (13–19). Six other patients
(nos. 9 –14) exhibited a CIV deficiency expressed in cell culture. As for
the majority of COX deficient patients, the disease-causing mutations in
these six patients are unknown (17, 20, 21) despite extensive efforts in
sequence analysis of candidate genes.
Cell culture
Skin fibroblast cell lines were established from patients 1–10, and
12 and 13. For patients 11 and 14, myoblasts derived from muscle
biopsy were studied. Fibroblasts and myoblasts were grown under
standard conditions (22). Bezafibrate (400 ␮M for fibroblasts or 200
␮M for myoblasts) or vehicle (dimethylsulfoxide), or PPAR-selective
compounds were added to the culture medium for the indicated time.
To determine optimal conditions for induction of RC enzyme by bezafibrate, COX activity was measured in dose-response and kinetics
experiments (Fig. 1). The PPAR-selective agonists used were GW␣
7647, GW␦ 0742, and GW␥ 7845 (obtained from Dr. Huet, GlaxoSmithKline, Les Ulis, France). The concentrations used, 1 ␮M for
GW␣ 7647, 10 nM for GW␦ 0742, and 1 ␮M for GW␥ 7845, are
selective for PPARs ␣, ␦, and ␥, respectively (22, 23).
Enzyme activity measurements
Fibroblasts pellets were homogenized, and enzyme activities were
measured by spectrophotometric methods described in Ref. 24. CI
was assayed after fibroblast treatment with digitonin and Percoll as
described in Ref. 25.
COX histochemistry
Myoblasts were grown in 24-well plates, and the medium was
replaced by 200 ␮l COX assay buffer [0.01% digitonin, 0.5 mg/ml
catalase, 1 mg/ml cytochrome c, and 0.5 mg/ml 3,3⬘-diaminobenzidine tetrahydrochloride, in 0.1 M 3-(N-morpholino)propanesulfonic
acid (pH 7.3). Myoblasts were incubated for 3 h at 37 C before
photography.
Real-time quantitative PCR (RT-QPCR)
The RT-QPCRs were performed as previously described in Ref. 22
using primers listed in supplemental Table 1, which is published as supplemental data on The Endocrine Society’s Journals Online web site at
http://jcem.endojournals.org. All primers spanned an intron/exon
boundary. The results were normalized for comparison by measuring
␤-actin mRNA levels in each sample.
Western blot analysis
Cell protein extracts (20 –25 ␮g) were resolved by 12% SDS-PAGE
and transferred to polyvinylidene fluoride membranes. Membranes
were blocked with 10% nonfat dry milk/0.1% Tween 20 and incubated with specific antibodies (Molecular Probes, Inc., Eugene, OR,
or MitoSciences, Eugene, OR). Secondary antibody was used; the
signals were revealed by chemiluminescence (enhanced chemiluminescence; Amersham Biosciences Inc., Piscataway, NJ) and quantified
by computerized analysis.
Polarographic studies
Respiration of digitonin-permeabilized fibroblasts (digitonin 0.01%
for 1 min) was measured using a Clark oxygen electrode at 37 C in 0.27
ml incubation medium containing: 10 mM KH2PO4, 300 mM mannitol,
10 mM KCl, 5 mM MgCl2, 1 mM ADP, 1 mg/ml BSA (pH 7.4), and 8 mM
pyruvate ⫹ 1 mM malate or 10 mM succinate as mitochondrial substrates
(26). The respiratory control ratio, i.e. maximal to basal respiration rates
ratio, was 4.5–5.5, indicating good mitochondrial integrity.
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
TABLE 1.
jcem.endojournals.org
1435
Mutations identified in patients
Patient
no.
1
2
3
4
5
6
7
8
Genes
CI NDUFS1
CI NDUFS3
CI NDUFS4
CI NDUFV1
CI NDUFV2
CIII BCS1
CIV SURF1
CIV COX10
Mutations
del222/D252G
T145I/R199W
ex 3– 4 del/ex 3– 4 del
Y204C/C206G
ex 2 del/ex 2 del
P99L/P99L
P183fsX189/P183fsX189
N204K/N204K
Potassium cyanide (KCN) titration of COX activity in
myoblasts
Respiration of digitonin-permeabilized myoblasts was measured
polarographically at 37 C in 0.27 ml incubation medium containing
2 mM ADP, 6 ␮M rotenone, and 10 mM succinate or 400 ␮M
N,N,N⬘,N⬘-tetramethyl-p-phenylenediamine (TMPD), 1 mM ascorbate, 2 mM ADP, and 0.6 ␮M antimycin (27, 28). Auto-oxidation of
TMPD plus ascorbate was negligible compared with the cell O2 consumption rates. Small amounts of 3 mM KCN solution were sequentially introduced at appropriate intervals to allow accurate measurements of constant slopes. The KCN-inhibited respiration measured
after each KCN addition was expressed as percentage of the uninhibited respiration (27, 28).
Protein
Ref.
RC enzyme subunit
RC enzyme subunit
RC enzyme subunit
RC enzyme subunit
RC enzyme subunit
RC assembly factor
RC assembly factor
RC assembly factor
14
15
17
14
13
16
19
18
Statistical analysis
Differences between vehicle and treated cells were analyzed by oneway ANOVA and the Fisher test, or by the unpaired t test. P ⬍ 0.05 was
considered significant.
Results
Bezafibrate can trigger significant increases of RC
enzyme activities
Cultured fibroblasts or myoblasts from 14 patients with mitochondrial disorders due to CI (n ⫽ 5), CIII (n ⫽ 1), or CIV (n ⫽ 8)
deficiency were exposed to bezafibrate for 72 h, and the activities of
the deficient RC complex were determined (Fig. 2). As shown in Fig.
2A, three out of five CI-deficient
cell lines (patients 2, 4, and 5) exhibited a marked increase in CI activity [reduced nicotinamide adenine
dinucleotide
(NADH)decylubiquinone oxidoreductase
(NQR)] in response to bezafibrate
(⫻1.4 to ⫻2), and a stimulatory effect of bezafibrate (⫻1.4) on NQR
activity was also observed in control fibroblasts. Furthermore,
pharmacological treatment resulted in the restoration of normal
NQR enzyme levels in patient 4
and 5 fibroblasts, which initially
exhibited a 40% enzyme defect
compared with control. CIII enzyme determinations [decylubiquinol cytochrome c reductase
(QCCR)] revealed a potent induction of this enzyme activity by
bezafibrate in control (⫻1.5), and
in CIII-deficient fibroblasts as well
(Fig. 2B). Indeed, more than a
2-fold increase in the QCCR level
was observed in patient 6 fibroFIG. 1. Dose-response and kinetics studies. The effects of bezafibrate on COX enzyme activity in control fibroblasts (A)
blasts treated by bezafibrate, leador myoblasts (B) were determined. For the dose-response, the cells were incubated 72 h. For the kinetics, the
ing to a complete correction of the
concentrations of 400 and 200 ␮M were used in fibroblasts and myoblasts, respectively. The determinations were
initial enzyme defect (⫺68% relaperformed in triplicate. Values are means ⫾ SD.
Mitochondrial Disorders and PPAR Agonists
A
NQR
B
QCCR
50
250
***
**
30
***
20
COX
***
150
100
50
0
150
***
200
Patient 5
Patient 4
Patient 2
Control
Patient 1
10
0
D
Vehicle
Bezafibrate
***
100
**
***
***
***
Control
***
50
Patient 14
Patient 13
Patient 12
Patient 11
Patient 8
Patient 10
Patient 9
0
Patient 7
Patient 14
Control
Cytochrome C oxidase activity
nmol/min/mg prot
Decylubiquinol cytochrome c
reductase activity
nmol/min/mg prot
40
Patient 3
NADH-decylubiquinone
oxidoreductase activity
nmol/min/mg prot
***
C
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
Patient 6
Bastin et al.
Control
1436
FIG. 2. Effect of bezafibrate on RC enzyme activities in human cells. A–C, Fibroblasts or myoblasts from
controls or RC-deficient patients were incubated for 72 h with bezafibrate (black bars) or vehicle only (white
bars) and then harvested for measurement of enzyme activities according to standard methods. The
following enzyme activities were measured: NQR, QCCR, and COX. Control values are means ⫾ SD of five to
eight different individuals measured in duplicate. Patient values are means ⫾ SD of at least three different
experiments. In each experiment, the determinations were performed in duplicate. **, P ⬍ 0.01; ***, P ⬍
0.001 vs. vehicle-treated cells. D, COX histochemistry of myoblasts from one control and patient 14. Cells
were incubated for 48 h with bezafibrate before staining.
tive to control). Exposure to bezafibrate also led to a marked (⫻1.5)
increase in CIV enzyme activity (COX) in control cells, whereas
various pharmacological profiles were found in the eight COXdeficient cell lines, in response to the drug (Fig. 2C). Thus, bezafibrate triggered a 1.4- to 2-fold increase in COX activity in patient
8 and 11–14 cells, and this led to restore normal
enzyme activity levels in four of these cell lines
(patients 11–14). In contrast, fibroblasts from patients 7, 9, and 10 showed no significant changes
in COX activity in response to bezafibrate. COX
histochemistry performed in control and patient
14 myoblasts (Fig. 2D) revealed a weak COX
staining in untreated patient myoblasts, which
was markedly increased after exposure to
bezafibrate.
PPAR␦-specific agonist restores RC
enzyme activities in CI and CIV-deficient
human cells
To determine which PPAR isoforms mediates the response to bezafibrate, cells were
treated for 48 h by agonists specific for PPAR ␣,
or ␦ or ␥. These experiments were performed in control, and in patients 5 and 14
that exhibited the highest increases in CI
and CIV enzyme activities, when treated by
bezafibrate (Fig. 3). Exposure to GW␣
7647 or GW␥ 7845 induced no significant
changes in RC enzyme activities. In contrast, treatment by GW␦ 0742 triggered a
1.3-fold increase in CI and COX enzyme
activities in controls. In patients 5 and 14,
GW␦ 0742 stimulated CI and COX enzyme activities by 1.7- and 1.6-fold, respectively, leading to the correction of the initial enzyme defects in both cell lines.
Bezafibrate up-regulates the
expression of genes encoding RC
subunits or assembly factors
Quantitative RT-PCR studies were performed to determine the effects of bezafibrate on mRNA levels of the disease-causing genes responsible for CI, CIII, or CIV
deficiency (Fig. 4). Treatment by bezafibrate increased all mRNA levels in control
fibroblasts (from ⫻1.4 to ⫻2.1 compared
with vehicle-treated cells). In the patient
cells, the mutated mRNAs were similarly
increased by bezafibrate, except in patients
3 and 1, consistent with the lack of increase
in CI enzyme activity in these two patients.
Changes in protein levels of RC
subunits or assembly factors in
response to bezafibrate
Studies were performed using the antibody specific of the mutated gene when available (NDUFS3,
SURF1) or antibodies representative of the mutated complex, i.e.
core 2 for CIII, or COX2 (mitochondrial-encoded subunit) and
FIG. 3. Effects of high-affinity PPAR agonists on NQR and COX enzyme activities in control and
patient cells. Fibroblasts (A) or myoblasts (B) from controls or RC-deficient patients were incubated for
72 h with GW␣ 7647, or GW␦ 0742 or GW␥ 7845, or vehicle only, and then harvested for
measurement of enzyme activities. Enzyme determinations were performed in triplicates. Values are
means ⫾ SD. **, P ⬍ 0.01 vs. vehicle-treated cells.
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
jcem.endojournals.org
1437
FIG. 5. Effect of bezafibrate on RC protein expression. Cells were incubated for
72 h with 400 ␮M bezafibrate (black bars) or with vehicle (white bars) before the
extraction. A–D, Histograms of protein amounts relative to porin. Values are
means ⫾ SD of at least three independent experiments. *, P ⬍ 0.05; **, P ⬍
0.01; ***, P ⬍ 0.001 vs. vehicle. A representative Western blot is shown.
FIG. 4. Effect of bezafibrate on mRNA levels in control and patient fibroblasts.
Fibroblasts from controls or patients were grown for 48 h in the presence of 400
␮M bezafibrate (filled bars) or vehicle (open bars) in the culture medium and then
harvested for mRNA level determinations. In each experiment, RT-QPCRs were
run in triplicate, and the results, expressed in arbitrary units, are means ⫾ SD. *,
P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001 vs. vehicle-treated cells.
COX4 (nuclear-encoded subunit) for CIV. Equal loading was
checked with an antibody against mitochondrial porin (Fig. 5).
Treatment by bezafibrate increased NDUFS3 protein levels in
control (⫻1.5) and patient 2 (⫻1.9) cells, and led to correct
partially the NDUFS3 protein defect in patient 2 cells (Fig. 5A).
The basal levels of core 2 protein were slightly but not significantly reduced in untreated patient 6 cells (Fig. 5B). Treatment by bezafibrate markedly increased (⬎2-fold) the amount
of core 2 protein in patient 6 and control fibroblasts. A severe
decrease in SURF1 protein was found in untreated patient 7
cells (Fig. 5C), down to less than 10% of normal level, and no
significant change occurred in response to bezafibrate. In con-
trol cells, in contrast, the SURF1 protein levels were strongly
inducible by bezafibrate. Altogether, the lack of changes in
SURF1 mutant protein in patient 7 cells treated with bezafibrate could account for the lack of increase in COX activity
(Fig. 2C), despite a drug-induced increase in SURF1 mRNA
(Fig. 4). Finally, the levels of COX2 and COX4 subunits (Fig.
5D) increased about 2-fold in control cells exposed to bezafibrate. In untreated patient 8 cells (COX10 deficient), both
COX2 and COX4 proteins were markedly reduced (⫺80 and
⫺50% relative to control, respectively). Treatment of patient
8 cells with bezafibrate induced a robust increase in COX2
and COX4 proteins (⫻4.7 and ⫻2.3, respectively), resulting
in the restoration of normal levels of both COX subunits.
Effects of bezafibrate on respiratory rates in control and
COX-deficient cells
To determine whether drug-induced changes in enzyme
activity of the various RC complexes impacted on mitochondrial respiration, we performed polarographic assays in control and in two COX-deficient cell lines that exhibited quite
1438
Bastin et al.
Mitochondrial Disorders and PPAR Agonists
TABLE 2. Polarographic assays of respiratory rates in
patients and control fibroblasts
Control
Control ⫹ bezafibrate
Patient 7
Patient 7 ⫹ bezafibrate
Patient 8
Patient 8 ⫹ bezafibrate
Pyruvate
(plus malate)
oxidation
Succinate
oxidation
6.7 ⫾ 0.3
8.9 ⫾ 0.8a
4.2 ⫾ 1.0
5.5 ⫾ 1.8
5.0 ⫾ 0.4
6.7 ⫾ 0.3a
10.6 ⫾ 1.5
15.4 ⫾ 1.9a
6.4 ⫾ 0,2
7.7 ⫾ 0.3
7.4 ⫾ 0.9
11.0 ⫾ 0.4a
The rates are the maximal rates measured in the presence of the various
substrates, corrected by the basal mitochondrial respiration rates. In all
experiments 0.25– 0.35 mg cell proteins were used. Oxygen consumption is
expressed as nanomol O2/min䡠mg protein. The values are means ⫾ SE of two or
three different experiments.
a
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
with control (⫺2.05) pointed to a rate limitation of respiration in
the COX-deficient myoblasts. However, exposure to bezafibrate
resulted in a marked increase in the slope (⫺1.86 vs. ⫺2.76 in untreated cells), which then reached the control range, reflecting an
increase in COX reserve within the mitochondria.
Bezafibrate induced transcription factors or nuclear
coactivators mRNA in human cells
Table 3 shows that PGC-1␣ mRNA levels were strongly increased (by 2- to 3-fold) by bezafibrate treatment in control and
patients cells. Interestingly, bezafibrate generally induced an increase in gene expression of NRF1, NRF2␣, and Tfam, key transcription factors of mitochondrial proteins.
P ⬍ 0.05 compared with dimethylsulfoxide-treated fibroblasts.
Discussion
Despite advances in the recognition and diagnosis of inborn RC
different response to bezafibrate, i.e. patients 7 (SURF1) and
deficiencies, the therapeutic approaches for these disorders re8 (COX10).
main quite limited (3, 32). In the present study, we sought to
In control fibroblasts, drug-induced increases in respiration
determine whether pharmacological activation of PPAR pathrates were observed (Table 2). Thus, pyruvate plus malate and
way could target to the mitochondrial RC, and could be efficient
succinate oxidation rates were significantly stimulated (⫹32 and
for correction of RC deficiency in human cells, by stimulating the
⫹45%; P ⬍ 0.05) after treatment by bezafibrate. In patient 7,
expression of the defective enzyme complex.
both oxidation rates were deficient, and no significant increases
occurred after treatment by bezafibrate. In patient 8, in contrast, the initial defects in substrate
oxidation (⫺25 to ⫺30% relative to control)
were fully corrected by bezafibrate, which induced a ⫻1.3 to ⫻1.5 increase in respiration.
To investigate further the functional relationship between COX activity and mitochondrial respiration in treated and untreated cells, we performed titration curves of cell respiration using
increasing amounts of KCN, a COX-specific inhibitor. Myoblasts from patient 14 were selected because they exhibited the largest drug-induced increase in COX activity (Fig. 2C). These experiments
were performed with succinate as a substrate, to
titrate COX as RC integrated step, or with TMPD
plus ascorbate, which directly feed the RC at the
level of CIV, allowing to titrate COX as an isolated
step (28, 29). Under these conditions it has been
shown that titration curves from COX-deficient
cells are steeper than those from healthy control,
due to the diminished COX reserve in the deficient
cells (30, 31). Accordingly, the slope of the titration
curves in patient myoblasts (⫺1.98, Fig. 6B) was
steeper (⫺20%) compared with control (⫺1.61,
Fig. 6A). Bezafibrate treatment increased the slopes
with: y ⫽ ⫺1.36x ⫹ 99.2, r ⫽ 0.995 and y ⫽
FIG. 6. KCN titration of ADP-stimulated oxygen consumption by digitonin-permeabilized myoblasts.
A and B, Substrate: 10 mM succinate. C and D, Substrates: 400 ␮M TMPD plus 1 mM ascorbate. The
⫺1.34x ⫹ 99.4, and r ⫽ 0.997 in patient and condata points are means ⫾ SE from two titration experiments with myoblasts from three controls. For
trol cells, respectively. This correction was even
the patient, the data are average value ⫾ SE of three titration experiments. Absence of error bars
more obvious when the experiments were perindicates that SE is smaller than symbol. The slopes were determined by the best-fitting linear
regression line of the initial part of the titration curves and were compared in all experimental
formed in the presence of TMPD plus ascorbate
conditions. The initial respiration rates in vehicle- or bezafibrate-treated cells were,
(Fig. 6, C and D). Indeed, the steeper slope in unrespectively:16.4 ⫾ 1.4 and 20.1 ⫾ 1 (A); 12.5 ⫾ 2 and 17.5 ⫾ 0.5 (B); 23 ⫾ 2 and 31.1 ⫾ 2.3 (C);
treated patient cells (⫺2.76, ⫺26%) compared
and 18 ⫾ 0.5 and 25.3 ⫾ 2 (D) expressed in nmol O2/min䡠mg protein.
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
TABLE 3.
jcem.endojournals.org
1439
Effects of bezafibrate on mRNA levels in control and patient cells
Fold change vs. vehicle
Controls
Patient 2
Patient 4
Patient 5
Patient 6
Patient 8
Patient 14
PGC1-␣
NRF1
NRF2␣
Tfam
2.35 ⫾ 0.40a
3.23 ⫾ 0.44b
2.89 ⫾ 0.10c
2.90 ⫾ 0.06c
2.76 ⫾ 0.40a
2.46 ⫾ 0.40a
1.84 ⫾ 0.30a
1.37 ⫾ 0.15a
1.40 ⫾ 0.07b
1.28 ⫾ 0.05b
1.36 ⫾ 0.16a
1.20 ⫾ 0.02b
1.22 ⫾ 0.02b
1.41 ⫾ 0.13a
1.79 ⫾ 0.12b
2.15 ⫾ 0.18b
1.87 ⫾ 0.13c
2.60 ⫾ 0.05c
1.45 ⫾ 0.03b
1.08 ⫾ 0.01
1.70 ⫾ 0.05b
1.98 ⫾ 0.35a
1.50 ⫾ 0.17a
1.72 ⫾ 0.05c
2.52 ⫾ 0.18b
1.15 ⫾ 0.03
1.30 ⫾ 0.02b
1.63 ⫾ 0.23b
Four genes were studied: PGC1-␣, NRF1, NRF2␣, and Tfam. In each experiment, RT-QPCRs were run in triplicate, and the results are means ⫾ SD from at least two
different experiments.
a
P ⬍ 0.05 compared with vehicle-treated cells.
b
P ⬍ 0.01 compared with vehicle-treated cells.
c
P ⬍ 0.001 compared with vehicle-treated cells.
The results obtained from control fibroblasts clearly indicate
a stimulatory effect of bezafibrate on RC, as shown by the druginduced increases in NQR, QCCR, and COX activities, and by
the parallel changes in the levels of proteins representative of CI,
CIII, and CIV. In addition, bezafibrate significantly increased the
transcript levels of a number of genes encoding RC subunits or
ancillary proteins. As an initial step in testing bezafibrate in RC
deficiencies, we studied a panel of patient cells carrying known
disease-causing mutations in nuclear genes of CI, CIII, and CIV,
as well as COX-deficient cells with unknown pathogenic
mutations.
Consistent with the observations made in control cells, bezafibrate was found to potentially stimulate the residual enzyme
activity of CI, CIII, and CIV in deficient cells. Indeed, nine out of
the 14 patient cell lines tested (64%) exhibited a significant increase in the activity of the deficient RC complex after treatment
by bezafibrate, and full pharmacological correction of the enzyme defect was achieved in seven cell lines initially presenting a
CI, CIII, or CIV deficiency. These increases in enzyme activities
were associated with parallel increases in mRNA levels of the
mutated genes, and in some instances we investigated the
changes in protein levels induced by bezafibrate. Altogether, our
results showed that the mutation T145I/R199W of NDUFS3
gene (patient 2) resulted in a clear reduction in NDUFS3 protein
level, which appeared consistent with the decreased CI activity.
Furthermore, in this patient’s cells, the drug-induced increase in
CI enzyme activity was clearly related to the parallel increases in
NDUFS3 mRNAs and protein levels. In patient 6, the homozygous P99L mutation of BCS1 gene led to a slight decrease in core
2 protein level, in keeping with the initial description of this
BCS1-deficient patient, which reported a reduced expression of
two other CIII proteins (Rieske and core 1 subunits) (16). In
patient 6, bezafibrate stimulated the expression of the BSC1 gene
and increased core 2 protein level, which might account for the
stimulation of CIII activity. The patient 8 cells, which harbored
a homozygous N204K mutation of COX10 assembly factor,
exhibited markedly low levels of COX2 and COX 4 proteins, as
previously described (18, 33). Treatment by bezafibrate stimulated the expression of the mutated COX 10 gene, restored
COX2 and COX 4 proteins to normal levels, and resulted in a
marked stimulation of COX activity. Interestingly, the partial
correction of COX deficiency by bezafibrate appeared sufficient
to restore normal respiratory rates in treated cells.
Three cell lines carrying identified gene mutations (patients 1,
3, and 7) did not respond to bezafibrate. Our data suggest that
this could be related to the lack of induction of mutated mRNA
in patient 1 and 3 cells in response to bezafibrate. In patient 7
cells, the homozygous SURF1 gene mutation P183fsX189, predicting a truncated protein, was associated with extremely low
levels of SURF1 protein, and consistent with a key role of SURF1
in CIV assembly, the residual COX activity was markedly reduced. Exposure to bezafibrate increased SURF1 mRNAs similarly in control and patient cells, and strongly up-regulated
SURF1 protein in control fibroblasts. However, the defective
SURF1 protein was not inducible by bezafibrate in patient 7 cells,
which likely explains the lack of changes in COX activity in the
treated cells.
Together, our data show that, in control human cells, bezafibrate can up-regulate the expression of genes encoding RC enzyme subunits or assembly factors, resulting in increases of the
corresponding RC enzyme activities. This drug effect might account for the pharmacological stimulation or correction of the
deficient-RC complex, which was achieved in some cells from
patients with mitochondrial disorders, when the disease-causing
mutations did not severely affect the production of mutated
mRNA or protein.
This, together with recent data in animal models (10, 11),
clearly suggests that PPARs could take part in OXPHOS regulation. However, the molecular mechanisms by which PPARs
might control RC genes are not fully elucidated. The simplest
explanation would be the presence of a functional PPAR response element (PPRE) on the 5⬘-flanking region of target genes.
However, PPREs have never been found in the promoter regions
of OXPHOS genes, despite extensive studies (5), suggesting that
PPAR-mediated regulation of RC is likely indirect. In support of
this, the present study showed that PPAR activation triggered a
coordinate up-regulation of several key OXPHOS transcription
regulators, and in particular of PGC-1␣. This marked induction
1440
Bastin et al.
Mitochondrial Disorders and PPAR Agonists
of PGC-1␣ is consistent with the recent identification of a PPRE
on the human PGC-1␣ gene promoter (34). Furthermore, forced
expression studies in mammalian cells demonstrated that increasing PGC-1␣ level induced the expression of NRF1 and
NRF2 (35). Therefore, induction of PGC-1␣ by bezafibrate
might account for the parallel up-regulation of NRF1 and NRF2
in the treated cells. Altogether, enhanced expression of the transcription factors NRF1 and NRF2, and of their coactivator PGC1␣, in response to bezafibrate, likely account for the stimulation
of RC nuclear genes.
CI, CIII, and CIV are multisubunit enzymes encoded by the
nuclear and the mitochondrial genomes, and parallel increases
in the mRNA of COX2 (mitochondrial gene) and COX4 (nuclear gene) were found in bezafibrate-treated cells (data not
shown), consistent with the Western blot data. Increases in
COX2 and possibly other mitochondrial mRNAs can be explained by the induction of the mitochondrial transcription
factor Tfam (6), in bezafibrate-treated cells. Interestingly,
stimulation of Tfam gene expression might result from multiple events because both NRF1 and NRF2 regulate the human
Tfam gene (5), and because overexpression of PGC-1␣ has
also been reported to increase Tfam expression (35). These
observations led us to propose that upon activation by bezafibrate, PPAR directly stimulates the expression level of PGC1␣, which in turn regulates NRF1, NRF2, and Tfam expression. This could ensure a coordinate stimulation of nuclear
and mitochondrial RC gene expression leading to subsequent
increases in RC proteins and enzyme activities.
Our experiments were performed using bezafibrate, a widely
used hypolipidemic drug considered as a pan (␣, ␦, and ␥) PPAR
agonist (36). Experiments using specific PPAR agonists indicate
that bezafibrate effects on RC gene expression are likely mediated by the PPAR␦ isoform, in agreement with recent literature
data. Indeed, PPAR␦ and not PPAR␣ was shown to activate
PGC-1␣ gene expression in muscle (12). In addition, targeted
overexpression of PPAR␦ in the mouse muscle results in a
marked increase in complex II enzyme activity (10) and in COX
mRNA (11). Interestingly, chronic mice treatment with a PPAR␦
agonist also resulted in the stimulation of OXPHOS gene expression in skeletal muscle, together with a global increase in
oxidative metabolism (10, 11, 37). Finally, recent data obtained
in mice in which PPAR␦ has been selectively ablated in skeletal
muscle indicate that PGC-1␣ mRNAs were decreased and that
transcript levels of several nuclear genes encoding components of
the RC complexes were significantly lower compared with wildtype animals (34).
Overall, our study brings new insights into the role of
PPAR in the control of energy metabolism by demonstrating
a functional link between activation of the PPAR pathway and
the regulation of mitochondrial RC in human cells. Exposure
to bezafibrate can correct RC deficiency in patient cells, indicating the relevance of OXPHOS control by PPAR in a physiopathological context. This might have implications in the
field of RC disorders but also in the context of common metabolic diseases like type 2 diabetes, for which OXPHOS regulation recently emerged as an important issue (38, 39), and
PPARs are considered as major therapeutic targets. Finally,
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
our approach suggests a new rationale to tackle the difficult
question of inborn RC disorder treatment, which still remains
largely unsolved.
Acknowledgments
Address all correspondence and requests for reprints to: Fatima Djouadi,
Ph.D., Centre National de la Recherche Scientifique Unité Propre de
Recherche 9078, Faculté Necker-Enfants Malades, 156 rue de Vaugirard, 75015 Paris, France. E-mail: [email protected].
This work was supported by a grant from the Association Française
contre les Myopathies.
Disclosure Statement: The authors have nothing to declare.
References
1. Chinnery PF, Turnbull DM 2001 Epidemiology and treatment of mitochondrial disorders. Am J Med Genet 106:94 –101
2. Zeviani M, Di Donato S 2004 Mitochondrial disorders. Brain 127:2153–2172
3. Dimauro S, Mancuso M 2007 Mitochondrial diseases: therapeutic approaches. Biosci Rep 27:125–137
4. Robinson BH 2000 Human cytochrome oxidase deficiency. Pediatr Res 48:
581–585
5. Kelly DP, Scarpulla RC 2004 Transcriptional regulatory circuits controlling
mitochondrial biogenesis and function. Genes Dev 18:357–368
6. Scarpulla RC 2006 Nuclear control of respiratory gene expression in mammalian cells. J Cell Biochem 97:673– 683
7. Finck BN, Kelly DP 2006 PGC-1 coactivators: inducible regulators of energy
metabolism in health and disease. J Clin Invest 116:615– 622
8. Feige JN, Gelman L, Michalik L, Desvergne B, Wahli W 2006 From molecular
action to physiological outputs: peroxisome proliferator-activated receptors
are nuclear receptors at the crossroads of key cellular functions. Prog Lipid Res
45:120 –159
9. Lee CH, Olson P, Evans RM 2003 Minireview: lipid metabolism, metabolic
diseases, and peroxisome proliferator-activated receptors. Endocrinology 144:
2201–2207
10. Luquet S, Lopez-Soriano J, Holst D, Fredenrich A, Melki J, Rassoulzadegan M,
Grimaldi PA 2003 Peroxisome proliferator-activated receptor ␦ controls muscle development and oxidative capability. FASEB J 17:2299 –2301
11. Wang YX, Zhang CL, Yu RT, Cho HK, Nelson MC, Bayuga-Ocampo CR,
Ham J, Kang H, Evans RM 2004 Regulation of muscle fiber type and running
endurance by PPAR␦. PLoS Biol [Erratum (2005) 3:e61] 2:e294
12. Hondares E, Pineda-Torra I, Iglesias R, Staels B, Villarroya F, Giralt M 2007
PPAR␦, but not PPAR␣, activates PGC-1␣ gene transcription in muscle. Biochem Biophys Res Commun 354:1021–1027
13. Benit P, Beugnot R, Chretien D, Giurgea I, De Lonlay-Debeney P, Issartel JP,
Corral-Debrinski M, Kerscher S, Rustin P, Rotig A, Munnich A 2003 Mutant
NDUFV2 subunit of mitochondrial complex I causes early onset hypertrophic
cardiomyopathy and encephalopathy. Hum Mutat 21:582–586
14. Benit P, Chretien D, Kadhom N, de Lonlay-Debeney P, Cormier-Daire V,
Cabral A, Peudenier S, Rustin P, Munnich A, Rotig A 2001 Large-scale deletion
and point mutations of the nuclear NDUFV1 and NDUFS1 genes in mitochondrial complex I deficiency. Am J Hum Genet 68:1344 –1352
15. Benit P, Slama A, Cartault F, Giurgea I, Chretien D, Lebon S, Marsac C,
Munnich A, Rotig A, Rustin P 2004 Mutant NDUFS3 subunit of mitochondrial complex I causes Leigh syndrome. J Med Genet 41:14 –17
16. de Lonlay P, Valnot I, Barrientos A, Gorbatyuk M, Tzagoloff A, Taanman JW,
Benayoun E, Chretien D, Kadhom N, Lombes A, de Baulny HO, Niaudet P,
Munnich A, Rustin P, Rotig A 2001 A mutant mitochondrial respiratory chain
assembly protein causes complex III deficiency in patients with tubulopathy,
encephalopathy and liver failure. Nat Genet 29:57– 60
17. Rotig A, Lebon S, Zinovieva E, Mollet J, Sarzi E, Bonnefont JP, Munnich A
2004 Molecular diagnostics of mitochondrial disorders. Biochim Biophys Acta
1659:129 –135
18. Valnot I, von Kleist-Retzow JC, Barrientos A, Gorbatyuk M, Taanman JW,
Mehaye B, Rustin P, Tzagoloff A, Munnich A, Rotig A 2000 A mutation in the
human heme A:farnesyltransferase gene (COX10) causes cytochrome c oxidase deficiency. Hum Mol Genet 9:1245–1249
19. Von Kleist-Retzow JC, Yao J, Taanman JW, Chantrel K, Chretien D, Cormier-
J Clin Endocrinol Metab, April 2008, 93(4):1433–1441
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
Daire V, Rotig A, Munnich A, Rustin P, Shoubridge EA 2001 Mutations in
SURF1 are not specifically associated with Leigh syndrome. J Med Genet
38:109 –113
Sacconi S, Salviati L, Sue CM, Shanske S, Davidson MM, Bonilla E, Naini AB,
De Vivo DC, DiMauro S 2003 Mutation screening in patients with isolated
cytochrome c oxidase deficiency. Pediatr Res 53:224 –230
Shoubridge EA 2001 Cytochrome c oxidase deficiency. Am J Med Genet 106:
46 –52
Djouadi F, Aubey F, Schlemmer D, Bastin J 2005 Peroxisome proliferator
activated receptor ␦ (PPAR␦) agonist but not PPAR␣ corrects carnitine palmitoyl transferase 2 deficiency in human muscle cells. J Clin Endocrinol Metab
90:1791–1797
Muoio DM, MacLean PS, Lang DB, Li S, Houmard JA, Way JM, Winegar DA,
Corton JC, Dohm GL, Kraus WE 2002 Fatty acid homeostasis and induction
of lipid regulatory genes in skeletal muscles of peroxisome proliferator-activated receptor (PPAR) ␣ knock-out mice. Evidence for compensatory regulation by PPAR␦. J Biol Chem 277:26089 –26097
Barrientos A 2002 In vivo and in organello assessment of OXPHOS activities.
Methods 26:307–316
Chretien D, Benit P, Chol M, Lebon S, Rotig A, Munnich A, Rustin P 2003
Assay of mitochondrial respiratory chain complex I in human lymphocytes and
cultured skin fibroblasts. Biochem Biophys Res Commun 301:222–224
Rustin P, Chretien D, Bourgeron T, Gerard B, Rotig A, Saudubray JM, Munnich A 1994 Biochemical and molecular investigations in respiratory chain
deficiencies. Clin Chim Acta 228:35–51
Kunz WS, Kudin A, Vielhaber S, Elger CE, Attardi G, Villani G 2000 Flux
control of cytochrome c oxidase in human skeletal muscle. J Biol Chem 275:
27741–27745
Villani G, Greco M, Papa S, Attardi G 1998 Low reserve of cytochrome c
oxidase capacity in vivo in the respiratory chain of a variety of human cell types.
J Biol Chem 273:31829 –31836
Villani G, Attardi G 2000 In vivo control of respiration by cytochrome c
oxidase in human cells. Free Radic Biol Med 29:202–210
Kuznetsov AV, Winkler K, Kirches E, Lins H, Feistner H, Kunz WS 1997
Application of inhibitor titrations for the detection of oxidative phosphory-
jcem.endojournals.org
31.
32.
33.
34.
35.
36.
37.
38.
39.
1441
lation defects in saponin-skinned muscle fibers of patients with mitochondrial
diseases. Biochim Biophys Acta 1360:142–150
Varlamov DA, Kudin AP, Vielhaber S, Schroder R, Sassen R, Becker A, Kunz
D, Haug K, Rebstock J, Heils A, Elger CE, Kunz WS 2002 Metabolic consequences of a novel missense mutation of the mtDNA CO I gene. Hum Mol
Genet 11:1797–1805
Chinnery P, Majamaa K, Turnbull D, Thorburn D 2006 Treatment for mitochondrial disorders. Cochrane Database Syst Rev CD004426
Williams SL, Valnot I, Rustin P, Taanman JW 2004 Cytochrome c oxidase
subassemblies in fibroblast cultures from patients carrying mutations in
COX10, SCO1, or SURF1. J Biol Chem 279:7462–7469
Schuler M, Ali F, Chambon C, Duteil D, Bornert JM, Tardivel A, Desvergne
B, Wahli W, Chambon P, Metzger D 2006 PGC1␣ expression is controlled in
skeletal muscles by PPAR␤, whose ablation results in fiber-type switching,
obesity, and type 2 diabetes. Cell Metab 4:407– 414
Wu Z, Puigserver P, Andersson U, Zhang C, Adelmant G, Mootha V, Troy A,
Cinti S, Lowell B, Scarpulla RC, Spiegelman BM 1999 Mechanisms controlling mitochondrial biogenesis and respiration through the thermogenic coactivator PGC-1. Cell 98:115–124
Tenenbaum A, Motro M, Fisman EZ 2005 Dual and pan-peroxisome proliferator-activated receptors (PPAR) co-agonism: the bezafibrate lessons. Cardiovasc Diabetol 4:14
Barish GD, Narkar VA, Evans RM 2006 PPAR␦: a dagger in the heart of the
metabolic syndrome. J Clin Invest 116:590 –597
Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J,
Puigserver P, Carlsson E, Ridderstrale M, Laurila E, Houstis N, Daly MJ,
Patterson N, Mesirov JP, Golub TR, Tamayo P, Spiegelman B, Lander ES,
Hirschhorn JN, Altshuler D, Groop LC 2003 PGC-1␣-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet 34:267–273
Patti ME, Butte AJ, Crunkhorn S, Cusi K, Berria R, Kashyap S, Miyazaki Y,
Kohane I, Costello M, Saccone R, Landaker EJ, Goldfine AB, Mun E, DeFronzo
R, Finlayson J, Kahn CR, Mandarino LJ 2003 Coordinated reduction of genes of
oxidative metabolism in humans with insulin resistance and diabetes: potential
role of PGC1 and NRF1. Proc Natl Acad Sci USA 100:8466 – 8471