MPV17 encodes an inner mitochondrial membrane protein and is mutated in infantile hepatic mitochondrial DNA depletion.

Abstract

The mitochondrial (mt) DNA depletion syndromes (MDDS) are genetic disorders characterized by a severe, tissue-specific decrease of mtDNA copy number, leading to organ failure. There are two main clinical presentations: myopathic (OMIM 609560) and hepatocerebral (OMIM 251880). Known mutant genes, including TK2, SUCLA2, DGUOK and POLG, account for only a fraction of MDDS cases. We found a new locus for hepatocerebral MDDS on chromosome 2p21-23 and prioritized the genes on this locus using a new integrative genomics strategy. One of the top-scoring candidates was the human ortholog of the mouse kidney disease gene Mpv17. We found disease-segregating mutations in three families with hepatocerebral MDDS and demonstrated that, contrary to the alleged peroxisomal localization of the MPV17 gene product, MPV17 is a mitochondrial inner membrane protein, and its absence or malfunction causes oxidative phosphorylation (OXPHOS) failure and mtDNA depletion, not only in affected individuals but also in Mpv17-/- mice.

Full text

Page 1

MPV17 encodes an inner mitochondrial membrane
protein and is mutated in infantile hepatic mitochondrial
DNA depletion
Antonella Spinazzola1, Carlo Viscomi1, Erika Fernandez-Vizarra1, Franco Carrara1, Pio D’Adamo2,
Sarah Calvo3–5, Rene ´ Massimiliano Marsano6, Claudia Donnini6, Hans Weiher7, Pietro Strisciuglio8,
Rossella Parini9, Emmanuelle Sarzi10, Alicia Chan11, Salvatore DiMauro12, Agnes Ro ¨tig10, Paolo Gasparini2,13,
Iliana Ferrero6, Vamsi K Mootha3–5, Valeria Tiranti1& Massimo Zeviani1
The mitochondrial (mt) DNA depletion syndromes (MDDS)
are genetic disorders characterized by a severe, tissue-specific
decrease of mtDNA copy number, leading to organ failure.
There are two main clinical presentations: myopathic (OMIM
609560) and hepatocerebral1(OMIM 251880). Known mutant
genes, including TK2 (ref. 2), SUCLA2 (ref. 3), DGUOK (ref. 4)
and POLG5,6, account for only a fraction of MDDS cases7. We
found a new locus for hepatocerebral MDDS on chromosome
2p21-23 and prioritized the genes on this locus using a new
integrative genomics strategy. One of the top-scoring
candidates was the human ortholog of the mouse kidney
disease gene Mpv17 (ref. 8). We found disease-segregating
mutations in three families with hepatocerebral MDDS and
demonstrated that, contrary to the alleged peroxisomal
localization of the MPV17 gene product9, MPV17 is a
mitochondrial inner membrane protein, and its absence or
malfunction causes oxidative phosphorylation (OXPHOS)
failure and mtDNA depletion, not only in affected individuals
but also in Mpv17–/–mice.
We enrolled three families with MDDS (families 1–3; Fig. 1a) in this
study. Individuals 1-1 and 1-3, originating from southern Italy, died of
liver failure during the first year of life, but liver transplantation at
1 year of age in individual 1-2 and dietary control of hypoglycemia in
individual 1-4 were effective in maintaining relative metabolic com-
pensation and long-term survival. However, growth in the surviving
children, who are 4 and 9 years old, has remained below the fifth
percentile, and the older child (individual 1-4) has developed neuro-
logical symptoms and multiple brain lesions documented by MRI
(Supplementary Fig. 1 online). Both probands 2-4 and 3-1 died in the
first months after birth of liver failure; 2-4 was the fourth affected
individual in a six-children sibship from first-cousin Moroccan
parents; 3-1 was the second child of unrelated parents from Canada;
her older brother is alive and well. We found marked mtDNA deple-
tion in liver in all probands, associated with defects of mtDNA-related
respiratory chain complexes (Supplementary Fig. 1). Normal or
mildly reduced levels of mtDNA content and respiratory chain
activities were found in muscle and cultured fibroblasts.
A genome-wide linkage analysis in family 1 gave a pairwise
maximum lod score of 3.45 at Y ¼ 0 with marker D2S390, and a
maximum multipoint location score of 5.65 within a 19-cM region,
between recombinant markers D2S2373 and D2S2259 on chromo-
some 2p21-23 (Fig. 1b). Haplotype reconstruction uncovered a
homozygous condition caused by identity-by-descent alleles in the
affected individuals (data not shown).
Because of the clinical and biochemical features of hepatocerebral
MDDS, we assumed that the responsible protein was likely to be
involved in mitochondrial metabolism and targeted to mitochondria.
In order to prioritize candidate genes, we relied on a method, dubbed
‘Maestro’, that integrates eight different types of genomic data to make
predictions about the human mitochondrial proteome and assign
every gene a score of likelihood of localization to the mitochondrion10.
Among the 151 genes annotated in the critical interval, Maestro
spotlighted several previously unstudied high-scoring candidates as
Received 7 November 2005; accepted 15 February 2006; published online 2 April 2006; doi:10.1038/ng1765
1Unit of Molecular Neurogenetics, Pierfranco and Luisa Mariani Center for the Study of Children’s Mitochondrial Disorders, National Neurological Institute ‘‘C. Besta’’,
Milan 20126, Italy.2Linkage Unit & Service, Telethon Institute for Genetic Medicine (TIGEM), Naples 80131, Italy.3Center for Human Genetic Research,
Massachusetts General Hospital, Cambridge, Massachusetts 02114, USA.4Department of Systems Biology, Harvard Medical School, Cambridge, Massachusetts, USA.
5Broad Institute of the Massachusetts Institute of Technology and Harvard University, Cambridge 02446, Massachusetts, USA.6Department of Genetics Anthropology
Evolution, University of Parma, Parma 43100, Italy.7University for Applied Sciences Bonn-Rhein-Sieg, Bonn, Germany 53359.8Department of Clinical and
Experimental Medicine, University of Catanzaro Magna Graecia, Catanzaro 88100, Italy.9Unit of Pediatrics, Pierfranco and Luisa Mariani Center for the Study of
Children’s Metabolic Disorders, University Hospital, Milano-Bicocca State University, Monza 20052, Italy.10Department of Genetics and INSERM Unit U393, Ho ˆpital
Necker-Enfants Malades, Paris 75015, France.11Department of Medical Genetics, University of Alberta, Edmonton, Alberta, Canada T6G 2H7.12Department of
Neurology, College of Physicians and Surgeons, Columbia University, New York, New York 10032, USA.13Department of Medical Genetics, University of Trieste,
Trieste 34137, Italy. Correspondence should be addressed to M.Z. (zeviani@istituto-besta.it).
570VOLUME 38 [ NUMBER 5 [ MAY 2006 NATURE GENETICS
LETTERS
© 2006 Nature Publishing Group http://www.nature.com/naturegenetics

Page 2

well as three genes already known to be mitochondrial (Fig. 1c).
Sequencing of these genes (HADHA, HADHAB and MRPL33) and of
ASXL2, COX7A2L and NM_018607, did not uncover any mutations
segregating in the affected individuals. The seventh candidate was a
ubiquitously expressed gene (Fig. 2a) annotated as the human
ortholog of mouse Mpv17 (ref. 8). We found a 149G-A (R50Q)
homozygous mutation in the probands of family 1 and a 498C-A
(N166K) homozygous mutation in the proband of family 2. The
proband of family 3 was a compound heterozygote harboring a
missense mutation in one allele (148C-T, R50W) and a 25-bp
deletion in the other (116–141del; Figs. 1a and 2b), which predicts
the synthesis of an aberrant and prematurely truncated polypeptide.
We found cosegregation of the mutations with the disease in all three
families (data not shown). We did not detect any mutation in 500
samples from unrelated Italians and 30 from unrelated Arabs.
The three missense mutations affected amino acid residues con-
served from the facultative aerobic yeast Saccharomyces cerevisiae
(Fig. 2c), in which the MPV17 ortholog is known as SYM1
(ref. 11), to humans. To validate the pathogenic significance of the
human mutations, we used a SYM1-defective yeast strain (Dsym1),
which fails to grow at 37 1C in medium containing Z2% ethanol11as
the only carbon source. This temperature-sensitive OXPHOS pheno-
type is rescued by reexpressing the wild-type SYM1 gene, or, albeit less
effectively, the mammalian MPV17 gene11(Fig. 3a). We obtained very
limited correction by reexpressing sym1R51Q, a variant harboring the
mutation equivalent to human R50Q, but we did not obtain any
correction with sym1R51Wand sym1N172K, equivalent to human R50W
and N166K. None of the human MPV17 mutant variants were able to
rescue the Dsym1 phenotype (Fig. 3a). These results indicate that the
human MPV17 mutations are deleterious in yeast. The sym1R51Qhad
less drastic effects, as does the R50Q in humans, which seems to cause
a relatively milder phenotype. mtDNA instability in S. cerevisiae is
associated with increased segregation of respiratory-deficient ‘petite’
mutants12. When the Dsym1 strains reexpressing the mutant SYM1
variants were grown on ethanol- and glucose-containing medium at
37 1C, they produced a significant increase of respiratory-deficient
mutants, suggesting a role for SYM1 in maintaining mtDNA integrity
and stability (Fig. 3b). To evaluate the nature of the petite mutations
(r–versus ro), we used a DNA blot to analyze the mtDNA of
sym1R51Q, sym1R51Wand sym1N172Krespiratory-deficient clones. Rear-
ranged mtDNA was present in all three (Fig. 3c), indicating that the
pathological alleles induced r–mutations.
The SYM1 gene product (Sym1p) has recently been shown to target,
and reside in, the yeast inner mitochondrial membrane (IMM)11.
Likewise, a high score for potential mitochondrial targeting was given
to the MPV17 N terminus by ad hoc algorithms, including Maestro
(see above). To test experimentally the hypothesis that mammalian
MPV17 is also a mitochondrial protein, we first transfected a construct
Family 1
149→A
(R50Q)
Family 2
498C→A
(N166K)
Family 3
116-141del / 148C→T
(R50W)
1
*
2
*
34
1234
***
**
**
*
R→Q
50
*
Q
36
L
37
V
38
E
39
R
40
P
41
D
42
V
163
I
164
W
165
N→K
166
S
167
Y
168
L
169
Q
47
R
48
G
49
T
51
L
52
T
53
Q
47
R
48
G
49
R→W
50
T
51
L
52
T
53
Parametric multipoint lod score
6
40
30
20
10
Maestro score
0
2126 3136 41 Mb
COX7A2LNM_018607
MPV17
ASXL2
MRPL33
HADHA
HADHB
–10
4
2
Location score
0
05 10 15
Position (cM)
20 25
D2S2373
D2S2259
–2
b
a
c
1
Figure 1 Linkage analysis, selection of candidate
genes and mutation analysis. (a) Pedigrees.
Filled symbols indicate affected individuals.
Asterisks indicate the individuals of family 1 that
participated in the genome-wide linkage analysis.
Arrows indicate the probands. Beside each
pedigree are the electropherograms of mutated
MPV17 sequences, with mutant nucleotides
highlighted in yellow. Amino acid changes are
indicated in red. Yellow arrow indicates the
starting point of the 116–141 deletion in family
3. (b) Multipoint analysis of the hepatocerebral
MDDS locus for family 1. An idiotype of
chromosome 2 reports the position of the locus
within the two flanking recombinant markers
D2S2373 and D2S2259. (c) Prioritization of
mitochondrial candidates within the D2S2373-
D2S2259 linkage interval. For each gene (filled
diamond) within the 19-Mb linkage peak, we plot
the genomic position (x-axis) and the Maestro
score (y-axis), which corresponds to the likelihood
of encoding a mitochondrial protein. In addition
to identifying the three known mitochondrial
genes (black), this method identified four
additional high-scoring candidates (red).
P
Mpv17 1 kb
β-actin
B H K M L Lu Pl
1
5′
R50Q R50Q
N166K
116-141del R50W
N166K
3′
NH2
COOH
234657
Human
Mouse
Rat
Fish
Yeast
a
b
c
Figure 2 MPV17 transcript, gene organization and protein. (a) RNA blot
analysis on polyA(+) RNA extracted from human pancreas (P), kidney (K),
muscle (M), liver (L), lung (Lu), placenta (Pl), brain (B) and heart (H).
(b) Schematic representation of the seven exons of the MPV17 gene,
transcript and protein, showing mutations found in our families. Darker
hues indicate the ORF sequences; lighter hues indicate the 5¢ and 3¢ UTRs.
(c) CLUSTALW multiple alignments of the MPV17 regions containing the
amino acids mutated in our families. Amino acid substitutions are shown
in color above the alignments.
NATURE GENETICS VOLUME 38 [ NUMBER 5 [ MAY 2006 571
LETTERS
© 2006 Nature Publishing Group http://www.nature.com/naturegenetics

Page 3

expressing MPV17HA, a recombinant human MPV17 tagged on the
C terminus with the hemagglutinin (HA) epitope of the influenza
virus. Transient and stable transfections were performed in COS7 and
HeLa cells, respectively. In both cell systems, the immunofluorescence
pattern specific to MPV17HAcoincided with that of the mitochondrial
single-stranded DNA binding protein13(Fig. 4a) or MitoTracker, a
mitochondrion-specific dye (data not shown).
To demonstrate active mitochondrial import, we incubated radi-
olabeled in vitro–translated MPV17 with freshly prepared, energized
HeLa cell mitochondria (Fig. 4b). Treatment with proteinase K did not
digest this translation product in intact mitochondria but completely
digested it after solubilization of mitochondrial membranes with Triton
X-100, indicating that the polypeptide was located inside the organelle.
The import process was dependent on the mitochondrial protonmotive
force (DC), as we obtained complete proteinase K digestion after
treatment with valinomycin, an ionophore that abolishes DC. These
results demonstrate that MPV17 is imported into mitochondria
through the DC-dependent TOM-TIM import machinery14,15, which
targets proteins to either the IMM or the mitochondrial matrix.
Notably, the MPV17-specific in vitro translation product had the
same electrophoretic mobility before and after internalization into
mitochondria (Fig. 4b). This result indicates that MPV17 is not cleaved
after the import process is completed, contrary to what is usually seen
for most IMM or mitochondrial matrix proteins. These proteins
contain an N-terminal presequence that is cleaved by the mitochondrial
processing peptidase16after mitochondrial internalization through the
TIM23 complex14. The absence of post-import cleavage is typical of
IMM carrier proteins, which are inserted into the IMM by the TIM22
complex14,15,17, but it has also been documented for a few other IMM
or mitochondrial matrix proteins18,19.
To clarify which mitochondrial compartment MPV17 belongs
to, we performed a series of protein blot experiments on different
cellular and suborganellar fractions, using an antibody to MPV17.
We detected material cross-reacting with MPV17 (CRM) in intact
HeLa cell mitochondria (Fig. 4c; same preparation as in Fig. 4b).
We detected enrichment of MPV17 CRM in both the mitochondrial
membrane fraction and in mitoplasts (that is, organelles in which
the outer mitochondrial membrane was disrupted and the content
of the intermembrane space (IMS) was washed out). We did not
detect any MPV17 CRM in the mitochondrial matrix (Fig. 4c).
Figure 4 Mitochondrial localization of MPV17. (a) Confocal immuno-
fluorescence on stably transfected HeLaMPV17cells and transiently
transfected Cos7MPV17cells. mtSSB, mitochondrial single-stranded DNA
binding protein. The two immunofluorescence patterns overlap (bottom row).
Magnification, 40?. (b) In vitro mitochondrial import of human MPV17,
as shown by35S autoradiography. Lane 1: MPV17 (translated in vitro); 2:
MPV17 (translated in vitro) after exposure to intact, energized mitochondria;
3: MPV17 is protected from proteinase K (pK) digestion; 4: MPV17 is
completely digested by pK after solubilization of mitochondria with Triton
X-100; 5: valinomycin abolishes mitochondrial DC; 6: mitochondrial import
is abolished, as demonstrated by complete pK digestion of the protein.
(c) Suborganellar localization of MPV17, as shown by protein blot analysis
on mitochondria isolated from HeLa cells. MPV17 CRM is present in the
HeLa cell lysate but is absent in the postmitochondrial supernatant (Sp)
and the mitochondrial matrix. Progressive enrichment of MPV17 CRM is
evident in isolated mitochondria, mitochondrial membranes and digitonin-
treated mitochondria (mitoplasts). Treatment of mitochondrial membranes
with 0.1% deoxycholate (DOC) did not release MPV17 from the pellet (Pt)
to the supernatant; very limited release was obtained using 1% DOC.
Overlapping results were obtained using an antibody against COX-IV,
a known component of the mitochondrial inner membrane. By contrast,
the CRM for ETHE1 is localized in the mitochondrial matrix12. (d) Protein blot analysis of MPV17 on mitochondria isolated from fibroblasts of individuals
1-2 and 1-4 and of an age-matched control individual (C). The specificity of the antibody to MPV17 was demonstrated by immunostaining of human
in vitro (i.v.)-translated MPV17.
Figure 3 Complementation studies in Saccharomyces cerevisiae.
(a) Oxidative growth phenotype at 37 1C. The BY4741 Dsym1 mutant
was transformed with either empty pFL38, pFL38 carrying SYM1, pFL38
carrying the pathological (sym1R51Q, sym1R51Wand sym1N172K) alleles,
pYEX carrying the human wild-type human allele (MPV17), or pYEX carrying
mutant human alleles (MPV17R50Q, MPV17R50Wand MPV17N166K) alleles.
Equal amounts of serial dilutions of cells from exponentially grown cultures
(105, 104and 103cells) were spotted onto YNB plates supplemented
with 2% glucose or YP plates supplemented with 2% ethanol. Growth was
scored after 5 d of incubation at 37 1C. (b) Respiratory-deficient mutant
accumulation in Dsym1 strains carrying mutant alleles. Black bars and
gray bars represent the percentage of respiratory-deficient and respiratory-
competent colonies, respectively. The percentage of the respiratory-deficient
colonies is shown at left for each recombinant strain. The diagnosis of
respiratory-deficient (petite) mutants was carried out on YNB plates
supplemented with 0.2% glucose and 2% ethanol. More than 1,000
colonies per strain were scored. All values are means of three independent experiments. In no case was the variation higher than 15%. (c) DNA blot analysis
showing the mtDNA patterns obtained from the parental wild-type (WT) strain and from respiratory-deficient clones of Dsym1 strains carrying mutant alleles.
Glucose
pFL38
SYM1
MPV17
MPV17R50Q
MPV17R50W
MPV17N166K
sym1R51Q
sym1R51Q
sym1R51W
sym1N172K
sym1R51W
sym1N172K
Ethanol
26.0
13.6
21.2
0
WT R51Q R51W N172K
20 406080 100
pFL38
SYM1
27.0
4.0
Percentage
respiratory-deficient
∆sym1
ab
c
HeLaMPV17
MPV17HA
mtSSB
Merge
Cos7MPV17
14
25
36
Mitochondria –+
–
–
–
+
+
–
–
+
+
–
+
+
–
+
–
+
+
+
–
–
–
–
pK
Valinomycin
Triton-X100
MPV17
MPV17
COX-IV
ETHE1
MPV17
COX-IV
i.v.#1-2 #1-4 C
12
Cell lysate
Post-mito Sp
Mitochondria
Matrix
MembranesMitoplasts
Pt Sp
0.1%
DOC
Isolated
mitochondria
1.0%
DOC
Pt
Sp
3456
ab
c
d
572VOLUME 38 [ NUMBER 5 [ MAY 2006 NATURE GENETICS
LETTERS
© 2006 Nature Publishing Group http://www.nature.com/naturegenetics

Page 4

Treatment of the mitochondrial membrane fraction with increasing
concentrations of deoxycholate (Fig. 4c) or with 0.1 M Na2CO3
(Supplementary Fig. 2 online), demonstrated that MPV17 is a tightly
bound membrane protein. Taken together, these experiments indicate
that MPV17 is an IMM protein. This conclusion is further supported
by the four hydrophobic, potential transmembrane domains found in
Sym1p (ref. 11), which are conserved in MPV17 (Fig. 2b). Finally, we
did not find any MPV17 CRM in mitochondria isolated from
fibroblasts of individuals 1-2 and 1-4 (Fig. 4d), suggesting that the
R50Q mutation causes protein instability and decay.
In contrast with previous studies that suggested that Mpv17 was
localized in the membrane of peroxisomes9, our MPV17HAimmuno-
fluorescence pattern consistently failed to match the pattern specific
to the peroxisomal membrane protein PMP70 (Fig. 5a). These
results, and those obtained by protein blot analysis on mouse liver
mitochondria and purified peroxisomes (Fig. 5b), clearly demonstrate
that MPV17 is largely, if not exclusively, confined to mitochondria.
The Mpv17 gene was first identified as the target of a transgene
insertion by a retroviral genome in a mouse germinal line, resulting in
Mpv17–/–animals9. Accordingly, we did not find any Mpv17 CRM in
mitochondria isolated from Mpv17–/–livers (Fig. 6). The mtDNA
content, measured by real-time PCR (Fig. 6), was markedly decreased
in liver of Mpv17–/–mice (approximately 4% ± 3 s.d. of mean control
values) and progressively less so in muscle (20% ± 8%), kidney (60%
± 19%) and brain (60% ± 18%). We obtained similar results by DNA
blot analysis (Fig. 6b). MtDNA-dependent respiratory chain activities
(complex I and complex IV) were consistently and significantly
decreased in liver of Mpv17–/–animals but were normal in muscle
and kidney (Fig. 6d).
The Mpv17–/–mice have been reported to develop age-dependent
hearing loss20and severe renal failure owing to glomerular sclerosis8.
However, after these initial observations, the kidney disease could no
longer be documented in Mpv17–/–mouse strains investigated at
different ages and in different research centers (M.Z., personal
observation). We are baffled by the disappearance of this phenotype.
A possible explanation is genetic selection of adaptive mechanisms
that markedly reduced the penetrance of the kidney-specific trait.
A less likely explanation is that this trait was due to a second mutation
in another gene that was lost over time during the selection of
Mpv17–/–and Mpv17+/–individuals.
Figure 6 Characterization of Mpv17–/–mice.
(a) Protein blot analysis of Mpv17 on liver
mitochondria and submitochondrial fractions
of 6-month-old Mpv17–/–and age-matched
Mpv17+/+control mice. Five animals were used
for each series. No Mpv17-CRM was detected
in the Mpv17–/–samples. (b) DNA blot analysis
of linearized mtDNA extracted from different
tissues of two Mpv17–/–mice versus one
Mpv17+/+mouse. The band corresponding to 18S
ribosomal DNA (rDNA) serves as a standard index
of the nuclear DNA content in each sample.
Histogram below the DNA blot shows the mtDNA/
18S rDNA ratios measured densitometrically in
the Mpv17–/–mice (black) and Mpv17+/+mouse
(gray), normalized to 100%. (c) Real-time PCR
quantification of the amount of mtDNA relative
to that of GAPDH, a nuclear gene used as a
standard. Five Mpv17–/–mice (black) were
compared with five age-matched Mpv17+/+mice
(gray), normalized to 100%. **, P o 0.001;
*, P o 0.01 (two-tailed unpaired Student’s
t-test). (d) Specific activities of complex I (CI),
complex IV (CIV) and succinate dehydrogenase (SDH), normalized to the specific activity of citrate synthase (CS), in kidney (K), liver (L), and muscle (M)
homogenates from five Mpv17–/–mice (black) and five age-matched Mpv17+/+mice (gray), normalized to 100%. SDH is a respiratory chain complex that is
encoded entirely in the nucleus. **, P o 0.001; *, P o 0.01 (two-tailed unpaired Student’s t-test).
HeLaMPV17
14
25
36
MPV17HA
PMP70
Merge
Cos7MPV17
PMP70
1234
Mpv17
COX-IV
ab
Figure 5 Mpv17 versus peroxisomal PMP70 immunostaining. (a) Confocal
immunofluorescence on stably transfected HeLaMPV17cells and transiently
transfected Cos7MPV17cells. The two immunofluorescence patterns do
not overlap (bottom row). Magnification, 40?. (b) Protein blot analysis
of different cell fractions from mouse liver, including a 2,000g fraction
containing ‘heavy’ mitochondria (lane 1), a 25,000g fraction enriched
in peroxisomes and ‘light’ mitochondria (lane 2), and two 100,000g
iodixanol gradient fractions separating ‘light’ mitochondria and microsomes
(lane 3) from purified peroxisomes (lane 4). Enrichment in the content of
peroxisomes can be seen in the increasing intensity of the PMP70-specific
band from left to right. No mitochondrial contamination is present in the
purified peroxisomal fraction (lane 4), as demonstrated by the virtual
absence of COX-IV CRM.
Mitochondria
Mpv17
Ethe1
150
140
120
100
80
60
40
20
0
100
50
0
BrainKidney Liver MuscleKL
CI/CS
MKL
CIV/CS
MKL
SDH/CS
M
+/+ –/–+/+–/– +/+ –/–
–/––/– –/– –/– –/––/– –/––/– +/++/++/++/+
MatrixMembranes
Brain
mtDNA
18S rDNA
100
50
0
KidneyLiver Muscle
**
**
**
*
*
*
ab
cd
NATURE GENETICS VOLUME 38 [ NUMBER 5 [ MAY 2006573
LETTERS
© 2006 Nature Publishing Group http://www.nature.com/naturegenetics

Page 5

Contrary to observations in humans, the absence of Mpv17 is
compatible with survival to adulthood in Mpv17–/–mice. However, in
the affected individuals used in this study, metabolic failure occurs, or
worsens substantially, during stress challenges such as fasting or febrile
episodes, conditions that are not experienced by laboratory animals.
Further studies, including stress challenges, are warranted to assess the
mitochondrial pathophysiology of Mpv17–/–mice.
The temperature-sensitive phenotype of the Dsym1 yeast strain
suggests a role for Sym1p in the cellular response to stress11. In
mouse fibroblasts, the absence of Mpv17 reduces, and its overexpres-
sion increases, the production of reactive oxygen species9. In addition,
our own findings indicate a role for MPV17 in controlling mtDNA
maintenance and OXPHOS activity in mammals and yeast. Taken
together, these results provide clues for future studies of the function
of this protein in mitochondrial homeostasis.
METHODS
Patients and DNA samples. All procedures were approved by the Review Board
of the National Neurological Institute ‘C. Besta’. We obtained informed consent
from parents of all probands and siblings before collecting blood for DNA
extraction or performing tissue biopsies. The treatment of animals was
according to the corresponding guidelines approved by the Ethical Committee
of the National Neurological Institute, Italy.
Linkage analysis. A genome-wide search was performed using the Applied
Biosystems (ABI) PRISM Linkage Mapping Set v. 2.5 on an ABI PRISM 3100
DNA sequencer, and results were processed by GENESCAN software. Alleles
were assigned using the GENOTYPER software. Statistical analysis was per-
formed assuming complete penetrance of a recessive disease. Pairwise linkage
analysis and multipoint analysis were performed using the LINKAGE computer
package and Simwalk2, respectively.
Identification of mitochondrial candidates. Every gene in the linkage region
(D2S2373–D2S2259), based on Ensembl annotations, was assigned a Maestro
score representing the likelihood of mitochondrial localization. The likelihood
scores were computed based on a naive Bayesian integration of eight genomic
data sets, where the contribution of each data set was weighted by its accuracy.
A significance threshold was determined based on a large training set of known
mitochondrial proteins.
Antibodies. The affinity-purified polyclonal antibody to Mpv17 was from
Proteintech, and the polyclonal antibody specific to the peroxisomal membrane
marker PMP70 was from Sigma. The mouse monoclonal antibody to the
hemagglutinin epitope of the influenza virus (anti-HA) was from Roche; the
monoclonal antibody against cytochrome c oxidase subunit IV was from
Molecular Probes/Invitrogen. Polyclonal antibodies against human mitochon-
drial single-stranded DNA binding protein and ETHE1 were raised in rabbit
and characterized as described13. Cross-reactivity of the antibody to Mpv17 was
confirmed by immunostaining of a human MPV17 recombinant protein
obtained by in vitro translation (see Fig. 4d).
Cell cultures and immunofluorescence studies. Mammalian cells were cul-
tured in DMEM + 10% fetal calf serum (FCS) at 37 1C in a 5% CO2
atmosphere. For immunofluorescence analysis, cells were plated on coverslips,
followed by fixation and incubation with primary and fluorescent dye–
conjugated secondary antibodies, as described13. In some experiments, cells
were preincubated with 100 nM MitoTracker Red dye (Molecular Probes) for
45 min at 37 1C. Fluorescence patterns were visualized by a confocal micro-
scope (BioRad).
Protein blot analysis. Approximately 2 ? 107cells were used to prepare
mitochondrial fractions in both HeLa cells and human fibroblasts21.
Equal amounts of non-collagen protein (50–100 mg protein per lane) were
used in SDS-polyacrylamide gel electrophoresis and protein blot analysis,
as described13.
Cell fractionation. Standard methods were used for the preparation of cell
lysates, mitochondrial and postmitochondrial fractions in cultured cells and
tissue homogenates21. For suborganellar localization, freshly isolated mitochon-
dria from HeLa cells were treated with three cycles of freezing-thawing and six
sonication strokes at 4 1C. The membrane and soluble fractions were then
separated by ultracentrifugation at 105g for 1 h at 4 1C. Mitoplasts were isolated
from mitochondria from HeLa cells by either digitonin treatment22or
hypoosmotic shock23. In some experiments, mitochondrial membranes were
treated with the nonionic detergent deoxycholate24or with 0.1 M Na2CO3,
pH 11.0 (ref. 25). Isolation of peroxisomes from mouse liver based on
iodixanol gradient centrifugation was performed using the Peroxisome Isola-
tion Kit (Sigma).
Biochemical assays. Specific activities of individual respiratory chain com-
plexes were measured on cell and tissue homogenates26. Specific activities of
each complex were normalized to that of citrate synthase (CS)26, an indicator of
the number of mitochondria.
In vitro import assay. A human MPV17 cDNA was amplified by PCR using
suitable pairs of primers flanking the coding region. The T7 phage promoter
was incorporated in the sense primer for in vitro transcription of the PCR
fragment. In vitro transcription and translation were performed using the TNT
T7 Quick Coupled Transcription/Translation System (Promega) in the presence
of 20 mCi [35S]methionine (Amersham). Fresh mitochondria were prepared
from 2 ? 107HeLa cells21and were resuspended with incubation buffer27to a
final protein concentration of 2 mg ml–1. The in vitro import assay was then
carried out as described27. Samples were resuspended in 20 ml solubilization
buffer24, boiled for 5 min and separated on a 16% SDS-polyacrylamide gel.
After fixation in 10% acetic acid/25% isopropanol, the gel was washed for
20 min in Amplify reagent (Amersham) and exposed using a phosphorimaging
screen (Biorad).
Generation of human MPV17 mutant variants. The QuikChange Site-
Directed Mutagenesis Kit (Stratagene) and the oligonucleotide pairs reported
in Supplementary Table 1 online were used to introduce three different point
mutations in the human MPV17 cDNA. After mutagenesis, sequences of inserts
were verified on both strands.
Yeast strains and culture media. Yeast strains were BY4741 (MATa
his3D1 leu2D0 lys2D0 ura3D0) and its isogenic Dsym1::kanMX4 mutant. Cells
were cultured in 0.67% Yeast Nitrogen Base (YNB) medium without amino
acids (Difco), supplemented with the appropriate amino acids and bases
to a final concentration of 40 mg ml–1. Various carbon sources were added
at 2% (wt/vol).
The SYM1 gene was amplified by PCR from the BY4741 wild-type strain and
cloned into the centromeric yeast plasmid pFL38. Site-directed mutagenesis,
performed by the overlap extension technique28, was used to introduce three
different point mutations in the S. cerevisiae SYM1 gene. The list of base
changes and corresponding modified primers used to generate mutated
alleles are listed in Supplementary Table 1. Sequences of mutant inserts were
verified on both strands. Both wild-type and mutated human MPV17 cDNAs
were cloned into the pYEX plasmid. The Dsym1 strain was transformed as
previously described29.
For respiratory-deficient diagnosis (petite colonies), cells from cultures
grown for seven generations in YNB supplemented with glucose and ethanol
at 37 1C were plated onto solid YNB medium containing 2% ethanol and 0.2%
glucose and maintained for 5 d at 28 1C. PCR-based analysis of mtDNA on
respiratory-deficient colonies was performed as previously described28.
Generation and cloning of MPV17HA. A full-length human MPV17 cDNA
clone was obtained from the RZPD consortium (clone IRAUp969G055D6). A
chimeric oligonucleotide primer containing the 3¢ end of the human MPV17
cDNA ORF in frame with HA encoding sequence13was used to obtain an HA-
tagged MPV17 cDNA fragment that was inserted into the eukaryotic expression
plasmid vector pcDNA3.1 (Invitrogen) using suitable restriction sites. The
recombinant plasmid was transfected by electroporation in Cos7 (for transient
expression) or in HeLa (for stable expression) cells as previously described13.
574VOLUME 38 [ NUMBER 5 [ MAY 2006 NATURE GENETICS
LETTERS
© 2006 Nature Publishing Group http://www.nature.com/naturegenetics

Page 6

DNA blot and real-time PCR. Total DNAwas extracted from human liver and
muscle biopsies and from different mouse tissues, following standard proce-
dures. For DNA blot analysis, mouse and human mtDNAs were linearized
using the restriction enzymes SacI and PvuII, respectively. Species-specific 1-kb
probes encompassing the human or mouse mtDNA sequence encoding
cytochrome c oxidase subunit I (COI) were used for mtDNA detection. A
1-kb probe encompassing part of the sequence encoding the mouse nuclear 18S
rDNA (Supplementary Table 2 online) was used as a standard index of the
nuclear DNA content in each sample for densitometric measurements. Both
probes were labeled with32P using the Decaprime II kit (Ambion). Yeast
mtDNA was extracted by rapid mitochondrial preparation, digested with
EcoRV (Amersham) and detected by DNA blot analysis, as described28.
Real-time quantitative PCR on mouse mtDNAwas carried out using an ABI
PRISM 7000 Sequence Detection System in a two-step reaction, essentially as
described previously30. Primers and detection probe specific to a region of the
murine COI gene were designed using ABI Primer Express software (Supple-
mentary Table 2). Each PCR reaction was performed in triplicate with the
following profile: one cycle at 50 1C for 2 min, one cycle at 95 1C for 10 min,
and then 40 cycles of 95 1C for 15 s and 60 1C for 1 min (two-step protocol).
Primers and detection probe of the genes encoding either human RNase-P or
mouse glyceraldehyde-3-phosphate dehydrogenase were used as nuclear gene
standard references, according the instructions of the manufacturer.
RNA blot. A pre-made multiple-tissue RNA blot containing 1 mg per lane of
purified polyA(+) RNA (Clontech) was incubated in ExpressHyb Hybridization
Solution (Clontech) for 30 min at 65 1C and then was hybridized for 2 h at
65 1C with a32P-labeled human MPV17 cDNA (Decaprime II, Ambion). The
multiple-tissue RNA blot filter was exposed overnight and visualized by a
phosphorimager apparatus (BioRad).
Genotyping. Nucleotide sequence analysis was carried out on a 3100 ABI
Automated Sequencer on samples prepared using the BigDye Termination kit
(Applied Biosystems). Data were elaborated using Secscape software (Applied
Biosystems). The oligonucleotide primers used for PCR amplification of the
seven exons of the human MPV17 gene are listed in Supplementary Table 3
online. PCR amplification conditions for exon 1, exons 4–6 and exon 7 were an
initial denaturation step at 94 1C for 3 min; followed by 32 cycles of 94 1C for
1 min, 54 1C for 30 s and 72 1C for 1 min; plus a final extension at 72 1C for
2 min. Exon 2 and exons 3+4 were amplified as above, except that the annealing
temperature was 58 1C.
Statistics. A two-tailed, unpaired Student’s t-test was used to calculate the
significance of biochemical and molecular data.
URLs.
mips.gsf.de/cgi-bin/proj/medgen/mitofilter; PSORT: http://psort.nibb.ac.jp/;
Predotar: http://www.inra.fr/predotar/french.html; TargetP: http://www.cbs.
dtu.dk/services/TargetP/; CLUSTALW: http://www.ebi.ac.uk/clustalw/; rzpd:
http://www.rzpd.de/; Ensembl: http://www.ensembl.org.
NCBI-Unigene:http://www.ncbi.nlm.nih.gov; Mitoprot2: http://
Note: Supplementary information is available on the Nature Genetics website.
ACKNOWLEDGMENTS
We are indebted to B. Geehan for revising the manuscript, E. Lamantea for
technical assistance and L. Palmieri for critical discussion. This work was
supported by Fondazione Telethon-Italy (grant GGP030039), Fondazione
Pierfranco e Luisa Mariani and MITOCIRCLE and EUMITOCOMBAT network
grants from the European Union Framework Program 6.
COMPETING INTERESTS STATEMENT
The authors declare that they have no competing financial interests.
Published online at http://www.nature.com/naturegenetics
Reprints and permissions information is available online at http://npg.nature.com/
reprintsandpermissions/
1. Spinazzola, A. & Zeviani, M. Disorders of nuclear-mitochondrial intergenomic signal-
ing. Gene 354, 162–168 (2005).
2. Saada, A. et al. Mutant mitochondrial thymidine kinase in mitochondrial DNA deple-
tion myopathy. Nat. Genet. 29, 342–344 (2001).
3. Elpeleg, O. et al. Deficiency of the ADP-forming succinyl-CoA synthase activity is
associated with encephalomyopathy and mitochondrial DNA depletion. Am. J. Hum.
Genet. 76, 1081–1086 (2005).
4. Mandel, H. et al. The deoxyguanosine kinase gene is mutated in individuals with
depleted hepatocerebral mitochondrial DNA. Nat. Genet. 29, 337–341 (2001).
5. Nguyen, K.V. et al. POLG mutations in Alpers syndrome. Neurology 65, 1493–1495
(2005).
6. Ferrari, G. et al. Infantile hepatocerebral syndromes associated with mutations in the
mitochondrial DNA polymerase-gammaA. Brain 128, 723–731 (2005).
7. Salviati, L. et al. Mitochondrial DNA depletion and dGK gene mutations. Ann. Neurol.
52, 311–317 (2002).
8. Weiher, H., Noda, T., Gray, D.A., Sharpe, A.H. & Jaenisch, R. Transgenic mouse model
of kidney disease: insertional inactivation of ubiquitously expressed gene leads to
nephrotic syndrome. Cell 62, 425–434 (1990).
9. Zwacka, R.M. et al. The glomerulosclerosis gene Mpv17 encodes a peroxisomal protein
producing reactive oxygen species. EMBO J. 13, 5129–5134 (1994).
10. Calvo, S. et al. Systematic identification of human mitochondrial disease genes
through integrative genomics. Nat. Genet., advance online publication 2 April 2006
(doi:10.1038/ng1776).
11. Trott, A. & Morano, K.A. SYM1 is the stress-induced Saccharomyces cerevisiae
ortholog of the mammalian kidney disease gene Mpv17 and is required for ethanol
metabolism and tolerance during heat shock. Eukaryot. Cell 3, 620–631 (2004).
12. Mounolou, J.C., Jakob, H. & Slonimski, P.P. Mitochondrial DNA from yeast ‘‘petite’’
mutants: specific changes in buoyant density corresponding to different cytoplasmic
mutations. Biochem. Biophys. Res. Commun. 24, 218–224 (1966).
13. Tiranti, V. et al. Ethylmalonic encephalopathy is caused by mutations in ETHE1, a gene
encoding a mitochondrial matrix protein. Am. J. Hum. Genet. 74, 239–252 (2004).
14. Koehler, C.M. New developments in mitochondrial assembly. Annu. Rev. Cell Dev. Biol.
20, 309–335 (2004).
15. Wiedemann, N., Frazier, A.E. & Pfanner, N. The protein import machinery of mito-
chondria. J. Biol. Chem. 279, 14473–14478 (2004).
16. Gakh, O., Cavadini, P. & Isaya, G. Mitochondrial processing peptidases. Biochim.
Biophys. Acta 1592, 63–77 (2002).
17. Rehling, P., Pfanner, N. & Meisinger, C. Insertion of hydrophobic membrane proteins
into the inner mitochondrial membrane – a guided tour. J. Mol. Biol. 326, 639–657
(2003).
18. Otsuka, M., Mizuno, Y., Yoshida, M., Kagawa, Y. & Ohta, S. Nucleotide sequence of
cDNA encoding human cytochrome c oxidase subunit VIc. Nucleic Acids Res. 16,
10916 (1988).
19. Hammen, P.K., Gorenstein, D.G. & Weiner, H. Structure of the signal sequences for two
mitochondrial matrix proteins that are not proteolytically processed upon import.
Biochemistry 33, 8610–8617 (1994).
20. Meyer zum Gottesberge, A.M., Reuter, A. & Weiher, H. Inner ear defect similar to
Alport’s syndrome in the glomerulosclerosis mouse model Mpv17. Eur. Arch. Otorhi-
nolaryngol. 253, 470–474 (1996).
21. Fernandez-Vizarra, E., Lopez-Perez, M.J. & Enriquez, J.A. Isolation of biogenetically
competent mitochondria from mammalian tissues and cultured cells. Methods 26,
292–297 (2002).
22. Gallet, P.F. et al. Transbilayer movement and distribution of spin-labelled phospholi-
pids in the inner mitochondrial membrane. Biochim. Biophys. Acta 1418, 61–70
(1999).
23. Ohba, M. & Schatz, G. Disruption of the outer membrane restores protein import to
trypsin-treated yeast mitochondria. EMBO J. 6, 2117–2122 (1987).
24. Tiranti, V. et al. Characterization of SURF-1 expression and Surf-1p function in normal
and disease conditions. Hum. Mol. Genet. 8, 2533–2540 (1999).
25. Fujiki, Y., Hubbard, A.L., Fowler, S. & Lazarow, P.B. Isolation of intracellular mem-
branes by means of sodium carbonate treatment: application to endoplasmic reticu-
lum. J. Cell. Biol. 93, 97–102 (1982).
26. Bugiani, M. et al. Clinical and molecular findings in children with complex I deficiency.
Biochim. Biophys. Acta 1659, 136–147 (2004).
27. Petruzzella, V. et al. Identification and characterization of human cDNAs specific
to BCS1, PET112, SCO1, COX15, and COX11, five genes involved in the formation
and function of the mitochondrial respiratory chain. Genomics 54, 494–504
(1998).
28. Ho, S.N., Hunt, H.D., Horton, R.M., Pullen, J.K. & Pease, L.R. Site-directed
mutagenesis by overlap extension using the polymerase chain reaction. Gene 77,
51–59 (1989).
29. Fontanesi, F. et al. Mutations in AAC2, equivalent to human adPEO-associated
ANT1 mutations, lead to defective oxidative phosphorylation in Saccharomyces
cerevisiae and affect mitochondrial DNA stability. Hum. Mol. Genet. 13, 923–934
(2004).
30. Winterthun,S. etal. Autosomalrecessive
due to mitochondrial polymerase gamma mutations. Neurology 64, 1204–1208
(2005).
mitochondrialataxic syndrome
NATURE GENETICS VOLUME 38 [ NUMBER 5 [ MAY 2006575
LETTERS
© 2006 Nature Publishing Group http://www.nature.com/naturegenetics