Brain (2000), 123, 93–104
Apoptosis in mitochondrial encephalomyopathies
with mitochondrial DNA mutations: a potential
pathogenic mechanism
Massimiliano Mirabella, Simone Di Giovanni, Gabriella Silvestri, Pietro Tonali and Serenella Servidei
Institute of Neurology, Catholic University, Rome, Italy
Correspondence to: Serenella Servidei, Istituto di
Neurologia, Università Cattolica del S. Cuore, Largo A.
Gemelli 8, 00168 Rome, Italy
Summary
Mitochondrial encephalomyopathies caused by mitochondrial DNA (mtDNA) defects are a genetically and
phenotypically heterogeneous group of disorders. The site,
percentage and distribution of mutations do not explain
the overall clinical heterogeneity that is found. Apoptosis
(programmed cell death) is an evolutionarily conserved
mechanism that is essential for tissue development and
homeostasis. Dysregulation of apoptosis has been
implicated in the pathogenesis of various human diseases,
such as cancer and autoimmune and neurodegenerative
disorders. Recent in vitro evidence has indicated the central
role of mitochondria in the apoptotic process. We
investigated the occurrence of apoptosis in muscle biopsies
of 36 patients carrying different mtDNA mutations and
four patients with inclusion body myositis and
mitochondrial abnormalities. Apoptotic features, mainly
localized in cytochrome c oxidase-negative fibres, were
observed in muscle fibres of patients carrying a high
percentage of single mtDNA deletions (>40%) and of tRNA
point mutations (>70%). By contrast, no apoptotic changes
were observed in inclusion body myositis and in patients
carrying mutations of mtDNA structural genes. Our study
suggests that apoptosis is not simply a means whereby cells
with dysfunctional mitochondria are eliminated, but that
it seems to play a role in the pathogenesis of mitochondrial
disorders associated with mtDNA defects affecting
mitochondrial protein synthesis. The imbalance and
relative abundances of nuclear-encoded and mtDNAencoded subunits may favour cytochrome c inactivation
and release. Cytochrome c, together with respiratory chain
dysfunction, could activate apoptotic pathways that, in
turn, inhibit the rate of mitochondrial translation and the
importation of nuclear-encoded mitochondrial protein
precursors. This vicious circle may amplify the biochemical
defects and tissue damage and contribute to the modulation
of clinical features.
Keywords: apoptosis; mitochondrial encephalomyopathies; mtDNA mutations
Abbreviations: COX ⫽ cytochrome c oxidase; LHON ⫽ Leber hereditary optic neuropathy; MELAS ⫽ mitochondrial
encephalomyopathy, lactic acidosis and stroke-like episodes; MERRF ⫽ myoclonic epilepsy and ragged red fibres; mtDNA ⫽
mitochondrial DNA; MNGIE ⫽ myogastrointestinal encephalopathy; NARP ⫽ neuropathy ataxia retinitis pigmentosa; PBS ⫽
phosphate-buffered saline; PEO ⫽ progressive external ophthalmoplegia; RRF ⫽ ragged red fibres; TUNEL ⫽ terminal
deoxynucleotidyl transferase-mediated dUTP nick end labelling
Introduction
Mitochondrial encephalomyopathies represent an expanding
group of clinically heterogeneous disorders associated with
mitochondrial DNA (mtDNA) mutations or nuclear gene
defects (Wallace, 1992; Morgan-Hughes, 1994; Di Mauro and
Bonilla, 1997; Di Mauro and Schon, 1998). The complexity of
mitochondrial metabolism, its central role in energy
production, its dual genetic control (mitochondrial and nuclear
DNA) and some unique features of mitochondrial genetics may
explain the exceptional clinical heterogeneity of mitochondrial
disorders. In sharp contrast to the rapid progress of knowledge
© Oxford University Press 2000
in molecular genetics, our understanding of the pathogenic
mechanisms that lead to the wide biochemical and clinical
variability is still far from clear. Heteroplasmy, the threshold
effect, and the site, percentage and inter- and intra-tissue
distribution of the mtDNA mutations contribute to the
phenotype but do not explain the overall clinical heterogeneity
that is also present within the same genetic defect. Moreover,
while a nuclear defect may lead to specific enzymatic
impairment, the correlation between mtDNA mutations,
respiratory chain dysfunction and clinical outcome is less
94
M. Mirabella et al.
obvious. The altered synthesis of mitochondrial proteins—the
consequence of mtDNA defects—may have various causes,
such as abnormal mRNA processing, the accumulation of
anomalous RNA transcripts, the impairment of transcript
binding to ribosomes, altered aminoacylation and incorrect
amino acid-tRNA conjugation (Di Mauro and Bonilla, 1997;
Schon, 1997; Di Mauro and Schon, 1998). Whatever the
mechanism, the final common step in mitochondrial
encephalomyopathies is a defect of energy production resulting
from respiratory chain impairment. Cellular necrosis is seldom
observed in these disorders. Other mechanisms, such as
incorrect assembly of the defective enzyme complex or
complexes with other mitochondrial proteins, the production
of free radicals and ways of cell death other than necrosis such
as apoptosis, may be postulated to explain the phenotypic
variability and the severity of mitochondrial diseases.
Apoptosis (programmed cell death) is an evolutionarily
conserved mechanism that is essential for tissue development
and homeostasis. It requires the activation of specific genes that
lead to a series of distinctive morphological and biochemical
features. These changes include the activation of cellular
proteases (caspases), mitochondrial depolarization, chromatin
condensation, oligonucleosomal DNA degradation and cell
volume loss or cell fragmentation without elicitation of an
inflammatory response (White, 1996; Salvasen and Dixit,
1997; Hetts, 1998).
Apoptotic features in human pathology are documented in
neoplasms, autoimmune diseases, stroke and some neurodegenerative disorders such as Alzheimer’s disease,
Parkinson’s disease and familial amyotrophic lateral sclerosis
(Hetts, 1998).
In human muscle pathology, apoptotic nuclei and apoptosisrelated proteins have been demonstrated in atrophic fibres in
neurogenic muscle atrophy (Fidziamska et al., 1990; Mirabella
et al., 1996; Tews and Goebel, 1997).
Mitochondria have recently been found to have a leading
role in the triggering and mediation of apoptosis. Experimental
studies in vitro have shown that the disruption of the
mitochondrial transmembrane potential (∆Ψ) and the release
of some mitochondrial proteins (cytochrome c and apoptosis
inducing factor) into the cytoplasm are able to initiate and
activate different apoptotic pathways (Liu et al., 1996;
Kroemer et al., 1997; Kluck et al., 1997; Zhivotosky et al.,
1998).
In order to discover whether apoptosis plays a part in
mediating tissue damage in human mitochondrial diseases,
we investigated the presence of DNA fragmentation and the
expression of apoptosis-associated proteins (Fas, p75 and
caspase-3) in muscle biopsies of patients with mtDNA point
mutations and deletions.
Material and methods
Patients
We studied muscle biopsies obtained, with informed consent,
from 36 patients of different ages (2 months to 69 years)
affected by various mitochondrial disorders with heterogeneous mtDNA defects, four patients with inclusion body
myositis and mitochondrial abnormalities and 10 patients
who had proved to be free of muscle disease. Diagnosis of
all mitochondrial patients was based on clinical, biochemical
and molecular genetic studies (their clinical, morphological,
biochemical and genetic features are summarized in Table 1).
Five patients had MERRF (myoclonic epilepsy and ragged
red fibres) with an mtDNA mutation at nucleotide 8344 in
the tRNALys gene (patients 1–5); six patients had the A3243G
transition in the tRNALeu (UUR) gene, three of them with typical
mitochondrial encephalomyopathy, lactic acidosis and strokelike episodes (MELAS) (patients 6, 7 and 11), one with
progressive external ophthalmoplegia (PEO) (patient 8), one
with cardiomyopathy (patient 9) and one with PEO and
cardiomyopathy (patient 10); 16 patients had a single large
mtDNA deletion associated with multisystem Kearns–Sayre
syndrome (patients 12–14), Pearson’s syndrome (patient 15),
encephalopathy (patient 16) or PEO (patients 17–27); four
patients (patients 28–31) had PEO with multiple mtDNA
deletions, two with autosomal dominant transmission and
two sporadic cases; one patient with myoneurogastrointestinal
encephalopathy (MNGIE) syndrome carried a 1 bp deletion
(patient 32); two patients (patients 33 and 34) had PEO
associated with a T4285C (tRNAIle) and G5521A (tRNATrp)
point mutation, respectively; patient 35 had neuropathy ataxia
retinitis pigmentosa (NARP) syndrome associated with a
T8993C point mutation in the ATPase 6 gene; patient 36 had
Leber hereditary optic neuropathy (LHON) associated with
a G11778A mutation in the ND4 gene.
Mitochondrial myopathy, ranging from absent (NARP and
LHON) to marked (Table 1), was defined morphologically by
the presence of a variable number of succinate dehydrogenase
strongly reactive ragged red fibres (RRF) and of cytochrome
c oxidase (COX)-negative non-RRF.
The four inclusion body myositis patients with typical
nuclear and cytoplasmic filamentous inclusions in vacuolated
fibres (Griggs et al., 1995) were chosen because of the
presence of numerous RRF and COX-negative fibres. In
these patients, electron microscopy showed proliferation in
the size and number of mitochondria with abnormal shape
and paracrystalline inclusions. A very low number of mtDNA
deletions was demonstrated in inclusion body myositis
patients by the PCR (polymerase chain reaction), but not by
Southern blotting.
TUNEL and immunohistochemistry
The terminal deoxynucleotidyl transferase-mediated dUTP
nick end labelling (TUNEL) technique was used for the
detection of nuclear DNA fragmentation in situ. Frozen
muscle sections from patients and controls were incubated
under the same coverslip with TUNEL reaction mixture, and
incorporated fluorescein-dUTP was detected by using alkaline
phosphatase-conjugated anti-fluorescein antibodies according
to the manufacturer’s instructions (In Situ Cell Death
62 y
36 y
40 y
2y
30 y
52 y
18 y
34 y
31 y
34 y
26 y
10 y
34 y
26 y
2m
20 y
43 y
42 y
14 y
36 y
46 y
65 y
69 y
69 y
58 y
36 y
49 y
27 y
25 y
67 y
47 y
24 y
15 y
68 y
9y
18 y
1FD
2F
3F
4F
5F
6F
7F
8F
9F
10 M
11 M D
12 M
13 M D
14 F
15 F D
16 F
17 M
18 M
19 M
20 F
21 F
22 F
23 M
24 M
25 F
26 F
27 F
28 M
29 F
30 F
31 F
32 M
33 F
34 M
35 F
36 M
MERRF
MERRF
MERRF
MERRF
MERRF
MELAS
MELAS
PEO
Cardiomyopathy
PEO
cardiomyopathy
MELAS
KSS
KSS
KSS
Pearson syndrome
Encephalopathy
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
PEO
MNGIE
PEO
PEO/myopathy
NARP
LHON
Phenotype
Mitochondrial‡
myopathy
⫹⫹⫹
⫹⫹
⫹
–
⫹
⫹⫹
⫹⫹
⫹
⫹/–
⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹⫹
⫹
–
⫹/–
⫹/–
⫹
⫹
⫹/–
⫹
⫹
⫹⫹
⫹⫹
⫹⫹⫹
⫹
⫹
⫹⫹
⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹
–
–
Severity of
phenotype
⫹⫹⫹⫹
⫹⫹
⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹⫹
⫹⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹⫹⫹
⫹
⫹
⫹⫹⫹
⫹
Normal
IV 55%
I 0%; IV 11%
Normal
I 40%; IV 39%
IV 64%
Normal
IV 60%
Normal
IV 56%
I 36%; IV 48%
Normal
Normal
Normal
I 42%; IV 47%
Normal
Normal
IV 67%
IV 61%
IV 88%
Normal
IV 58%
IV 67%
IV 15%
Normal
Normal
IV 73%
IV 50%
IV 68%
Normal
Normal
I 20%
I 50%
Normal
Normal
I 24%, IV 39%
Biochemistry§
Maternal
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
Sporadic
AD
AD
Sporadic
Sporadic
Sporadic
Sporadic
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Maternal
Inheritance
tRNALys
tRNALys
tRNALys
tRNALys
tRNALys
tRNALeu(UUR)
tRNALeu(UUR)
tRNALeu(UUR)
tRNALeu(UUR)
tRNALeu(UUR)
A3243G tRNALeu(UUR)
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Single deletion
Multiple deletions
Multiple deletions
Multiple deletions
Multiple deletions
1 bp deletion
T4285C tRNA-Ile
G5521A tRNA-Trp
T8993C ATPase 6
G11778A ND4
A8344G
A8344G
A8344G
A8344G
A8344G
A3243G
A3243G
A3243G
A3243G
A3243G
mtDNA defect
91%
78%
97%
100%
90%
55%
62%
48%
78%
44%
25%
39%
31%
33%
30%
10%
35%
45%
40%
50%
65%
73%
87%
76%
77%
85%
86%
80%
54%
80%
78%
% of
mutant
genomes
⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹⫹⫹
⫹
–
–
–
–
–
–
–
⫹
⫹
⫹⫹
⫹⫹⫹
⫹/–
⫹/–
⫹/–
⫹/–
⫹⫹⫹
⫹
⫹⫹
–
–
⫹⫹⫹
⫹⫹
⫹
⫹
⫹
⫹⫹
⫹⫹
–
⫹⫹
⫹⫹
TUNEL *
⫹⫹
⫹
⫹⫹⫹
⫹
⫹⫹
⫹
–
–
–
–
–
–
–
⫹
⫹
⫹⫹
⫹⫹
⫹/–
⫹/–
⫹/–
⫹/–
⫹⫹
⫹
⫹⫹
⫹
–
⫹⫹
⫹⫹
–
⫹
⫹
⫹
⫹⫹
–
⫹
⫹
p75 *
⫹⫹
⫹
⫹⫹
⫹
⫹⫹⫹
⫹
–
–
–
–
–
–
–
⫹
⫹
⫹⫹
⫹⫹
⫹/–
⫹/–
⫹/–
⫹/–
⫹⫹
⫹
⫹⫹
⫹
–
⫹⫹
⫹⫹
–
⫹
⫹
⫹⫹
⫹⫹
⫹
⫹⫹
⫹⫹
Fas *
⫹⫹
⫹⫹
⫹⫹
⫹
⫹⫹⫹
⫹
–
–
–
–
–
–
–
⫹
⫹
⫹
⫹⫹
⫹/–
⫹/–
⫹/–
⫹/–
⫹⫹
⫹
⫹
–
–
⫹⫹
⫹⫹
–
–
⫹
⫹⫹
⫹
–
⫹⫹
⫹⫹
Cas-3 *
M ⫽ male; F ⫽ female; D ⫽ deceased; y ⫽ years; m ⫽ months. †Severity of phenotype: mild (⫹) to severe (⫹⫹⫹⫹). ‡Mitochondrial myopathy: – ⫽ absent; ⫹/– ⫽ very mild;
⫹ ⫽ mild; ⫹⫹ ⫽ moderate; ⫹⫹⫹ ⫽ marked. §Respiratory chain: I ⫽ complex I; II ⫽ complex II; IV⫽ complex IV or cytochrome c oxidase; residual activity as % of normal.
*Percentage of muscle fibres positive for TUNEL, p75, Fas and caspase-3: – ⫽ 0–3%; ⫹/– ⫽ 3–5%; ⫹ ⫽ 5–25%; ⫹⫹ ⫽ 25–50%; ⫹⫹⫹ ⫽ 50–75%; ⫹⫹⫹⫹ ⫽ 75–100%
Age at
biopsy
Patient
Table 1 Clinical, morphological and genetic features of patients with mitochondrial disorders
Apoptosis in mitochondrial disorders
95
96
M. Mirabella et al.
Detection Kit, Boehringer, Mannheim, Germany). Negative
experimental controls were incubated with label solution
without terminal transferase instead of TUNEL reaction
mixture, while a positive control was set up by preincubating
muscle sections with DNase I for 10 min before the TUNEL
procedure. TUNEL-positive nuclei were counted in at least
100 muscle fibres per section and biopsies were classified in
six groups depending upon the percentage of muscle fibres
harbouring apoptotic nuclei (0–3%, 3–5%, 5–25%, 25–50%,
50–75%, 75–100%). Moreover, the number of TUNELpositive nuclei was correlated with the age of the patient at
the time of biopsy, the presence of myopathy, the severity of
the phenotype, the biochemical defect and the type and
percentage of mtDNA mutation (Table 1). In order to evaluate
the relative number of TUNEL-positive nuclei within
individual fibres and their correlation with COX-negative
fibres, TUNEL was also performed on muscle sections after
histochemistry for COX and was followed by nuclear staining
with Hoechst 33258. Unfixed 10 µm muscle sections adjacent
to those analysed by TUNEL were processed for
immunocytochemistry as follows. Sections were dried at
room temperature, fixed in cold acetone and pretreated with
0.3% H2O2 in PBS (phosphate-buffered saline) to quench
endogenous peroxidase activity, rinsed in PBS and incubated
with 10% normal serum (goat or rabbit depending on the
secondary antibodies used) for 60 min to mask non-specific
adsorption sites. Sections were then incubated for 1 h at
room temperature with one of the following antibodies: mouse
monoclonal antibody against human Fas/APO-1 (Calbiochem,
Cambridge, Mass., USA, diluted 1 : 50); rabbit polyclonal
antibody against human Fas (C-20, Santa Cruz, Calif., USA,
diluted 1 : 200) and p75-NTR (Promega, Madison, Wis.,
USA, diluted 1 : 100); and goat polyclonal anti-human
caspase-3 antibody (Santa Cruz, diluted 1 : 200). In control
experiments the primary antibodies were omitted or replaced
by preimmune sera. After several rinses in PBS, the sections
were incubated with the appropriate biotinylated secondary
antibodies (goat anti-mouse, goat anti-rabbit or rabbit antigoat IgG), washed in PBS and then incubated with the avidin–
biotin peroxidase complex according to the manufacturer’s
instructions (Vectastain ABC; Vector Laboratories,
Burlingame, Calif., USA). Peroxidase staining was obtained
by incubating the sections in 0.075% DAB (3,3diaminobenzidine) and 0.002% H2O2 in 50 mM Tris buffer
(pH 7.6) for 10 min.
Electron microscopy
For electron microscopy, muscle samples were fixed with
2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4),
postfixed in 1% osmium tetraoxide, dehydrated in a graded
ethanol series and embedded in Epon 812. Ultrathin sections
were cut and stained with uranyl acetate and lead citrate, and
examined with a Philips 208S electron microscope.
Results
TUNEL labelling showed that myonuclei containing doublestranded DNA fragments in situ, indicative of apoptosis,
were present in variable amount in the muscle biopsies of
patients with mtDNA point mutations and deletions (Table 1
and Fig. 1). Apoptotic nuclei were commonly observed in
fibres with a non-pathological appearance; the few TUNELpositive nuclei located in the rare muscle fibres exhibiting
histological changes characteristic of necrosis were not
included in the counts. No TUNEL-positive myonuclei were
observed in negative experimental controls, and only rare,
isolated positive nuclei (⬍2%) were present in normal muscle.
The presence of TUNEL-positive myonuclei significantly
correlated with increased p75, Fas and caspase-3
immunoreactivity in all biopsies studied (Table 1). Immunoreactivities for p75 (Fig. 2A) and Fas (Fig. 3) were localized
mainly in the plasmalemma, while intense caspase-3
immunostaining was observed in the cytoplasm (Fig. 2C–F).
In addition to the surface membrane staining, diffuse or
granular cytoplasmic immunopositivity for both p75 and Fas
was observed within myofibres (Figs 2A and 3). In all
biopsies, including normal muscles, p75 and Fas
immunoreactivities were detected in intramuscular nerve
twigs (Fig. 2B) and blood vessels, respectively. In control
experiments, when primary antibodies were omitted or
replaced by non-immune sera, the immunoreaction did not
take place. In the 10 control biopsies we did not find any
significant apoptotic features. In four inclusion body myositis
patients with marked mitochondrial abnormalities, there were
no TUNEL-positive myonuclei, while Fas was expressed on
the surface membrane of a variable number (5–25%) of
muscle fibres and inflammatory cells (not shown).
Patients carrying a single mtDNA deletion with ⬍40% of
mutated mtDNA molecules had only sporadic muscle fibres
with apoptotic traits, and all these patients showed a mild
phenotype (patients 17–23). When the percentage of mutated
mtDNA exceeded 40%, apoptotic features were present in a
significant number of fibres. TUNEL-positive nuclei, Fas,
p75 and caspase-3 were expressed in 25–50% of fibres which
had 40–62% of mutated mtDNA molecules (patients 12–
14,16, 24–26); in one patient with 65% mutated genomes,
apoptotic nuclei were expressed in 50–75% of fibres (patient
27) (Figs 1B, 2A and D and 3C); in another patient with
78% mutated genomes and a very severe phenotype, virtually
all fibres appeared apoptotic (patient 15) (Figs 1A, 2C and
3A). One patient with MNGIE carrying a 1 bp deletion also
showed 50–75% TUNEL-positive fibres and 25–50% of
fibres positive for Fas, p75 and caspase-3 (patient 32).
Four patients with PEO syndrome and multiple mtDNA
deletions, with mild neurological impairment and myopathy,
showed TUNEL-positive nuclei and Fas, p75 and caspase-3
expression in 3–5% of fibres (patients 28–31).
tRNA point mutations causing mainly MERRF, MELAS
and PEO syndromes, with a moderate to severe neurological
phenotype, were associated with a major degree of apoptosis
Apoptosis in mitochondrial disorders
97
Fig 1 In situ labelling of nuclear DNA fragmentation (TUNEL). (A) Almost all muscle fibres exhibited TUNEL-positive nuclei in a
2-month-old patient with Pearson’s syndrome (patient 15). (B) TUNEL-positive nuclei were present in 75% of the muscle fibres of a
patient carrying a single mtDNA deletion with 65% of genomes mutated (patient 27). (C) More than 25% of myofibres contained
TUNEL-positive nuclei in a patient with 80% of the A3243G point mutation in tRNALeu(UUR) (patient 7). (D) Only a few isolated
TUNEL-positive myonuclei, similar to those seen in normal controls, were present in the muscle biopsy of a patient with NARP and
97% of the T8993C point mutation in the ATPase 6 gene (patient 35). Original magnification: A, ⫻500; B and D, ⫻250; C, ⫻125.
in the biopsies when the percentage of mutated genomes was
⬎73% (patients 1–7 and 9–11) (Figs 1C and 2E and F). The
most severely affected patient with MERRF, who carried
73% of mutated mtDNAs (patient 1) (Fig. 3D) had ⬎50%
of muscle fibres displaying TUNEL, FAS, p75 and caspase3 positivity. In one patient with mild PEO and 54% mutated
mtDNA, there were no clear signs of apoptosis; apoptotic
nuclei were rare, immunostaining for caspase-3 was negative
and ⬍25% of fibres expressed Fas and p75 immunoreactivities (patient 8).
T8993C and G11778A point mutations, involving the
structural genes for ATPase 6 and ND4, respectively, causing
NARP and LHON syndromes, were not associated with
significant apoptotic signs, even if expressed in high amounts
(97–100%); only rare apoptotic nuclei were present, and Fas
and p75 positivity was seen in a minority of fibres (patient
34) (Fig. 1D).
Major apoptotic features correlated with severe dysfunction
of the respiratory chain, mainly affecting complexes I and
IV, but they were present even in patients with a mild
biochemical impairment if the muscle fibres carried a high
percentage of mutated mtDNA (Table 1).
By performing TUNEL, COX histochemistry and nuclear
staining with Hoechst 33258 (Fig. 4) on the same sections,
either single or multiple TUNEL-positive nuclei were
observed within individual muscle fibres on a given section,
but all nuclei visible with Hoechst 33258 were TUNELlabelled only in rare COX-negative areas of longitudinal
fibres (Fig. 4A and B). The majority of TUNEL-positive
nuclei were present within COX-negative fibres, especially
98
M. Mirabella et al.
Fig. 2 p75 and caspase-3 immunocytochemistry. (A) p75 immunoreactivity was detected on the
surface membrane and in the form of a granular cytoplasmic staining in ~50% of the muscle fibres
of patient 27. (B) p75 immunoreactivity was localized exclusively around intramuscular nerve twigs
in a patient with PEO and 10% of single mtDNA deletion (patient 22). Cytoplasmic caspase-3
immunoreactivity was detected in almost 100% of the muscle fibres of patient 15 (C) and in
⬎25% of fibres of patient 27 (D). (E and F) High-power photomicrographs illustrating the strong
cytoplasmic caspase-3 immunostaining in the form of multiple dots in patients 10 (E) and 6 (F),
both of whom carried the A3243G point mutation. Original magnification: A, B and D, ⫻125; C
and E, ⫻500; F, ⫻250.
Apoptosis in mitochondrial disorders
99
Fig. 3 Fas immunocytochemistry. (A) Fas immunoreactivity was present both over the surface membrane and within the cytoplasm of
50–75% of muscle fibres in patient 15. (B, C and D) Fas-immunoreactive muscle fibres (25–50%) were observed in patient 7, who had
the A3243G point mutation (B), in patient 27 (C) and in a patient with the A8344G point mutation in the tRNALys gene (patient 1) (D).
Original magnification: A, ⫻500; B and D ⫻250; C, ⫻125.
in patients with a high percentage of single mtDNA
deletions or tRNA point mutations. However, TUNELpositive nuclei were also seen in COX-positive fibres
(Fig. 4C and D) and RRF did not express apoptotic signs
at a significantly higher rate than non-RRF. Moreover, in
spite of numerous COX-negative fibres, the rare apoptotic
nuclei observed in patients with multiple mtDNA deletions
were not preferentially located in the COX-negative fibres
(Fig. 4E and F).
Sex and age at the time of muscle biopsy did not appear
to influence the apoptotic phenotype.
In TUNEL-positive muscle biopsies, electron microscopy
showed, in addition to the abnormalities of number and
structure of mitochondria, myofibres containing nuclei with
highly condensed chromatin (Fig. 5). There were no signs
of necrosis in the muscle fibres with abnormal nuclei.
Interestingly, there was no constant correlation between
mitochondrial abnormalities and myonuclei with morphological changes indicative of apoptosis. In fact, myonuclei
with irregular shape and condensed chromatin were present
either within muscle fibres with an otherwise normal
morphology or in association with an increased number of
dense mitochondria.
Discussion
In muscle biopsies from a large group of patients with
mitochondrial diseases and heterogeneous mtDNA defects
we demonstrated the significant presence of apoptotic features
100
M. Mirabella et al.
Fig. 4 TUNEL, COX histochemistry and Hoechst 33258 staining on the same sections. (B, D and F are double-exposed photographs).
(A and B) Most TUNEL-positive nuclei were present within COX-negative fibres (patient 13, with 62% of single mtDNA deletions and
severe biochemical COX deficiency). Single or multiple TUNEL-positive nuclei were observed within individual muscle fibres on a
given section, and some COX-negative areas in longitudinal fibres showed TUNEL labelling in all nuclei visible with Hoechst 33258
(fluorescent blue). (C and D) A few TUNEL-positive nuclei (arrowheads) were also seen in COX-positive fibres or in fibres with only
mild reduction of COX staining. Note that some COX-negative fibres did not contain TUNEL-positive nuclei. (E and F) In spite of
numerous COX-negative fibres, the rare apoptotic nuclei observed in patient 28 with multiple mtDNA deletions were not preferentially
located in the COX-negative fibres. Original magnification: ⫻250.
associated with specific phenotypes and a high percentage of
mutated mitochondrial genomes.
Mitochondria are the main cellular source of ATP, the
primary generators of reactive oxygen species and an
important storage site for calcium homeostasis. In recent
years several lines of evidence have indicated a critical role
Apoptosis in mitochondrial disorders
101
Fig. 5 Electron micrograph of a muscle fibre displaying a nucleus with condensed chromatin. Original
magnification: ⫻16 000 (bar ⫽ 300 nm).
for mitochondria and mitochondrial proteins in the control
of the apoptotic process: (i) the ∆Ψ disruption precedes by
far the typical nuclear signs of apoptosis—the condensation
and margination of chromatin and the cleavage and
fragmentation of genomic DNA (Zamzami et al., 1956;
Marchetti et al., 1996; Kroemer et al., 1997); (ii) mitochondria
release proteins, such as cytochrome c and AIF, that activate
proteases responsible for DNA fragmentation (Liu et al.,
1996; Kluck et al., 1997; Higuchi et al., 1997; Yang et al.,
1997); (iii) the uncoupling of oxidative phosphorylation and
the respiratory chain inhibitors induce apoptosis in vitro
(Wolvetang et al., 1994; Marton et al., 1997); (iv) excessive
free radical production (reactive oxygen species can alter the
external mitochondrial membrane and facilitate the release
of cytochrome c and AIF) increases apoptosis (Richter et al.,
1995; Slater et al., 1995; Yoneda et al., 1995); (v) the inner
and external membranes of mitochondria host a set of
proteins belonging to the Bcl-2 superfamily that are important
regulators of the apoptotic machinery (Yang et al., 1997;
Kluck et al., 1997; Narita et al., 1998).
In mitochondrial myopathies, apoptosis may represent
a selective mechanism for the elimination of cells with
dysfunctional mitochondria and excessive free radical
production or, on the contrary, may produce or amplify
cell damage.
The impact of nuclear apoptosis on the functionality and
viability of single cells is more difficult to evaluate in
multinucleated and partly regenerating tissues such as muscle
than in mononucleated, highly regenerating tissues. In fact,
DNA fragmentation within isolated nuclei in syncytial muscle
cells does not imply apoptosis of the entire fibre cell, and
the time course of the apoptotic changes may differ from the
classical apoptosis observed in other cell types, such as
lymphocytes, thymocytes, fibroblasts and tumour cells.
However, it is conceivable that individual myofibres may be
severely affected when apoptotic myonuclei exceed a certain
critical number. Unlike muscular dystrophies and
inflammatory myopathies, the occurrence of muscle necrosis
and regeneration in mitochondrial myopathies is quite rare.
This strengthens the significance of apoptosis as a pathogenic
mechanism of tissue damage in mitochondrial disorders, and
the focal distribution of TUNEL-positive nuclei may explain
the long survival of the affected fibres.
A variable percentage of TUNEL-positive fibres has been
demonstrated previously by us and others in muscle of
patients with mitochondrial myopathies (Mirabella et al.,
102
M. Mirabella et al.
1998; Monici et al., 1998). However, TUNEL positivity per
se does not always indicate apoptosis. The expression on
muscle cells of other markers of the apoptotic machinery,
such as p75, Fas and caspase-3, further supports the specificity
of the findings obtained with TUNEL. The overexpression
of Fas and caspase-3 observed in TUNEL-positive fibres of
patients with specific mtDNA defects suggests that the
apoptotic pathway starting with Fas and ending with
downstream activation of caspases may be a real model of
cell death in some human mitochondrial diseases.
In fact, in patients carrying single mtDNA deletions or point
mutations in tRNA genes (tRNALys, tRNALeu(UUR), tRNAIle,
tRNATrp) the degree of apoptosis in muscle matched the
number of mutated genomes and the severity of both
mitochondrial myopathy and the neurological phenotype.
MERRF-8344 (tRNALys), MELAS-3243 (tRNALeu(UUR)) and
disorders associated with single mtDNA deletions account
for the great majority of mitochondrial encephalomyopathies.
The absence or reduction of the translation of all mtDNAencoded mRNAs is the consequence of mtDNA defects
involving at least one tRNA (Schon, 1997). By contrast, we
found only modest or no signs of apoptosis in mitochondrial
diseases associated with point mutations in structural genes,
such as NARP and LHON, in spite of the presence of an
extremely high percentage of mutated genomes in muscle
(97–100%). In these disorders, mtDNA mutations do not
affect overall mitochondrial protein synthesis but selectively
involve a single subunit of a specific enzyme complex—
ATPase in NARP and complex I in LHON.
This evidence suggests that only mtDNA abnormalities that
impair mitochondrial protein synthesis can induce apoptosis
when the percentage of mutated mtDNAs exceeds a threshold:
~40% in the case of single deletions and 70% in the case of
tRNA point mutations. Our in vivo data are in agreement
with a previous in vitro demonstration of the increased
susceptibility to apoptosis, mediated by overexpression of
Fas, of cybrids with a high percentage of mtDNAs carrying
a single deletion or a point mutation in the tRNAIle gene
(Asoh et al., 1996).
The paucity of apoptotic features in myopathies associated
with multiple mtDNA deletions is puzzling. Multiple
deletions, in fact, also impair mitochondrial protein synthesis
and are detected in high amounts in muscle. The formation
of multiple deletions of mtDNA in mitochondrial disorders
that are inherited in a Mendelian manner is still unclear, and
is probably due to a defective nuclear gene that appears to
increase the frequency of mtDNA rearrangements (Schon,
1997). It is intriguing to hypothesize that the faulty
communication between nuclear and mitochondrial genomes
in these disorders may also induce intracellular proteins to
inhibit apoptosis that would otherwise be activated by the
presence of multiple mtDNA deletions.
Low percentages of multiple deletions and COX-negative
fibres in muscle are often observed as accompanying
phenomena in inclusion body myositis (Griggs et al., 1995),
and have also been considered to be an age-dependent
manifestation in normal old individuals. We did not find
any apoptotic features in four patients with inclusion body
myositis who had a level of mitochondrial abnormalities
similar to or higher than that in the experimental group.
Furthermore, there was no correlation between apoptosis and
age at biopsy in all series of patients and controls.
Apoptosis is an active process that requires energy for its
accomplishment, but a reduction in the level of ATP makes
cells more susceptible to programmed cell death (Richter
et al., 1996; Leist et al., 1997). There is, in fact, evidence
that the induction of selective respiratory chain deficiencies
makes cells more vulnerable to Fas-mediated apoptosis (Asoh
et al., 1996) and that respiratory chain inhibitors induce
apoptosis in vitro (Wolvetang et al., 1994; Marton et al.,
1997; Higuchi et al., 1998).
In mitochondrial disorders associated with deletions or
tRNA mutations there is partial, multiple involvement of
all complexes of respiratory chain bearing mitochondrial
subunits, although this involvement is sometimes so slight
as to be hardly detectable by biochemical assays in vitro
(Table 1). Our data demonstrate that, with few exceptions,
in all patients with respiratory chain dysfunction, mainly
involving complexes I and IV, muscle fibres are more prone
to undergo apoptosis. However, even in the absence of a
clear biochemical deficiency, apoptotic features are present
when the percentage of mutated mtDNA exceeds the
threshold.
In positive patients, although a rough correlation between
the severity of mitochondrial myopathy and apoptosis could
be established, RRF did not show more apoptotic features
than non-RRF. However, especially in patients with a high
percentage of single mtDNA deletions or tRNA point
mutations and severe biochemical impairment, apoptotic
nuclei were mainly located in COX-negative fibres.
Nevertheless, apoptotic nuclei were scarce in multiple
mtDNA deletions in spite of the presence of numerous COXnegative fibres. These findings, along with the negative results
obtained in inclusion body myositis, suggest that apoptosis
is not simply a way to eliminate cells that accumulate
abnormal mitochondria, but is specifically induced in muscle
fibres of patients with certain mtDNA abnormalities.
It is possible, in fact, that in mtDNA mutations affecting
protein synthesis the imbalance and relative abundance of
nuclear-encoded versus mtDNA-encoded subunits favour
cytochrome c inactivation and release. Cytochrome c is a
soluble protein loosely attached to the inner mitochondrial
membrane and is an essential component of the mitochondrial
respiratory chain. It has been recognized to play an important
role in apoptosis signalling (Krippner et al., 1996; Liu et al.,
1996) and has been demonstrated to be able to induce
apoptosis-activating caspase-3 directly (Yang et al., 1997;
Zou et al., 1997; Scaffidi et al., 1998; Zhivotovsky et al.,
1998), independently of mitochondrial transmembrane
depolarization (Liu et al., 1996; Yang et al., 1997, BossyWetzel et al., 1998).
Cytochrome c, reactive oxygen species and dysfunction of
Apoptosis in mitochondrial disorders
the respiratory chain may all contribute to apoptosis, which
in turn further inhibits the maturation of mitochondrial
protein precursors encoded in the nucleus and the rate of
mitochondrial translation (Mignotte et al., 1990; Vayssière
et al., 1994; Mitsui et al., 1996), thus amplifying the
biochemical defect, the production of reactive oxygen species
and mitochondrial damage. This vicious circle is maintained
and amplified in postmitotic non-regenerating tissues. Thus,
apoptosis, which may represent a cure for mitochondrial
disorders (Wolvetang et al., 1994; Asoh et al., 1996) through
the specific killing of cells that accumulate mutant mtDNA,
appears instead to be deleterious in perennial tissues, such
as muscle, brain and heart. In agreement with this view is
the example of Pearson’s syndrome. This disorder, associated
with a single deletion of mtDNA, is characterized by
refractory sideroblastic anaemia and the vacuolization of
bone marrow precursors. Death occurs in early childhood
and the few surviving patients later develop Kearns-Sayre
syndrome. The clinical improvement of blood dyscrasia in
these patients is probably due to the elimination of affected
blood cell precursors that carry a high percentage of mtDNA
deletions; however, some remain to accumulate in muscle,
the heart and the brain, giving rise to the Kearns-Sayre
syndrome (Di Mauro and Bonilla, 1997). We demonstrated
that virtually all nuclei were apoptotic in muscle from a
patient with Pearson’s syndrome. This suggests that apoptosis
could be the mechanism by which dysfunctional, highly
proliferative blood cell precursors are eliminated, unlike
muscle and postmitotic cells, which cannot be easily removed
and replaced.
In mitochondrial encephalomyopathies associated with
mtDNA mutations, selected areas of a given tissue or specific
subsets of cells are more prone to be affected clinically, even
in the presence of the same number of mutant genomes.
Apoptosis may be triggered by a variety of cell death signals,
and there are multiple paths of commitment to death and
different susceptibilities of diverse cells to the various stimuli
(Hetts, 1998; Peter and Krammer, 1998; Scaffidi et al., 1998).
Thus, apoptosis may help to explain the uneven tissue
involvement of mitochondrial disorders.
In conclusion, apoptosis seems to represent a potential
pathogenic mechanism of muscle tissue damage and to
play a role in modulating the clinical expression of some
mitochondrial disorders. Further studies are needed to clarify
the sequence of the different events and to establish whether
a single or several apoptotic pathways contribute to the
heterogeneity of phenotypes even among patients carrying
the same mtDNA mutation.
Acknowledgements
103
Bhagwati S, Leung B, Shafiq SA, Ghatpande A. Myotonic dystrophy:
decreased levels of myotonin protein kinase (Mt-PK) lead to
apoptosis in muscle cells. Exp Neurol 1997; 146: 277–81.
Bossy-Wetzel E, Newmeyer DD, Green DR. Mitochondrial
cytochrome c release apoptosis occurs upstream of DEVD-specific
caspase activation and independently of mitochondrial
transmembrane depolarization. EMBO J 1998; 17: 37–49.
DiMauro S, Bonilla E. Mitochondrial encephalomyopathies. In:
Rosenberg RN, Prusiner SB, DiMauro S, Barchi RL, editors. The
molecular and genetic basis of neurological disease. 2nd ed. Boston:
Butterworth-Heinemann; 1997. p. 201–35.
Di Mauro S, Schon EA. Mitochondrial DNA and diseases of the
nervous system: the spectrum. Neuroscientist 1998; 4: 53–63.
Fidzianska A, Goebel HH, Warlo I. Acute infantile spinal muscular
atrophy. Muscle apoptosis as a proposed pathogenetic mechanism.
Brain 1990; 113: 433–45.
Griggs RC, Askanas V, DiMauro S, Engel A, Karpati G, Mendell
JR, et al. Inclusion body myositis and myopathies. [Review]. Ann
Neurol 1995; 38: 705–13.
Hetts SW. To die or not to die. An overview of apoptosis and its
role in disease. [Review]. JAMA 1998; 279: 300–7.
Higuchi M, Aggarwal BB, Yeh ET. Activation of CPP32-like
protease in tumor necrosis factor-induced apoptosis is dependent
on mitochondrial function. J Clin Invest 1997; 99: 1751–8.
Higuchi M, Proske RJ, Yeh ET. Inhibition of mitochondrial
respiratory chain complex I by TNF results in cytochrome c release,
membrane permeability transition, and apoptosis. Oncogene 1998;
17: 2515–24.
Kluck RM, Bossy-Wetzel E, Green DR, Newmeyer DD. The release
of cytochrome c from mitochondria: a primary site for Bcl-2
regulation of apoptosis. Science 1997; 275: 1132–6.
Krippner A, Matsuno-Yagi A, Gottlieb RA, Babior BM. Loss of
function of cytochrome c in Jurkat cells undergoing Fas-mediated
apoptosis. J Biol Chem 1996; 271: 21629–36.
Kroemer G, Zamzami N, Susin SA. Mitochondrial control of
apoptosis. [Review]. Immunol Today 1997; 18: 44–51.
Leist M, Single B, Castoldi AF, Kühnle S, Nicotera P. Intracellular
adenosine triphosphate (ATP) concentration: a switch in the decision
between apoptosis and necrosis. J Exp Med 1997; 185: 1481–6.
Liu X, Kim CN, Yang J, Jemmerson R, Wang X. Induction of
apoptotic program in cell-free extracts: requirement for dATP and
cytochrome c. Cell 1996; 86: 147–57.
Marchetti P, Castedo M, Susin SA, Zamzami N, Hirsch T, Macho
A, et al. Mitochondrial permeability transition is a central
coordinating event of apoptosis. J Exp Med 1996; 184: 1155–1160.
This work was supported by Telethon-Italy, grant 1121 and
MURST ex 40%.
Marton A, Mihalik R, Bratincsák A, Adleff V, Petak I, Vegh
M, et al. Apoptotic cell death induced by inhibitors of energy
conservation—Bcl-2 inhibits apoptosis downstream of a fall of ATP
level. Eur J Biochem 1997; 250: 467–75.
References
Asoh S, Mori T, Hayashi J-I, Ohta S. Expression of the apoptosismediator Fas is enhanced by dysfunctional mitochondria. J Biochem
(Tokyo) 1996; 120: 600–7.
Mignotte B, Larcher JC, Zheng DQ, Esnault C, Coulaud D, Feunteun
J. SV40 induced cellular immortalization: phenotypic changes
associated with the loss of proliferative capacity in a conditionally
immortalized cell line. Oncogene 1990; 5: 1529–33.
104
M. Mirabella et al.
Mirabella M, Engel WK, Passinetti G, Finch CE, Askanas V.
Denervation of adult human muscle fibers induces apoptosis,
evidenced by fragmentation of nuclear DNA, and increased
expression of the clusterin (ApoJ) gene [abstract]. Neurology 1996;
46 (2 Suppl): A270.
Mirabella M, Di Giovanni S, Silvestri G, Ricci E, Bertini E,
Tonali P, et al. Apoptosis and mitochondrial encephalomyopathies:
correlation with phenotype [abstract]. In: Proceedings AACR.
Molecular Mechanisms of Apoptosis Regulation; 1998 Jan 9–13.
Indian Wells (CA).
Mitsui T, Kawai H, Nagasawa M, Kunishige M, Akaike M, Kimura
Y, et al. Oxidative damage to skeletal muscle DNA from patients
with mitochondrial encephalomyopathies. J Neurol Sci 1996; 139:
111–6.
Monici MC, Toscano A, Girlanda P, Aguennouz M, Musumeci O,
Vita G. Apoptosis in metabolic myopathies. Neuroreport 1998; 9:
2431–5.
Morgan-Hughes JA. Mitochondrial diseases. In: Engel AG, FranziniArmstrong C, editors. Myology. 2nd ed. New York: McGraw-Hill;
1994. p. 1610–60.
Schon EA. The mitochondrial genome. In: Rosenberg RN, Prusiner
SB, DiMauro S, Barchi RL, editors. The molecular and genetic
bases of neurological disease. 2nd ed. Boston: ButterworthHeinemann; 1997. p. 189–200.
Slater AF, Nobel CS, Orrenius S. The role of intracellular oxidants
in apoptosis. [Review]. Biochim Biophys Acta 1995; 1271: 59–62.
Tews DS, Goebel HH. Apoptosis-related proteins in skeletal muscle
fibers of spinal muscular atrophy. J Neuropathol Exp Neurol 1997;
56: 150–6.
Vayssiere JL, Petit PX, Risler Y, Mignotte B. Commitment to
apoptosis is associated with changes in mitochondrial biogenesis
and activity in cell lines conditionally immortalized with simian
virus 40. Proc Natl Acad Sci USA 1994; 91: 11752–6.
Wallace DC. Diseases of the mitochondrial DNA. [Review]. Annu
Rev Biochem 1992; 61: 1175–212.
White E. Life, death, and the pursuit of apoptosis. [Review]. Genes
Dev 1996; 10: 1–15.
Wolvetang EJ, Johnson KL, Krauer K, Ralph SJ, Linnane AW.
Mitochondrial respiratory chain inhibitors induce apoptosis. FEBS
Lett 1994; 339: 40–4.
Narita M, Shimizu S, Ito T, Chittenden T, Lutz RT, Matsuda H,
et al. Bax interacts with the permeability transition pore to induce
permeability transition and cytochrome c release in isolated
mitochondria. Proc Natl Acad Sci USA 1998; 95: 14681–6.
Yang J, Liu X, Bhalla K, Kim CN, Ibrado AM, Cai J, et al.
Prevention of apoptosis by Bcl-2: release of cytochrome c from
mitochondria blocked. Science 1997; 275: 1129–32.
Peter EM, Krammer PH. Mechanisms of CD95 (Apo-1/Fas)mediated apoptosis. [Review]. Current Opin Immunol 1998; 10:
545–51.
Yoneda M, Katsumata K, Hayakawa M, Tanaka M, Ozawa T.
Oxygen stress induces an apoptotic cell death associated with
fragmentation of mitochondrial genome. Biochem Biophys Res
Commun 1995; 209: 723–9.
Richter C, Gogvadze V, Laffranchi R, Schlapbach R, Schweizer M,
Suter M, et al. Oxidants in mitochondria: from physiology to
diseases. [Review]. Biochim Biophys Acta 1995; 1271: 67–74.
Zamzami N, Susin SA, Marchetti P, Hirsch T, Gòmez-Monterrey I,
Castedo M, et al. Mitochondrial control of nuclear apoptosis. J Exp
Med 1996; 183: 1533–44.
Richter C, Schweizer M, Cossarizza A, Franceschi C. Hypothesis.
Control of apoptosis by the cellular ATP level. [Review]. FEBS
Lett 1996; 378: 107–10.
Zhivotovsky B, Orrenius S, Brustugun OT, Doskeland SO. Injected
cytochrome c induces apoptosis [letter]. Nature 1998; 391: 449–50.
Salvesen GS, Dixit VM. Caspases: intracellular signaling by
proteolysis. [Review]. Cell 1997; 91: 443–6.
Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human
protein homologous to C. elegans CED-4, participates in cytochrome
c-dependent activation of caspase-3. Cell 1997; 90: 405–13.
Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ
et al. Two COD95 (APO-1/Fas) signaling pathways. EMBO J 1998;
17: 1675–87.
Received June 25, 1999. Accepted July 26, 1999