ABSTRACT
Mitochondria are essential organelles in all eukaryotic cells where cellular ATP
is generated through the process of oxidative phosphorylation. Protein
components of the respiratory assembly are gene products of both mitochondrial
and nuclear genes. The mitochondrial genome itself encodes several protein and
nucleic acid components required for such oxidative phosphorylative processes,
but the vast majority of genes encoding respiratory chain components are
nuclear. Similarly, the processes of replication and transcription of
mitochondrial DNA rely exclusively upon RNA and protein species encoded by
nuclear genes. We have analyzed two key nuclear-encoded proteins involved in mitochondrial DNA replication and
transcription as a function of the presence or absence of mitochondrial DNA.
Mitochondrial DNA polymerase (DNA polymerase
[gamma]
), the nuclear-encoded enzyme which synthesizes mtDNA, is expressed and translated in
cells devoid of mitochondrial DNA itself. In contrast, mitochondrial
transcription factor A protein levels are tightly linked to the mtDNA status of
the cell. These results demonstrate that the DNA polymerase
[gamma]
protein is stable in the absence of mitochondrial DNA, and that there appears
to be no regulatory mechanism present in these cells to alter levels of this
protein in the complete absence of mitochondrial DNA. Alternatively, it is
possible that this enzyme plays an additional, as yet undefined, role in the
cell, thereby mandating its continued production.
Energy-generating mitochondria contain their own double-stranded circular genome, mitochondrial DNA (mtDNA). In mammals,
mtDNA constitutes ~1% of total cellular DNA. MtDNA copy number is maintained remarkably
constant, ranging from 1 to 10 000 copies per cell, a signature of the
particular cell or tissue type (
1
). Replication and transcription of mtDNA requires the participation of several
nuclear-encoded enzymes, some of which consist of both protein and nucleic acid
components (
2
). Copy number control then must depend upon careful orchestration of processes
which regulate mitochondrial nucleic acid metabolism, soliciting the
cooperation of both nuclei and mitochondria.
Several notable features distinguish the replication of mtDNA and chromosomal
DNA. In contrast with chromosomal (nuclear) DNA, mtDNA synthesis does not
appear to be restricted to any particular phase of the cell cycle (
3
). Moreover, mtDNA may be replicated more than once or not at all within a given
cell generation time (
4
). Replication of chromosomal DNA is heavily monitored and corrected through
elaborate DNA repair mechanisms, whereas limited DNA repair occurs within the
mitochondrion (
2
). In keeping with the different requirements for synthesis of mtDNA,
replication and transcription of the organellar genome is thought to occur
within the mitochondrial matrix through the activity of the mitochondrial-specific replication and transcription machineries. Both mtDNA polymerase
(mtDNA polymerase [gamma]) and mitochondrial RNA (mtRNA) polymerase are encoded by nuclear genes
in vertebrates (Ropp and Copeland, submitted; Garman and Clayton, unpublished
observations), as are all known transcription factors which interact with DNA
sequences present in nuclear or mitochondrial genes involved in either
oxidative phosphorylation or mtDNA replication and transcription processes (
5
-
7
).
Mammalian cells in culture can be completely and irreversibly depleted of mtDNA
by long-term exposure to the intercalating agent ethidium bromide (EB) (
8
). Alternatively, short-term exposure to very low concentrations of EB leads to reversible
inhibition of mtDNA synthesis, and consequent depletion of mtDNA (
9
). Cells devoid of mtDNA ([rho]
o
cells) can be readily maintained in culture, given the presence of exogenous
pyruvate and uridine (
8
). [rho]
o
cells exhibit uridine auxotrophy due to a deficiency in respiratory chain-dependent dihydroorotate dehydrogenase (
10
) and rely exclusively on glycolysis for energy. In the presence of pyruvate and
uridine, however, the growth rate of [rho]
o
cells approaches that of the parental cell line from which they were derived (
11
). [rho]
o
cells are incapable of producing energy through oxidative phosphorylation due
to the absence of all mitochondrially-encoded proteins, which together represent a relatively small percentage
of all proteins present in the inner membrane enzyme complexes (
12
). Apparently the remaining contingent of nuclear-encoded respiratory subunits is sufficient to maintain a nearly normal
membrane potential (
13
).
Previous studies in the yeast
Saccharomyces cerevisiae
have employed [rho]
o
cells as a model system to investigate potential compensatory responses
elicited by nuclear genes in the absence of mtDNA (
14
-
17
). Results from such studies suggest that, at least in single-cell eukaryotes, there may exist some forms of genetic communication
between the mitochondrial and nuclear organelles. While [rho]
o
strains of
S.cerevisiae
have been used extensively over the years to study factors affecting mtDNA
metabolic processes, significantly less is known concerning mammalian [rho]
o
cells. A few recent studies, however, have employed subtractive hybridization
strategies to investigate molecular differences between [rho]
+
and [rho]
o
cells (
13
,
18
).
In an effort to understand further interorganellar communication between nuclei
and mitochondria, we sought to analyze certain nuclear-encoded proteins involved in maintenance of mtDNA copy number in either
the presence or absence of mtDNA. Our results suggest that the presence, and
presumed activity, of proteins which play a role in the replication and
transcription of mtDNA appear not to be universally dependent on the mtDNA
status of the cell.
143B (TK-) human osteosarcoma cells (ATCC CRL 8303, kindly provided by Dr Eric A.
Shoubridge) were maintained in a humidified 5% CO
2
atmosphere in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 5%
fetal bovine serum (FBS) (Gibco-BRL, Bethesda, MD), 1.25 [mu]g/ml gentamycin (Sigma, St Louis, MO), 50 [mu]g/ml uridine and 100 [mu]g/ml (333 [mu]M) BrdU (both from US Biochemicals, Cleveland, OH). 143B
(TK-) cells containing ([rho]
+
) or lacking mtDNA ([rho]
o
) were cultured under identical conditions. Two independent [rho]
o
clones (#4 and #14) were used and gave equivalent results in all experiments.
HeLa human cervical carcinoma cells (ATCC CCL 2) were grown as monolayer
cultures in DMEM (Gibco-BRL, Bethesda, MD) containing high glucose, L-glutamine and 110 mg/l sodium pyruvate, and supplemented with 5% FBS
and 1.25 [mu]g/ml gentamycin. For ethidium bromide (EB) treatments, HeLa cells were first
cultured onto glass coverslips and then treated for 3 days with 20 or 100 ng/ml
EB (Sigma), diluted into DMEM from a 1 [mu]g/ml stock solution.
Whole-cell extracts from 143B (TK-) cells were prepared as previously described (
19
). Briefly, ~10
7
cells were centrifuged and washed with phosphate buffered saline (PBS). Cells
were then washed twice in 10 mM HEPES, pH 7.9, 1.5 mM MgCl
2
, 10 mM KCl and 0.5 mM DTT, pelleted, and resuspended in 20 [mu]l of 20 mM HEPES, pH 7.9, 25% glycerol, 0.42 mM NaCl, 0.2 mM EDTA, 0.5 mM
DTT, 0.5 mM PMSF, 1 [mu]g/ml pepstatin A, 1 [mu]g/ml leupeptin, 1 [mu]g/ml aprotinin and 0.1% Nonidet P-40. The suspension was incubated on ice for 10 min, mixed
briefly, and pelleted at 14 000
g
. The lysed cell supernatant was diluted with 20-50 [mu]l buffer D, containing 20 mM HEPES, pH 7.9, 20% glycerol, 50 mM KCl,
0.2 mM EDTA, 0.5 mM DTT, 0.5 mM PMSF, 1 [mu]g/ml pepstatin A, 1 [mu]g/ml leupeptin and 1 [mu]g/ml aprotinin. Protein concentrations were determined by the method of
Bradford (
20
), and extracts were stored at -80oC. Extracts (10-20 [mu]g) were resolved by SDS-PAGE, using 10% linear polyacrylamide gels. Prior
to electrophoresis, proteins were denatured by boiling in SDS (0.5%) at 100oC for 5 min, and disulfide bonds in proteins were reduced by treatment with
[beta]-mercaptoethanol. Following electrophoresis, proteins were
transferred to a nitrocellulose membrane by electroblotting for 1 h at 100 V in
6.25 mM Tris-HCl, pH 7.5, 1.3% glycine and 20% methanol. The membrane was blocked at 4oC for 15 h in TBS-T (20 mM Tris-HCl, pH 7.5, 136 mM NaCl, 0.1% Tween-20), supplemented with 5% non-fat dried milk. After washing briefly with
TBS-T, the membrane was incubated at 25oC for 60 min with primary antibody [polyclonal anti-DNA polymerase [gamma], pre-immune serum (both diluted 1:500 in TBS], or
monoclonal anti-actin (1:2000 in TBS) (Amersham, Buckinghamshire, UK). After repeated
washes with TBS-T, the membrane was incubated at 25oC for 60 min with secondary antibody [<HRP> swine anti-rabbit IgG (DAKO, Carpinteria, CA) or <HRP> sheep anti-mouse IgG (Amersham)]. The membrane was then washed
repeatedly with TBS-T, and immunoreactive species were detected by autoradiography with
chemiluminescence, according to the manufacturer's instructions (Amersham UK,
Buckinghamshire, UK).
Monolayers of 143B (TK-) and HeLa cells were seeded onto autoclaved glass coverslips and
cultured in 60 mM plastic dishes (Becton Dickinson, Lincoln Park, NJ) in
supplemented DMEM. Prior to fixation, cells were labeled
in vivo
with the mitochondrial membrane potential-sensitive dye Mitotrackertm (Molecular Probes, Eugene, OR). Briefly, cells were incubated in
the presence of the dye (100 nM, diluted in DMEM) for 15 min at 37oC, followed by an identical incubation in DMEM, except in the absence of
dye. Labeling and all subsequent steps were performed in the absence of light.
Cells were then washed twice with PBS, pH 7.4. Cells were first fixed with 4%
paraformaldehyde [diluted from freshly prepared 20% paraformaldehyde (Sigma)]
for 10 min at 25oC, followed by fixation/dehydration with -20oC acetone. Cells were rehydrated for 3 min (25oC) to 20 h (4oC) in PBS. Both primary and secondary antibodies were
diluted in PBS/0.1% BSA at the following concentrations: anti-DNA polymerase [gamma] (1:250), pre-immune serum (DNA polymerase [gamma]) (1:250), anti-mtTFA (1:1000), <Cy5>anti-rabbit IgG (1:500). Primary antibody
incubations were performed for 60 min at 25oC, and secondary antibody incubations were perfomed for 30 min at 25oC. Cells were washed repeatedly between primary and secondary antibody
incubations as well as after the secondary antibody incubation period. Finally,
cells on coverslips were mounted onto glass slides with Cytoseal 60 mounting
medium (Stephens Scientific, Riverdale, NJ). Microscopy was performed using a
BioRad MRC 1000 laser-scanning confocal microscope, equipped with CoMOS imaging software.
Digital images were merged using Adobe Photoshop software (Adobe Systems Inc.,
Mountain View, CA).
Total RNA was isolated from 143B (TK-) cells ([rho]
o
and [rho]
+
) with TRIzol reagent (Life Technologies) as previously described (
21
). Non-isotopic ribonuclease protection assays were performed with the BrightStar
BIOTINscript, RPA II and BrightStar BioDetect kits (Ambion, Austin, TX). The
ribonuclease protection assay biotin-labeled probe was prepared from the human
DNA pol
[gamma] plasmid cDNA clone K12 (Ropp and Copeland, submitted) as follows. The
K12 clone was digested with
Kpn
I, and the 3.1 kb fragment was isolated. This fragment contains the pBluescript
vector and a 184 bp portion of the
DNA pol
[gamma] cDNA. The fragment was religated to generate DK12
184
BS. This was linearized with
Pvu
II, and a biotin-labeled RNA transcript was synthesized with T7 RNA polymerase in the
presence of biotin-14-CTP. Template DNA was removed by DNase I treatment, and the
transcript was gel-purified. For the ribonuclease protection, 20 [mu]g of total cell RNA from 143B (TK-) ([rho]
+
), 143B (TK-) ([rho]
o
, #4) and 143B (TK-) ([rho]
o
, #14) was mixed with 2 or 4 ng of the
DNA pol
[gamma] probe or 0.2 ng of a human [beta]-actin probe (total volume <15 [mu]l) and 20 [mu]l of solution A (80% formamide, 100 mM sodium
citrate, pH 6.4, 300 mM sodium acetate, pH 6.4, 1 mM EDTA) and incubated at 90oC for 4 min. The tubes were then incubated at 42oC for 48 h followed by the addition of 0.5 U RNase A and 20 U Rnase T1
to remove single-stranded RNA. The sample was incubated at 37oC for 30 min followed by precipitation of the DNA at -20oC and centrifugation at 14 000
g
for 15 min at 4oC. The pellet was resupended in 8 [mu]l gel loading buffer (95% formamide, 0.025% xylene cyanol, 0.025%
bromophenol blue, 0.5 mM EDTA, 0.025% SDS), heated at 90oC for 4 min, and separated through a 5% acrylamide-8 M urea gel at 250 V for 60 min. The RNA was transferred to a
BrightStar Plustm membrane by semi-dry electroblotting in 0.5* TBE at 200 mA for 60 min, and the RNA was fixed to the
membrane by UV-crosslinking. The membrane was washed three times (5 min each) in wash
buffer, twice (5 min each) in blocking buffer, once (30 min) in blocking
buffer, once (30 min) in conjugate solution (1:1000 dilution of streptavidin-alkaline phosphatase conjugate in blocking buffer), once (10 min) in
blocking buffer, three times (5 min each) in wash buffer, twice (5 min each) in
assay buffer and once (5 min) in CDP-Startm. The membrane was wrapped in plastic wrap and exposed to film for
30 min. The RNA protection assay was repeated three times, each yielding the
same results.
The mtDNA content of 143B (TK-) and 143B (TK-) [rho]
o
cell lines were determined by PCR of a portion of the
Leu tRNA
gene of the mitochondrial genome. At the same time, the trinucleotide repeat
region of the
DNA pol
[gamma] gene was also amplified as an internal genomic DNA control. Total DNA
was prepared from ~1 * 10
5
cells by treating with 50 [mu]g/ml proteinase K in 10 mM Tris-HCl, pH 7.5, 1 mM EDTA in a total volume of 50 [mu]l at 65oC for 2 h followed by incubation at 95oC for 15 min. Each reaction mixture contained in a
total volume of 10 [mu]l: 5 pmol each primer (5'mt: 5'-GAT GGC AGA GCC CCG GTA ATC GC and 3'mt: 5'-TAA GCA TTA GGA ATG CCA TTG CG for
the
mtLeu tRNA
gene and H42: 5'-CCC TCC GAG GAT AGC ACT TGC GGC and H43: 5'-AGC GAC GGG CAG CGG CGG CGG CA for the
DNA pol
[gamma]
gene), 0.2 mM each dNTP, 10 [mu]Ci [[alpha]-
32
P]dCTP (3000 Ci/mmol), 10 mM Tris-HCl, pH 8.3, 1.5 mM MgCl
2
, 50 mM KCl, 0.25 U
Taq
DNA polymerase and 1 [mu]l of the total DNA sample. The reaction mixtures were incubated at 95oC for 2 min followed by 35 cycles of 94oC for 0.5 min, 55oC for 0.5 min and 72oC for 0.5 min. To each reaction was added 10 [mu]l of stop solution (95% formamide, 20 mM EDTA, 0.05%
bromophenol blue, 0.05% xylene cyanol) and 3 [mu]l was loaded onto an 8% sequencing gel containing 40% formamide (
22
).
Oligonucleotide primers specific to the human
DNA pol
[gamma] cDNA were used to amplify the polymerase domain of the human gene (Ropp
and Copeland, submitted). The amplified products were subcloned into the
E.coli
expression vector pGEX2T (Pharmacia, Uppsala, Sweden) to express the DNA pol [gamma] peptide, G714 to L1061, as a glutathione-
S
-transferase fusion protein (CD-7). The CD-7 antigen was expressed and purified from
E.coli
DH5[alpha] as described (
23
). Anti-CD7 antibodies were purified over protein-A agarose as described (
24
).
In order to investigate potential intracellular communication links between
mitochondria and nuclei, we employed a previously characterized human cell line
[143B(TK-)] with ([rho]
+
) or without ([rho]
o
) mtDNA (
11
). 143B(TK-) [rho]
o
cells were isonuclear with the parental 143B (TK-) cell line since they were generated from the parental cells through
long-term exposure to EB as previously described (
8
), and exhibited characteristic uridine- and pyruvate-dependence for growth in culture (see Materials and Methods). To
confirm that long-term exposure of the 143B (TK-) cells to EB had completely depleted the cells of mtDNA, total
cellular DNA was extracted from the cells and probed simultaneously for both
mtDNA (mtLeu) and genomic DNA (DNA pol [gamma]) by PCR (Fig.
1
). The mtLeu gene was chosen as a probe because it lies in an area of infrequent
deletions and/or duplications (
25
), however repeat experiments with additional mtDNA probes produced identical
results (data not shown). MtDNA was only detectable in the parent 143B (TK-) cell line as well as in a control cell line (LS180), both of which are [rho]
+
(Fig.
1
, lanes 1 and 4). The two [rho]
o
isolates did not contain any detectable mtDNA even though approximately two to
three times more of these samples were loaded than the [rho]
+
sample, based on the intensity of the genomic signal (Fig.
1
, lanes 2 and 3). In addition, these cells exhibited a phenotype (pyruvate- and uridine-dependency) expected of [rho]
o
cells.
Mitochondria are semi-autonomous organelles containing a resident genome which encodes ribosomal
RNAs, transfer RNAs and proteins, all of which function within the organelle.
However, all mammalian mitochondrially-encoded proteins function as components of the respiratory chain assembly.
The processes of replication and transcription of the mitochondrial genome,
therefore, are accomplished entirely through the activities of nuclear-encoded enzymes. In this study, we have begun to analyze issues relating
to communication between the nucleus and the mitochondrion, two distinct but
interdependent organelles. [rho]
o
cells serve as a model system in which to investigate potential changes in
nuclear-encoded genes involved in mtDNA copy number control as a function of the
presence or absence of mtDNA. Our results demonstrate that cells which have
lost mtDNA either recently or chronically continue to express the
DNA pol
[gamma] gene which encodes the mtDNA polymerase [gamma] protein responsible for synthesizing mtDNA.
Previous studies in both yeast and mammalian cells have addressed the issue of
potential nuclear compensatory responses in [rho]
o
cells which are completely lacking in mtDNA (
13
-
16
,
18
). These works have demonstrated variable changes in the expression of genes
encoding documented regulators of mitochondrial function (
13
,
16
,
28
) or genes encoding proteins involved in cellular roles heretofore unrelated to
mitochondrial function (
13
-
16
,
18
). It remains unclear what the functional implications of these changes are
within the mitochondrion, since, at least in mammals, [rho]
o
mitochondria appear grossly normal and maintain a membrane potential.
We have shown that mtDNA polymerase [gamma] is expressed and translated in the absence of mtDNA, and hence, in the
absence of mtDNA replication. A paucity of nuclear genes encoding regulators of
mtDNA copy number have been cloned in mammalian cells (
6
; Ropp and Copeland, submitted; Garman and Clayton, unpublished observations).
Of these, mtTFA is a mitochondrially-imported transcription factor which is absolutely required (at least
in vitro
) for specific transcription initiation at mammalian mitochondrial promoters (
27
). Since in mitochondria, replication and transcription are functionally linked
(
2
), this protein is likely a major regulator of mtDNA copy number. Previous
studies have demonstrated that the stability and/or activity of mtTFA is
tightly associated to the presence of mtDNA within the cell, since cells
depleted of mtDNA are characterized by the absence of detectable levels of
mtTFA protein (
28
). Interestingly, however, the mtTFA gene is expressed at normal levels in [rho]
o
cells (
28
). Our data corroborate those results, since we observed that even with low dose
EB-mediated short-term depletion of mtDNA, mtTFA protein levels dropped severely. It
thus appears that the activity of mtTFA is regulated stringently at a
translational level.
In contrast, there appears to be no such control exerted over the mtDNA
polymerase [gamma] protein. Although the
DNA pol
[gamma] gene is also expressed at normal levels in [rho]
o
cells, levels of mtDNA polymerase [gamma] protein are indistinguishable from those in [rho]
+
cells. Furthermore, we observed that the subcellular distribution of this
enzyme in [rho]
o
cells closely resembles that of the parental [rho]
+
cell line. These results are consistent with at least two possibilities. First,
differences in protein levels between the two proteins may simply be an issue
of stability. MtTFA, a member of the HMG box class of proteins, avidly binds
DNA and may require such an interaction to maintain its stability within the
cell.
A second possibility is that mtDNA polymerase [gamma] may play some as yet undefined additional role. By virtue of double-label confocal microscopy, we have detected the mtDNA polymerase [gamma] protein only within mitochondria of the intact human cells
used in this study. Previous studies employing biochemical fractionation
techniques have identified nuclear (
29
-
31
) or extramitochondrial (
32
) forms of DNA polymerase [gamma]. However, it is presently unclear whether or not such proteins are
genetically distinct from the mtDNA polymerase [gamma], since those reports preceded the recent cloning of this gene. At the
moment, the concept that DNA polymerase [gamma] plays any other role in the cell other than mtDNA synthesis remains, at
best, speculative. However, we cannot rule out the possibility that the DNA
polymerase [gamma] protein might be maintained in [rho]
o
cells to assist in maintaining the structural integrity of the [rho]
o
mitochondrial membrane, although this appears unlikely since mtDNA polymerase [gamma] is not among the most abundant proteins within the organelle.
Nonetheless, it is clear from these results that there appears to be no
effective feedback mechanism to control the presence and presumed activity of
mtDNA polymerase [gamma] protein in the absence of mtDNA. Since [rho]
o
cells have been depleted of mtDNA for hundreds of generations, it would appear
extremely uneconomical for the cell to continue to produce and maintain a
protein which it does not use. It is interesting to note, however, that mtDNA
polymerase [gamma] has been detected in sea urchin sperm, in which mtDNA replication does
not occur (
33
). We do not know whether a similar scenario is operative in [rho]
o
cells of
S.cerevisiae
, since although the gene which encodes mtDNA polymerase [gamma] in the yeast
S.cerevisiae
(
MIP1
) has been cloned (
34
), analagous experiments to analyze for the presence of this protein in the
absence of mtDNA have not yet been reported. The reciprocal experiment,
expectedly, gave unequivocal results; chromosomal disruption of the
MIP1
gene resulted in complete loss of mtDNA in
S.cerevisiae
cells (
34
).
Our results do not implicate mtDNA polymerase [gamma] as a major regulator of mtDNA copy number, and in fact strengthen the
notion that other nuclear-encoded proteins such as mtTFA may play a significant role in that
capacity. Potential differences in [rho]
o
cells in the abundance, activity or stability of other auxiliary proteins
presumably involved in mtDNA replication such as mitochondrial single-stranded DNA-binding proteins (mtSSBs), primases and helicases, have not been
measured. With the exception of human SSB (
35
), none of these nuclear genes has been cloned in human cells. The very recent
cloning of DNA polymerase [gamma] from the yeast
Schizosaccharomyces pombe
(
36
) and various types of vertebrate cells (Ropp and Copeland, submitted) should
herald the beginning of the identification and characterization of some of
these other proteins which function in the mechanics of mammalian mtDNA
replication.
This work was supported by grants from the NIH (to W.C.C) and grant GM33088-25 from the National Institute of General Medical Sciences (to D.A.C.).
A.F.D. was supported by a postdoctoral fellowship from the Evelyn Neizer Fund
of Stanford University.
REFERENCES
Return