 |
INTRODUCTION |
The compartmentalization of mitochondrial DNA
(mtDNA)1 in distinct
organelles and the changes in the distribution of the mitochondrial genome in a cell brought about by mitochondrial growth, division, and
movement and, in some cells, by mitochondria fusion account for the
fact that the organization of these organelles plays a critical role in
determining the segregation and complementation behavior of wild-type
and mutant mtDNA, when they co-exist in a cell.
In simpler eukaryotic cells, like Saccharomyces
cerevisiae (1, 2), Chlamydomonas species (3) and slime
mold (4), and in plants (5, 6), there is ample evidence of mtDNA
complementation and recombination, which is made possible by the well
established capacity of mitochondria to undergo fusion in these
organisms. These fusion events are genetically regulated (2, 4, 7). In
animal cells, studies using time-lapse bright-field, phase-contrast, interference and fluorescence microscopy and electron microscopy of
serial sections have revealed in some cell types a mitochondrial organization in the form of a network of interconnected tubular organelles, with suggestion of fusion events (8-11). Recently, the
occurrence of a developmentally regulated fusion of mitochondria in
postmeiotic spermatids of Drosophila melanogaster, which is mediated by the transient synthesis of a novel GTPase encoded in the
fuzzy onions (fzo) gene, has been described (12).
Homologs of this gene have been found in mammals, nematodes, and
S. cerevisiae (12-14). However, mitochondria of mammalian
spermatids have been reported not to fuse but to form structurally
distinct contacts (15). Similar contacts have also been described in
rat cardiac tissue (16). A developmentally regulated formation of a
reticulum of tightly joined mitochondria has been reported for rat
diaphragm muscle (17).
Observations on cultured human cells have so far failed to provide an
unambiguous picture of the extent of mtDNA interactions in these cells.
In investigations carried out in this laboratory (18), two mtDNAs, each
carrying a recessive mutation in a different mitochondrial tRNA gene,
i.e. the tRNALys gene mutation associated with
the myclonic epilepsy and ragged red fiber (MERRF) encephalomyopathy
(19) and the tRNALeu(UUR) gene mutation associated with the
mitochondrial myopathy, encephalopathy, lactic acidosis, and
stroke-like episodes (MELAS) encephalomyopathy (20, 21), were
independently introduced into the same human mtDNA-less
(
0) cell (143B.0206) by enucleated cell
(cytoplast) × 0 cell fusion (22). After growing
the fusion products under conditions non-selective for respiratory
competence, no evidence of transcomplementation of the mutations was
found in five clones carrying both types of parental mtDNA, even 3 months after cell fusion. Other investigators, analyzing by
fluorescence microscopy the fusion products of enucleated wild-type
human cells and human 0 cells, obtained results that
were interpreted to indicate the occurrence of a rapid and extensive
fusion of host and exogenous mitochondria and subsequent rapid
diffusion of mtDNA and its transcripts throughout the organelles (23).
More recently, the same investigators (24, 25), after constructing
cybrids carrying two types of mtDNA with appropriate markers within
distinct organelles and culturing them under conditions either
selective or non-selective for recovery of respiratory capacity,
isolated a few clones that showed evidence of translational complementation.
In the present work, in order to obtain a more conclusive answer to the
question of the genetic autonomy of mitochondria in cultured human
cells, large scale fusion experiments were carried out between cells
and cytoplasts (or cells) carrying different recessive mtDNA mutations,
in particular either the tRNALys MERRF mutation (26) or a
frameshift mutation in the NADH dehydrogenase ND4 subunit gene (27).
The determination of both the efficiency of cell fusion and the
frequency of transcomplementation between the two mutations showed
unambiguously that events of intergenomic complementation did occur,
but only very rarely and sluggishly in cells carrying both parental
mtDNAs, with evidence of slow growth and significant cell death of the
cell fusion products.
| |
EXPERIMENTAL PROCEDURES |
Cell Lines and Media--
The pT1 human cell line, carrying in
nearly homoplasmic form the A to G transition at position 8,344 in the
mitochondrial tRNALys gene, which is associated with the
MERRF syndrome (19), and the C4T cell line, carrying in homoplasmic
form a cytidine insertion in a stretch of six Cs at positions 10,947 to
10,952 in the mitochondrial ND4 gene (27), had been
previously isolated by transfer into human mtDNA-less
143B.0206 cells of mitochondria from myoblasts of a
MERRF patient (26) and, respectively, from a mutant of the
VA2B human cell line (27) by cytoplast × cell fusion.
They were grown in Dulbecco's modified Eagle's medium (DMEM),
supplemented with 5% dialyzed fetal calf serum (FCS) and 100 µg of
bromodeoxyuridine (BrdUrd) per ml. The parental line of
0206, 143B.TK
(22) was grown in DMEM with
5% FCS and 100 µg of BrdUrd per ml. The 013.1 cell
line, a mtDNA-less derivative of an adenine phosphoribosyltransferase (APRTase)-less mutant of 143B.TK (26) was grown in DMEM
supplemented with 5% FCS and containing 50 µg of uridine, 100 µg
of BrdUrd, and 50 µg of 8-azadenine per ml. Transformants of the
013.1 cell line, obtained by cytoplast × cell
fusion-mediated transfer of mitochondria from C4T cells (hereafter
designated C4.13-E, -I, or -J), were grown in DMEM supplemented with
5% dialyzed FCS, 100 µg of BrdUrd, and 50 µg of 8-azadenine per ml
(26). The population doubling times of the various cell lines were
determined as described previously (29).
Isolation of Double Transformants--
Unless other conditions
are indicated, 1-5 × 105 cells of the chosen
mitochondrial donor cell line were enucleated by centrifugation in the
presence of cytochalasin B, and the predominantly enucleated cells were
then fused, using polyethylene glycol 1500 (PEG), with 1-5 × 105 cells of the chosen recipient cell line, as described
previously (30). After 1-6 days of incubation, as specified below, in
non-selective medium, double transformants carrying both the MERRF
tRNALys gene 8344 mutation and the ND4
frameshift mutation were selected and thereafter grown in "special
DMEM/galactose" medium. This medium contained 0.9 mg/ml galactose
instead of glucose and 0.50 mg/ml pyruvate (28) and was supplemented
with 10% dialyzed FCS, 100 µg of BrdUrd per ml, and, when C4.13-E,
or -I, or -J were used as recipient cells, also with 50 µg of
8-azaadenine per ml. Colonies were recognized between 3 and 4 weeks
after plating, and those that appeared healthy were picked up for analysis.
Karyotype Determination--
Karyotype determination was done,
as described previously (31), in cells arrested in metaphase by
treatment with 0.5 µg of colchicine per ml for 3 h.
DNA Analysis--
Total DNA was isolated from cells as described
previously (32). Quantification of mtDNA mutations was carried out in
most cases by allele-specific termination of primer extension (33, 34).
For this purpose, a 346-base pair mtDNA segment between positions
10,762 and 11,108, containing the ND4 mutation, and a
173-base pair segment between positions 8191 and 8364 of mtDNA, containing the MERRF tRNALys mutation, were amplified with
Taq polymerase. The two amplified fragments were separated
from free nucleotides on a Tris borate/EDTA, 1.5% agarose gel, and
eluted from the gel by the QIAEX II gel extraction kit (Qiagen). Then
each amplified product was used for allele-specific primer extension
termination, using Sequenase (Amersham Pharmacia Biotech) and the
corresponding 32P-5'-end-labeled primer (33, 34).
Nucleotide concentrations were 33 µM dGTP and 500 µM ddTTP for the quantification of the C insertion in the
ND4 gene, and 33 µM TTP, 33 µM
dGTP, and 500 µM ddCTP for the quantification of the
mutation in the tRNALys gene. The mixtures were heated to
95 °C for 3 min and annealed at 50 °C for 10 min. After addition
of Sequenase, they were incubated for 6 min at 45 °C.
Electrophoretic analysis of the products and quantification of the
intensity of the bands were carried out as described previously (33,
34). In some experiments, quantification of the MERRF mutation was
carried out by NaeI restriction enzyme digestion of a mtDNA
segment that had been amplified by polymerase chain reaction, using a
mismatched primer generating a new NaeI site (18).
Mitochondrial Protein Synthesis and O2 Consumption
Analysis--
The cultures were labeled with
[35S]methionine (>800 Ci/mmol, 25 µCi/ml) for 2 h
in the presence of 100 µg of emetine per ml, and translation products
were electrophoresed as described previously (26). O2
consumption measurements were made as detailed earlier (22).
| |
RESULTS |
Characterization of Parental Cell Lines--
To verify the
homoplasmic nature of the tRNALys gene mutation in the pT1
human cell line and of the ND4 gene mutation in the C4T cell
line (see "Experimental Procedures"), allele-specific primer
extension experiments were carried out, using a previously described
protocol for ND4 (33) and a protocol illustrated
schematically in Fig. 1a for
the tRNALys gene (34). As shown in Fig. 1b, the
two mutations appear to be homoplasmic. The MERRF mutation in pT1 has
been shown to produce a defect in the aminoacylation of the
mitochondrial tRNALys, which causes a severe impairment of
protein synthesis due to premature termination of translation (35).
This results in an almost complete loss of synthesis of the largest
polypeptides (ND5, ND4, and CYTb) (Fig. 1c) and the
accumulation of premature translation termination products, prominent
among which are pMERRF, a truncated COI product, and a
truncated ND2 product (indicated by small arrows
in Fig. 1c). The frameshift mutation in C4T causes a
complete loss of synthesis of ND4, with resulting lack of assembly of
the membrane arm of the NADH dehydrogenase and loss of enzyme activity
(27). As shown in Table I, in both the
pT1 and C4T cell lines, as a result of their mtDNA mutations,
O2 consumption is dramatically decreased to a level close
to that previously observed in 143B.0206 cells (22).
View larger version (54K):
[in this window]
[in a new window]
|
Fig. 1.
Amounts of wild-type (WT)
and mutant (MT) tRNALys and ND4
genes and patterns of mitochondrial translation products in the
pT1 and C4T cell lines and in three transcomplementing clones derived
from the enpT1xrecC4T cell fusion experiments. a,
scheme showing the principle of the sequence-dependent
termination of primer extension to quantify the A8344G transition in
the mitochondrial tRNALys gene. The same approach was used
to quantify the ND4 gene mutation (33). b,
fluorogram, after electrophoresis through a sequencing gel, of the
32P-5'-end-labeled primer extension products obtained from
samples of total cellular DNA purified from the indicated cell lines.
The separation of the mutant and the wild-type versions of the
tRNALys gene, as well as of the ND4 gene, is
clearly illustrated. The relative proportions of the two mutant mtDNAs
are shown in Table I. c, fluorogram, after electrophoresis
through an SDS-polyacrylamide gradient gel, of the mitochondrial
translation products of pT1 and C4T cells and of three
transcomplementing clones, labeled with [35S]methionine
for 2 h in the presence of emetine. COI, COII, and
COIII, subunits I, II, and III of cytochrome c
oxidase; ND1 ND2, ND3, ND4, ND4L, ND5, and ND6,
subunits 1, 2, 3, 4, 4L, 5, and 6 of the respiratory chain NADH
dehydrogenase; A6 and A8, subunits 6 and 8 of
H+-ATPase; Cyt b, apocytochrome b.
The small arrows indicate premature translation termination
products characteristic of the pT1 cells, and the large
arrowheads indicate the two polymorphic variants (ND3
and ND3') of the ND3 polypeptide.
|
|
View this table:
[in this window]
[in a new window]
|
Table I
Mitochondrial genotype, chromosome number, oxygen consumption rate, and
doubling time of different cell lines
|
|
Selective Conditions for Full mtDNA Functionality--
For a large
scale screening of transcomplementing clones among mitochondrial
cybrids and hybrids containing two types of mitochondria carrying
either the tRNALys or the ND4 gene mutation, a
medium containing galactose instead of glucose, which was capable of
selecting against the two parental defective cell lines, was used. Such
type of medium has been previously shown to curtail severely the growth
of cells deficient in oxidative metabolism (28, 36-40).
As shown in Table I, cells carrying wild-type mtDNA
(143B.TK (22)) and the mutant C4T and pT1 cell lines grew
well in regular DMEM containing glucose and supplemented with 5% FCS.
When glucose was substituted by galactose, the cells carrying wild-type
mtDNA (143B.TK) continued to grow, although at a reduced
rate, whereas both types of mutant cell lines not only stopped growing
but showed extensive cell death. This death was indicated by the
progressive cell detachment from the plates. To confirm that the normal
expression of mtDNA was absolutely required for the survival and growth
of the 143B.TK cells in special DMEM/galactose medium, as
suggested by the results described above, mitochondrial protein
synthesis was specifically inhibited by adding 40 µg/ml
chloramphenicol (CAP) to the incubation medium; furthermore, the medium
was supplemented with 50 µg/ml uridine, since it is known that cells
without full mitochondrial gene expression become auxotrophic for
pyrimidines (22, 36, 41). 143B.TK cells grew well in the
presence of CAP in glucose-containing medium supplemented with uridine
but stopped growing and became detached from the plate in special
DMEM/galactose supplemented with CAP and uridine (data not shown).
Evidence of Reciprocal Complementation of tRNALys and
ND4 Mutations in Double Transformants--
Advantage was taken of the
selective pressure provided by growth in special DMEM/galactose medium
in order to isolate clones exhibiting transcomplementation of the
tRNALys and ND4 gene mutations. For this
purpose, 105 pT1 cells were subjected to enucleation (22),
and the predominantly enucleated cells were then fused with 5 × 105 C4T cells. After a 2-day incubation in non-selective
medium, a portion of the fusion mixture, corresponding to 2.5 × 105 recipient cells, was transferred to special
DMEM/galactose medium in a Petri dish. After 3-4 weeks, eight clones
were detected. Two of them subsequently died, and the other six were
picked up. However, of them, only three (T3, T6, and T7) kept growing
and were utilized for further analysis (Table I and Fig. 1). An
equivalent portion of the fusion mixture was transferred to special
DMEM/galactose medium after a 6-day incubation in non-selective medium
and yielded initially six clones which, however, were not picked up for
further growth.
Primer extension experiments revealed that all three clones contained
both the mtDNA carrying the A8344G MERRF mutation and the mtDNA
carrying the ND4 mutation (Fig. 1b), with the
former one being strongly predominant. A quantification by
PhosphorImage analysis showed that the proportion of the
tRNALys mutation-carrying mtDNA varied between 71 and
~88% and that of the ND4 mutation-carrying mtDNA between
10.5 and ~22% of the total mtDNA, and that the two mutant types of
mtDNA accounted for all mtDNA, within experimental error (Table I). The
latter observation indicated that these clones did not contain any
significant amount of residual wild-type mtDNA from either parental
cell line. Furthermore, the proportion of wild-type tRNALys
genes present in ND4 gene mutation-carrying mtDNA exceeded
the minimum previously shown to be able to complement the MERRF
mutation (~10%) (18). Therefore, one expected that, if there was
complete mixing of the gene products in the three clones, there would
be full complementation of the MERRF mutation. There are no data on the
quantitative effects of a null mutation in the ND4 genes on
their expression. However, if the same pattern of control applies to
these genes as recently observed for the human ND5 genes
(42), one would expect in the three clones between 70 and 90% of the normal rate of synthesis of the ND4 protein.
The analysis of mitochondrial protein synthesis in the three isolated
clones fully confirmed the predictions mentioned above. In fact, as
shown in Fig. 1c, the three clones exhibited a near to
normal rate of ND5, ND4, and CYTb polypeptide synthesis. Furthermore, there was no clear evidence of premature translation termination, with
the exception of a small amount of the main premature translation termination product, pMERRF (35), migrating just ahead of ND4 in T6.
The occurrence in C4T mtDNA, as in the VA2B parental cell mtDNA, of an ND3 polymorphism previously described in HeLa cells, which
causes a faster migration of this subunit in a polyacrylamide gel (43),
offered an additional useful marker of transcomplementation. Thus, in
the protein synthesis patterns of Fig. 1c, the relative labeling of the ND3 and ND3' bands, characteristic of pT1 and C4T
mtDNA, respectively, showed a strong prevalence of ND3 in clones T3 and
T6 and nearly equal labeling of the two bands in T7. Considering the
large excess of MERRF mutation-carrying mtDNA in the three
transformants and the expected low rate of ND3 synthesis in homoplasmic
MERRF mutant cells relative to the wild-type cells (~35% (35)), the
observed high ratio of ND3 to ND3' labeling in T3 and T6 is consistent
with a substantial level of transcomplementation in these clones of the
tRNALys mutation by ND4 mutation-carrying
mtDNA.
Chromosome count in the three clones showed that T7 had an average
chromosome number very similar to those of the parental cell lines, as
expected for cybrids, whereas T3 and T6 had a chromosome number nearly
double, strongly suggesting that these two clones represented cell
hybrids (Table I). Oxygen consumption measurements (Table I) showed a
recovery of the normal rate in all three clones, when compared with the
rate expected in 143B cells (5-6 fmol/min/cell (18, 44)), in agreement
with the conclusion that a reciprocal complementation of the two
mutations had occurred in them. The calculated absolute frequency of
transcomplementation in this experiment was thus ~1.2 × 105 (Table II). However, a
surprising finding was that the three clones grew in regular DMEM
considerably more slowly than 143B cells (Table I). Also in special
DMEM/galactose medium, they grew more slowly than 143B cells (Table I).
Furthermore, in the latter medium, they showed evidence of a
significant rate of cell death, as revealed by the detachment of cells
from the plate (Table I). It was particularly intriguing that these
clones kept growing at a low rate and exhibited an abnormally high rate
of cell death even after 18 weeks of continuous culturing since their
isolation.
Residual Wild-type mtDNA in the Parental pT1 Cell Line--
The
very low frequency of transcomplementing clones in the experiments
described above raised the possibility that the pT1 cell line and/or
the C4T cell line contained a very minor residue of wild-type mtDNA,
not detectable by the primer extension assays, and distributed in a
non-uniform way among the cells. Rare cells in one or the other
parental cell line might conceivably produce, by intramitochondrial
complementation of the mutation, clones mimicking true
transcomplementing clones. To exclude this possibility, 107
C4T cells were maintained for 6 weeks in special DMEM/galactose medium.
As expected, massive cell death was observed, but not a single C4T
clone able to grow in galactose medium was detected. A similar
experiment carried out with pT1 cells also revealed a massive cell
death. Surprisingly, however, surviving clones were detected, at a very
low frequency, i.e. of ~2 per 105 original
cells. In order to investigate further this phenomenon, eight of the
surviving clones were isolated, and their mtDNA was analyzed. In all
cases, these clones were found to carry the MERRF mutation in
heteroplasmic form (Fig. 2a),
exhibiting a minimum of 9% wild-type mtDNA, which was sufficient to
restore mtDNA functionality in MERRF mutation-carrying cells (18).
View larger version (43K):
[in this window]
[in a new window]
|
Fig. 2.
Evidence of the presence of a residual minor
amount of wild-type (WT) mtDNA in pT1 cells.
a, shows a fluorogram, after electrophoresis through a
sequencing gel, of the 32P-5'-end-labeled primer extension
products obtained from samples of total cellular DNA extracted from the
parental pT1 and C4T cell lines and eight pT1 clones able to grow in
galactose-containing medium. b, the yield of subclones
growing in galactose-containing medium from 106 cell
samples of 10 independent clones isolated in glucose-containing medium
from the original pT1 cell line is compared with the yield obtained
from the original pT1 cell population. Mean indicates the average yield
of subclones from the 10 clones ± S.E.
|
|
In order to analyze the distribution of the residual wild-type mtDNA in
the MERRF mutant cell population, 10 clones were isolated from this
population in regular DMEM medium, and 106 cells of each
clone were plated in special DMEM/galactose medium. As shown in Fig.
2b, the frequency of the subclones growing in this medium
varied greatly among the original 10 pT1 clones, with the cumulative
frequency (2.4 × 105) being very similar to that
observed in the original cell population (2 × 105).
The experiment described above allowed the isolation of pT1 clones
carrying the MERRF mutation in reliably homoplasmic form, as judged
from the fact that 106 of their cells did not generate any
surviving subclones under strictly selective conditions. This
conclusion was confirmed for clone 3 by an experiment in which
107 cells were replated in special DMEM/galactose, and no
surviving cells were detected after 5 weeks of culture. As expected,
this clone had maintained the respiratory-deficient phenotype of the pT1 cell line (Table I). Clone 3 (hereafter designated pT1.C3) was
therefore chosen for all subsequent experiments.
Role of the Nuclear Background in Transcomplementation--
In the
transcomplementation experiment described above, a surprising finding
was that in the transcomplementing clones, including the cybrid T7, the
MERRF mutation-containing mtDNA was in large excess over the mtDNA
carrying the ND4 gene frameshift mutation. In order to test
the possible role of the nuclear background in this phenomenon, as well
as explaining the low frequency of transcomplementation, another cell
fusion experiment was carried out, this time using 5 × 105 predominantly enucleated C4T cells as mitochondria
donors and 5 × 105 pT1.C3 cells as recipients. The
use of the pT1.C3 cells was also expected to provide a test of the
possible influence of the small amount of endogenous wild-type
tRNALys genes in the previous experiment. A portion of the
fusion mixture, corresponding to ~1.7 × 105
recipient pT1.C3 cells, after 24 h incubation in non-selective medium, was transferred to special DMEM/galactose in two 10-cm dishes.
Of 31 detected clones, only 10 continued to grow, and among these, five
clones (F31, F34, F35, F39, and F310) were randomly selected for
further analysis.
Primer extension analysis revealed that all five clones contained both
the MERRF mutation-carrying mtDNA and the ND4
mutation-carrying mtDNA (Fig. 3), with
the former being again predominant, although somewhat less than in the
first experiment (Table I). The ND4 mutation-carrying mtDNA,
with the MERRF-mutant mtDNA, again accounted, within ±<3%, for the
total mtDNA (Table I). Karyotype analysis indicated the likely cybrid
nature of the F31 and F310 clones, with an increased parental
chromosome number, and the probable hybrid nature of the F34, F35, and
F39 clones (Table I).
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 3.
Amounts of tRNALys gene or
ND4 gene mutation-carrying mtDNAs in the
transcomplementing clones derived from enC4T × recpT1.C3 and
enpT1.C3 × recC4.13 cell fusions. a and b,
fluorograms, after electrophoresis through a sequencing gel, of the
32P-5'-end-labeled primer extension products obtained from
samples of total cellular DNA extracted from the indicated cell lines.
The relative proportions of the two mutant mtDNAs in each cell line,
estimated from the fluorograms, are shown in Table I. WT,
wild type; MT, mutant.
|
|
As observed for the clones isolated in the first experiment, the new
five selected clones grew very slowly in special DMEM/galactose medium
(Table I), again with evidence of a significant rate of cell death as
revealed by cell detachment from the plates (data not shown). The
recovery of mtDNA functionality in these clones was indicated by the
re-establishment of full respiratory capacity (Table I) and of a nearly
normal mitochondrial protein synthesis pattern, as well as by the
substantial reduction of the premature translation termination products
pMERRF and truncated ND2 polypeptide and by the relatively high rate of
labeling of the ND3 subunit. These results indicated that
transcomplementation of the two mutations had occurred in these clones,
although the evidence suggested a less effective expression of the C4T
mtDNA than in the previous experiment. The calculated absolute
complementation frequency was ~6 × 105,
i.e. considerably higher than that measured in the first
experiment (Table II), presumably reflecting the higher ratio of
mitochondrial donor to recipient cells used in the cell fusion. In
conclusion, the results obtained in this experiment, which utilized
pT1.C3 cells as recipients, showed that the small residue of wild-type mtDNA in pT1 cells was not responsible for the transcomplementation observed in the previous experiment. Furthermore, they indicated that
the nuclear background of the recipient cells did not play any
significant role in the predominance of the MERRF mutation-carrying mtDNA in the transcomplementing clones.
Frequency of Transcomplementation Relative to Cell Fusion
Frequency--
A critical factor in evaluating the significance of the
low absolute frequency of transcomplementation in the experiments described in the previous sections was the efficiency of cytoplast or
cell × cell fusion. In order to estimate the frequency of the fusion events, growth of the products under conditions non-selective for respiration competence, i.e. in regular
glucose-containing DMEM, was required. The quantification in this type
of experiment, in turn, was expected to be greatly facilitated by the
elimination from the mitochondrial donors, in this case pT1.C3 cells,
of non-enucleated cells. These, in fact, could grow in medium
non-selective for respiration competence and thus interfere with the
identification of the A8344G mutation-carrying mtDNA in the cybrids or
hybrids, to be carried out by NaeI digestion of total cell
DNA (see "Experimental Procedures"). In order to eliminate the
non-enucleated mitochondrial donor cells, a C4T derivative carrying a
second drug resistance marker, besides BrdUrd resistance, was isolated.
For this purpose, the ND4 mutation-carrying mitochondria
were transferred, as detailed under "Experimental Procedures," into
a cell line different from 143B.0206, i.e.
143B.013.1. This cell line had been previously isolated
by long term exposure to a low concentration of ethidium bromide of an
APRTase-less mutant, which was, therefore, resistant to 8-azaadenine,
of 143B.TK cells (26). Several clones resistant to BrdUrd
and 8-azaadenine were thus isolated, and three of these (C4.13E,
C4.13J, and C4.13I) were utilized as recipients for fusion with
enucleated pT1.C3 cells in subsequent experiments. It was verified that
they contained pure mutant ND4 gene and had a
respiratory-deficient phenotype, as shown for clones C4.13J and C4.13I
in Fig. 3 and Table I.
Several fusions between C4.13 clones and enucleated pT1.C3 cells were
carried out, as detailed in Table II. In two experiments (experiments 7 and 8), a 5-fold and, respectively, 25-fold higher ratio of cytoplasts
to recipient cells was used, as compared with the other experiments. In
each case, after 24 h incubation in non-selective medium, a
portion of the fusion mixture was plated in an appropriate number of
96-microwell plates, at a concentration of ~0.2 cell/microwell, in
regular DMEM containing 100 µg of BrdUrd and 50 µg of 8-azaadenine
per ml. Another portion of the fusion mixture (30 to ~100%, as
specified in Table II) was plated, either in one Petri dish or in
96-microwell plates (in general three), in special DMEM/galactose
medium containing BrdUrd and 8-azaadenine, and the remainder was frozen.
Total cell DNA from 88 to 212 clones, isolated in different experiments
from a portion of the cell fusion mixture grown under conditions
non-selective for respiration competence (in regular DMEM), was tested
by restriction enzyme NaeI digestion for the presence and
proportion of the A8344G tRNALys gene mutation in mtDNA. In
three of the fusion mixtures analyzed (experiments 3, 4, and 8 in Table
II), a small fraction of the clones tested showed the presence of the
tRNALys gene A8344G mutation in a proportion varying
between 26 and 90% of total mtDNA. Because of the apparent complete
absence of wild-type tRNALys genes in pT1.C3 cells and of
wild-type ND4 genes in C4.13 cells (Fig. 3), in the clones
which carried the tRNALys gene mutation in a portion of
their mtDNA, the remainder of mtDNA was considered to represent mtDNA
with mutant ND4 genes. Accordingly, the percentage of clones
with mutant tRNALys genes provided an estimate of the
frequency of fusion of the two parental lines. This varied between 1.4 and 3.4% (Table II). In three of the fusion experiments, no clone
containing the tRNALys gene A8344G mutation was found, due
presumably to the very low frequency of cell fusion.
A few clones derived from the pT1.C3xC4.13 clone fusions, which grew in
galactose medium containing 8-azaadenine, were analyzed for genotype
(Fig. 3), chromosome number, oxygen consumption rate (Table I), protein
synthesis (Fig. 4), and growth rate. All
the clones tested contained both types of mutant mtDNA, with the
proportion of MERRF mutation-carrying mtDNA varying between 60 and
82%, and with the two mutant forms accounting, within 1-3%, for the
totality of mtDNA (Fig. 3). These clones exhibited a chromosome number indicating their cybrid nature, as expected from the recessive nature
of the APRTase null mutation, which would have killed any hybrid cell.
Furthermore, they were fully respiratory-competent, as expected from
the selection medium used (Table I). Their protein synthesis patterns
(Fig. 4) revealed the reacquisition of the capacity to synthesize the
ND5, ND4, and CYTb polypeptides and the strong reduction or absence of
the abnormal premature translation termination products pMERRF and
truncated ND2 polypeptide. In addition, they exhibited again a
preponderant labeling of the ND3 polypeptide characteristic of pT1.C3
mtDNA (Fig. 4). As observed in the previous experiments, the clones
analyzed grew in general very slowly in galactose medium (Table I) and
exhibited a significant rate of cell death (data not shown).
View larger version (125K):
[in this window]
[in a new window]
|
Fig. 4.
Fluorogram, after electrophoresis through an
SDS-polyacrylamide gradient gel, of the mitochondrial translation
products of the indicated cell lines labeled with
[35S]methionine for 2 h in the presence of
emetine. Symbols are as described in the legend of Fig. 1.
|
|
The evidence presented above indicated clearly that the clones growing
in galactose medium were true transcomplementing clones. As shown in
Table II, the absolute transcomplementation frequency among the clones
isolated in the experiments with high cell fusion frequency
(experiments 3, 4, and 8) was similar to that previously observed for
the enC4T × pT1.C3 fusion (experiment 2), i.e. 5.4 and
8.4 × 105 (experiments 3 and 4), or much higher
(5.5 × 104), in the case of the fusion utilizing
the largest excess of cytoplasts (experiment 8). It was, on the
contrary, considerably lower (1.2 to 3 × 105) in
the experiments with low cell fusion frequency (experiments 5-7), as
expected. The absolute frequency of the transcomplementing clones,
divided by the frequency of cell fusion, provided an estimate of the
relative efficiency of transcomplementation. As shown in Table II, this
varied between 3.0 × 103 and 3.8 × 103 in experiments 3 and 4, and 1.6 × 102 in experiment 8. These results indicated clearly the
intrinsically very low frequency of the transcomplementation events.
| |
DISCUSSION |
In the present work, large scale fusion experiments between
cultured human cell lines carrying two non-allelic recessive mtDNA mutations have allowed the detection of stable transcomplementation between the two mutations under growth conditions selective for recovery of respiratory capacity. However, the phenomenon occurred only
very rarely among the double transformants, with a 0.3 to 1.6%
frequency, accounting for the failure to detect it in earlier experiments not involving selection for respiratory competence (18).
The transcomplementing clones became established very sluggishly and
exhibited in general slow growth and a substantial rate of cell death.
The present results confirm the previous observations of
transcomplementation of mtDNA mutations in human cell cultures (24,
25). However, they are in striking contrast to the general model that
has been proposed of human mitochondria functioning as a single dynamic
unit in a living cell, with rapid diffusion of mtDNA and/or its
products throughout the organelles (23). They support the conclusion
that the capacity of mitochondria to fuse and mix their genetic
contents is not an intrinsic general property of these organelles in
mammalian cells, although it may be susceptible to developmental and
physiological regulation.
Previously, the occurrence of apparent mitochondrial fusion in
mammalian cultured cells, independent of cell fusion, has been reported
by others (9, 10, 11) on the basis of studies using optical or electron
microscopy. Although fusion is necessary for the occurrence of genetic
complementation, any physical evidence of fusion would not necessarily
imply a mixing of the mtDNAs and/or their products from different
organelles. It is in fact quite possible that only the external
mitochondrial membranes fuse or that the mtDNAs and their transcripts
remain compartmentalized in fully fused mitochondria.
Mechanism Underlying the Transcomplementation of the Two
Mutations--
In the present work, one must assume that some form of
mitochondrial fusion event produced the observed transcomplementation of the two mtDNA mutations analyzed. However, nothing is known about
the mechanism involved. The very low frequency of reciprocal complementation of the two mutations detected in the cytoplast or
cell × cell fusion products argues strongly against a programmed phenomenon. One cannot exclude that, in a small fraction of the transformed recipient cells, a gene homologous to the fzo
gene transiently activated during spermatid development in D. melanogaster (12) and to its equivalent found in S. cerevisiae (13, 14) and the genes encoding other components of the
mitochondrial fusion machinery (13) are expressed at a low level.
However, a more likely possibility, which has not been excluded in any
of the experiments reported so far in which transcomplementation has been observed (23-25), is that the transient exposure of the cells to
polyethylene glycol (PEG) used to induce cell fusion may be responsible
for the fusion of mitochondria. The low frequency of stable
transcomplementing clones observed in the present and previous work
would imply that these mitochondrial fusion events, which would produce
heteroplasmic organelles, would be followed by a reversal of
transcomplementation, due to division of the organelles and
re-segregation of the two mtDNAs. According to this hypothesis, only
very infrequently would transformants with stable intramitochondrial
heteroplasmy be selected, accounting for the great rarity of stable
transcomplementing clones observed. A similar hypothesis had been
previously proposed, which associated mitochondrial fusion with a
transient response to a cell fusion stimulus, like during myotube
formation (45). One may further elaborate this alternative model by
assuming that the transient response to a cell fusion stimulus would
involve an activation of the fzo gene and related genes. If
an extensive reversible mitochondrial fusion event during
cell fusion does indeed occur, the real in vivo frequency of
transcomplementation events may be much lower than observed here. This
possibility is open to experimental test through the introduction of
donor mitochondria into the recipient cell by microinjection (46).
Factors Affecting the Efficiency of Transcomplementation--
In
the scenario discussed above of mitochondrial fusion being transiently
activated during cell fusion, the very rare, slowly established
transcomplementing clones detected in this work and previously by
others (24, 25) would represent those few clones that underwent no
segregation or only partial intermitochondrial segregation of the two
mutant mtDNAs. The considerable proportion of dead cells that was
observed in the present work during growth of the surviving clones
after cell fusion may well be the result of an equilibrium situation
between selection of cells having reached the threshold of respiratory
competence associated with the minimum necessary number of
heteroplasmic organelles and continuous loss of this competence due to
segregation of the two defective mtDNAs by mitochondrial division.
It has been suggested that the very low frequency of
transcomplementation between homoplasmic recessive mtDNA mutations in cell fusion products grown under conditions selective for respiration may result from lack of production of sufficient energy to overcome the
energy barrier of membrane fusion (25). However, this suggestion is not
supported by the results of previously mentioned work from this
laboratory (18), which had shown that the non-transcomplementing clones
isolated under conditions non-selective for respiration grew at a high
rate (doubling times between 22 and 28 h) and were presumably able
to produce adequate ATP from glycolysis to carry out mitochondrial
fusions. In the present work, the possibility that the observed very
low frequency of transcomplementation was due to the fact that the
incubation of the cell fusion products under non-selective conditions
was too short to allow adequate mitochondrial fusion, and genetic
complementation is not supported by the available evidence. In fact, in
the first experiment, a 2-day and a 6-day incubation of the fusion
products under non-selective conditions yielded about the same number
of clones surviving under selective pressure. Furthermore, an argument
against the idea of a quick death of respiratory-deficient cells in
selective medium is the observation that the proportion of dead cells
after a 2-day incubation in such medium of the two parental cell lines
was only 12-40%. Finally, the observation that, in ND4
gene frameshift mutation-carrying mitochondria, all proteins other than
ND4 exhibited a normal rate of synthesis and the expectation that these
organelles carried an ~2-fold excess of aminoacylated
tRNALys over that required to support normal translation
(35) strongly suggests that, if they underwent a fusion with
tRNALys mutation-carrying mitochondria, there would be a
rapid complementation of the latter mutation.
The finding in the present work that, independently of whether
tRNALys gene mutation- or ND4 gene
mutation-carrying cells were used as recipients for cell fusion, the
MERRF mutation-carrying mtDNA was always predominant (60-90% in the
various complementing clones) presumably reflected the need for the
cell to have an adequate rate of NADH dehydrogenase activity. It is
known, in fact, that Complex I-dependent O2
consumption is usually the rate-limiting step in respiration (47,
48).
Implications for Disease-causing or Aging-related mtDNA
Mutations--
The present observations have important implications
for understanding the segregation and complementation behavior of
disease-causing mtDNA mutations in man and the transmission of such
diseases (49), as well as for potential therapeutic approaches
involving mitochondria-mediated gene transfer. Another area where the
observed tendency of mammalian mitochondria to remain genetically
autonomous has significant implications is that of
aging-dependent mtDNA damage. There is substantial evidence
of an aging-related occurrence in mtDNA of oxidative derivatives of
nucleotides (50), of small deletions and insertions, and of large
deletions (51-54). Most significantly, very recently an
aging-dependent large accumulation of specific mutations in
a critical control region for mtDNA replication has been demonstrated
in human fibroblasts (55). The functional effects of these mutations
depend not only on their frequency but also on whether or not
mitochondria can fuse and mix their mtDNA products, so as to allow
wild-type genes to complement mutant genes. Additivity of the
aging-dependent dysfunctions of the individual organelles
in a cell, as opposed to transcomplementation between mitochondria,
would magnify considerably the overall damage to the cell.
§¶,
,
TOP
ABSTRACT
INTRODUCTION
Present address: Dpto. Bioquímica y Biología
Molecular A., Universidad de Murcia, 30071 Murcia, Spain.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
