Nuclear-recessive mutations of factors involved in mitochondrial translation are responsible for age-related respiration deficiency of human skin fibroblasts.

We addressed the question of whether both mitochondrial and cytoplasmic translation activities decreased simultaneously in human skin fibroblasts with the age of the donors and found that the age-related reduction was limited to mitochondrial translation. Then, to determine which genome, mitochondrial or nuclear, was responsible for this age-related, mitochondria-specific reduction, pure nuclear transfer was carried out from mitochondrial DNA (mtDNA)-less HeLa cells to four fibroblast lines, two from aged subjects, one from a fetus, and one from a patient with cardiomyopathy, and their nuclear hybrid clones were isolated. A normal fibroblast line from the fetus and a respiration-deficient fibroblast line from the patient were used as a positive and a negative control, respectively. Subsequently, the mitochondrial translation and respiration properties of the nuclear hybrid clones were compared. A negative control experiment showed that this procedure could be used to isolate even nuclear hybrids expressing overall mitochondrial respiration deficiency, whereas no respiration deficiencies were observed in any nuclear hybrids irrespective of whether their mtDNAs were exclusively derived from aged or fetal donors. These observations suggest that nuclear-recessive mutations of factors involved in mitochondrial translation but not mtDNA mutations are responsible for age-related respiration deficiency of human fibroblasts.

It has been presumed that somatic mutations accumulate in mitochondrial DNA (mtDNA) much faster than in nuclear DNA because mitochondria are highly oxygenic organelles due to their function in producing energy, mtDNA lacks histones protecting it from mutagenic damage, and its repair systems are limited (1). Therefore, it has been proposed that the accumulation of various somatic mutations in mtDNA and the resultant decrease in mitochondrial respiratory function could be involved in aging processes in mammals (2)(3)(4). There have been many reports that the respiration capacity of mitochondria in highly oxidative tissues decreases during aging (4). Moreover, the accumulation of somatic and pathogenic mtDNA mutations, which have been shown to cause various kinds of mitochondrial encephalomyopathies (5)(6)(7)(8), was also shown to increase with age in normal subjects (9,10). However, as the nuclear genome encodes most mitochondrial proteins including factors necessary for replication and expression of the mitochondrial genome, it is possible that only mutations in the nuclear genome contribute to the age-related decline of mitochondrial respiratory function. In fact, there is no convincing evidence that mtDNA somatic mutations are responsible for this age-related phenotype.
Previously, we observed age-related reduction of cytochrome c oxidase (COX) 1 activity and mitochondrial translation in cultured human skin fibroblasts isolated from donors of various ages (0 -97 years), and in studies on their mtDNA transfer to mtDNA-less ( 0 ) HeLa cells, we showed that mtDNA mutations were not responsible for the observed age-related mitochondrial dysfunction of human skin fibroblasts (11). Recently, using an mtDNA transfer system similar to ours (11), i.e. mtDNA transfer from human skin fibroblasts to 0 human cells, Laderman et al. (12) reported contradictory results, suggesting the presence of age-related heritable alterations in fibroblast mtDNA. However, the procedure for isolating cybrids by the transfer of fibroblast mtDNA to 0 cells is not appropriate for unambiguous determination of whether accumulation of mtDNA mutations is involved in age-related mitochondrial dysfunction, because during selection to exclude 0 cells using medium without uridine and/or pyruvate, cybrids expressing respiration deficiency due to accumulation of mtDNA mutations might also be eliminated preferentially, so that only cybrids with normal mitochondrial respiratory function are isolated. Thus in these conditions (11,12), defective cybrids with mutant mtDNA from both young and aged subjects may have been excluded preferentially.
Recently, to determine whether the mitochondrial or nuclear genome is responsible for mitochondrial diseases, we developed a system for delivery of a normal nuclear genome from 0 HeLa cells to fibroblasts from patients with respiration deficiency by isolating nuclear hybrids (13). In these nuclear hybrid clones, the nuclear genome was derived from both parental 0 HeLa cells and fibroblasts, whereas the mitochondrial genome is exclusively from fibroblasts. Thus, if mitochondrial dysfunction is restored by the introduction of pure 0 HeLa nuclei, it should be attributed to nuclear-recessive mutations, and if not, it should be due to either mtDNA mutations or nuclear-dominant mutations. Moreover, since HeLa nuclei completely free from mtDNA can be introduced into fibroblasts simply by fusion of the fibroblasts with 0 HeLa cells followed by hypoxanthine/ aminopterin/thymidine (HAT) and ouabain (Oua) selection, this nuclear genome delivery system does not impose any selective pressure upon mitochondrial respiratory function to remove 0 parental cells. Therefore, it should be effective for isolating nuclear hybrid clones, even when they express mitochondrial respiration deficiency like 0 cells.
In the present work, we did not use mtDNA transfer techniques but used nuclear transfer techniques that do not eliminate clones expressing complete respiration deficiency, so that we could exclude the possibility of preferential selection of respiratory-competent clones. By this method, we showed that nuclear-recessive mutations involved in mitochondrial translation are responsible for age-related mitochondrial respiration deficiency, and that mtDNA mutations, if they occurred, did not play any significant role in expression of the phenotype. human skin fibroblast lines TIG3S,  TIG2S, TIG1, TIG104, TIG105, TIG106, TIG107, TIG101, and TIG102 were obtained from the Department of Biology, Tokyo Metropolitan Institute of Gerontology (TIG). Fibroblast lines TIG3S, TIG2S, and TIG1 were from fetuses, whereas TIG104, TIG105, TIG106, TIG107, TIG101, and TIG102 were derived from 70, 72, 80, 81, 86, and 97-yearold subjects, respectively. For the study of aging, all TIG fibroblast lines were carefully established in TIG from normal subjects with no clinical abnormalities. TIG3S was used as a positive control of mitochondrial respiratory function in nuclear transfer experiments. As a negative control, we used a respiration-deficient fibroblast line CM derived from a 18-year-old patient with fatal cardiomyopathy with a tRNA Ile 4,269 pathogenic mtDNA mutation (14,15), which was obtained from Dr. S. Okada, the Department of Pediatrics, Osaka University Medical School. The 0 HeLa cells are resistant to 20 M 6-thioguanine and 2 mM Oua (6). The fibroblast lines and 0 HeLa cells were grown in normal medium (RPMI 1640, 0.1 mg/ml pyruvate, 50 g/ml uridine, 15% fetal bovine serum).

Cells and Cell Culture-Normal
Intercellular Transfer of the HeLa Nuclear Genome-Intercellular transfer of HeLa nuclei to fibroblast lines was achieved by fusion of the fibroblasts with 0 HeLa cells using polyethylene glycol 1500 (Boehringer Mannheim) as described previously (13). Briefly, cells in the fusion mixtures were cultivated in selective medium (RPMI 1640, 0.1 mg/ml pyruvate, 50 g/ml uridine, 15% fetal bovine serum, HAT (Sigma), 2 mM Oua). On day 10 after fusion, colonies grown in the selection medium were cloned by the cylinder method, and the nuclear hybrid clones were then cultivated in the normal medium.
Southern Blot Analysis-Total DNA (2g) extracted from 2 ϫ 10 5 cells was digested with the single-cut restriction enzyme XhoI to estimate the contents of mtDNA in fibroblasts from fetal and elderly donors or digested with HhaI to determine whether mtDNA of the cybrids was derived from mtDNA of the donor cells. The fragments separated by agarose gel electrophoresis were then transferred to nitrocellulose membranes and hybridized with [␣-32 P]dATP-labeled HeLa mtDNA. For quantitation of mtDNA of normal size in the fibroblasts, the membranes were exposed to imaging plates (Fuji Film, Tokyo) for 30 min, and radioactivity was measured with a bioimaging analyzer, Fujix BAS 2000 (Fuji Film). Quantitation of the mtDNA content was normalized by the use of [␣-32 P]dATP-labeled 18 S rDNA as an internal control.
Northern Blot Analysis of Transcripts of mtDNA-Total cellular RNA was extracted with an isogen RNA isolation kit (Nippon Gene, Toyama, Japan). Total denatured RNA (10 g) was electrophoresed in 1.1% agarose gel containing formaldehyde and then transferred to a nytran membrane. The membrane was hybridized with [␣-32 P]dATP-labeled oligonucleotide probes of mtDNA-coded COI (nucleotide positions 5,847-7,746) and 16 S rRNA (1,740 -3,130) and with [␣-32 P]dATPlabeled cDNA probes of nuclear DNA-coded mitochondrial transcription factor A and ATPase F 1 ␣ (13). The radioactivities of the bands were measured with a bioimaging analyzer, Fujix BAS 2000. Values for RNA contents were normalized by the use of [␣-32 P]dATP-labeled ␤-actin as an internal control.
PCR Analysis-For detection of a small amount of a common deletion mutant mtDNA, ⌬mtDNA4977, amplification was carried out with 20 ng of total DNA in 10 l of solution containing 0.5 M of a primer set and 0.25 unit of EX-Taq polymerase (Takara, Japan), as described previously (11,16,17) with slight modifications. Briefly, three sets of oligo-nucleotide primers were used for amplification: F1 (nucleotide positions 7,901-7,920) on the light strand and R1 (14,220 -14,201) on the heavy strand; F2 (8,220) and R2 (13,851-13,832); F3 (8,305) and R3 (13,650 -13,631). The first round of PCR employed the outer primer set F1 and R1. After incubation for 5 min at 94°C for complete denaturation of the DNA, 30 cycles were carried out at 94°C for 30 s for denaturation, 54°C for 30 s for annealing, and 72°C for 75 s for extension using a DNA thermal cycler (Thermal sequencer TRS-300; Iwaki Garasu, Japan). The second round of PCR employed the inner primer set F3 and R3, and an additional 20 cycles were run at 94°C for 30 s for denaturation, 60°C for 25 s for annealing, and 72°C for 25 s for extension. For detection of very small amounts of deletion mutant mtDNA molecules, the primer set F2 and R2 was used in the second round, and the primer set F3 and R3 was used in the last round (18). The amplified products were separated in 2.5% agarose gels containing ethidium bromide (0.1 g/ml).
DNA Sequencing of PCR Products-PCR products purified using a Qiaex II gel extraction kit (Qiagen, Germany) were sequenced directly with a Taq polymerase kit (Prism) with fluorescent primers and dideoxynucleotides. Sequencing reactions were analyzed with an Applied Biosystems model 377 automatic sequencer.
Measurements of Mitochondrial and Cytoplasmic Translation Activities-Mitochondrial translation products were labeled with [ 35 S]methionine as described previously (13). Briefly, semiconfluent cells in a dish were incubated in methionine-free medium containing 10% fetal bovine serum for 1 h at 37°C. Then, the cells were labeled with [ 35 S]methionine for 2 h in the presence of emetine (0.15 mg/ml) to inhibit cytoplasmic translation. The mitochondrial fraction was obtained by homogenization in 0.25 M sucrose, 1 mM EGTA, 10 mM Hepes-NaOH, pH 7.4, followed by differential centrifugation. Proteins in the mitochondrial fraction (20 g/lane) were separated by SDS, 12% polyacrylamide gel electrophoresis. The dried gel was exposed to an imaging plate for 24 h, and the labeled polypeptides were located and measured with a bioimaging analyzer. For measurement of cytoplasmic translation activity, the cells were labeled with [ 35 S]methionine for 2 h in the presence of chloramphenicol (0.1 mg/ml) to inhibit mitochondrial translation, and proteins in the whole cells (50 g/lane) were separated by SDS, 12% polyacrylamide gel electrophoresis. The dried gel was exposed to an imaging plate for 12 h, and the labeled total polypeptides were located and measured with a bioimaging analyzer.
Analysis of COX Activity-For biochemical analysis, log-phase cells were harvested, and COX activity was measured as the rate of cyanidesensitive oxidation of reduced cytochrome c as described before (19).
Measurement of Oxygen Consumption-Oxygen consumption rate was measured by trypsinizing cells (1.5 ϫ 10 7 ), incubating the suspension in phosphate-buffered saline, and recording oxygen consumption in a polarographic cell (1.0 ml) at 37°C with a Clark-type oxygen electrode (Yellow Springs Instrument Co., OH) (20).

RESULTS
First, we compared the COX activities of three fibroblast lines from fetuses and six from aged subjects (70, 72, 80, 81, 86, and 97 years old) and confirmed that the COX activities of the lines from aged subjects were only about 20 -40% those of the fetal lines (Fig. 1), consistent with our previous observations (11). To determine the reasons for the decrease in COX activity in aged subjects, we compared the levels of mtDNA and its transcripts in the fibroblasts. Southern blot and Northern blot analyses showed that their levels did not change substantially with the age of the fibroblast donors (Fig. 2). Then we compared the mitochondrial translation activities of the fibroblasts by measuring [ 35 S]methionine incorporation into mtDNA-encoded polypeptides in the presence of emetine to inhibit cytoplasmic translation. Results showed that mitochondrial translation activity decreased with aging (Fig. 3a). In contrast, we found that the cytoplasmic translation activity in fibroblasts remained constant, as shown by [ 35 S]methionine incorporation into nuclear DNA-encoded polypeptides in the presence of chloramphenicol to inhibit mitochondrial translation. All fibroblast lines from both fetal and aged donors showed similar [ 35 S]methionine incorporation activity (Fig. 3b), suggesting that agerelated reduction of translation activity was limited to mitochondria. Therefore, the observed age-related decrease of COX activity could be due at least partly to reduction of mitochon-drial translation activity.
Since mutations in both nuclear and mitochondrial genomes contribute to reduction of mitochondrial translation activity (21,22), we examined which genome was responsible for the age-related, mitochondria-specific reduction. For this, we used our recently developed procedure for delivery of pure normal nuclear genomes from 0 HeLa cells to fibroblasts by isolating nuclear hybrids (13). As 0 HeLa cells have been shown to possess no mtDNA (6) and to be resistant to both 6-thioguanine and Oua, HeLa nuclei free from mtDNA could be introduced into fibroblasts simply by fusion of fibroblasts with 0 HeLa cells followed by cultivation in selective medium with Oua ϩ HAT (Table I). Oua and HAT were used to exclude unfused parental fibroblasts and 0 HeLa cells, respectively (see "Materials and Methods"). Accordingly, the nuclear genome of the nuclear hybrids was derived from both parental 0 HeLa cells and fibroblasts, whereas the mitochondrial genome was exclusively from fibroblasts.
First, it was necessary to show unambiguously that our nuclear genome delivery system did not exclude clones expressing complete respiration deficiency, even though our system does not have to use selective pressure upon mitochondrial respiratory function for removal of parental 0 HeLa cells. To demonstrate the reliable isolation of respiration-deficient clones, we carried out nuclear transfer from 0 HeLa cells to respiration-deficient fibroblasts containing 90% mtDNA with a pathogenic mutation (A to G) in tRNA Ile at 4,269 derived from a patient with fatal cardiomyopathy (14,15). Then we examined the content of the mtDNA with the tRNA Ile 4,269 mutation in all 12 nuclear hybrid clones by analysis of the SspI restriction pattern of the PCR products as described previously (15). The results showed that seven nuclear hybrid clones contained more than 95% mutant mtDNA (Table I). We compared mitochondrial translation activity by [ 35 S]methionine incorporation into the polypeptides synthesized in mitochondria using clones containing predominantly the mutant or wild-type mtDNA. Fig. 4 shows that [ 35 S]methionine incorporation into all 13 polypeptides encoded by mtDNA was reduced in clones NHCM1 and 3 containing more than 95% pathogenic mutant mtDNA to the level in 0 HeLa cells, resulting in overall loss of COX activity (Fig. 5). These results proved the reliability of our nuclear delivery system for isolating clones with complete respiration deficiency.
Using this procedure, we introduced pure HeLa nuclei into the fibroblast lines TIG3S (fetal), TIG106 (80 years old), and TIG102 (97 years old). After selection with Oua ϩ HAT, 12 nuclear hybrid clones growing in the selective medium were isolated from each fusion mixture (Table I). The possibility that the clones we isolated were hybrids between the fibroblasts and contaminating presumptive revertant 0 HeLa cells with recovered HeLa mtDNA was excluded by the fact that no HeLa mtDNA was present in these nuclear hybrid clones, as demonstrated by HhaI digestion of the PCR products (data not shown), a procedure that can distinguish HeLa mtDNA from other human mtDNAs, as described previously (23). Therefore, the mitochondrial genome of these clones was derived exclusively from the fibroblasts.
Using these nuclear hybrid clones, we compared mitochondrial translation activity by analysis of [ 35 S]methionine incorporation into mitochondrially synthesized polypeptides. As shown in Fig. 4, all nuclear hybrid clones showed similar mitochondrial translation activities irrespective of whether their mtDNA was derived from fetal or aged subjects, suggesting that all mtDNA-encoded factors necessary for the translation in mitochondria, such as 22 mitochondrial tRNAs and 2 rRNAs, were intact in fibroblasts from the aged subjects. Then, we compared COX activities and found that these phenotypes were also completely restored by the introduction of HeLa nuclei (Fig. 5). Similar results were obtained on comparison of oxygen consumption rate (data not shown). These observations suggest that accumulation of nuclear-recessive mutations of factors involved in mitochondrial translation was responsible for the defects but that all mtDNA-encoded factors necessary for the formation of functional respiration complexes as well as those necessary for the mitochondrial translation were intact. Therefore, mtDNA in the fibroblasts of aged subjects did not contribute to the observed age-related decline of mitochondrial respiratory function in the aged fibroblasts.
Recently, very small amounts of mtDNA molecules with large scale deletion mutations that are not detectable by Southern blot analysis were observed in human (9, 10) and mouse brain (24,25) by the PCR technique. Since a common mutant mtDNA with a 4,977-bp deletion, ⌬mtDNA 4977 , was found to accumulate preferentially in human brain with increase in age  (9, 10), we examined using the PCR technique whether ⌬mtDNA 4977 also accumulated in fibroblast lines from aged subjects. We found that the 392-bp fragment amplified from ⌬mtDNA 4977 was not detectable in DNA samples of all six fibroblast lines from the aged donors and from fetuses (Fig. 6a). On the other hand, when the sensitivity of the amplification conditions was significantly increased, a 279-bp fragment derived from a deletion mutant mtDNA other than ⌬mtDNA 4977 was observed in one fetal fibroblast line (Fig. 6b). Sequence analysis showed that the deletion was 5,090 bp long with a break point from nucleotide positions 8,450 to 13,541, and that the deletion was flanked by a 4-bp direct repeat (5Ј-AATAT-3Ј). Thus, even if a very small amount of deletion mutant mtDNA is present in fibroblast lines, it is not associated with aging. DISCUSSION Since the discovery of the intrinsic limitation for population doubling in cultured human diploid fibroblasts (26), they have been used extensively as models for investigating in vitro cellular aging (27,28). However, decrease of mitochondrial respiratory function has not been observed in human diploid fibroblasts during in vitro cellular aging, i.e. during increase of their population doubling level of fibroblast cultivation (29) but has been observed during aging in vivo, i.e. with age of the fibroblast donors (11). In fact, increase of the population doubling level did not affect both COX and mitochondrial translation activities in fibroblast lines from fetus and aged subjects (11). This seems consistent with the observation of Goldstein et al. (29) that there was no gross deficit in energy metabolism at increased population doubling level when fibroblasts from the same donor with different population doubling level were com-pared. The apparent discrepancy between in vivo and in vitro aging could be attributed to the difference of their time scales; in vitro cellular aging lasted only for several months, whereas in vivo aging lasted up to 70 -100 years. Therefore, the agerelated mitochondrial dysfunction observed in human skin fibroblasts is the phenomenon restricted to in vivo aging.
In this study, we showed that this in vivo aging-related mitochondrial dysfunction could be due at least partly to reduction of mitochondrial translation activity. We then examined whether similar reduction could be observed in cytoplasmic translation activity in fibroblasts from the aged subjects and found that the age-related reduction was limited to translation in mitochondria (Fig. 3, a and b). Then, to determine which genome, mitochondrial or nuclear genome, is responsible for this age-related, mitochondria-specific defects, pure nuclear transfer was carried out from 0 HeLa cells to fibroblast lines from fetal and aged subjects. Results showed that the agerelated, mitochondria-specific defects observed in human skin fibroblasts were due to nuclear-recessive mutations of the factors involved in mitochondrial translation, although we could not completely rule out the possibility of contribution of nonnuclear DNA-and non-mtDNA-encoded factors in 0 HeLa cells to the correction of the age-related defects.
We previously reported that age-related reduction of COX activity in human skin fibroblasts inherited in a nuclear-recessive way, based on the observation that the reduction was restored by the introduction of pure HeLa nuclei (11). However, in our previous work we used only one nuclear hybrid clone isolated by the fusion of aged fibroblasts with 0 HeLa cells and did not examine positive control clones using fetal fibroblasts (11). Therefore, it was possible that we may have picked up by chance a respiration-competent clone or that the apparent restoration of COX activity by the introduction of HeLa nuclei was not sufficient to exclude the involvement of mtDNA somatic mutations in age-related mitochondrial dysfunction. Moreover, we did not prove that respiration-deficient clones could be isolated in selection medium with Oua ϩ HAT using respiration-deficient fibroblasts as negative controls.
In this study, pure HeLa nuclear transfers to normal fibroblasts from a fetus were carried out as a positive control and to respiration-deficient fibroblasts from a patient with cardiomyopathy as a negative control. The negative control experiment showed that we could isolate nuclear hybrid clones expressing no mitochondrial translation or COX activity (Figs. 4 and 5). Thus, our system for isolation of nuclear hybrids is suitable for isolating respiration-deficient clones and excludes the possibility of preferential selection of respiratory-competent clones. By this method, we showed that the activities of both mitochondrial translation and respiratory function of fibroblasts from all aged subjects were restored to comparable levels to those of fetal fibroblasts by the introduction of pure HeLa nuclei, suggesting that mtDNA in fibroblasts from aged subjects is functionally intact.
Recently, by isolating cybrid clones using a similar mtDNA transfer system to that which we reported previously (11), Laderman et al. (12) reported contradictory observations, suggesting that mtDNA is involved in age-related mitochondrial dysfunction in human fibroblasts. They claimed the occurrence of age-related accumulation of mtDNA mutations in human fibroblasts based on the observations that 5% cybrid clones with mtDNA imported from fibroblasts of elderly subjects showed slightly lower mitochondrial respiratory function than those from younger subjects. However, they did not prove the presence of any mtDNA mutations in the clones with a lower respiratory function. Moreover, since the other 95% clones with mtDNA from elderly subjects showed comparable respiratory  TIG3S, TIG2S, TIG1, TIG106, TIG107, and TIG102, respectively. The amplified products were separated in 2.5% agarose gels containing ethidium bromide (0.1 g/ml). The fragment of 392 bp is a PCR product amplified from ⌬mtDNA 4977 . a, PCR analysis using two sets of primers. In the first round of PCR, the outer primer set F1 and R1 (see "Materials and Methods") were employed. In the second round of PCR, the inner primer set F3 and R3 were employed. b, PCR analysis using three sets of primers. For detection of a very small amount of deletion mutant mtDNA, the primer set F2 and R2 was used for second round, and then the primer set F3 and R3 was used for the last round. Note that 0 HeLa cells did not show any signals. function to those with mtDNA from younger subjects, their observations could be reinterpreted as showing that most mtDNA molecules in fibroblasts from aged donors do not have more mutations than those from younger subjects. Furthermore, the presence of only 5% clones with a lower respiratory function could hardly explain the age-related shift to 60 -80% reduction of the respiratory function in fetal fibroblasts (Fig. 1). Therefore, the apparent discrepancy between the report of Laderman et al. (12) and ours (11) could be due simply to a difference in the interpretation of observations.
In these mtDNA transfer techniques, however, the selection medium without uridine has to be used for isolation of cybrids to exclude parental 0 cells (11,12). This medium could also exclude cybrids expressing respiration deficiency, resulting in the selective isolation of respiratory-competent cybrids. Therefore, in the present work, we did not use mtDNA transfer techniques but used nuclear transfer techniques that do not eliminate clones expressing complete respiration deficiency (Figs. 4 and 5). Then, we compared both mitochondrial translation activity and mitochondrial respiratory function and showed that no respiration-deficient clones were present among 36 nuclear hybrid clones irrespective of whether their mtDNA were derived exclusively from fetal or aged donors (Fig.  5). These observations completely excluded the possibility that accumulation of mtDNA with somatic mutations plays a role in the age-related respiration defects observed in human skin fibroblasts.
Recently, a deletion mutant ⌬mtDNA 4977 was found to accumulate preferentially in human brain with increase in age (9,10). In this study, however, we showed that no ⌬mtDNA 4977 was detected in mtDNA of any human fibroblast lines from aged or fetal subjects by PCR amplification (Fig. 6a), whereas one fragment derived from a very small amount of mtDNA with a deletion mutation other than ⌬mtDNA 4977 was observed in a fibroblast line from a fetus by PCR amplification using three sets of primers (Fig. 6b). These observations suggest that large scale deletion mutant mtDNA molecules, if they do occur, do not accumulate in mitotic cells with age, partly due to selection against the surviving cells containing these deletion mutants (5). On the other hand, they could be propagated in specific conditions, such as in blood cells of patients with Pearson syndrome (30) and in cybrid clones (6). Although the observations of PCR experiments do not exclude the possibility of accumulations of various other unidentified somatic mutations in the mtDNA populations of fibroblasts from aged subjects, our nuclear transfer experiment completely ruled out the possibility of their involvement, at least in age-related respiration defects in fibroblasts.
We recently proposed the idea that mitochondria and the mitochondrial genome function as a single dynamic cellular unit in living human cells by the presence of exchanging mtDNA and its products between mitochondria (31). This was supported from the evidence for the coexistence and cooperation of mutant HeLa mtDNA with chloramphenicol resistance and mutant mtDNA with a large scale deletion originated from organelles of different cells (32). As mtDNA mutations in different genes can complement each other (32), the presence of various kinds of mutant mtDNA molecules in single cells would not have serious additive influence on mitochondrial respiratory function.
All these considerations indicate that age-related accumulation of somatic mutations in mtDNA, even if it occurs in fibroblasts, is not responsible for the age-related decrease in mitochondrial respiratory function observed in human fibroblasts. As the nuclear genome encodes most mitochondrial proteins including factors necessary for expression of the mitochondrial genome (33) and as the defects were at least in part ascribable to reduction of mitochondrial translation activity, we are now investigating nuclear-coded factors particularly involved in mitochondrial translation to understand the precise mechanisms of the age-related mitochondrial dysfunction in human skin fibroblasts.