Mitochondrial respiratory dysfunction caused by a heteroplasmic mitochondrial DNA mutation blocks cellular reprogramming

Abstract

Mitochondrial dysfunction caused by pathogenic mutations in mitochondrial tRNA genes emerges only when mutant mitochondrial DNA (mtDNA) proportions exceed intrinsic pathogenic thresholds; however, little is known about the actual proportions of mutant mtDNA that can affect particular cellular lineage-determining processes. Here, we mainly focused on the effects of mitochondrial respiratory dysfunction caused by m.3243A>G heteroplasmy in MT-TL1 gene on cellular reprogramming. We found that generation of induced pluripotent stem cells (iPSCs) was drastically depressed only by high proportions of mutant mtDNA (≥90% m.3243A>G), and these proportions were strongly associated with the degree of induced mitochondrial respiratory dysfunction. Nevertheless, all established iPSCs, even those carrying ∼100% m.3243A>G, exhibited an embryonic stem cell-like pluripotent state. Therefore, our findings clearly demonstrate that loss of physiological integrity in mitochondria triggered by mutant mtDNA constitute a roadblock to cellular rejuvenation, but do not affect the maintenance of the pluripotent state.

Introduction

Mitochondrial dysfunction caused by pathogenic mutations in mitochondrial tRNA genes emerges only when mutant mitochondrial DNA (mtDNA) proportions exceed intrinsic pathogenic thresholds (1), and thus, mitochondrial diseases cause a wide spectrum of clinical symptoms that are associated with mitochondrial dysfunction (2,3). Pathogenic mutations of mtDNA-specific tRNA genes can result in loss of various functions at a molecular level (4), and such loss can lead to symptomatic appearance of mitochondrial diseases. Degree of m.3243A>G heteroplasmy, the mtDNA mutation most often responsible for mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes (MELAS) (5,6), frequently determines the trajectory of disease progression and phenotypic severity in affected individuals. The estimated pathogenic threshold level of m.3243A>G heteroplasmy required to induce mitochondrial respiratory dysfunction is ∼90% in artificial trans-mitochondrial cellular systems (7); however, these cancer-cell-line-based models do not faithfully recapitulate patient-specific pathophysiology of affected tissues and organs.

Breakthroughs in cellular reprogramming technology include generation of embryonic stem cell (ESC)-like induced pluripotent stem cells (iPSCs) from various human somatic cells driven by ectopic expression of transcription factors required for pluripotency (8,9). Patient-specific iPSCs that carry mutant mtDNA hold enormous promise for disease modeling and for drug discovery in mitochondrial medicine. Additionally, several recent studies demonstrate that mitochondria are gradually reconstructed to an ESC-like ‘quiescent state’ during cellular reprogramming; this state comprises rounded mitochondrial morphology with poor cristae structure, less active mitochondrial respiration and reduced mtDNA copy number (10–13). Nevertheless, the effects of mitochondrial respiratory dysfunction triggered by mutant mtDNA on cellular lineage-determining processes, including cellular reprogramming or differentiation, still remain unexamined.

In this study, we mainly focused on the effects of mitochondrial respiratory dysfunction caused by m.3243A>G heteroplasmy in MT-TL1 gene on cellular reprogramming and on maintenance of the pluripotent state. We clearly demonstrated that induced mitochondrial respiratory dysfunction caused by over pathogenic threshold level of m.3243A>G heteroplasmy constituted a roadblock to cellular reprogramming, but did not affect maintenance of the pluripotent state.

Results

A wide variety of m.3243A>G heteroplasmy levels among patient-derived primary fibroblasts and corresponding iPSCs

First, we examined whether and how pathogenic levels of m.3243A>G heteroplasmy affected cellular reprogramming and maintenance of the pluripotent state. We selected three patient-derived primary fibroblasts, which differed in m.3243A>G proportions, to generate single-cell-derived fibroblast clones and to determine the distributions of m.3243A>G heteroplasmy at the single-cell level (Fig. 1A). We found that primary fibroblasts of each patient comprised multiple cell populations, and each fibroblast clone carried different m.3243A>G proportions. Patient 1 showed quasi-homoplasmy (>95% of all clones possessed only mutant mtDNA); in contrast, Patients 2 and 3 showed a wide variety of m.3243A>G proportions (Fig. 1B). Such cellular heterogeneity was also confirmed by cytochemical staining of cytochrome c oxidase, an indicator of mitochondrial respiration activity (Supplementary Material, Fig. S1). Patient-derived iPSCs were then successfully generated from primary fibroblasts of all patients, and the distributions of m.3243A>G heteroplasmy for these iPSCs closely reflected those of the corresponding primary fibroblasts (Fig. 1C). Notably, there was no correlation between cellular reprogramming period and m.3243A>G proportions (Fig. 1D).

Figure 1.

Wide variation in m.3243A>G heteroplasmy levels among patient-derived primary fibroblasts and corresponding iPSCs. (A) Experimental design used to measure heterogeneity in m.3243A>G proportions among patient-derived primary fibroblasts and corresponding iPSCs. (B) Histograms showing cellular heterogeneity in m.3243A>G proportion among single-cell-derived fibroblast clones from cultures of primary fibroblasts from each patient. The averaged m.3243A>G heteroplasmy levels of all fibroblast clones are expressed as a mean with a standard deviation. (C) Histograms showing heterogeneity in m.3243A>G proportion among iPSCs generated from primary fibroblasts of each patient. (D) Relationship between m.3243A>G proportion and reprogramming period required for iPSCs; data for each patient are shown.

No drastic segregation of m.3243A>G heteroplasmy levels during fibroblast proliferation and iPSC generation

Next, we asked whether m.3243A>G heteroplasmy often changed during fibroblast proliferation, cellular reprogramming, or both. We used several single-cell-derived clonal fibroblast lines derived from Patients 2 and 3, which differed in m.3243A>G proportions, to estimate the distribution of m.3243A>G heteroplasmy and to generate iPSCs (Fig. 2A). Note that Patient 1 was excluded from these experiments because of its poor growth ability of fibroblast clones. Clonal expansion apparently did not affect cellular proliferation potential of Patients 2 and 3, which was confirmed by cytochemical staining of senescence-associated β-galactosidase, a representative marker of cellular senescence (Supplementary Material, Fig. S2). During ∼20 rounds of fibroblast doubling, m.3243A>G proportions among fibroblast subclones tested gradually segregated within a very narrow-range encompassed by the original range of m.3243A>G heteroplasmy (Fig. 2B). All iPSC colonies generated from several clonal fibroblast lines of different patients were sampled to determine m.3243A>G proportions and showed no significant alteration in the distributions of m.3243A>G heteroplasmy compared with the corresponding clonal fibroblast lines (Fig. 2C). Therefore, we concluded that m.3243A>G proportions did not change during cellular reprogramming (Fig. 2D).

Figure 2.

Absence of drastic segregation of m.3243A>G heteroplasmy levels during fibroblast proliferation and iPSC generation. (A) Experimental design used to identify segregation of m.3243A>G heteroplasmy levels during fibroblast proliferation and iPSC generation. (B) Histograms showing the distribution of m.3243A>G proportions after ∼20 additional doublings from single-cell-derived fibroblast clones; data originated from Patients 2 and 3 are shown. For each histogram, the averaged m.3243A>G heteroplasmy level of all fibroblast clones is expressed as a mean with a standard deviation. (C) Histograms showing stochastic variation in m.3243A>G proportions among iPSCs generated from single-cell-derived cloned fibroblast lines; data originated from Patients 2 and 3 are shown. (D) Graphical summary showing our model of segregation of m.3243A>G heteroplasmy levels during fibroblast proliferation and iPSC generation.

Induced mitochondrial respiratory dysfunction and depressed iPSC generation caused by m.3243A>G heteroplasmy levels above the pathogenic threshold

We also addressed the relationship between iPSC generation and mitochondrial respiratory dysfunction caused by different degrees of m.3243A>G heteroplasmy. We used single-cell-derived clonal fibroblast lines derived from Patients 2 and 3, which differed in m.3243A>G proportions, to analyze mitochondrial respiratory function and to generate iPSCs (Fig. 3A). Again, Patient 1 was unsuitable for clonal expansion of fibroblasts due to its deteriorated growth potential, and only clonal fibroblast lines with similar cellular proliferation potential were examined to minimally reduce the confounding effects of cellular senescence on our experiments of mitochondrial respiratory function and of iPSC generation. The intrinsic pathogenic threshold level of m.3243A>G heteroplasmy with regard to mitochondrial respiratory function was ∼90% for both patients (Fig. 3B), and this finding was consistent with those from artificial trans-mitochondrial cellular systems (7). Nevertheless, no obvious difference in intracellular ATP level was observed among clonal fibroblast lines tested, most likely because of upregulated anaerobic glycolysis and the resulting compensatory ATP synthesis (Supplementary Material, Fig. S3). Most surprisingly, clonal fibroblast lines carrying ≥90% m.3243A>G proportions also had a propensity for drastically depressed iPSC generation (Fig. 3C). Although the reprogramming period was extended for up to 6 weeks, only a few ESC-like colonies were barely able to emerge from such clonal fibroblast lines of both patients. These results strongly indicated that induced mitochondrial respiratory dysfunction caused by above pathogenic threshold level of m.3243A>G heteroplasmy actually blocked cellular reprogramming (Fig. 3D).

Figure 3.

Induced mitochondrial respiratory dysfunction and depressed iPSC generation caused by m.3243A>G heteroplasmy levels above the pathogenic threshold. (A) Experimental design used to determine the relationship between mitochondrial respiratory dysfunction triggered by m.3243A>G heteroplasmy and iPSC generation. (B) Several single-cell-derived cloned fibroblast lines each carrying different m.3243A>G heteroplasmy levels were used to assess the relationship between m.3243A>G proportion and mitochondrial respiratory complexes activities; data originated from Patients 2 and 3 are shown. Measurements were performed in triplicate; data were expressed as means with standard deviations. (C) Several single-cell-derived cloned fibroblast lines each carrying different m.3243A>G heteroplasmy levels were used to assess the relationship between m.3243A>G proportion and cellular reprogramming efficiency; data originated from Patients 2 and 3 are shown. Insets indicate marked depression of iPSC generation only for clonal lines with ≥90% m.3243A>G heteroplasmy (3 cloned fibroblast lines), but not for those with <90% m.3243A>G heteroplasmy (5 cloned fibroblast lines). Statistical significance was evaluated by unpaired, two-tailed t-test. (D) Graphical summary showing the relationship between induced mitochondrial respiratory dysfunction and depressed iPSC generation caused by over pathogenic threshold level of m.3243A>G heteroplasmy.

Characterization of the established iPSC lines each carrying different m.3243A>G proportions

We evaluated the effects of m.3243A>G proportions on maintenance of the pluripotent state in the established iPSC lines. All iPSC lines tested, even those carrying ∼100% m.3243A>G proportions, showed typical ESC-like characteristics with regard to pluripotency markers expression (Fig. 4A) and embryoid body (EB)-mediated in vitro differentiation into three germ layers (Fig. 4B). We then analyzed enzymatic activities of mitochondrial respiratory complexes in patient-derived iPSC lines. A trend quite similar to the corresponding primary fibroblasts (Patient 1) and clonal fibroblast lines (Patients 2 and 3), patient-derived iPSC lines derived from all patients those carrying ∼100% m.3243A>G proportions also exhibited low mitochondrial respiratory activities (Fig. 4C). Interestingly, patient-derived iPSC lines derived from Patients 2 and 3 those carrying 60–70% m.3243A>G proportions (Heteroplasmy-1 and Heteroplasmy-2 of both patients) displayed significantly decreased mitochondrial respiration capacity than the corresponding clonal fibroblast lines those carrying similar m.3243A>G proportions, which were below pathogenic threshold level of m.3243A>G heteroplasmy (see also Fig. 3B). These results probably indicated that patient-derived iPSC lines acquired an ESC-like ‘quiescent’ mitochondrial state during cellular reprogramming (10–13), and therefore, we concluded that m.3243A>G proportions minimally influenced maintenance of the pluripotent state in patient-derived iPSC lines.

Figure 4.

Characterization of the established iPSC lines each carrying different m.3243A>G proportions. (A) Representative images of the established iPSC lines; OCT4 (red), NANOG (red), TRA-1-60 (green), TRA-1-81 (green). (B) Representative images of the embryoid body (EB)-mediated in vitro spontaneous differentiation; TUJ1 (ectoderm, red), αSMA (mesoderm, red), AFP (endoderm, red). Cell nuclei were co-stained with Hoechst 33342 (blue). (C) Influence of m.3243A>G proportions on enzymatic activities of mitochondrial respiratory complexes in patient-derived iPSC lines. Measurements were performed in triplicate; data were expressed as means with standard deviations. Statistical significance was evaluated by unpaired, two-tailed t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant compared with the corresponding primary fibroblasts (Patient 1) and clonal fibroblast lines (Patients 2 and 3).

Clone-dependent fluctuation behavior of m.3243A>G proportions during iPSC self-renewal

We were surprised that some iPSC lines derived from Patients 2 and 3 showed gradual elevation of m.3243A>G proportions without drastic changes in relative mtDNA abundance during iPSC proliferation, while other iPSC lines derived from Patients 2 and 3 stably maintained their intrinsic m.3243A>G proportions (Fig. 5A). We performed mtDNA mutation analysis to assess the influence of newly introduced mtDNA variant(s) on elevated m.3243A>G proportions; however, all iPSC lines were completely identical to corresponding primary fibroblasts with regard to entire mtDNA sequence of each patient (data not shown). We also evaluated gene expression levels to estimate time-dependent alteration of mitochondrial biogenesis and mtDNA replicative turnover during long-term maintenance of iPSCs. As the passage number increased, mitochondrial biogenesis and mtDNA replicative turnover were upregulated in iPSC lines those exhibiting elevated m.3243A>G proportions and, vice versa, were downregulated in iPSC lines those exhibiting maintained m.3243A>G proportions (Fig. 5B). Overall, Patient 3 showed a clear propensity for this phenomenon than Patient 2. These results possibly implied that more frequent mitochondrial biogenesis and more rapid mtDNA replicative turnover might induce the accelerated segregation of m.3243A>G heteroplasmy toward homoplasmic state (Fig. 5C).

Figure 5.

Clone-dependent fluctuation behavior of m.3243A>G proportions during iPSC self-renewal. (A) Time-dependent alteration of m.3243A>G proportions and mtDNA copy number during long-term maintenance of iPSC lines (Heteroplasmy-1 and Heteroplasmy-2), each of which originated from Patients 2 and 3. Measurements were performed in triplicate; data were expressed as means with standard deviations. (B) Time-dependent alteration of gene-expression levels associating with mitochondrial biogenesis (PPARGC1A and NRF1), mtDNA replication (POLG1) and mtDNA degradation (DNASE2) during long-term maintenance of iPSC lines (Heteroplasmy-1 and Heteroplasmy-2), each of which originated from Patients 2 and 3. Measurements were performed in triplicate; data were expressed as means with standard deviations. (C) Graphical summary showing two distinct fluctuation behavior of m.3243A>G heteroplasmy levels during iPSC self-renewal. (Left) Significant fluctuations of m.3243A>G proportions. The detailed molecular machinery required for the acquisition of replicative advantage by either wild-type or mutant mtDNA molecules remains unclear. (Right) Stably maintained intrinsic heteroplasmy levels.

Cell-type-specific and patient-specific variation in the molecular pathogenic potential of mutant m.3243A>G

We further investigated cell-type-specific molecular pathogenic potential of mutant m.3243A>G. We used only clonal fibroblast lines and iPSC lines derived from Patients 2 and 3 those carrying ∼100% m.3243A>G proportions (Fig. 6A). In comparison with control samples, relatively high mtDNA copy number (Fig. 6B) and significantly accumulated RNA19 (Fig. 6C), a mitochondrial RNA processing intermediate associating with MELAS pathogenesis (14,15), were observed in both clonal fibroblast lines and iPSC lines. Such RNA19 accumulation exhibited a trend quite similar to that previously reported for MELAS patients' tissues (15). Furthermore, most parts of mtDNA transcripts levels were markedly lower in clonal fibroblast lines (Fig. 6D), and these results strongly associated with our presenting data on induced mitochondrial respiratory dysfunction and on depressed iPSC generation (see also Fig. 3B and C). Overall, Patient 2 showed more severe molecular pathogenic phenotypes than Patient #3, and iPSCs possessed a less damaged mitochondrial status by mutant m.3243A>G than fibroblasts (see also Fig. 4C). Therefore, we concluded that the molecular pathogenic influences of mutant m.3243A>G actually changed during cellular lineage-commitment processes along with the degree of mitochondrial maturation, and between the two patient-derived cell lines, there must be some differences in mtDNA polymorphic variation and/or in nuclear DNA genetic background (Fig. 6E).

Figure 6.

Cell-type-specific and patient-specific variation in the molecular pathogenic potential of m.3243A>G mutation. (A) Experimental design used to identify differences in the molecular pathogenic potential of ∼100% mutant m.3243A>G between cloned fibroblast lines and iPSC lines. (B) Relative mtDNA abundance in cloned fibroblast lines and iPSC lines from Patients 2 and 3 that carried ∼100% mutant m.3243A>G. Measurements were performed in triplicate; data were expressed as means with standard deviations. Statistical significance was evaluated by unpaired, two-tailed t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant compared with control fibroblasts or iPSCs. (C) RNA19 levels in cloned fibroblast lines and iPSC lines from Patients 2 and 3 that carried ∼100% mutant m.3243A>G. Measurements were performed in triplicate; data were expressed as means with standard deviations. Statistical significance was evaluated by unpaired, two-tailed t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant compared with control fibroblasts or iPSCs. (D) mtDNA transcripts levels in cloned fibroblast lines and iPSC lines from Patients 2 and 3 that carried ∼100% mutant m.3243A>G. Measurements were performed in triplicate; data were expressed as means with standard deviations. Statistical significance was evaluated by unpaired, two-tailed t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant compared with control fibroblasts or iPSCs. (E) Graphical summary showing the differences between Patients 2 and 3 with regard to cell-type-specific molecular pathogenic potential of mutant m.3243A>G.

Discussion

Mitochondria play some crucial roles in cellular physiology, and damaged mitochondria can disrupt cellular homeostasis, accelerate senescence and induce apoptosis. Each of these cellular perturbations closely links to clinical symptoms associated with mitochondrial diseases. Here, we clearly demonstrate that induced mitochondrial respiratory dysfunction caused by over pathogenic threshold level of mtDNA heteroplasmy is an obstacle to cellular reprogramming, but do not affect maintenance of the pluripotent state. Our findings are the first evidence that physiologically damaged mitochondria can directly affect cellular lineage-determining processes. Other recent studies have also reported generation of patient-derived iPSCs and their derivatives carrying various heteroplasmic mtDNA mutations (16–19); however, these reports did not adequately demonstrate whether these cells actually exhibited mutation-specific mitochondrial pathophysiology. On the other hand, directed differentiation of patient-derived iPSCs into specific cell types representing mitochondrial respiratory dysfunction would be an important technical step for studying overall pathomechanisms in various mitochondrial diseases triggered by mtDNA heteroplasmy. We clearly demonstrate cell-type-specific variation in the molecular pathogenic potential of mutant m.3243A>G in a patient-specific manner, and also, Hämäläinen et al. (19) have displayed the possibility for in vitro recapitulation of neuronal pathophysiology in MELAS using iPSC-derived neurons. Thus, an iPSC-based modeling of disease progression and phenotypic severity in affected tissues and organs would greatly facilitate mitochondrial medicine.

Rapid and drastic segregation of mtDNA heteroplasmy was thought to occur during cellular reprogramming (20), and recently, some researchers have noted the possibility of such segregation behavior (16,19). Nevertheless, previously published data did not sufficiently document cellular heterogeneity in mtDNA heteroplasmy among patient-derived primary cell lines tested. Here, we used several clonal fibroblast lines derived from primary fibroblasts of different patients and clearly documented that significant segregation of m.3243A>G proportions did not occur at least during iPSC generation. We also observed that m.3243A>G heteroplasmy gradually segregated within a narrow-range during fibroblast proliferation. This segregation can be explained by the ‘mitotic mtDNA segregation model’ (21) and is probably due to autonomic mtDNA replication and mitochondrial fusion/fission, independently from cell proliferation. This model should be applicable to self-renewal of iPSCs; however, we were surprised that some iPSCs actually exhibited significant elevation in m.3243A>G proportions without drastic changes in mtDNA copy number during iPSC proliferation. Note that a similar phenomenon was previously reported in cancer-cell-line-based artificial trans-mitochondrial cellular systems (22). Although we found that more frequent mitochondrial biogenesis and more rapid mtDNA replicative turnover might contribute to the accelerated segregation of m.3243A>G heteroplasmy toward homoplasmic state, further studies are needed to clarify the detailed molecular machinery responsible for acquisition of this replicative advantage of mutant mtDNA molecules.

In conclusion, we clearly demonstrate that cellular rejuvenation is strongly inhibited by loss of physiological integrity in mitochondria triggered by mutant mtDNA. Our finding may also imply that an appropriate mitochondrial remodeling is essential for bona fide cellular reprogramming. Therefore, cellular models representing various mitochondrial dysfunctions might also shed light on adaption of the intracellular organelle to an ESC-like environment as differentiated somatic cells acquire pluripotency.

Materials and Methods

Patients

This study was approved by our institutional review board and was stringently conducted according to ethical principle ‘Declaration of Helsinki’. Each patient's biopsy was performed for diagnostic purposes only after we received a signed informed consent with permission to study patient-derived iPSCs.

Fibroblast culture

Primary fibroblasts were established from patient-derived skin biopsies via a standard protocol. Patient-derived fibroblasts were maintained in DMEM/F12 (Gibco) supplemented with 10% FBS (Gibco), 100 units/ml penicillin (Gibco), 100 µg/ml streptomycin (Gibco). During establishment of primary fibroblasts, 0.5 µg/ml MC210 (DS Pharm) as a mycoplasmacidal reagent and 2.5 µg/ml fungizone (Gibco) as a fungicidal reagent were also added to culture medium.

For fibroblast cloning experiment, patient-derived fibroblasts (0–1 cells/100 µl/well) were seeded onto 96-well culture plates and were maintained at 37°C under humidified atmosphere of 5% CO₂. Each cloned fibroblast line was collected to determine m.3243A>G heteroplasmy levels, to analyze mitochondrial and cellular physiological functions, to generate iPSCs, or to use for subcloning. Cells from cloned fibroblast lines (200 cells/mm²) were also seeded onto 4-well culture slides and were maintained at 37°C under humidified atmosphere of 5% CO₂. After 3 days in culture, SA-β-gal staining kit (Cell Signaling) was used according to manufacturer's instructions along with an optical microscope (BX50 System; Olympus) to evaluate cellular senescence.

Cytochemical COX staining

Patient-derived fibroblasts (200 cells/mm²) were seeded onto 4-well culture slides and were maintained at 37°C under humidified atmosphere of 5% CO₂. After 3 days in culture, cells were stained with reaction buffer [pH 5.5; 100 mM sodium acetate, 0.1% MnCl₂, 0.001% H₂O₂, 10 mM diaminobenzidine] at 37°C for 1 h and were incubated with 1% CuSO₄ at 37°C for 5 min. Cell nuclei were co-stained with hematoxylin. Stained cells were rinsed, fixed, and dehydrated, and samples were then observed under an optical microscope (BX50 System; Olympus).

Generation of patient-derived iPSCs with episomal vector

Patient-derived iPSCs were generated using episomal vectors as described previously (23) with modifications: briefly, each 1 µg of episomal plasmid vectors (Plasmid #27077, #27078, #27080; Addgene) were electroporated into patient-derived fibroblasts (5 × 10⁵ cells) with an electroporator (Neon; Invitrogen). Transformed patient-derived fibroblasts (1 × 10⁵ cells) were reseeded onto mouse embryonic fibroblasts (MEF; ReproCELL) 4 days after electroporation. The next day, culture medium was replaced with primate ESC culture medium (ReproCELL) supplemented with 10 ng/ml bFGF (ReproCELL), 100 units/mL penicillin (Gibco), 100 µg/ml streptomycin (Gibco), and transformed patient-derived fibroblasts were maintained at 37°C under humidified atmosphere of 5% CO₂. Emergent colonies with ESC-like morphology were manually picked up to establish patient-derived iPSCs, and these iPSCs were expanded for long-term maintenance as described elsewhere (8).

To determine the relationship between m.3243A>G proportions in each iPSC and cellular reprogramming period (Fig. 1D), we set the minimum period necessary for emergent colonies with ESC-like morphology to become over 1 mm²-size as ‘the reprogramming period’. Each primary iPSC colony generated in this experiment was sampled to extract DNA for measurement of m.3243A>G heteroplasmy level.

Characterization of patient-derived iPSCs

Immunocytochemical staining of patient-derived iPSCs to detect pluripotency markers was performed as follows: briefly, cultured and harvested patient-derived iPSCs were transferred onto MEF-seeded six-well culture plates and were maintained at 37°C under humidified atmosphere of 5% CO₂. After 3 days in culture, iPSCs were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and then blocked at room temperature for 30 min. Staining with fluorophore-conjugated primary antibody was performed at room temperature for 2 h. Fluorophore-conjugated primary antibodies used were as follows: 5 µg/ml Cy3-conjugated anti-OCT4 (Millipore), 5 µg/ml Cy3-conjugated anti-NANOG (Millipore), 5 µg/ml AlexaFluor 488-conjugated anti-TRA-1-60 (Millipore), 5 µg/ml AlexaFluor 488-conjugated anti-TRA-1-81 (Millipore). Cell nuclei were co-stained with 0.5 µg/ml Hoechst 33342 (Molecular Probes). Stained iPSCs were observed under a fluorescent microscope (IX71 System; Olympus).

Successive in vitro differentiation of iPSCs into EB-mediated three germ layers was monitored via immunocytochemical staining as follows: briefly, cultured and harvested patient-derived iPSCs were transferred onto ultra-low-adherent culture dishes (HydroCell; CellSeed) and were maintained in primate ESC culture medium without bFGF at 37°C under humidified atmosphere of 5% CO₂. After 7 days in floating culture, emergent EBs were transferred onto gelatin-coated six-well culture plates and were maintained in primate ESC culture medium without bFGF at 37°C under humidified atmosphere of 5% CO₂. After 14 additional days in adherent culture, EBs were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 and then blocked at room temperature for 30 min. Primary antibody probing was performed at room temperature for 2 h. Primary antibodies used were as follows: 5 µg/ml anti-TUJ1 (Abcam), 5 µg/ml anti-αSMA (Abcam), 5 µg/ml anti-AFP (Abcam). Secondary antibody probing was performed with 2.5 µg/ml Alexa Fluor 568 (Molecular Probes) at room temperature for 1 h. Cell nuclei were co-stained with 0.5 µg/ml Hoechst 33342 (Molecular Probes). Stained EBs were observed under a fluorescent microscope (IX71 System; Olympus).

Analysis of mtDNA mutation

Extracted DNA (1 ng) was used as template for quantitative PCR with TaqMan Universal PCR Master Mix kit (Applied Biosystems) according to the manufacturer's instructions. A sequence detection system (ABI PRISM 7900HT; Applied Biosystems) was used, and a calibration curve was created using several copy-number standards with plasmids containing the amplified mtDNA fragments of either wild-type or mutant m.3243A>G. Primers and allele-specific TaqMan probes used are listed in Supplementary Material, Table S1.

Long PCR-based whole mtDNA sequencing was performed as described elsewhere (24) with modifications for each patient to eliminate any adverse results arising from pseudo-sequences in nuclear DNA: briefly, extracted DNA as template (10 ng for iPSCs, 50 ng for fibroblasts) was amplified via mtDNA-specific long-range PCR and the following mtDNA-specific nested PCR with a thermal cycler (GeneAmp PCR System 9700; Applied Biosystems). The amplified mtDNA fragments were sequenced with a DNA analyzer (ABI PRISM 3130xl; Applied Biosystems). The obtained mtDNA sequence data from each patient was compared with the WEB-databases of ‘Human Mitochondrial Genome Database (MITOMAP)’ and ‘Human Mitochondrial Genome Polymorphism (mtSNP)’ (25) to find any genetic variants.

Analysis of mtDNA copy number

Extracted DNA (1 ng) was used as template for quantitative PCR with Power SYBR Green PCR Master Mix kit (Applied Biosystems) according to the manufacturer's instructions. A Sequence detection system (ABI PRISM 7900HT; Applied Biosystems) was used to measure mtDNA copy number per cell. The averaged threshold cycle number for mtDNA genes (MT-ND1 and MT-CYB) and for nuclear DNA genes (FBXO15 and GAPDH) were adopted for ΔΔC_T-based relative quantification. Primers used are listed in Supplementary Material, Table S1.

Quantitative PCR for gene-expression analysis

Reverse transcription was performed with PrimeScript RT Master Mix kit (TaKaRa Bio) according to the manufacturer's instructions. Quantitative PCR was performed with Power SYBR Green PCR Master Mix kit (Applied Biosystems) according to the manufacturer's instructions. After reverse transcription of extracted total RNA, total cDNA (10 ng) was used as template for quantitative PCR with a sequence detection system (ABI PRISM 7900HT; Applied Biosystems). The averaged threshold cycle number for housekeeping genes (ACTB, PPIA, TBP) were adopted for ΔΔC_T-based relative quantification. Primers used are listed in Supplementary Material, Table S1.

Measurement of mitochondrial respiratory complex activity

For cloned fibroblast lines, mitochondrial respiratory complex activity was measured with Complex I Human Enzyme Activity Microplate Assay kit (Abcam) and with Complex IV Human Enzyme Activity Microplate Assay kit (Abcam) according to the manufacturer's instructions, respectively. Cell extracts from fibroblasts or from iPSCs (150 μg for complex I, 50 μg for complex IV) were used to measure time-dependent absorbance alterations on a multi-well plate reader (SPECTROstar Nano; BMG Labtech).

Measurement of intracellular ATP level

For cloned fibroblast lines, intracellular ATP level was measured with CellTiter-Glo Luminescent Cell Viability Assay kit (Promega) according to the manufacturer's instructions. Cultured and harvested fibroblasts were used for measurements. Intracellular ATP level was monitored on a chemiluminescent multi-well plate reader (Centro LB 960; Berthold Technologies). A calibration curve was created with several ATP concentration standards, and the obtained data were normalized relative to the number of fibroblasts used.

Measurement of extracellular lactate level

For cloned fibroblast lines, extracellular lactate level was measured with Lactate Assay kit (BioVision) according to the manufacturer's instructions. Supernatants from cultured fibroblasts were filtered with Amicon 10kD Molecular Weight Filters (Millipore) and were used for measurements. Extracellular lactate level was monitored on a multi-well plate reader (SPECTROstar Nano; BMG Labtech). A calibration curve was created with several lactate concentration standards, and the obtained data were normalized relative to the number of fibroblasts used.

Supplementary Material

Supplementary Material is available here.

Authors’ Contributions

H.H. conceived the study. H.H. and Y.G. supervised the study. M.Y. and H.H. designed experiments. M.Y., S.O. and Y.O. performed experiments. M.Y. and H.H. analyzed and interpreted data. M.Y., H.H. and Y.G. wrote the manuscript.

Funding

This study was financially supported in part by a Grant-in-Aid for Young Scientists (B) (Grant No. 25860732 to M.Y.) from the Japan Society for the Promotion of Science; a Grant-in-Aid for Research on Intractable Diseases (Mitochondrial Disorder) from the Ministry of Health, Labour and Welfare, Japan; and AMED-CREST from the Japan Agency for Medical Research and Development.

Acknowledgements

We deeply appreciate all patients and their families for participating in this study. We also thank Junko Takei, Yumiko Ondo, Yoshie Sawano and Miyuki Kanazawa (NCNP, Japan) for helpful experimental assistance.

Conflict of Interest statement. None declared.