Mitophagy-driven mitochondrial rejuvenation regulates stem cell fate

Our understanding on how selective mitochondrial autophagy, or mitophagy, can sustain the archetypal properties of stem cells is incomplete. PTEN-induced putative kinase 1 (PINK1) plays a key role in the maintenance of mitochondrial morphology and function and in the selective degradation of damaged mitochondria by mitophagy. Here, using embryonic fibroblasts from PINK1 gene-knockout (KO) mice, we evaluated whether mitophagy is a causal mechanism for the control of cell-fate plasticity and maintenance of pluripotency. Loss of PINK1-dependent mitophagy was sufficient to dramatically decrease the speed and efficiency of induced pluripotent stem cell (iPSC) reprogramming. Mitophagy-deficient iPSC colonies, which were characterized by a mixture of mature and immature mitochondria, seemed unstable, with a strong tendency to spontaneously differentiate and form heterogeneous populations of cells. Although mitophagy-deficient iPSC colonies normally expressed pluripotent markers, functional monitoring of cellular bioenergetics revealed an attenuated glycolysis in mitophagy-deficient iPSC cells. Targeted metabolomics showed a notable alteration in numerous glycolysis- and TCA-related metabolites in mitophagy-deficient iPSC cells, including a significant decrease in the intracellular levels of α-ketoglutarate -a key suppressor of the differentiation path in stem cells. Mitophagy-deficient iPSC colonies exhibited a notably reduced teratoma-initiating capacity, but fully retained their pluripotency and multi-germ layer differentiation capacity in vivo. PINK1-dependent mitophagy pathway is an important mitochondrial switch that determines the efficiency and quality of somatic reprogramming. Mitophagy-driven mitochondrial rejuvenation might contribute to the ability of iPSCs to suppress differentiation by directing bioenergetic transition and metabolome remodeling traits. These findings provide new insights into how mitophagy might influence the stem cell decisions to retain pluripotency or differentiate in tissue regeneration and aging, tumor growth, and regenerative medicine.

abnormal mitophagy has a deleterious impact on cell homeostasis and may lead to cell death, which causally contributes to the pathogenesis of degenerative disorders [10][11][12][13][14][15][16]. Although our knowledge of mitophagy in somatic cell physiology is extensive, the role of mitophagy in the physiology of stem cells, which have the unique ability to self-renew and differentiate into various cell types, is less understood. Thus, while mitophagy is believed to play a pivotal role in stem cell functions during aging, tissue regeneration, and cancer [17][18][19][20][21][22][23][24], our current understanding on how mitophagy can sustain the archetypal properties of stem cells is rudimentary.
Mitochondria appear to play crucial roles during stemness factor-mediated nuclear reprogramming of somatic cells into induced pluripotent stem cells (iPSCs), a convenient "in a dish" model that allows a comprehensive understanding of stem cell biology. Functional metamorphosis of somatic oxidative phosphorylation into glycolytic metabolism plays a causal role in enabling the reprogramming process of acquisition and maintenance of stemness to occur [25][26][27][28][29][30][31][32][33][34][35]. It is also apparent that the intrinsic metabolic demands that drive reprogramming to stemness involve substantial structural mitochondrial reorganization, transforming mitochondria into a cristae-poor, immature phenotype [36][37][38][39][40][41][42][43][44][45]. Paradoxically, the establishment of induced pluripotency requires a transient and early energy-demanding metabolic state characterized by increased mitochondrial oxidative phosphorylation and hyperactive mitophagy [46,47]. Because the unique metabolic state required to achieve cell plasticity is accompanied by significant temporal changes in mitochondrial function, composition, structure, and maturation, it might appear elementary to suggest that mitophagy is a prerequisite of induced pluripotency. Nonetheless, recent studies have shed light on how interlinked processes critical for mitochondrial health, including mitochondrial fragmentation and mitochondrial fission/fusion, significantly alter the efficiency and speed of induced pluripotency [48][49][50][51], but little information is available on the role of mitophagy in the acquisition and maintenance of stemness.
PTEN-induced putative kinase 1 (PINK1) encodes a key mitochondrial protein that specifically identifies and commits mitochondria to degradation via selective autophagy [52][53][54][55][56][57][58][59][60][61][62][63]. Using embryonic fibroblasts from PINK1 gene-knockout (KO) mice, we here tested the hypothesis that mitophagy is a pivotal mechanism of cell-fate plasticity by converting functionally mature mitochondrial networks into immature states and vice versa during nuclear reprogramming to stemness and commitment to differentiation, respectively. By examining the ability of mitophagy to causally modulate cell fate decisions during the entry to and exit from pluripotency, we have identified a hitherto unrecognized role of the mitophagy pathway as a critical mitochondrial switch that directs bioenergetic transition and metabolome remodeling traits to ultimately determine the efficiency and quality of nuclear reprogramming and stemness transition in somatic cells.

PINK1-mediated mitophagy is necessary for efficient nuclear reprogramming of somatic cells into iPSCs
Because the initial stages of reprogramming trigger a stress response involving repression of mitochondrial functions and oxidative stress [36][37][38][39][40][41][42][43][44][45]66], we hypothesized that the critical ability of PINK1 to identify and selectively trim impaired mitochondria from the mitochondrial network might determine the efficiency of reprogramming. To test whether PINK1-KO mouse embryonic fibroblasts (MEFs) constitute a useful model to dissect the role of mitophagy in the establishment of induced pluripotency, we first mimicked mitochondrial damage by experimentally depolarizing mitochondria with the uncoupler carbonyl cyanide m-chlorophenylhydrazone (CCCP) and then monitoring loss of MitoTracker staining after mitophagy stimulation [62,67]. Collectively, our findings show that PINK1 is necessary to efficiently drive the mitophagic digestion of damaged mitochondria (Supplemental data; supplemental Fig. 1).
We then used PINK1 -/-MEFs to explore whether PINK1-mediated mitophagy might constitute part of the molecular roadmap facilitating reprogramming. To do this, we compared iPSC generation in early-passage PINK +/+ and PINK -/-MEFs using a three-factor induction protocol (Oct4, Sox2, and Klf4, hereafter referred to as OSK). We transduced MEFs with OSK at a 1:1:1 ratio on day 0 and repeated the transduction up to four times (one infection every 12 h using the same batch of all three retroviruses), after which the regular media was replaced with standard mESC media supplemented with the knockout serum replacement (KSR). As early as 7 days after transduction, clearly recognizable flat, packed, tight colonies characteristic of ES-like cells appeared in OSK-transduced PINK1 +/+ MEF cultures (Fig. 1A). Conversely, OSK-transduced PINK1 -/-MEFs mostly failed to display the typical compact ES cell colony morphology (Fig. 1A). Indeed, using a parallel live cell-imaging 96-well-plate-based screening assay to rapidly assess the expression of the www.impactaging.com pluripotency-associated surface marker Ssea-1 during reprogramming, we found that the appearance of Ssea-1 + clusters was delayed by 3-4 days in PINK1 -/-MEFs compared with PINK1 +/+ isogenic counterparts (data not shown). We combined the observations of ES cell-like morphological changes (e.g., defined boundaries and high nucleus-to-cytoplasm ratio within individual cells) with alkaline phosphatase (AP) activity, a commonly used pluripotency indicator, to quantify bona fide iPSC colonies. From 50,000 PINK1 +/+ MEFs transduced, 150±10 colonies were AP-positive at day 14 after transduction, resulting in an iPSC generation efficiency of 0.3% (Fig. 1A). In contrast, only 30±4 colonies were generated from an equivalent number of PINK1 -/-MEFs, equivalent to an iPSC generation efficiency of 0.06% (Fig. 1A). Concerning the transduction efficiency, we did not observe any significant differences between the two groups (data not shown), thus confirming that the observed decrease in reprogramming efficiency is due to the absence of PINK1. These findings show that loss of PINK1dependent mitophagy is sufficient to dramatically decrease the speed and efficiency of nuclear reprogramming.  www.impactaging.com

Loss of PINK1-mediated mitophagy destabilizes the undifferentiated state in iPSCs
To verify that loss of PINK1-driven mitophagy dampens the reprogramming process, we picked random colonies from the previous experiment and established PINK1 +/+ -iPSC and PINK1 -/--iPSC clonal cell lines on pre-seeded MEF feeder layers, which are known to encourage induced stem cells to remain in an undifferentiated state. Immunofluorescence staining analysis of early-passage iPSCs failed to reveal major differences in the expression level of Oct3/4, Nanog, Sox2, and Ssea-1 of individual cells within PINK1 -/-and PINK1 +/+ -iPSC colonies (Fig. 1B, right). Indeed, immunoblotting procedures demonstrated a similar expression profile of the pluripotency markers Oct3/4 and Sox2 in early-passage PINK1 -/--and PINK1 +/+ -iPSCs (Fig. 1B, middle). Both early-and late-passage PINK1 +/+ -iPSC colonies were large and well-rounded and stained strongly for AP, reflective of a high percentage of non-differentiated cells (Fig. 1B). The majority of relatively late-passage PINK1 -/--iPSC colonies, however, were slightly smaller upon colony expansion, with very few showing a smooth, circular and distinct edge, and in many cases displaying irregular morphologies and undefined edges. Indeed, late-passage PINK1 -/--iPSC colonies tended to rapidly lose AP activity and failed to stably retain the compact colony morphology typical of undifferentiated iPSCs [50,68] (Fig. 1B, right).
Given that the apparent destabilization of the undifferentiated state of PINK1 -/--iPSCs occurred without drastic changes in the expression of several pluripotency markers, we explored whether loss of PINK1-driven mitophagy significantly altered the wellknown ability of nuclear reprogramming to transform the mitochondrial infrastructure and induce iPSCassociated bioenergetic transition and metabolome remodeling traits [30].

Mitophagy deficient-iPSCs cannot "rejuvenate" the morphological characteristics of the mitochondria network
We first confirmed that loss of PINK1 induces a moderate mitochondrial fragmentation in MEFs [52] as well as a more prominent accumulation of mitochondrial aggregates due to impaired mitophagy ( Fig. 2A). Given that cell reprogramming leads to mitochondrial structural and functional alterations described as "rejuvenation" [36-42, 44, 45, 66], we used transmission electron microscopy (TEM) to examine whether the decreased capacity of PINK1 -/--iPSCs to maintain their undifferentiated state involves alterations in the morphology of the mitochondrial network. Although we could detect a small decrease in the length and area of mitochondria in PINK1 -/-MEFs, together with an increase in the number of mitochondria per cell (Fig. 2B), the majority of mitochondria in both MEF populations had a similar morphology characterized by mature mitochondrial networks with tubular structures and densely-packed cristae (Fig. 2B).
PINK1 +/+ -iPSC mitochondria exhibited a dramatically decreased long diameter and increased short diameter relative to PINK1 -/-MEFs (Fig. 2B). Interestingly, though nuclear reprogramming of PINK1 +/+ -iPSCs led to the acquisition of an immature mitochondrial phenotype characterized by a rounded morphology with sparse cristae, it was noteworthy that reprogramming failed to fully reset the mitochondrial morphology of PINK1 -/-MEFs to a bona fide embryonic-like state. Con-sequently, PINK1 -/-iPSCs accumulated larger, irregular mitochondria containing different inclusions and more cristae (Fig. 2B). These findings, altogether, strongly suggest that mitophagy deficient-iPSCs fail to fully rejuvenate the morphological characteristics of the mitochondrial network.

Loss of PINK1-driven mitophagy impairs the bioenergetic transition associated with nuclear reprogramming
Mitochondrial rejuvenation is a key mechanism to protect cells from reprogramming factor-induced oxidative stress and reactive oxygen species (ROS) accumulation, a well-known roadblock to reprogramming [36-42, 44, 45, 66]. We therefore speculated that blockade of PINK1-driven mitophagy might lead to a detrimental accumulation of ROS during the initial stages of reprogramming (Supplemental data; supplemental Fig. 2). Only when forced expression of c-Myc, which can override the cell cycle checkpoints imposed in response to ROS accumulation [69][70][71], was combined with the exogenous addition of vitamin C PINK1 -/-MEFs reached reprogramming efficiencies equivalent to those observed in OSK-transduced PINK1 +/+ MEFs in the absence of this antioxidant (Supplemental data; supplemental Fig. 2). Because the incapacity of the ROS scavenger and epigenetic regulator vitamin C [72][73][74][75] to fully bypass the reprogramming roadblock imposed by the loss of PINK1 suggested that mitophagy-driven remodeling of the mitochondria network into an immature state might constitute a critical barrier during somatic reprogramming, we performed a multimodal metabolic characterization of PINK1 -/and PINK1 +/+ iPSCs at the level of cellular bioenergetics and intracellular metabolome.

www.impactaging.com
First, a well-validated extracellular flux technology was employed to establish functional monitoring of cellular bioenergetics in PINK1 -/and PINK1 +/+ iPSCs. Measurement of the extracellular acidification rate (ECAR) enabled real-time assessment of the glycolytic phenotype associated with iPSCs ( Fig. 3A). Sequential supplementation of the glycolytic fuel glucose, the ATP synthase complex V mitochondrial inhibitor oligomycin, and the competitive inhibitor of glucose 2deoxy-glucose (2-DG), dissected key parameters of glycolytic function, including glycolysis (i.e., the ECAR rate reached by iPSCs after the addition of saturating www.impactaging.com amounts of glucose), glycolytic capacity (i.e., the maximum ECAR rate reached upon blockade of oxidative phosphorylation), and glycolytic reserve (i.e., the capability of iPSCs to respond to an energetic demand). PINK1 -/--iPSCs were found to be significantly less glycolytic than PINK1 +/+ -iPSCs ( Fig. 3A), suggesting that the reduced capacity of mitophagy deficient-iPSCs to efficiently drive mitochondrial metamorphosis translate into a reduced capacity to bioenergetically transitioning from somatic cellular respiration to glycolysis during iPSC derivation [30].

Loss of PINK1-driven mitophagy impairs the metabolome remodeling associated with nuclear reprogramming
We utilized our recently developed targeted metabolomics platform coupling gas chromatography with quadrupole time-of-flight mass spectrometry and an electron impact source (GC-EI-QTOF-MS), which allows the simultaneous measurement of selected metabolites representative of the catabolic and anabolic status of key metabolic nodes. These metabolites include not only representatives of glycolysis and the mitochondrial tricarboxylic acid (TCA) cycle, but also other biosynthetic routes such as pentose phosphate pathway, amino acid metabolism and de novo fatty acid biogenesis [76,77]. Metabolite-based clustering obtained by partial least squares-discriminant analysis (PLS-DA) model revealed a clear and significant separation between PINK1 -/--iPSCs and PINK1 +/+ -iPSCs in two-dimensional (2D) score plots (Fig. 3B). Profiling of the intracellular metabolome supported a PINK1 -/-iPSCs signature distinct from the PINK1 +/+ iPSCs counterpart, which apparently involved a notable decrease in a majority of the measured glycolysis-and TCA-related biochemicals (Fig. 3C).
Heatmap visualization, commonly used for unsupervised clustering, likewise revealed distinct segregation of metabolites in PINK1 -/--iPSCs and PINK1 +/+ -iPSCs groups, pointing to an altered metabolic signature associated with the loss of PINK1dependent mitophagy in iPSCs (Fig. 4). Unsupervised hierarchical clustering of all pairwise comparisons among individual metabolites revealed several "hot spots" of highly correlated metabolites in a correlation matrix (Fig. 4). When VIP scores ≥ 1 in the PLS-DA model were used to maximize the difference of metabolic profiles between PINK1 -/--and PINK1 +/+ -iPSCs, the TCA metabolite α-ketoglutarate was the metabolite majorly impacted in mitophagy-deficient PINK1 -/--iPSCs (Fig. 5). Quantitative assessment of metabolite concentrations confirmed that PINK1 -/--iPSCs significantly accumulated > 3-fold less αketoglutarate than PINK1 +/+ -iPSCs ( Table 1), suggesting that the reduced capacity of mitophagy deficient-iPSCs to efficiently drive mitochondrial metamorphosis during nuclear reprogramming translate into a reduced capacity to achieve the embryonic stem cell-like metabolome that characterizes generated iPSCs. MetaboAnalyst's data annotation tool [103]. Rows: metabolites; columns: samples; color key indicates metabolite expression value (blue: lowest; red: highest). Right. Heatmap of correlations between iPSCs metabolites. Each square represents the Spearman's correlation coefficient between the metabolite of the column with that of the row. Metabolite order is determined as in hierarchical clustering using the distance function 1-correlation. Self-self correlations are identified in dark brown.

Mitophagy-deficient iPSC colonies exhibit a significantly reduced teratoma-initiating capacity
The ability to derive the three germ layers generated during development, (ectoderm, mesoderm and endoderm), is the gold standard for determining whether potential iPSC candidates are fully pluripotent [78][79][80]. We thus examined the teratoma initiating and differentiation potential of PINK1 +/+ -and PINK1 -/--iPSCs in vivo. To exclude the possibility that any uncoupling of pluripotent capacity from tumorigenesis might be dose-dependent, i.e., with differentiation occurring at lower cell doses but tumors forming at higher cell doses, we injected 4-to 5-week-old athymic nude mice subcutaneously either with 5 x 10 5 undifferentiated iPSCs or with 2 x 10 6 cells, the latter being a saturating concentration to ensure the development of teratoma masses within a few weeks [64,81]. We then analyzed efficiency, latency, and histology of teratoma composition. The rate of teratoma formation in PINK1 +/+ -iPSCs was 100% (6/6 mice in each group) regardless of the number of cells injected (Fig. 6A). In contrast, we observed a cell numberindependent reduction in the rate of teratoma formation following the injection of PINK1 -/--iPSCs (60%; 4/6 mice in each group). Thus, the time required for 50% of animals to develop palpable teratomas was lengthened by 161% (from 26 to 68 days) upon injection of 5 x 10 5 PINK1 -/--iPSCs, and by 166% (from 21 to 56 days) upon injection of 2 x 10 6 PINK1 -/--iPSCs in (Fig. 6A). Indeed, injection of PINK1 -/--iPSCs resulted in drastically smaller teratomas than those observed upon injection of PINK +/+ -iPSCs, i.e., the lesions in PINK1 -/--iPSC-injected mice were 15-fold smaller in size compared to the mean teratoma size observed in the PINK +/+ -iPSC group (78 mm 3 versus 1166 mm 3 , respectively) (Fig. 6B).  www.impactaging.com PINK1 -/--teratomas was not noticeably different from equivalent PINK +/+ -teratomas at the histological level. Hematoxylin-eosin staining showed various tissue derivatives of the three germ layers, including neural rosettes (ectoderm), gut-like epithelial tissues (endoderm), and smooth muscle, adipocytes, bone, and cartilage (mesoderm), in both teratoma groups (Fig. 7), confirming the full pluripotency and multi-germ layer differentiation potential of the iPSCs regardless of the PINK1-mediated mitophagy status. These findings, together with the fact that PINK1 -/--iPSCs normally expressed pluripotent markers confirmed that mitophagy-deficient iPSCs re-established pluripotency at the molecular and cellular level.
Although PINK1 +/+ -and PINK1 -/--iPSCs gave rise to teratomas composed of various recognizable tissue elements, we observed striking differences in the embryonal carcinoma (EC)-like component of poorly differentiated, primitive-appearing, blast-like teratocarcinoma stem cells. Accordingly, the large teratomas originating from PINK1 +/+ -iPSCs displayed extensive areas of undifferentiated tissue, e.g., abundant embryonic-appearing neuroepithelium with a high number of mitotic figures, which were considered as malignant based on the examination by a pathologist (Fig. 7). Conversely, PINK1 -/--iPSCs developed teratomas consisting almost exclusively of fully committed adult tissues, forming very small, morpho- www.impactaging.com logically benign, mature, and well-differentiated cystic lesions. Using TEM, we found that the cytoplasm of the extensive undifferentiated regions in PINK1 +/+ -iPSCgenerated teratomas exhibited a simple architecture typical of embryonic-like cells in the early stages of development. Consequently, these regions were devoid of most organelles except for the presence of numerous ribosomes, a well-developed Golgi apparatus, and rough endoplasmic reticulum (Fig. 8). These regions contained few mitochondria and, when found, presented a globular shape with poorly developed cristae and electron-lucid matrix, and perinuclear localization, all indicative of functionally immature mitochondria. Conversely, tera-tomas from PINK1 -/--iPSCs generated tissues with conspicuous and numerous mitochondria, possessing a complex morphology with well-developed cristae, denser matrix, and elongated or branched appearance (Fig. 8).
Although it might be argued that, because mitophagydeficient iPSC colonies tended to rapidly differentiate in vitro and exhibited a significantly reduced teratomainitiating capacity, the percentage of fully reprogrammed cells might be significantly lower within PINK1 -/--iPSC colonies, these findings are also consistent with a more rapid differentiation of mitophagy-deficient iPSCs in vivo. www.impactaging.com

DISCUSSION
Here we provide the first demonstration that mitophagy is a necessary mechanism for the conversion of somatic cells to a pluripotent cell fate with maximum efficiency. Our discovery that mitophagy-driven mitochondrial rejuvenation is required for induction and maintenance of stem cell pluripotency and that the mitophagy pathway plays a critical mitochondrial switch that determines the efficiency and quality of somatic reprogramming, illustrates how mitophagy can play a pivotal role in stem cell functions during aging and tissue regeneration.
We first addressed the question of whether mitophagy is a crucial process during nuclear reprogramming. Our findings reveal that the sole loss of PINK1-dependent mitophagy was sufficient to dramatically decrease the efficiency (∼80% reduction) and speed of the nuclear reprogramming process (Fig. 9). Deficiency of PINK1regulated mitochondrial quality control constitutes a previously unrecognized barrier to reprogramming. This fact, taken together with recent studies showing that whereas activation of DRP1-driven mitochondrial fragmentation contributes to the acquisition and maintenance of stem cell pluripotency [48,50], deficiency of mitofusins Mfn1 and Mfn2 (which co-

Figure 9. Mitophagy-regulated nuclear reprogramming of somatic cells into pluripotent stem cells. Mitophagy
is part of the roadmap during nuclear reprogramming of somatic cells to pluripotency and, as such, its blockade is sufficient to dramatically alter the speed and efficiency of iPSC reprogramming by "elevating" the "reprogramming barriers" of the epigenetic landscape and decreasing the size of the stem cell state basin of attraction, which results in the deceleration (i.e., lower efficiency and slower kinetics) of the nuclear reprogramming process. This conceptual figure represent cells stabilized in an initial non-pluripotent, somatic attractor and how nuclear reprogramming can make cells exceed the "reprogramming barriers", represented as a wall of interlocking bricks, easier or harder in the presence or absence of PINK1-dependent mitophagy, respectively, and fall down in a final attractor of fully reprogrammed, induced pluripotent stem cell states. The cellular reprogramming process is presented as a colored line from the initial to the final cellular state.
www.impactaging.com ordinately regulate mitochondrial fusion) instead elicits mitochondrial metabolic reprogramming to pluripotency [49], bolsters the notion that the ability of mitochondrial fission/fusion and mitophagy to restructure mitochondrial dynamics is central for the control of cell-fate plasticity. Accordingly, it was reported that a restricted 2-day burst of autophagy (causing mitophagy) at the early stages of reprogramming was vital for iPSC generation [47].
We next focused on whether mitophagy is the driving mechanism for the conversion of functionally mature mitochondria to an immature state and vice versa, during reprogramming to stemness and commitment to terminal differentiation, respectively. The fact that loss of PINK1-dependent mitophagy impeded the full rejuvenation of the mitochondrial network during reprogramming indicates that mitophagy operates as a key mechanism in generating the immature mitochondrial structure commonly found in stem cells. Because a high level of mitophagy has recently been found to be a requisite for the high quality of mitochondria required for the stem cell state [82], and given that the number of AP + colonies during early stages of reprogramming is used as an initial indicator of successful reprogramming of cells, the fact that cells within PINK1 -/-iPSC colonies appeared to inherit a mixture of mature ("old") and immature ("young") mitochondria when compared with PINK1 +/+ -iPSCs colonies whose cells almost exclusively inherited immature mitochondria might suggest that the ratio of fully reprogrammed colonies among AP + colonies is significantly lower in the absence of PINK1-dependent mitophagy. Furthermore, because AP expression levels is a less sensitive measure to differentiate between undifferentiated and early differentiating cells, the fact that mitophagy-deficient iPSC colonies normally expressed the pluripotent markers Oct4 and Sox2 together with their strong tendency to spontaneously differentiate and form heterogeneous populations of cells strongly suggest that PINK1-dependent mitophagy might be necessary for the iPSCs to remain in undifferentiated state. Indeed, with growing evidence for remodeling of energy metabolism in cell fate decisions, the fact that functional monitoring of cellular bioenergetics revealed an attenuated glycolytic capacity in mitophagy-deficient iPSC cells strongly suggest that the mitophagy pathway is an operating mechanism of mitochondrial switching that directs bioenergetic transition from somatic oxidative in somatic cells to glycolysis in iPSCs. While mitophagy might ultimately determine the efficiency and quality of nuclear reprogramming and stemness transition in somatic cells by participating in the bioenergetic conversion for establishing functional pluripotency, it remains to be unambiguously defined whether the ultimate role of mitophagy in stem cells is to regulate the preferential, asymmetric apportion of younger mitochondria during self-renewal [82].
The main characteristics of iPSC mitochondria are their rounded morphology with condensed cristae and their poor oxidative activity due to the low membrane potential [36][37][38][39][40][41][42][43][44][45]65]. Given that mitophagy is triggered by mild oxidative stress in a mitochondrial fissiondependent manner [83], we hypothesized that abnormal mitophagy might lead to the accumulation of "old", ROS-generating mitochondria, and this, in turn, might impair the efficiency of reprogramming. Our results showed that the mitophagy deficiency-imposed roadblock for reprogramming is bypassed, in part, by the ROS scavenger vitamin C upon the inclusion of oncogenic c-Myc, which is a key inducer of glycolytic reconfiguration [31,84]. Interestingly, c-Myc also functions as the major contributor of reprogrammingmediated oxidative stress [38,85]. Therefore, while it seems likely that ROS partially contribute to the lower reprogramming efficiency of PINK1 -/-MEFs, other ROS-independent mitochondrial changes imposed by the loss of PINK1-driven mitophagy seem to operate as a dominant roadblock during reprogramming. Because histone demethylases have been shown to be the direct downstream effectors of vitamin C-dependent enhancement of cell reprogramming, in addition to its antioxidant activity [72][73][74][75], it is possible that mitophagy deficiency might impede reprogramming by inhibiting histone demethylation. In this regard, it was noteworthy that the iPSC metabolite majorly impacted by loss of PINK1-dependent mitophagy was αketoglutarate, a key mitochondrial metabolite that is siphoned from the TCA cycle to support rapid cell proliferation via lipid and amino acid biosynthesis that can exit also the mitochondria to function as a cofactor for dioxygenase enzymes including Jumonji-family histone demethylases, TET-family DNA hydroxylases, and prolyl hydroxylases [86][87][88]. Indeed, the intracellular α-ketoglutarate levels have been shown to contribute to the maintenance of cellular identity and have a mechanistic role in the transcriptional and epigenetic state of stem cells [89]. Further research, however, is needed to evaluate the precise epigenetic modifications associated with mitophagy-related changes in mitochondrial biogenesis, structure, and function, i.e., how mitophagy might drive the acquisition and maintenance of stemness by metabolically regulating the epigenetic landscape of the nuclear genome [90][91][92].
The fact that PINK1 -/--iPSCs possessed a normal capacity to differentiate into mature tissue confirmed www.impactaging.com that deficiency of PINK1-driven mitophagy does not interfere with the pluripotent quality of iPSCs. Because the kinetics of teratoma formation is dependent on the number of remaining pluripotent stem cells during the differentiation procedure [93][94][95][96][97][98], it might be tempting to suggest that loss of PINK1-driven mitophagy facilitates the depletion of residual undifferentiated pluripotent cells (teratocarcinoma-initiating cells) during in vivo teratoma formation. Moreover, because the retention of the embryonal character is considered the basis for continuous and progressive growth of malignant teratomas, loss of biological aggressiveness in PINK1 -/--iPSC-derived teratomas apparently suggest that mitophagy deficiency impedes the retention of undifferentiated pluripotent stem cells by promoting their differentiation and inhibiting potential dedifferentiation of committed cells. Accordingly, the highly significant more active mitochondrial state of tissues from PINK1 -/teratomas might be indicative of rapid, most committed differentiation of otherwise pluripotent PINK1 -/--iPSCs. Conversely, it might be argued that these findings reflect that the percentage of fully reprogrammed cells might be significantly lower within mitophagy-deficient PINK1 -/--iPSC colonies. Forthcoming studies should evaluate whether rapidly differentiating, mitophagy-deficient heterogeneous colonies of PINK1 KO-iPSCs might illuminate new mitochondria-centered mechanisms aimed to restore or stimulate a differentiation checkpoint capable of limiting the aberrant self-renewal of life-threatening cancer stem cells in tumor tissues.
We are beginning to dissect the roles of mitochondria in the establishment and homeostasis of stemness, which may help to uncover novel insights into our understanding of a wide variety of degenerative diseases, aging, and aging-related diseases including cancer. Although we are still far from a comprehensive understanding of the physiological functions of mitochondria in stem cells, our findings extend previous studies into the causal mechanism behind the wellrecognized metabolic switch during the establishment of pluripotency, which is accompanied by significant changes in mitochondrial function, composition, structure, maturation, and signaling. Mitophagy appears to be a crucial cellular process for the conversion of functionally mature mitochondria to an immature state and vice versa during reprogramming and differentiation, respectively. In this regard, mitophagy may ensure a metabolic transition to meet the specific energetic and anabolic demands of the stemness state, e.g., mitophagy-induced repression of mitochondrial functions including mitochondrial clearance might accelerate the onset of the glycolytic metabolism [99]. Furthermore, mitophagy-driven mitochondrial rejuvena-tion might contribute to the ability of stem cells to suppress differentiation by orchestrating the mitochondria function as signaling organelles of diverse biological functions [100,101], including not only bioenergetic transitions but, perhaps more importantly, metabolome remodeling traits connecting mitochondrial metabolites with epigenetics [100][101][102]. Further studies are warranted to determine the causal role of mitophagy-driven mitochondrial rejuvenation as part of the mechanism involved in the maintenance and asymmetric transmission of the pluripotency and differentiation fate of stem cells. Our discovery that mitophagy-controlled mitochondrial quality is a critical director of cell-fate plasticity and stem-cell fate should provide new insights into how mitophagy might influence the stem cell decisions to retain pluripotency or differentiate in tissue regeneration and aging, tumor growth, and regenerative medicine.

MATERIALS AND METHODS
PINK1-knockout mouse embryonic fibroblasts. PINK1knockout mice were generated by targeted deletion of exon 1 as described [52]. Loss of Pink1 mRNA expression in primary embryonic fibroblasts (MEFs) was confirmed by quantitative RT-PCR [52].
To generate feeder-free iPSC cultures for teratoma assays, culture plates were coated with 0.3 mg/mL Matrigel (growth factor-reduced, BD Biosciences, San Jose, CA) at 4 o C overnight. Unbound Matrigel was aspirated, and the cells were washed with DMEM/F12 medium. iPSCs were seeded on Matrigel-coated plates in MEF-conditioned ES cell medium supplemented with leukemia inhibitory factor (LIF) and bFGF (4 ng/mL). The medium was changed every day.
Immunofluorescence staining. High-content confocal imaging was performed in 96-well clear bottom imaging tissue culture plates (BD Biosciences) optimized for automated imaging applications. Triton ® www.impactaging.com X-100 permeabilization and blocking, primary antibody staining, secondary antibody staining using Alexa Fluor ® 488 goat anti-rabbit/mouse IgG (Invitrogen, Molecular Probes, Eugene, OR) and counterstaining (Hoechst 33258; Invitrogen) were performed following BD Biosciences protocols. Images were captured in different channels for Alexa Fluor® 488 (pseudocolored green) and Hoechst 33258 (pseudocolored blue) on a BD PathwayTM 855 Bioimager System (BD Biosciences) with 20× or 40× objectives (NA 075, Olympus). Merged images were obtained according to the Recommended Assay Procedure using BD Attovision TM software.
Reactive Oxygen Species (ROS) detection. Cells were incubated for 60 min with 10 μmol/L 2',7'-dihydrodichlorofluorescein-diacetate (H2DCF-DA) (Invitrogen, Molecular Probes) at 37˚C. Cellular green fluorescence was then measured by flow cytometry. Cell-permeant non-fluorescent H 2 DCF-DA, upon cleavage of the acetate moiety by intercellular esterases and oxidation by ROS, is converted to strongly fluorescent DCF and thus reports the ROS abundance.
Extracellular flux bioenergetic assays. Extracellular acidification rates were measured using an XF Extracellular Flux Seahorse Analyzer (Seahorse Bioscience). XFp Glycolysis Stress tests were performed in accordance with manufacturer's instructions. Each plotted value is the mean of at least 6 replicates and was normalized to Hoechst signal in each well.
Targeted metabolomics and data analysis. Measurements of bioenergetics metabolites obtained from PINK1 +/+ -iPSC and PINK1 -/--iPSC clonal cell lines were performed by employing a previously described simple and quantitative method based on gas chromatography coupled to quadrupole-time of flight mass spectrometry and an electron ionization interface (GC-EI-QTOF-MS) [76,77].
Raw data were processed and compounds were detected and quantified using the Qualitative and Quantitative Analysis B.06.00 software (Agilent Technologies), respectively. MetaboAnalyst 3.0 (http://www.metaboanalyst.ca) was used to generate scores/loading plots, Heatmaps, and correlation maps [103].
Teratoma assays. To form teratomas, iPSCs were harvested from Matrigel-coated culture dishes and injected subcutaneously (s.c.) into the dorsal flank of female athymic nude mice (four-to five-weeks-old, 23-25g; Harlan Laboratories, France). Mice were weighed once per week. Teratomas were measured daily with electronic calipers and tumour volumes were calculated using the formula: volume (mm3) = length × width 2 × 0.5. General health of the mice in response to teratoma development (e.g., subcutaneous teratomas cause ulceration on the skin) was monitored daily by a specialized veterinarian. Teratomas were carefully dissected and removed in entirety, fixed in 10% phosphate buffered formalin (3.6% formaldehyde) for 24 hours, and paraffin-embedded. For histopathological analysis, consecutive sections (4 μm) were cut and stained with haematoxylin and eosin according to standard procedures.
The Institutional Animal Care and Use Committee (IACUC) of the Institut d'Investigació Biomèdica de Bellvitge (IDIBELL; Animal Use Protocol #6302 authorized by the Animal Experimental Commission from the Catalan Government, Barcelona, Spain) approved the experiments.
Transmission electron microscopy. Small pieces of teratomas were fixed in a 2% glutaraldehyde solution in 0.1 M cacodylate buffer, pH 7.4. Samples were then post-fixed in 1% osmium tetroxide (OsO 4 ) for 2 h and dehydrated through a graded series of acetone prior to impregnation in increasing concentrations of resin in acetone over a 24 h period. Semi-thin sections (500 nm) were stained with 1% toluidine blue. Ultrathin sections (70 nm) were subsequently cut using a diamond knife, double-stained with uranyl acetate and lead citrate, and examined with a transmission electron microscope (Hitachi, Tokyo, Japan).
Statistical analysis. The results are presented as the mean ± SD of at least three repeated individual experiments for each group. The analyses were performed using XLSTAT 2010 (Addinsoft TM ). A Pvalue ≤ 0.05 was considered statistically significant. www.impactaging.com