mTOR has a developmental stage-specific role in mitochondrial fitness independent of conventional mTORC1 and mTORC2 and the kinase activity

The mammalian target of rapamycin (mTOR), present in mTOR complex 1 (mTORC1) and mTORC2, is a serine/threonine kinase that integrates nutrients, growth factors, and cellular energy status to control protein synthesis, cell growth, survival and metabolism. However, it remains elusive whether mTOR plays a developmental stage-specific role in tissue development and whether mTOR can function independent of its complexes and kinase activity. In this study, by inducible genetic manipulation approach, we investigated the role of mTOR and its dependence on mTOR complexes and kinase activity in mitochondrial fitness of early, progenitor stage (lineage-negative; Lin-) versus later, lineage-committed stage (lineage-positive; Lin+) of hematopoietic cells. We found that oxidative phosphorylation (OXPHOS), ATP production and mitochondrial DNA synthesis were decreased in mTOR-/- Lin- cells but increased in mTOR-/- Lin+ cells, suggesting that mTOR plays a developmental stage-specific role in OXPHOS, ATP production and mitochondrial DNA synthesis. In contrast to mTOR deletion, simultaneous deletion of Raptor, a key component of mTORC1, and Rictor, a key component of mTORC2, led to increased mitochondrial DNA in Lin- cells and decreased mitochondrial DNA and ATP production in Lin+ cells, suggesting that mTOR regulates mitochondrial DNA synthesis in Lin- and Lin+ cells and ATP production in Lin+ cells independent of mTORC1 and mTORC2. Similar to mTOR deletion, deletion of Raptor alone attenuated glycolysis and increased mitochondrial mass and mitochondrial membrane potential in Lin- cells and increased mitochondrial mass and OXPHOS in Lin+ cells, whereas deletion of Rictor alone had no effect on these mitochondrial parameters in Lin- and Lin+ cells, suggesting that mTOR regulates glycolysis and mitochondrial membrane potential in Lin- cells, OXPHOS in Lin+ cells, and mitochondrial mass in both Lin- and Lin+ cells dependent on mTORC1, but not mTORC2. Either Raptor deficiency or Rictor deficiency recapitulated mTOR deletion in decreasing OXPHOS in Lin- cells and glycolysis in Lin+ cells, suggesting that mTOR regulates OXPHOS in Lin- cells and glycolysis in Lin+ cells dependent on both mTORC1 and mTORC2. Finally, mice harboring a mTOR kinase dead D2338A knock-in mutant showed decreased glycolysis in Lin+ cells, as seen in mTOR-/- Lin+ cells, but no change in glycolysis in Lin- cells, in contrast to the decreased glycolysis in mTOR-/- Lin- cells, suggesting that mTOR regulates glycolysis in Lin+ cells dependent on its kinase activity, whereas mTOR regulates glycolysis in Lin- cells independent of its kinase activity.


Introduction
The mammalian target of Rapamycin (mTOR) is a serine/threonine kinase that is present in two molecular complexes: mTOR complex 1 (mTORC1) and mTORC2. mTORC1 is comprised of Raptor, mLST8, PRAS40 and Deptor, and mTORC2 is comprised of Rictor, mLST8, mSIN1 and Protor [1]. In response to extracellular stimuli such as nutrients and growth factors and intracellular stimuli such as elevated energy state, mTORC1 is activated by phosphatidylinositol-3-OH (PI 3) kinase, PDK1 and Akt [1,2]. Activated mTORC1 regulates ribosome biogenesis, protein synthesis, cell growth, and autophagy through S6K1 and 4E-BP. mTORC2, on the other hand, promotes cell survival through Akt and/or PKC-θ [2,3] Mitochondria are the main site for generating ATP, the major energy source for many cellular activities including cell proliferation and survival. Mitochondrial ATP production is driven by mitochondrial metabolism such as oxidative phosphorylation (OXPHOS) and glycolysis [4]. Proper mitochondrial function relies on the homeostatic maintenance of mitochondrial mass, mitochondrial membrane potential and mitochondrial DNA contents. It is well known that mitochondrial metabolism is regulated by mTOR [4,5]. mTORC1 signaling regulates glucose, glutamine, lipid, amino acid and nucleic acid metabolism in a variety of cells including fibroblasts, T lymphocytes, epithelial cells and hepatocytes [6][7][8][9]. Although less studied, mTORC2 is also shown to regulate glycolysis, lipogenesis and OXPHOS in hepatocytes [4,10,11] It is generally thought that mTOR acts through mTOR complexes. In support, mTOR regulates oxidative muscle integrity via mTORC1 and endothelial cell proliferation through both mTORC1 and mTORC2 [12,13]. However, it remains unknown whether mTOR has a complex-independent role in regulating cell behaviors. mTOR kinase activity is presumably required for mTOR function. In this aspect, mTOR kinase activity is responsible for phosphorylation of S6K, 4E-BP and AKT, and chemical inhibitors of mTOR kinase activity suppress normal and cancer cell growth and/or survival [14][15][16]. However, genetic evidence of the importance of mTOR kinase activity is lacking and it is elusive whether mTOR has kinase activity-independent function. Moreover, it remains unclear whether mTOR plays a developmental stage-specific role in tissue development. In this study, by single deletion of mTOR, Raptor or Rictor, simultaneous deletion of Raptor and Rictor, and knock-in of a mTOR kinase dead (KD) mutant, we have examined mTOR complex and kinase activity dependence of mTOR in the regulation of mitochondrial fitness (e.g. ATP production, OXPHOS, glycolysis, mitochondrial mass, mitochondrial membrane potential and mitochondrial DNA synthesis) of hematopoietic cells. We have found that mTOR may regulate mitochondrial fitness dependent or independent of mTOR complexes and kinase activity. By comparing the effects of mTOR deletion on mitochondrial fitness of early, progenitor stage (lineage-negative; Lin -) with that of later, lineage-committed stage (lineage-positive; Lin + ) of hematopoietic cells, we show that mTOR has a developmental stage-specific role in regulating mitochondrial fitness.

Materials and methods Mice
Animal study was in compliance with the Cincinnati Children's Hospital Medical Center Animal Care and Use Committee protocols and with National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978). Animal study was performed under specific ethics approval obtained from the Cincinnati Children's Hospital Medical Center Animal Care. Conditional gene-targeted mTOR loxp/loxp , Raptor loxp/loxp and Rictor loxp/loxp mice were generated as described previously [12,17,18]. The flox allele contains loxp sites flanking exon 1-5 of the mTOR gene, exon 6 of the Raptor gene, or exon 3 of the Rictor gene. To delete mTOR, Raptor, or Rictor in vivo in hematopoietic cells, mTOR loxp/loxp ;Mx-Cre + , Raptor loxp/loxp ; Mx-Cre + or Rictor loxp/loxp ;Mx-Cre + mice were generated by breeding mTOR loxp/loxp , Raptor loxp/loxp or Rictor loxp/loxp mice with the Mx-Cre + transgenic mice carrying a bacteriophage Cre recombinase driven by an interferon-α-inducible Mx1 promoter. The expression of Cre was induced by 5 i.p. injections of polyinosine-polycytidine (poly I:C) (Amersham Pharmacia Biotech, Piscataway, NJ) into the Mx-Cre + mice at 2-day intervals with each injection of 10 μg/g of body weight. Control, Mx-Cremice also received poly I:C injections. The mice were utilized 7-10 days after last poly I:C injection. mTOR KD mutant D2338A knock-in mice were generated by CRISPR/Cas9 technology. Because mTOR KD/KD mice were embryonically lethal and mTOR +/KD mice didn't show abnormalities [19], we generated mTOR loxp/KD mice and crossed with Mx1-Cre + mice. When the resultant mTOR loxp/KD ; Mx1-Cre + mice are treated with poly I:C, mTOR is ablated from the floxed allele. Thus, mTOR kinase activity is not only lost on the D2338A mutant knock-in allele but also on the floxed allele. All of the mice were housed under specific pathogen-free conditions in the animal facility at Cincinnati Children's Hospital Research Foundation. Mice were anesthetized when necessary, using ketamine (80-100 mg/kg im), aceptromazine (4-6 mg/kg im) and atropine (0.1mg/kg im). Anesthesia was maintained using ketamine (30 mg/kg im) as needed. During the course of experiments, mice were isolated in microisolator cages and cared for in the Laboratory Animal Resource Center by a trained technician and two veterinarians. Animals were checked daily by qualified personnel in the lab. The method of euthanasia used was CO 2 euthanasia. The animal study was in compliance with the Cincinnati Children's Hospital Medical Center Animal Care and Use Committee protocols and with the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978) Immunoblot Bone marrow cells were lysed and protein content was normalized by Bradford assay. Lysates were separated by 10% SDS-polyacrylamide gel electrophoresis. The expression of Raptor and Rictor and phosphorylation of S6K, 4E-BP and Akt (S473) were probed by Western blot using corresponding antibodies (Cell Signaling Technology).

Transcript expression analysis
RNA was isolated from Linand Lin + cells using RNeasy Micro Kit (QIAGEN) and converted to cDNA using a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems Inc). Quantitative Real-time PCR was performed with SYBR Green dye or the Taqman assay on a 7900HT Real-Time machine (Applied Biosystems Inc, Foster City, CA). Data were analyzed using SDS 2.3 software (Applied Biosystems Inc, Foster City, CA) and normalized to GAPDH.

Analysis of mitochondrion mass and mitochondrial membrane potential
Bone marrow cells were incubated for 20 min at room temperature with antibodies against lineage markers Gr1, Mac1, Ter119, B220, CD3, CD4, and CD8 (BD Biosciences). For the measurement of Mitochondrion mass and mitochondrial membrane potential, the cells were incubated with 100 nM Mitotracker Green and 50 nM DilC-5 (Invitrogen), respectively, according to the manufacturer's protocols. The Immunolabeled cells were analyzed by flow cytometry on a FACSCanto system using FACSDiVa software (BD Biosciences, Franklin Lakes, NJ) [21].

Analysis of mitochondrial DNA
Purified Linand Lin + cells were lysed and homogenized and mitochondrial DNA was then purified with a Mitochondrial DNA isolation Kit (Biovision), according to the manufacturer's manual. The purified mitochondrial DNA was quantified by quantitative real-time PCR with SYBR Green dye. Mitochondrial DNA content was represented by mitochondrial cyclo-oxygenase (Cox) 2 normalized to the nuclear intron of β-globin. The primer sequences were: Cox2: GCCGACTAAATCAAGCAACA (forward), CAATGGGCATAAAGCTATGG (reverse); β-globin:

Analysis of ATP
Purified Linand Lin + cells were lysed with a cytosol extraction buffer supplied in the Mitochondrial DNA isolation Kit, homogenized and then ATP was determined by an ATP Determination Kit (Molecular Probes), according to the manufacturer's protocol.

Statistical analysis
The data are expressed as the mean ± standard deviation (SD). The data were analyzed by Student's unpaired t-test for single comparisons and by one-way ANOVA followed by Bonferroni test for multiple comparisons. Significance was accepted at p<0.05.
OXPHOS and glycolysis generate ATP. We found that ATP was decreased in mTOR -/-Lincells but increased in mTOR -/-Lin + cells (Fig 5A and 5B), suggesting that mTOR promotes energy production in Lincells but inhibits it in Lin + cells and thus mTOR plays a developmental stage-specific role in regulating energy production. Deletion of Raptor or simultaneous deletion of Raptor and Rictor, but not deletion of Rictor, recapitulated mTOR deletion in decreasing ATP in Lincells (Fig 5A), suggesting that mTOR regulates energy production in Lincells depending on mTORC1. In contrast to the increased ATP in mTOR -/-Lin + cells, ATP was decreased in Raptor -/-Rictor -/-Lin + cells (Fig 5B), suggesting that mTOR regulates energy production in Lin + cells independent of mTORC1 and mTORC2. Interestingly, in both Linand Lin + cells, deletion of Raptor decreased ATP, whereas deletion of Rictor increased ATP (Fig 5A and 5B). These data suggest that mTORC1 and mTORC2 play an opposing role in regulating ATP production and that the effect of Raptor deficiency on ATP production prevails over that of Rictor deficiency, leading to a net decrease in ATP in Raptor -/-Rictor -/-Linand Lin + cells. In addition, the decreased ATP in mTOR -/-, Raptor -/and Raptor -/-Rictor -/-Lincells may reflect a combined effect of the decreased OXPHOS and glycolysis. The increased ATP in mTOR -/-Lin + cells may reflect a prevailing effect of the increased OXPHOS over the decreased glycolysis, whereas the decreased ATP in Raptor -/and Raptor -/-Rictor -/-Lin + cells may reflect a prevailing effect of the decreased glycolysis over the increased OXPHOS. On the other hand, the increased ATP in Rictor -/-Linand Lin + cells can't be explained by the decreased OXPHOS in Rictor -/-Linand the decreased glycolysis in Rictor -/-Lin + cells. It may reflect a net effect of OXPHOS, glycolysis and/or other metabolic programs such as lipid and amino acid metabolism. A role of mTOR, mTORC1 and mTORC2 in mitochondrial mass, mitochondrial membrane potential and mitochondrial DNA synthesis Mitochondrial mass, mitochondrial membrane potential and mitochondrial DNA synthesis are important for mitochondrial metabolism. We found that depletion of mTOR or Raptor or simultaneous deletion of Raptor and Rictor increased mitochondrial mass in both Linand Lin + cells, whereas deletion of Rictor had no effect (Fig 6A and 6B), suggesting that mTOR maintains mitochondrial homeostasis in Linand Lin + cells dependent on mTORC1. Depletion of mTOR or Raptor or simultaneous deletion of Raptor and Rictor increased mitochondrial membrane potential in Lincells, whereas deletion of Rictor had no effect (Fig 6C), suggesting that mTOR inhibits mitochondrial membrane potential in Lincells dependent on mTORC1. Similar to that in Lincells, depletion of mTOR increased mitochondrial membrane potential in Lin + cells. However, while simultaneous deletion of Raptor and Rictor recapitulated mTOR deletion, deletion of Raptor or Rictor had no effect on mitochondrial membrane potential in Lin + cells (Fig 6D). These results suggest that mTORC1 and mTORC2 play a redundant but important role in controlling mitochondrial membrane potential of Lin + cells and that mTOR inhibits mitochondrial membrane potential of Lin + cells depending on both mTORC1 and mTORC2. Deletion of mTOR decreased mitochondrial DNA in Lincells but increased it in Lin + cells (Fig 6E and 6F), suggesting a developmental stage-specific role of mTOR in regulating mitochondrial DNA synthesis. In contrast to mTOR deletion, simultaneous deletion of Raptor and Rictor increased mitochondrial DNA in Lincells and decreased Regulation of mitochondrial fitness by mTOR it in Lin + cells (Fig 6E and 6F), suggesting that mTOR regulates mitochondrial DNA synthesis independent of mTORC1 and mTORC2. Similar to simultaneous deletion of Raptor and Rictor, deletion of Raptor increased mitochondrial DNA in Lincells and decreased it in Lin + cells (Fig 6E and 6F). In contrast to simultaneous deletion of Raptor and Rictor and deletion of Raptor, deletion of Rictor decreased mitochondrial DNA in Lincells and increased it in Lin + cells (Fig 6E and 6F). These data suggest that Raptor and Rictor play an opposing role in regulating mitochondrial DNA synthesis in Linand Lin + cells and that the effect of Raptor deficiency on mitochondrial DNA synthesis prevails over that of Rictor deficiency, leading to a net increase in mitochondrial DNA in Raptor -/-Rictor -/-Lincells and a net decrease in mitochondrial DNA in Raptor -/-Rictor -/-Lin + cells.

A role of mTOR kinase activity in mitochondrial fitness
To define the role of mTOR kinase activity in mitochondrial fitness, we generated mTOR KD D2338A mutant [12,19,24] knock-in mice by CRISPR-Cas9 technology (Fig 7A). Because mTOR KD/KD mice were embryonically lethal and mTOR +/KD mice didn't show abnormalities [19], we generated mTOR loxp/KD mice and crossed with Mx1-Cre + mice. When the resultant mTOR loxp/KD ;Mx1-Cre + mice are treated with poly I:C, mTOR is ablated from the floxed allele. Thus, mTOR kinase activity is not only lost on the D2338A mutant knock-in allele but also on the floxed allele, whereas protein expression is preserved on the D2338A mutant knock-in allele. By immunoblotting, we confirmed the preservation of mTOR protein expression and the loss of mTOR kinase activity in phosphorylating S6, 4E-BP and Akt in mTOR loxp/KD ; Mx1-Cre + bone marrow cells, upon poly I:C induction (Fig 7B). The loss of mTOR kinase activity recapitulated mTOR deficiency in decreasing basal and mitochondrial uncoupler

FCCP-induced maximal OXPHOS in Lincells and in increasing basal and maximal OXPHOS
in Lin + cells (Fig 7C-7F), suggesting that mTOR kinase activity is required for mTOR to regulate OXPHOS in Linand Lin + cells. The loss of mTOR kinase activity also recapitulated mTOR deficiency in decreasing glycolysis in Lin + cells. However, it had no effect on glycolysis in Lincells, in contrast to mTOR deficiency that caused decreased glycolysis in Lincells ( Fig  7G-7J). These data suggest that mTOR kinase activity is required for mTOR to regulate glycolysis in Lin + cells but not in Lincells.

Discussion
In this study, by employing hematopoietic cells at the Linand Lin + stages of hematopoiesis as a model system, we demonstrate that mTOR has a development stage-specific role in the regulation of mitochondrial fitness, and mTOR regulates mitochondrial fitness in complex-and kinase activity-dependent and -independent manner. Specifically, direct comparison of the Regulation of mitochondrial fitness by mTOR effects of mTOR deletion on mitochondrial fitness of Lincells with that of Lin + cells reveals that mTOR plays a development stage-specific role in regulating OXPHOS, ATP production and mitochondrial DNA synthesis (Table 1). By comparing the effects of mTOR deletion with that of simultaneous deletion of Raptor and Rictor on mitochondrial fitness, we show that mTOR regulates ATP production in Lin + cells and mitochondrial DNA synthesis in both Linand Lin + cells independent of mTORC1 and mTORC2 (Table 2). It remains elusive how mTOR regulates ATP production and mitochondrial DNA synthesis in Linand/or Lin + cells. Nonetheless, given that mTOR can bind to IκB kinase (IKK) [25], IKK, in complex with mTOR, may underlie mTOR-regulated ATP production and mitochondrial DNA synthesis in Regulation of mitochondrial fitness by mTOR Linand/or Lin + cells. By comparing the effects of mTOR deletion with that of Raptor or Rictor deletion on mitochondrial fitness of Linand Lin + cells, we demonstrate that mTOR regulates glycolysis, mitochondrial mass and mitochondrial membrane potential in Lincells and mitochondrial mass and OXPHOS in Lin + cells depending on mTORC1. In contrast, mTOR regulates OXPHOS in Lincells and glycolysis in Lin + cells depending on both mTORC1 and mTORC2 (Table 3). By comparing the effects of mTOR deletion with that of mTOR KD mutant knock-in on mitochondrial fitness, we unveil that mTOR regulates OXPHOS in both Linand Lin + cells and glycolysis in Lin + cells depending on its kinase activity, whereas mTOR regulates glycolysis in Lincells independent of its kinase activity.
One interesting observation is that while mTOR deficiency increases mitochondrial membrane potential in Lin + cells, neither Raptor nor Rictor deficiency has an effect ( Table 3). Consistent with our findings, it has previously been shown that while mTOR deficiency reduces muscle dystrophin and mTOR knockdown dampens translational efficiency of terminal oligopyrimidine (TOP) mRNAs, Raptor or Rictor deficiency shows marginal effect [12,26]. It is seemingly true that mTOR regulates mitochondrial membrane potential in Lin + cells, dystrophin expression in muscle cells and TOP mRNA translation independent of mTORC1 and mTORC2. However, the fact that simultaneous deletion of Raptor and Rictor recapitulates mTOR deficiency in increasing mitochondrial membrane potential in Lin + cells ( Table 2) rather suggests that mTOR depends on mTOR complexes in the regulation of mitochondrial Regulation of mitochondrial fitness by mTOR membrane potential in Lin + cells. These data also suggest that mTORC1 and mTORC2 play an important but redundant role in regulating mitochondrial membrane potential in Lin + cells. Thus, whether mTOR functions through its complexes can only be revealed by comparing mTOR perturbation with simultaneous intervention of mTORC1 (e.g., Raptor) and mTORC2 (e.g., Rictor), and whether mTORC1 and mTORC2 are redundant but important for a cell function can only be revealed by comparing the intervention of mTORC1 or mTORC2 alone with simultaneous intervention of mTORC1 and mTORC2. mTOR is a serine/threonine kinase and the kinase activity is expected to be responsible for its function. However, by knock-in an mTOR KD mutant, we show that mTOR regulates glycolysis in Lincells independent of its kinase activity. By overexpression of the same mTOR KD mutant in muscle cells, a recent study has shown that mTOR has a kinase activity-independent role in regulating muscle dystrophin expression [12]. However, the non-specific nature of the overexpression approach may confound the conclusion of kinase activity independence of mTOR in regulating muscle dystrophin expression. In this context, our study unequivocally demonstrates that mTOR can function independent of its kinase activity in regulating mitochondrial fitness.
In conclusion, by using primary hematopoietic cells, our study has revealed that mTOR plays a developmental stage-specific role in the regulation of mitochondrial fitness. Furthermore, mTOR has a complex-and kinase activity-independent role in regulating mitochondrial fitness. Whether these observations can be extended to other cellular events and other tissues remains to be determined. Furthermore, as mitochondrial activity has been associated with hematopoietic cell behaviors [22, 27, 28], it will be interesting to see whether mTOR-regulated mitochondrial fitness has a role in hematopoiesis.