Free Radic Biol Med 60: 29-40

PMC: PMC3654039

PMID: 23376468

doi: 10.1016/j.freeradbiomed.2013.01.013

PI3K/AKT Signaling Regulates Bioenergetics in Immortalized Hepatocytes

Introduction

Mitochondria are major cellular energy-transducing organelles that integrate metabolic processes and signaling pathways and are closely involved in the regulation of cell survival and apoptotic pathways. The phosphotidylinositol-3 kinase (PI3K) signaling pathway–activated by the binding of insulin, growth factors, and cytokines to the cell surface receptor [1]–regulates glucose and lipid metabolism [2], processes linked to transfer of reducing equivalents–generated by the tricarboxylic acid cycle–through the respiratory chain and coupled to oxidative phosphorylation. The enzymatic activity of PI3K leads to the conversion of phosphatidylinositol-4,5-phosphate (PIP₂) to phosphatidylinositol-3,4,5-phosphate (PIP₃) [3]. Binding of PIP₃ to pleckstrin homologue (PH) domains of downstream molecules serves as a major signaling node that transmits cell growth and hormone signals from cell surface to inside the cells [4, 5]. In muscle, insulin is a major regulator for mitochondrial ATP production partly by promoting the synthesis of resident mitochondrial genes and proteins [6, 7]. Consistent with a role of PI3K signaling in oxidative phosphorylation, deletion of forkhead transcriptional factor FoxO1 (downstream of PI3K) increased electron transport chain activity and normalized NAD^{/NADH values in the IRS1/2 double knockout mouse liver [}8]. In mouse models of resistance to insulin and activation of PI3K, such as the ob/ob mice [8, 9], mitochondrial electron-transfer chain activity and determinants of NAD^{/NADH values are downregulated. Oxidative phosphorylation genes are also deregulated when liver cells fail to respond to insulin, such as in individuals with insulin resistance [}8]. PI3K signaling also plays a key role in cell survival, partly due to its effect on regulating the release of mitochondrial cytochrome c [10]. Activation of PI3K [2] or inhibition of a negative regulator for the pathway, PTEN (Phosphatase and Tensin Homologue deleted on Chromosome 10) results in accelerated cell growth and better survival [11–14]. Insulin signals through PI3K to induce lipogenesis and glucose uptake/metabolism in both muscle and adipose tissues [15]. Deletion of PTEN results in improved insulin sensitivity, increased lipogenesis and glucose uptake/metabolism [16, 17].

A major target of PI3K activated by binding of PIP₃ is the serine/threonine kinase AKT [18, 19]. Activation of PI3K by insulin and other growth factors leads to phosphorylation and activation of AKT [18, 19]. PTEN removes the 3′ phosphate moiety from PIP₃ and converts it to PIP₂ [20, 21], thereby inhibiting signaling transduction. In cells lacking PTEN, AKT is ubiquitously phosphorylated [12, 14, 16, 17, 22–25]. Activation of AKT and the consequential inhibition of FoxO lead to lipogenesis [17, 26, 27]. AKT was found to regulate the mitochondrial intrinsic cell death signaling through direct or indirect interactions with pro-apoptotic factors of the Bcl-2 family (i.e., Bad and Bax) and glycogen synthase kinase 3 [10]. Bad and GSK3 are the two best-characterized substrates that are directly phosphorylated and inhibited by AKT [28]. Thus, glucose/lipid metabolism and cell survival/apoptosis–both regulated by PI3K/AKT signaling–converge at mitochondria.

How AKT may control mitochondrial respiratory function, however, is not clear. AKT may phosphorylate hexokinase II and strengthen its association with the outer mitochondrial membrane voltage-dependent anion channel (VDAC), thus activating the first step in glucose metabolism [28, 29]. AKT may have indirect effects on mitochondrial respiration in terms of increased substrate supply (pyruvate) to mitochondria as well as direct effects that entail the translocation of AKT to mitochondria [30, 31]. The latter is associated with an increased phosphorylation (inhibition) of a mitochondrial constitutive form of GSK3β in neuroblastoma cells [31]. However, the biological function of AKT translocation and phosphorylation of mitochondrial GSK3β is not clear. Association of AKT with other mitochondrial proteins, such as ATP synthase, was also reported [31].

This study examines the role of the PI3K/AKT pathway on mitochondrial bioenergetics by experimental approaches–performed on immortalized hepatocytes–entailing (a) genetic approaches (specific deletion of Pten), (b) stimulation of the PI3K/AKT route by IGF-1, and (c) specific inhibition of PI3K. The results indicate that AKT activation (phosphorylation) enhances cellular bioenergetics by mechanisms that increase both substrate supply to- and the catalytic efficiency of mitochondria.

Materials and Methods

Cell lines

Immortalized wild type control (Control) and Pten null (Pten^−/−) hepatocytes were established from 1-month old Pten^{; Alb-Cre}⁻ and Pten^{; Alb-Cre}⁺ mouse liver, respectively [17, 25]. The cell lines were cultured in DMEM supplemented with 10% FBS, 1% penicillin/streptomycin, 5 μg/ml insulin, 10 ng/ml EGF in a humidified 5% CO₂ atmosphere. Primary hepatocytes were cultured from 3-month old Pten/Akt2 double mutant (Dm) (Pten-^{; Akt2}^{; Alb-Cre}⁺), Pten^−/− (Pten^{; Alb-Cre}⁺), and control (Pten^{; Alb-Cre}⁻) mouse liver by collagenase perfusion.

Chemicals

Oligomycin, rotenone, and carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) were purchased from MP Medicals (Solon, OH). Lithium chloride, sodium pyruvate, D-(+)-Glucose, insulin-like growth factor 1 (IGF-1), LY294002, and protease/phosphatase inhibitor cocktail were from Sigma Chemical Co (St Louis, MO).

XF24 extracellular metabolic flux analysis

Control and Pten^{hepatocytes were seeded on XF24 plates at a density of 2×10}^{cells/well. The plate was incubated in a humidified 5% CO}₂ atmosphere for 24 h. For IGF-1 treatment, cells were further incubated in the serum free DMEM for 6 h followed by IGF-1 treatment for 1 h. For LY294002 treatment, cells were treated with 40 μM LY294002 dissolved in unbuffered DMEM for 1 h. In both IGF-1 and LY294002 treatments, unbuffered DMEM was used as control. Four basal measurements of oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were made followed by the addition of mitochondrial inhibitors. Four readings were made after the injection of each inhibitor. Three mitochondrial inhibitors were used: oligomycin (1 μM) was added to inhibit the ATP synthase followed by the measurement of ATP turn over, FCCP (1 μM) was added to uncouple the mitochondria followed by the measurement of the maximal mitochondrial respiratory capacity, and rote-none (1 μM) was added to inhibit the complex I followed by the measurement of non-mitochondrial respiration.

Mitochondria isolation

Mitochondria were isolated from control and Pten^{immortalized hepatocytes by using a Thermo Scientific Kit (Rockford, IL) by using dounce homogenization followed by centrifugation; the mitochondrial pellet was re-suspended in 15% Percoll and topped on to a pre-formed discontinuous Percoll gradient. The purified mitochondrial fraction was collected after centrifugation. The purity of mitochondrial fraction was assessed by measuring markers of microsomal (NADPH-cytochrome P450 reductase) and cytosolic (lactate dehydrogenase) contamination. In control hepatocytes, NADPH-cytochrome P450 reductase activity was 18.7 nmol/min/mg protein in the cell lysates and 1.19 nmol/min/mg protein in the mitochondrial fraction; lactate dehydrogenase activity was 16.2 μmol/min/mg protein in the cytosol and 2.17 μmol/min/mg protein in the mitochondrial fraction. The activity of lactate dehydrogenase in the cytosol of}Pten^{cells was 25 μmol/min/mg protein and that in the mitochondrial fraction, 3 μmol/min/mg protein; NADPH-cytochrome P450 reductase activity in}Pten^{cell lysates was 19 nmolmin/mg protein and in the mitochondrial fraction 2.5 nmol/min/mg protein.}

Proteinase K treatment of mitochondria

Mitochondria from control and Pten^{hepatocytes were treated with proteinase K (50 μg/ml) at 4°C for 20 min. Phenylmethylsulfonyl fluoride (2 mM) was added to stop the reaction and mitochondria were garnered by centrifugation. Mitochondria were then lysed in 2% CHAPS containing protease and phosphatase inhibitors.}

Western blot analysis

Control and Pten^{hepatocytes treated with various doses of IGF-1, LY294002, or LiCl were lysed in RIPA buffer (50 mM Tris/HCl, pH 7.4, 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, phosphatase inhibitor and protease inhibitor cocktail). Mitochondrial proteins were extracted by incubating isolated mitochondria for 30 min at 4°C in 2% CHAPS containing protease and phosphatase inhibitors. Different amounts (10, 20, 30 or 60 μg) of cell lysates or mitochondrial proteins were loaded in each well of 10–12% sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) gel followed by electrophoresis and proteins were transferred to a polyvinylidine difluoride (PVDF) membrane. The antibodies used for immunoblotting are anti-pAKT, anti-AKT, anti-pGSK3, and anti-GSK3 from Cell Signaling (Danvers, MA); anti-pPDH-E}_1α from EMD Millipore (Billerica, MA) and anti-PDH-E_1α from MitoSciences (Eugene, OR). Band intensities were analyzed by the VersaDoc system (Bio-Rad, Hercules, CA).

Enzyme activity assay

Mitochondrial ATP synthase activity was measured by monitoring ATP hydrolysis (ATPase activity) in the reaction buffer containing 60 mM sucrose, 50 mM triethanolamine-HCl, 50 mM KCl, 4 mM MgCl₂, 2 mM ATP, 2 mM EGTA, 1 mM KCN (pH 8.0, adjusted by KOH) in the presence of 100μM NADH, 5 units/ml pyruvate kinase and 5 units/ml lactate dehydrogenase at 37°C. The reaction was monitored for 5 min and the ATP synthase activity was determined by the changes of absorbance at 340 nm [32, 33]. Mitochondrial PDH activity was measured in PDH assay buffer (35 mM KH₂PO₄, 2 mM KCN, 0.5 mM EDTA, 5 mM MgCl₂, pH 7.25 with KOH) in the presence of 200 mM sodium pyruvate, 2.5 mM rotenone, 4 mM sodium CoA (co-enzyme A), 40 mM TPP (thiamine pyrophosphate) at 37°C. The conversion of NAD (nicotinamide adenine dinucleotide) to NADH (nicotinamide adenine dinucleotide reduced form) was monitored by the change of absorbance at 340 nm [34]. Activity for complexes I, II, and IV was determined using microplate assay kits (Abcam) in which the respective complexes were immunocaptured in the microplate wells, and the enzyme activity was determined by following the change in absorbance at 450, 600, and 550 nm, respectively.

Immunoprecipitation

Mitochondrial proteins (1 mg/ml) prepared from control and Pten^{immortalized hepatocytes were incubated with phospho-AKT-substrate (PAS) antibody conjugated beads (Cell Signaling) for 48 h. The beads with the immune complexes were washed 3 times with DPBS, mixed with reducing loading sample buffer (Thermo Scientific, Rockford, IL), and heated for 5 min at 100°C. Proteins were separated by a 12% SDS-PAGE gel. The gel was stained with Coomassie blue (Biorad, Hercules, CA). To detect the AKT contact and phosphorylation of ATP synthase, mitochondrial proteins (1 mg/ml) were immuno-precipitated with anti-phospho-AKT-substrate antibody conjugated beads and anti-AKT antibody plus protein A/G agarose beads respectively followed by SDS-PAGE electrophoresis. The membrane is incubated with MitoProfile}^{Total OXPHOS Rodent WB Antibody Cocktail (MS604; Mitosciences, Eugene, OR) containing anti-Complex I subunit NDUFB8 (CI-20), anti-Complex II subunit 30kDa (CII-30), anti-Complex III subunit Core 2 (CIII-core2), anti-Complex IV subunit I (CIV-I), and anti-ATP synthase subunit alpha (CV-alpha).}

LC/MS/MS analysis

Proteins were extracted from the Coomassie blue-stained gel followed by LC/MS/MS analysis at the USC Keck School of Medicine Proteomics Core Facility. In-gel trypsin digest: Protein bands from SDS-PAGE were excised from the gels using biopsy punches (Acuderm, Lauderdale, FL). In-gel tryptic digest was carried out using trypsin that was reductively methylated to reduce autolysis. The digestion reaction was carried out overnight at 37 °C. Digestion products were extracted from the gel with a 5% formic acid, 50% (v/v) acetonitrile solution (2×), and one acetonitrile extraction followed by evaporation using an APD SpeedVac (ThermoSavant, Milford, MA). Digested peptide mixture was analyzed by an online tandem LC/MS system. The LC/MS system consists of an Eksigent NanoLC-2D connected to an LTQ XL linear ion trap mass spectrometer equipped with a nanospray ion source (Thermo Fisher Scientific, Waltham, MA). Peptides were separated on a 10 cm reveres phase (RP) spray tip column connected to a 10 cm RP extended column through a MicroTee with in-union high voltage contact as described [35, 36]. Sample was loaded to the column using a 2 ml sample loop. Peptides were eluted at a flow rate of 500 nl/min into the mass spectrometer using the following gradients: 5–50% ACN + 0.1% formic acid over 75 minutes and 50–90% ACN + 0.1% formic acid over 40 min. The LTQ was operated at 2.6 kV spray voltage and 275°C capillary temperature. Full scan was generated and 10 most intensive ions above minimum signal threshold (500 counts) in each full scan were subjected for MS/MS fragmentation. MS/MS experiments were performed with normalized collision energy of 35% and isolation width of 2 m/z. The resulting spectra were searched against Uniport protein database using SEQUEST in BioWorks Browser 3,2EF1. The results were filtered using Δcn and cross-correlation score (Xc) versus charge state. Δcn was always set at ≥ 0.1 and Xc was set at 1.50 for +1 charged peptides, 2.00 for +2 charged peptides, and 2.50 for +3 or higher charged peptides [35, 36].

Statistical analysis

Statistical analysis was performed using student’s t test for unpaired data or ANOVA. P < 0.05 was considered significant.

Cell lines

Chemicals

XF24 extracellular metabolic flux analysis

Mitochondria isolation

Proteinase K treatment of mitochondria

Western blot analysis

Enzyme activity assay

Immunoprecipitation

LC/MS/MS analysis

Statistical analysis

Statistical analysis was performed using student’s t test for unpaired data or ANOVA. P < 0.05 was considered significant.

Results

PI3K/AKT signaling activation induces mitochondrial respiration

Liver is a major organ where lipid and glucose metabolism converge with mitochondrial respiration. The role of PI3K/AKT signaling in mitochondrial respiration was assessed with a hepatocyte cell line established from mouse liver. Treatment of hepatocytes with insulin growth factor-1 (IGF-1)–that induces the activation of the PI3K/AKT pathway–at 50 and 100 ng/ml for 1 h induced a robust activation of AKT as indicated by the increase in its phosphorylation at Ser⁽Fig. 1A). A concomitant increase of the phosphorylated forms of GSK3α and β at Ser^{and Ser}^{, respectively, indicated increased AKT activity in the IGF-1-treated samples (}Fig. 1A).

An external file that holds a picture, illustration, etc.
Object name is nihms441394f1.jpg

Fig. 1

Effect of IGF-1 on AKT phosphorylation status and cellular bioenergetics

(A) Western blots of AKT and GSK3 and their phosphorylated forms. Hepatocytes were incubated in serum-free medium for 6 h and treated with IGF-1 for 1 h. β-actin was used as protein loading controls. (B) Effect of IGF-1 on oxygen consumption rate (OCR). Immortalized hepatocytes (2×10^cells/well) were pre-incubated with IGF-1 for 1 h before starting the measurement of basal respiration. Subsequent additions: 1 μM oligomycin; 1 μM FCCP; 1 μM rotenone. Four readings were made after the injection of each inhibitor. (●) vehicle; (●) plus 50 ng IGF-1/ml; (●) plus 100 ng IGF-1/ml. (C) Quantitation of basal and maximum OCR in vehicle- and IGF-1-treated samples (data from Fig. 1C). Data are the average of 4 time-points each from 4 independent samples. (D) Effect of IGF-1 on extracellular acidification rate (ECAR): assay conditions as in Fig. 1B. All experiments were performed in quadruplicate. Other assay conditions as described in the Materials and Methods section. **P < 0.01.

IGF-1 treatment of hepatocytes resulted in an increase in oxygen consumption rates (OCR) and extracellular acidification rates (ECAR), the former reflecting mitochondrial respiration and the latter anaerobic glycolysis (i.e., lactate formation). IGF-1 pre-treatment for 1 h did not affect the basal OCR, whereas it elicited a substantial increase of maximal respiratory capacity (following the addition of the uncoupler FCCP) (Fig. 1B). At an IGF-1 concentration of 50 ng/ml, the maximal OCR increased by 90.61 ± 2.71 % (compared to the vehicle-treated cells) (Fig. 1B); a further increase was observed following treatment of hepatocytes with IGF-1 at a concentration of 100 ng/ml (128.18 ± 7.79 % higher than vehicle-treated cells) (Fig. 1B). IGF-1 treatment also increased ECAR (30.64 ± 8.64 % and 56.50 ± 5.78 % by 50 ng/ml and 100 ng/ml IGF-1, respectively) indicating an enhanced anaerobic glycolysis (Fig. 1D). These data suggest that IGF-1/PI3K/AKT signaling positively regulates anaerobic glycolysis and mitochondrial maximal respiratory capacity; up-regulation of glycolysis reflects increased substrate supply to mitochondria. Dranka et al [37] proposed an intermediate turnover state in cells (between state 3 and state 4 respiration), termed State_apparent that provides information on the mitochondrial workload and is calculated from the equation State_apparent = 4 − [(Basal − Oligo)/(FCCP − Oligo)] [37, 38]; based on this equation, values of State_apparent for IGF-1-treated hepatocytes were 3.48 ± 0.03 and 3.62 ± 0.03 (for 50- and 100 ng IGF-1/ml, respectively). Likewise, the mitochondrial reserve capacity [37, 38] was substantially increased by IGF-1 treatment (50 ng/ml, 134.30 ± 11.02 pmoles/min; 100 ng/ml, 206.75 ± 19.21 pmoles/min). The low maximal respiratory capacity value (lower than basal OCR) in vehicle-treated hepatocytes was due to low level of active AKT after the 6 h incubation in serum-free medium (a requirement when treating cells with IGF-1) (Fig. 1 A). The data indicate that AKT activation is critical to stimulate and maintain the mitochondrial reserve capacity (measured by the addition of mitochondrial uncoupler, FCCP). When immortalized control hepatocytes are starved with serum free medium for 6 h, AKT phosphorylation level is low and there is almost no reserve capacity. The cell density in the IGF-1 treatment groups was the same as that in the vehicle group. The reserve capacity was restored in the presence of IGF-1.

Inhibition of PI3K/AKT signaling attenuates mitochondrial respiration

Inhibition of PI3K by LY294002 resulted in a decreased AKT and GSK3 phosphorylation (Fig. 2A), the former indicating a lack of activation and the latter release of the inhibitory effect exerted upon phosphorylation at Ser^{of GSK3β. Accordingly, basal and maximal OCRs were reduced (38.8 ± 4.6 and 50.6 ± 2.9, respectively) by treatment with LY294002 as compared with vehicle-treated cells (}Fig. 2B,C). Decreased basal and maximal OCR indicates that the inhibition of the PI3K/AKT signaling by LY294002 treatment leads to the inhibition of coupled and maximal mitochondrial respiration. The inhibition of ECAR by LY294002 treatment implied that anaerobic glycolysis was not compensating for the decrease in mitochondrion-driven O₂ consumption (Fig. 2D). LY294002 treatment decreased the mitochondrial reserve capacity [37, 38] from 161.53 ± 27.51 pmoles/min to 59.87 ± 11.71 pmoles/min) and S_apparent from 3.64 ± 0.05 to 3.51 ± 0.08). These data, along with those following IGF-1 treatment, strengthen the importance of PI3K/AKT signaling on mitochondrial bioenergetics.

An external file that holds a picture, illustration, etc.
Object name is nihms441394f2.jpg

Fig. 2

Effect of PI3K inhibition on AKT and GSK phosphorylation status and cellular bioener-getics

(A) Western blots of AKT and GSK3α/β and their phosphorylated forms. Cells were incubated with 40 μM LY294002 for 1 h. β-actin is detected as protein loading controls. (B) Effect of PI3K inhibitor on oxygen consumption rate (OCR) (assay performed on 2 × 10^{cells/well). Basal OCR rate was measured 30 min after the addition of LY294002. (●) vehicle; (●) plus 40 μM LY294002. (C) Quantitation of basal and maximal OCR in vehicle- and LY29002-treated samples. (D) Effect of LY294002 on extracellular acidification rate (ECAR): assay conditions as in Fig. 2B. All experiments were performed in quadruplicates. Quantitative data in (C) are average of 4 time-points each from 4 independent samples. **}P < 0.01.

Mitochondrial bioenergetics in Pten^{immortalized hepatocytes}

An isogenic murine hepatocyte cell line with constitutive activation of PI3K/AKT signaling was used in order to confirm the role of PI3K/AKT in mitochondrial respiration. The Pten null (Pten^−/−) hepatocyte cell line was established from a 1 month-old mouse carrying a liver-specific deletion of Pten (Pten^{; Alb-Cre}^{) [}25] (Fig. 3A). Immunoblotting analysis revealed that AKT activity–indicated by its phosphorylation status–was substantially increased in the Pten^−/− hepatocytes as compared with the control cell lines with intact Pten (Fig. 3B). Accordingly, GSK3β phosphorylation at Ser^{was increased (}Fig. 3B). Fig. 3C shows a time course of AKT and GSK3α and GSK3β phosphorylation in control- and Pten^−/− hepatocytes following PI3K inhibition by LY294002.

An external file that holds a picture, illustration, etc.
Object name is nihms441394f3.jpg

Fig. 3

AKT and GSK3 phosphorylation status in immortalized hepatocytes from wild- and Pten^{knockout mice}

(A) PTEN protein expression in control- and Pten^{hepatocytes. (B) Western blots of AKT and GSK3 and their phosphorylated forms in cell lysates from control- and}Pten^{hepatocytes. Different amounts of proteins (10-, 30-, and 60 μg) were loaded in three sets of samples. (C) Effect of LY294002 on the time course of p-AKT, p-GSK3α, and p-GSK3β in control- and}Pten^{hepatocytes. β-actin as protein loading controls. Densitometry reading of p-AKT to β-actin ratio (pAKT/β-actin) for each lane is shown at the bottom of the blots.}

Basal OCR was significantly higher in the Pten^−/− than in the control hepatocytes (289.76 ± 3.74 and 154.88 ± 2.88 pmoles/min, respectively). Following the addition of FCCP, maximal hepatocytes (393.32 ± 56.64 pmoles/min) remained significantly higher than OCR in the Pten^−/− that in the controls (213.50 ± 8.14 pmoles/min) (Fig. 4 A,B). The number of control cells and Pten^−/− cells in Fig. 4A was the same, as indicated by similar values of non-mitochondrial respiration (following the addition of rotenone) and measurements of protein concentration in both cell types at the end of the experiment. Basal ECAR was higher in the Pten^−/− hepatocytes (40.48 ± 1.43 mpH/min) than in the controls (26.99 ± 1.99 mpH/min). The higher anaerobic glycolysis level in Pten^−/− hepatocytes indicates a higher substrate supply from glycolysis to mitochondria. Compared with control hepatocytes, Pten^{hepatocytes showed an enhanced mitochondrial reserve capacity (108.8 ± 11.02 pmoles/min) than control hepatocytes (38.35 ± 22.16 pmoles/min;}Pten^{). Taken together, these data confirmed that the PTEN-regulated PI3K/AKT pathway controls bioenergetics by modulating both substrate supply (ECAR) and mitochondrial catalytic/energy transducing capacity (OCR).}

An external file that holds a picture, illustration, etc.
Object name is nihms441394f4.jpg

Fig. 4

Bioenergetics of control- and Pten^hepatocytes

(A) Oxygen consumption rates (OCR) of control- and Pten^{hepatocytes (2 × 10}^{cells/well). Additions (oligomycin, FCCP, and rotenone) as in}Fig. 1B. (●) control hepatocytes; (●) Pten^{hepatocytes. (B) Quantitative data of basal and maximal OCR of control (●)- and}Pten^{(●) hepatocytes (data from Fig. 4A). (C) Extracellular acidification rate (ECAR) of control- and}Pten^{hepatocytes (2 × 10}^{cells/well) (assay conditions as in Fig. 4A). (D) Quantitative data of basal ECAR of control- and}Pten^{hepatocytes. All experiments were performed in quadruplicate. Data in (B) and (D) are the average of 4 time-points from 4 different samples. **}P < 0.01.

The involvement of AKT in the regulation of cellular bioenergetics was further confirmed genetically in primary hepatocytes from a double knockout (Pten^{, Akt2}^−/−) model, which show a reduced level of basal OCR (312.18 ± 18.64 pmoles/min) as compared with Pten^−/− hepatocytes (424.71 ± 16.34 pmoles/min) (not shown). The slight decrease of maximal OCR in Pten^{, Akt2}^−/− hepatocytes (480.82 ± 29.55 pmoles/min) compared with Pten^{hepatocytes (507.19 ± 50.95 pmoles/min) was not statistically significant. An explanation for this observation is that AKT2 mainly regulates the ATP synthase activity and no difference can be observed in the presence of FCCP (not shown).}

Activated AKT translocates to mitochondria

Consistent with previous reports on the translocation of AKT to mitochondria in NIH/3T3 [30] and neuroblastoma [31] cells, a small amount of AKT translocated to mitochondria in both control- and Pten^−/− hepatocytes. Pten^−/− hepatocytes–with a more robust AKT phosphorylation–showed significantly more AKT in mitochondria (Fig. 5A). Expectedly, IGF-1 treatment also induced a higher translocation of AKT to mitochondria. This is more evident in the Pten^−/− cell lines, in which the pAKT/AKT ratio is already high without IGF-1 treatment. When treated with IGF-1, even more AKT translocates to the mitochondria in the Pten^−/− hepatocytes (Fig. 5B). Fig. 5C shows a time-dependent increase in pAKT (Ser^{) levels in both control and}Pten^−/− cell lines in response to IGF-1 treatment; compared with control, Pten^−/− cell lines showed significantly higher increase in pAKT (Ser^{) levels in response to IGF-1.}

An external file that holds a picture, illustration, etc.
Object name is nihms441394f5.jpg

Fig. 5

Mitochondrial localization of AKT

(A) AKT expression in the mitochondrial fraction isolated from control and Pten^{hepatocytes. (B) Time course of AKT expression in the mitochondrial fraction following incubation of hepatocytes with 50 ng IGF-1/ml. Cells (control and}Pten^{hepatocytes) were incubated for 0, 5, 15, and 30 min previous to the isolation of the mitochondrial fraction. (C) Time course p-AKT expression in the cell lysate following incubation for different times with 50 ng IGF-1/ml. L, light exposure; D, dark exposure. (D) Effect of proteinase K on AKT and p-AKT expression in the mitochondrial fraction isolated from control and}Pten^{hepatocytes. TOM20, a mitochondrial outer membrane protein was used as control for proteinase K activity. (E) Relative levels of mitochondrial AKT in control and}Pten^{hepatocytes (data from (D)). COX IV, an inner membrane protein is used for control of mitochondrial protein loading and β-actin used for total cell lysate loading.}

Treatment of isolated mitochondria with proteinase K (that degrades proteins on the outside of the outer mitochondrial membrane) revealed that AKT translocates to mitochondria and localizes in mitochondrial inter-membrane space and mitochondrial matrix (Fig. 5D,E). The sensitivity of TOM20 (a mitochondrial outer-membrane protein) to proteinase K indicates that the experimental condition (incubation with 50 μg/ml of proteinase K for 20 min at 4°C) is sufficient to degrade mitochondrial outer-membrane proteins.

AKT may regulate mitochondrial respiration through GSK3β and ATP synthase

GSK3β which was proposed as a constitutive mitochondrial protein, is a well-characterized AKT phosphorylation target in these organelles [31]. GSK3β was shown to catalyze the phosphorylation (and inactivation) of the pyruvate dehydrogenase complex (PDH) [39], a critical mitochondrial enzyme complex that couples anaerobic glycolysis (in cytosol) to the oxidative decarboxylation of pyruvate to acetyl-CoA and its entry in the tricarboxylic acid cycle. Thus, GSK3β is a potential target of mitochondrial AKT that may be partly responsible for PI3K/AKT induced mitochondrial respiration. Treatment of hepatocytes were treated with LiCl, a reagent commonly used to inhibit GSK3β activity [40], led to increased phosphorylation (inactivation) of GSK3β (Fig. 6A) and a dose-dependent reduction in phosphorylation (inactivation) of PDH-E_1α (Fig. 6B), indicating activation of the enzyme complex. Accordingly, LiCl treatment increased mitochondrial PDH activity in both control (Fig. 7A) and Pten^{hepatocytes (}Fig. 7B). Consistent with the role of PDH as a crucial mitochondrial enzyme, LiCl treatment increased basal and maximal respiration in control (Fig. 8A) and Pten^{hepatocytes (}Fig. 8B). This observation supported a potential role for a mitochondrial AKT-GSK3β-PDH signaling circuitry in regulating mitochondrial basal and maximal respiration. Consistent with the role of GSK3β in PI3K/AKT signaling, the mitochondrial reserve capacity in Pten^−/− hepatocytes increased from 45.25 ± 34.65 pmoles/min to 330.33 ± 96.03 pmoles/min) with LiCl treatment; no significant differences were observed in the mitochondrial reserve capacity in control hepatocytes in response to LiCl treatment. Thus, mitochondrial respiration may be regulated by the coordinated balance between GSK3β (inhibitory effect) and AKT (stimulatory effect). In accordance with the increased phosphorylation level of GSK3β at Ser^{(inactive), phosphorylation level of mitochondrial PDH-E}_1α at Ser^{(inactive) decreased in}Pten^{hepatocytes (}Fig. 9A), correlates with the enhanced PDH activity (Fig. 9B). Therefore, these data confirm that the AKT signaling affects mitochondrial bioenergetics in part through GSK3β regulation of mitochondrial PDH.

An external file that holds a picture, illustration, etc.
Object name is nihms441394f6.jpg

Fig. 6

Effect of LiCl on p-GSK3β and p-PDH in the cell lysate and mitochondrial fraction, respectively

(A) Effect of different concentrations of LiCl on p-GSK3β (Ser^{) and GSK3β expression. Hepatocytes were treated with the different concentrations of LiCl for 4 h. β-actin as protein loading controls. (B) Effect of different concentrations of LiCl on p-PDH-E}_1α (Ser293) and PDH-E_1α expression. COX IV, an inner membrane protein is used for control of mitochondrial protein loading.

An external file that holds a picture, illustration, etc.
Object name is nihms441394f7.jpg

Fig. 7

Effects of LiCl treatment on mitochondrial PDH activity in control and Pten−/− immortalized hepatocytes

The effects of increasing concentration of LiCl on mitochondrial PDH activity in control (A) hepatocytes and Pten^{(B) hepatocytes. Hepatocytes were incubated with difference concentrations (10 and 20 mM) of LiCl for 4h. Mitochondria were isolated after LiCl treatment and PDH activity was measured as described. **}P < 0.01.

An external file that holds a picture, illustration, etc.
Object name is nihms441394f8.jpg

Fig. 8

Effect of LiCl on the bioenergetics of control- and Pten^hepatocytes

(A) Effect of LiCl on the oxygen consumption rate (OCR) of control hepatocytes. Cells (2 × 10^{/well) were incubated for 4 h with 20 mM LiCl before starting the measurement of basal respiration. Additions (oligomycin, FCCP, and rotenone) as in}Fig. 1B. Graph on the right: quantitative data indicating the effect of LiCl on basal- and maximal respiration. (B) Effect of LiCl on the oxygen consumption rate (OCR) of Pten^{hepatocytes. Assay conditions as in (A). Graph on the right: quantitative data indicating the effect of LiCl on basal- and maximal respiration of}Pten^{hepatocytes. **}P < 0.01.

An external file that holds a picture, illustration, etc.
Object name is nihms441394f9.jpg

Fig. 9

Pyruvate dehydrogenase phosphorylation status and activity in mitochondria from control and Pten^{immortalized hepatocytes}

(A) Western blots of mitochondrial pyruvate dehydrogenase and its phosphorylated form (Ser^{) in control- and Pten}^{immortalized hepatocytes. Mitochondria were isolated from control- and Pten}^{immortalized hepatocytes. 20 μg mitochondrial proteins were loaded in each well; (B) Measurement of PDH activity in mitochondria from control- and Pten}^{immortalized hepatocytes. (●)control hepatocytes; (●)Pten}^{hepatocytes. **P < 0.01.}

Other potential mitochondrial AKT phosphorylation targets were examined by pull-down experiments from mitochondrial lysates prepared from control- and Pten^−/− hepatocytes. Immuno-precipitation with pAKT substrate antibody identified 4 bands at ~100, 90, 80, 50 kDa that were significantly stronger in Pten^−/− mitochondrial lysates than in the controls (Fig. 10A). LC/MS/MS analyses revealed several potential targets for phosphorylation by AKT (Table 1): the 100 kDa band was identified as leucine-rich PPR motif-containing protein; the 90 kDa band was identified as monofunctional C1-tetrahydrofolate synthase, lon protease, and NAD(P) transhydrogenase; the 80 kDa band as aconitase, and the 50 kDa band as glutamate dehydrogenase 1, aldehyde dehydrogenase, and ATP synthase subunit α and subunit β. Consistent with these observations, ATP synthase activity was higher in the Pten^−/− hepatocytes than in control cells (Fig. 10B).

An external file that holds a picture, illustration, etc.
Object name is nihms441394f10.jpg

Fig. 10

AKT phosphorylation targets in mitochondrial fraction isolated from control- and Pten^hepatocytes

(A) Commassie blue staining of protein lysates extracted from p-AKT substrate antibody immunoprecipitation assay. Left lane: control hepatocytes; right lane: Pten^{hepatocytes. The arrows indicate the bands subjected to LC/MS/MS analyses:}a: leucine-rich PPR motif-containing protein; b: monofunctional C1-tetrahydrofolate synthase, lon protease homolog, NAD(P)H transhydrogenase; c: aconitase; d: glutamate dehydrogenase 1, aldehyde dehydrogenase, ATP synthase subunit α, ATP synthase subunit β (see Table 1). (B) ATP synthase activity of control- and Pten^{hepatocytes (assay conditions as described in the Materials and Methods section). **}P < 0.01. (C) Structural features of ATP synthase α protein sequence. Triangles indicate ATP synthase α–β interacting interface. A reverse substrate consensus sequence for AKT substrate is identified from residue 166–171.

Table 1

LC/MS/MS analyses. Xc and Δcn values were calculated using TurboSequest software. Expect scores below 0.05 were considered significant.

	Xc	Δcn	Expect scores
Leucine-rich PPR motif-containing protein
K.DESSDNFGSFFLR.H	3.82	0.49	1.07 × 10⁻⁹
R.SCGSLLPELSLAER.T	3.6	0.52	1.33 × 10⁻⁴
Monofunctional C1-tetrahydrofolate synthase
K.LIAEESVSLLAAALR.I	3.98	0.49	2.4 × 10⁻⁵
K.YVLVAGITPTPLGEGK.S	4.28	0.52	7.8 × 10⁻⁵
R.FQFLYDVQLPIVEK.I	3.84	0.44	7.50 × 10⁻¹¹
Lon protease homolog
K.AQLSAAVLTLLIK.Q	4.82	0.53	2.2 × 10⁻⁶
K.HVMDVVDEELSK.L	4.07	0.44	1.85 × 10⁻⁵
K.LALLDNHSSEFNVTR.N	4.42	0.61	5.43 × 10⁻⁸
NAD(P)H transhydrogenase
K.ILIVGGGVAGLASAGAAK.S	4.19	0.53	2.9 × 10⁻⁵
Aconitase
K.IVYGHLDDPANQEIER.G	4.69	0.55	2.2 × 10⁻¹⁰
K.LTGSLSGWTSPK.D	3.47	0.56	1.9 × 10⁻⁵
K.QGLLPLTFADPSDYNK.I	3.2	0.55	2.0 × 10⁻⁵
ATP synthase subunit α
K.TSIAIDTIINQK.R	3.75	0.42	2.83 × 10⁻⁶
R.ILGADTSVDLEETGR.V	4.83	0.64	1.03 × 10⁻¹²
R.NVQAEEMVEFSSGLK.G	3.8	0.57	1.68 × 10⁻⁷
R.TGAIVDVPVGEELLGR.V	4.48	0.57	1.22 × 10⁻⁹
ATP synthase subunit β
R.FTQAGSEVSALLGR.I	4.23	0.62	6.59 × 10⁻⁸
R.IMNVIGEPIDER.G	4.14	0.37	9.76 × 10⁻⁵
R.IPSAVGYQPTLATDMGTMQER.I	4.69	0.65	4.59 × 10⁻¹¹

ATP synthase subunit α contains the conserved sequence to the consensus RxRxxS/T sequence for a potential AKT substrate (Table 2). This sequence is located at a structure characteristic where the subunit α of ATP synthase may interact with the β subunit (Fig. 10C). Sequence analysis for ATP synthase α revealed that the consensus sequence segment is conserved in several species analyzed including human, mouse, rat, and E. coli. ATP synthase β has similar sequence features as a substrate for AKT (Table 2) and had been previously shown to be a potential substrate for AKT [31]. These analyses suggest that the ATP synthase enzymes are likely substrates for AKT and that their phosphorylation might be partly responsible for the higher OCR measurements after FCCP treatment in the Pten^−/− and IGF-1 treated cultured cells.

Table 2

Sequence alignment for ATP synthase subunits

Subunit α sequence T/SxxRxR^*
166–171	kgpigSktRrRvglkapgii	mouse
166–171	kgpvgSkiRrRvglkap	rat
166–171	kgpigSktRrRvglkapg	human
247–252	valmgEyfRdRgedaliiy	E. coli

Subunit β sequence T/SxxRxR
453–458	seedklTvsRaRkiqrfl	mouse
399–404	seedklTvsRaRkiqrfl	rat
453–458	seedklTvsRaRkiqrfl	human
390–406	seedklVvaRaRkiqrfl	E. coli

^{Reserved sequence of AKT substrate consensus sites}

The assumption that AKT contacts and phosphorylates ATP synthase α was confirmed by immuno-precipitation and western-blotting analysis (Fig. 11A). Using a phospho-AKT substrate antibody (Fig. 11A, lanes 1 and 2) and total AKT antibody (Fig. 11A, lanes 3 and 4) to immune-precipitate substrates for AKT from the mitochondrial lysate followed by immunoblotting, a clear association of complex V ATP synthase 5α with AKT was observed. The levels of ATP synthase 5α associated with AKT were higher in mitochondria from Pten^−/− (Fig. 11A, lane 4) than those from control hepatocytes (Fig. 11A, lane 3).

An external file that holds a picture, illustration, etc.
Object name is nihms441394f11.jpg

Fig. 11

Activation of AKT alters the respiration complexes

(A) Mitochondrial proteins (1mg/ml) were immunoprecipitated with anti-phospho-AKT-substrate antibody (PAS, lane 1 and 2, control and Pten^{) and anti-AKT antibody (lane 3 and 4, control and}Pten^{) followed by SDS-PAGE electrophoresis. The membrane was incubated with MitoProfile}^{Total OXPHOS Rodent WB Antibody Cocktail (MS604) containing anti-Complex I subunit NDUFB8 (CI-20), anti-Complex II subunit 30kDa (CII-30), anti-Complex III subunit Core 2 (CIII-core2), anti-Complex IV subunit I (CIV-I), and anti-ATP synthase subunit α (CV-α). Mitochondrial proteins were loaded as control (lane 5 and 6, control and}Pten^{). (B) Expression of mitochondrial complexes in total cell lysates. Western blots of mitochondrial complexes in control- and}Pten^{immortalized hepatocytes. Mitochondria were isolated from control- and}Pten^{immortalized hepatocytes. 20 μg mitochondrial proteins were loaded in each well. The membrane was incubated with MitoProfile}^{Total OXPHOS Rodent WB Antibody Cocktail (MS604); Top, boiled samples; bottom, unboiled samples. (C) Measurement of complexes I, II, and IV activities in mitochondria from control- and}Pten^{immortalized hepatocytes. (●) control hepatocytes; (●)}Pten^{hepatocytes. **}P < 0.01.

In addition, complex I expression was higher in mitochondrial lysate from Pten^−/− cells (Fig. 11A, lane 6) than that from control hepatocytes (Fig. 11A, lane 5). This was further confirmed with immunoblotting of the respiratory chain complexes (Fig. 11B) showing a higher expression of complex I in the mitochondrial fraction from Pten^{cells than that from control hepatocytes; furthermore, complex I activity was significantly higher in}Pten^{hepatocytes (}Fig. 11C); no statistically significant changes were observed in complex II and complex IV activities. Thus, AKT activation is associated an increased mitochondrial electron transfer capacity independent of its direct kinase activity in mitochondria. A transcriptional regulation may be involved in this mechanism.

PI3K/AKT signaling activation induces mitochondrial respiration

Fig. 1

Effect of IGF-1 on AKT phosphorylation status and cellular bioenergetics

Inhibition of PI3K/AKT signaling attenuates mitochondrial respiration

Fig. 2

Effect of PI3K inhibition on AKT and GSK phosphorylation status and cellular bioener-getics

Mitochondrial bioenergetics in Pten^{immortalized hepatocytes}

Fig. 3

AKT and GSK3 phosphorylation status in immortalized hepatocytes from wild- and Pten^{knockout mice}

Fig. 4

Bioenergetics of control- and Pten^hepatocytes

Activated AKT translocates to mitochondria

Fig. 5

Mitochondrial localization of AKT

AKT may regulate mitochondrial respiration through GSK3β and ATP synthase

Fig. 6

Effect of LiCl on p-GSK3β and p-PDH in the cell lysate and mitochondrial fraction, respectively

Fig. 7

Effects of LiCl treatment on mitochondrial PDH activity in control and Pten−/− immortalized hepatocytes

Fig. 8

Effect of LiCl on the bioenergetics of control- and Pten^hepatocytes

Fig. 9

Pyruvate dehydrogenase phosphorylation status and activity in mitochondria from control and Pten^{immortalized hepatocytes}

Fig. 10

AKT phosphorylation targets in mitochondrial fraction isolated from control- and Pten^hepatocytes

Table 1

LC/MS/MS analyses. Xc and Δcn values were calculated using TurboSequest software. Expect scores below 0.05 were considered significant.

	Xc	Δcn	Expect scores
Leucine-rich PPR motif-containing protein
K.DESSDNFGSFFLR.H	3.82	0.49	1.07 × 10⁻⁹
R.SCGSLLPELSLAER.T	3.6	0.52	1.33 × 10⁻⁴
Monofunctional C1-tetrahydrofolate synthase
K.LIAEESVSLLAAALR.I	3.98	0.49	2.4 × 10⁻⁵
K.YVLVAGITPTPLGEGK.S	4.28	0.52	7.8 × 10⁻⁵
R.FQFLYDVQLPIVEK.I	3.84	0.44	7.50 × 10⁻¹¹
Lon protease homolog
K.AQLSAAVLTLLIK.Q	4.82	0.53	2.2 × 10⁻⁶
K.HVMDVVDEELSK.L	4.07	0.44	1.85 × 10⁻⁵
K.LALLDNHSSEFNVTR.N	4.42	0.61	5.43 × 10⁻⁸
NAD(P)H transhydrogenase
K.ILIVGGGVAGLASAGAAK.S	4.19	0.53	2.9 × 10⁻⁵
Aconitase
K.IVYGHLDDPANQEIER.G	4.69	0.55	2.2 × 10⁻¹⁰
K.LTGSLSGWTSPK.D	3.47	0.56	1.9 × 10⁻⁵
K.QGLLPLTFADPSDYNK.I	3.2	0.55	2.0 × 10⁻⁵
ATP synthase subunit α
K.TSIAIDTIINQK.R	3.75	0.42	2.83 × 10⁻⁶
R.ILGADTSVDLEETGR.V	4.83	0.64	1.03 × 10⁻¹²
R.NVQAEEMVEFSSGLK.G	3.8	0.57	1.68 × 10⁻⁷
R.TGAIVDVPVGEELLGR.V	4.48	0.57	1.22 × 10⁻⁹
ATP synthase subunit β
R.FTQAGSEVSALLGR.I	4.23	0.62	6.59 × 10⁻⁸
R.IMNVIGEPIDER.G	4.14	0.37	9.76 × 10⁻⁵
R.IPSAVGYQPTLATDMGTMQER.I	4.69	0.65	4.59 × 10⁻¹¹

Table 2

Sequence alignment for ATP synthase subunits

Subunit α sequence T/SxxRxR^*
166–171	kgpigSktRrRvglkapgii	mouse
166–171	kgpvgSkiRrRvglkap	rat
166–171	kgpigSktRrRvglkapg	human
247–252	valmgEyfRdRgedaliiy	E. coli

Subunit β sequence T/SxxRxR
453–458	seedklTvsRaRkiqrfl	mouse
399–404	seedklTvsRaRkiqrfl	rat
453–458	seedklTvsRaRkiqrfl	human
390–406	seedklVvaRaRkiqrfl	E. coli

^{Reserved sequence of AKT substrate consensus sites}

Fig. 11

Activation of AKT alters the respiration complexes

Discussion

The PI3K/AKT signaling pathway regulates both growth/survival and metabolism; however, it is not clear how these two functions are integrated. One potential link may be surmised to be integrated in mitochondria, organelles involved energy-transducing process and the regulation of apoptotic and survival pathways. These functions render mitochondria viable candidates for integrating the PI3K/AKT signaling pathway as well as other signaling pathways. This study showed that PI3K/AKT signaling positively regulates mitochondrial bioenergetics in experimental models entailing IGF-1-driven stimulation of the PI3K/AKT route (Fig. 1), inhibition of PI3K (Fig. 2), and a Pten^−/− cell model with a robust insulin signaling (Figs. 3,,44).

The increased mitochondrial respiratory function correlated with the inhibition of GSK3β by AKT. Inhibition of GSK3β by LiCl exerted a decrease in phosphorylation (inactivation) of PDH-E_1α at Ser^{and an increased mitochondrial basal and maximal respiration, indicating that GSK3β negatively regulates mitochondrial bioenergetics through modulating PDH activity. AKT, therefore, may partly regulate mitochondrial bioenergetics through inhibiting GSK3β. In rat hippocampal neurons, GSK3β was found to be a mitochondrial resident protein that phosphorylated and inhibited PDH [}39]. It may be surmised that mitochondrial pAKT phosphorylates and inhibits GSK3β, thus changing the pPDH_inactive/PDH_active ratio. This function of AKT may represent a feed forward reaction that prevents the product of cytosolic glycolysis and mitochondrial PDH substrate–pyruvate–from buildup while simultaneously AKT activation induces glucose uptake, hexokinase-2/VDAC interaction, and glucose metabolism to pyruvate. PDK1, the enzyme that phosphorylates AKT on the second site Thr^{to achieve full activation, has been shown previously to be constitutively active in the mitochondrial inter-membrane space [}30]. This observation further confirmed a role of PDK1-AKT-GSK3β signaling axis in regulating mitochondrial function. However, stress conditions and aging are associated with translocation of JNK to mitochondria in primary cortical neurons [41, 42]; at variance with AKT (this study and [30, 31]), JNK associated with the outer mitochondrial membrane (proteinase K-sensitive) and triggered a phosphorylation cascade that involved PDK1 and resulted in phosphorylation and inhibition of PDH [41, 42].

The increased mitochondrial bioenergetics may also be attributed to the translocation of pAKT to mitochondria where it phosphorylates a number of other mitochondrial proteins including the two isoforms of ATP synthase, α and β. ATP synthase is a multi-subunit complex composed by an ATPase and a proton channel and it converts the membrane potential generated by the proton gradient to ATP. The α and β subunits are part of the ATPase complex and function as dimers; the sequence of ATP synthase, subunit α phosphorylated by AKT is part of the α–β interacting interfaces and its phosphorylation correlates with an increased ATPase activity. These observations imply that AKT-mediated phosphorylation of ATP synthase α affects its ability to interact with ATP synthase β and thus may affect directly its function. Though not an AKT phosphorylation substrate, complex I expression and activity were increased in mitochondrial lysates from Pten^{cells (as compared to control hepatocytes). How a robust PI3K/AKT signaling and the effects of AKT translocation to mitochondria affect complex I expression and activity remains to be elucidated.}

Recent evidence suggests that the ‘Warburg effect’ (enhanced glycolytic rates in aerobic conditions) is a hallmark of tumor cells and may contribute to the cause of tumorigenesis [43]. Upregulation of PI3K/AKT signaling–particularly because loss of PTEN function–is a common occurrence in human tumor specimens. However, how metabolic changes observed as Warburg effects contribute to this process are still unclear. The PI3K/AKT signaling pathway may be viewed as a major candidate accounting for some of the Warburg effects observed in cancer cells due to its dual role in growth and metabolism [2]. Transfection of Rat1a cells with a constitutively active form of AKT robustly increased the association of hexokinase with outer mitochondrial membrane VDAC [44], an effect that was critical for the anti-apoptotic function of AKT [45]. The ability of AKT to inhibit apoptosis is dependent on this association of hexokinase with the mitochondrial membrane [44, 45]. These studies, thus, suggest that mitochondria are critical for linking the metabolic and cell survival effects of PI3K/AKT signaling.

Taken together, this study demonstrated that PI3K/AKT activation leads to increased mitochondrial respiration through: 1) increasing substrate supply to mitochondria (regulation of cytosolic processes by AKT); 2) enhancing the mitochondrial catalytic machinery (changes in pPDH/PDH ratios and a stronger metabolism of pyruvate (PDH activity)) upon inhibition of GSK3β and 3) up-regulation of mitochondrial electron transfer and energy-transduction capacity, the former may entail the increased expression level and activity of complex I whereas the latter may involve a direct phosphorylation of ATP synthase by AKT.

Acknowledgments

This work was supported by P30DK48522 (to BLS) and AG16718 (to EC).

^{Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of Southern California, Los Angeles, CA 90089, USA}

^{Biochemistry and Molecular Biology, University of Southern California, Los Angeles, CA 90089, USA}

^{Pathology, Keck School of Medicine, University of Southern California, Los Angeles, CA 90089, USA}

^{Corresponding author at: University of Southern California, Pharmacology and Pharmaceutical Sciences, School of Pharmacy, Los Angeles, CA 90089, United States. Fax: +1 323 224 7473.}ude.csu@sanedac

Publisher's Disclaimer

Abstract

Regulation of cellular bioenergetics by PI3K/AKT signaling was examined in isogenic hepatocyte cell lines lacking the major inhibitor of PI3K/AKT signaling, PTEN (phosphatase and tensin homolog deleted on Chromosome 10). PI3K/AKT signaling was manipulated using the activator (IGF-1) and the inhibitor (LY 294002) of the PI3K/AKT pathway. Activation of PI3K/AKT signaling resulted in an enhanced anaerobic glycolysis and mitochondrial respiration. AKT, when phosphorylated and activated, translocated to mitochondria and localized within the membrane structure of mitochondria, where it phosphorylated a number of mitochondrial residence proteins including the subunits α and β of ATP synthase. Inhibition of GSK3β by either phosphorylation by AKT or lithium chloride resulted in activation of pyruvate dehydrogenase, i.e., decrease of its phosphorylated form. AKT-dependent phosphorylation of ATP synthase subunits α and β resulted in an increased complex activity. AKT translocation to mitochondria was associated with an increased expression and activity of complex I. These data suggest that the mitochondrial signaling pathway AKT-GSK3β-PDH, AKT-dependent phosphorylation of ATP synthase, and upregulation of mitochondrial complex I expression and activity are involved in the control of mitochondrial bioenergetics by increasing substrate availability and regulating the mitochondrial catalytic/energy-transducing capacity.

Keywords: Mitochondrial respiration, PI3K, AKT, PTEN, GSK3β, pyruvate dehydrogenase, ATP synthase, hepatocytes, bioenergetics

Abstract

Highlights

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Footnotes

References

1. Wymann MP, Marone RPhosphoinositide 3-kinase in disease: timing, location, and scaffolding. Curr Opin Cell Biol. 2005;17:141–149.[PubMed][Google Scholar]
2. Cantley LCThe phosphoinositide 3-kinase pathway. Science. 2002;296:1655–1657.[PubMed][Google Scholar]
3. Whitman M, Downes CP, Keeler M, Keller T, Cantley LType I phosphatidylinositol kinase makes a novel inositol phospholipid, phosphatidylinositol-3-phosphate. Nature. 1988;332:644–646.[PubMed][Google Scholar]
4. Harlan JE, Hajduk PJ, Yoon HS, Fesik SWPleckstrin homology domains bind to phosphatidylinositol-4,5-bisphosphate. Nature. 1994;371:168–170.[PubMed][Google Scholar]
5. Klarlund JK, Guilherme A, Holik JJ, Virbasius JV, Chawla A, Czech MPSignaling by phosphoinositide-3,4,5-trisphosphate through proteins containing pleckstrin and Sec7 homology domains. Science. 1997;275:1927–1930.[PubMed][Google Scholar]
6. Petersen KF, Befroy D, Dufour S, Dziura J, Ariyan C, Rothman DL, DiPietro L, Cline GW, Shulman GIMitochondrial dysfunction in the elderly: possible role in insulin resistance. Science. 2003;300:1140–1142.[Google Scholar]
7. Shelley P, Martin-Gronert MS, Rowlerson A, Poston L, Heales SJ, Hargreaves IP, McConnell JM, Ozanne SE, Fernandez-Twinn DSAltered skeletal muscle insulin signaling and mitochondrial complex II–III linked activity in adult offspring of obese mice. Am J Physiol Regul Integr Comp Physiol. 2009;297:R675–R681.[Google Scholar]
8. Cheng Z, Guo S, Copps K, Dong X, Kollipara R, Rodgers JT, Depinho RA, Puigserver P, White MFFoxo1 integrates insulin signaling with mitochondrial function in the liver. Nature Med. 2009;15:1307–1311.[Google Scholar]
9. Vianna CR, Huntgeburth M, Coppari R, Choi CS, Lin J, Krauss S, Barbatelli G, Tzameli I, Kim YB, Cinti S, Shulman GI, Spiegelman BM, Lowell BBHypomorphic mutation of PGC-1beta causes mitochondrial dysfunction and liver insulin resistance. Cell Metab. 2006;4:453–464.[Google Scholar]
10. Stiles BLPI-3-K and AKT: Onto the mitochondria. Adv Drug Deliv Rev. 2009;61:1276–1282.[PubMed][Google Scholar]
11. Galicia VA, He L, Dang H, Kanel G, Vendryes C, French BA, Zeng N, Bayan JA, Ding W, Wang KS, French S, Birnbaum MJ, Rountree CB, Stiles BExpansion of hepatic tumor progenitor cells in Pten-null mice requires liver injury and is reversed by loss of AKT2. Gastroenterology. 2010;139:2170–2182.[Google Scholar]
12. Stiles BEssential role of AKT-1/protein kinase B alpha in PTEN-controlled tumorigenesis. Mol Cell Biol. 2002;22:3842–3851.[Google Scholar]
13. Stiles B, Groszer M, Wang S, Jiao J, Wu HPTENless means more. Dev Biol. 2004;273:175–184.[PubMed][Google Scholar]
14. Stiles BL, Kuralwalla-Martinez C, Guo W, Gregorian C, Wang Y, Tian J, Magnuson MA, Wu HSelective deletion of Pten in pancreatic beta cells leads to increased islet mass and resistance to STZ-induced diabetes. Mol Cell Biol. 2006;26:2772–2781.[Google Scholar]
15. Khan CRThe molecular mechanism of insulin action. Annu Rev Med. 1985;36:429–451.[PubMed][Google Scholar]
16. Kurlawalla-Martinez C, Stiles B, Wang Y, Devaskar SU, Kahn BB, Wu HInsulin hypersensitivity and resistance to streptozotocin-induced diabetes in mice lacking PTEN in adipose tissue. Mol Cell Biol. 2005;25:2498–2510.[Google Scholar]
17. Stiles B, Wang Y, Stahl A, Bassilian S, Lee WP, Kim YJ, Sherwin R, Devaskar S, Lesche R, Magnuson MA, Wu HLiver-specific deletion of negative regulator Pten results in fatty liver and insulin hypersensitivity. Proc Natl Acad Sci USA. 2004;101:2082–2087.[Google Scholar]
18. Franke TF, Yang SI, Chan TO, Datta K, Kazlauskas A, Morrison DK, Kaplan DR, Tsichlis PNThe protein kinase encoded by the Akt proto-oncogene is a target of the PDGF-activated phosphatidylinositol 3-kinase. Cell. 1995;81:727–736.[PubMed][Google Scholar]
19. Kohn AD, Takeuchi F, Roth RAAkt, a pleckstrin homology domain containing kinase, is activated primarily by phosphorylation. J Biol Chem. 1996;271:21920–21926.[PubMed][Google Scholar]
20. Hopkin KA surprising function for the PTEN tumor suppressor. Science. 1998;282:1027–1230.[PubMed][Google Scholar]
21. Maehama T, Dixon JEThe tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-trisphosphate. J Biol Chem. 1998;273:13375–13378.[PubMed][Google Scholar]
22. Rountree CB, Ding W, He L, Stiles BExpansion of CD133-expressing liver cancer stem cells in liver-specific phosphatase and tensin homolog deleted on chromosome 10-deleted mice. Stem Cells. 2009;27:290–299.[Google Scholar]
23. Stanger BZ, Stiles B, Lauwers GY, Bardeesy N, Mendoza M, Wang Y, Greenwood A, Cheng KH, McLaughlin M, Brown D, Depinho RA, Wu H, Melton DA, Dor YPten constrains centroacinar cell expansion and malignant transformation in the pancreas. Cancer Cell. 2005;8:185–195.[PubMed][Google Scholar]
24. Stiles BLPhosphatase and tensin homologue deleted on chromosome 10: extending its PTENtacles. Int J Biochem Cell Biol. 2009;41:757–761.[Google Scholar]
25. Xu X, Kobayashi S, Qiao W, Li C, Xiao C, Radaeva S, Stiles B, Wang RH, Ohara N, Yoshino T, LeRoith D, Torbenson MS, Gores GJ, Wu H, Gao B, Deng CXInduction of intrahepatic cholangiocellular carcinoma by liver-specific disruption of Smad4 and Pten in mice. J Clin Invest. 2006;116:1843–1852.[Google Scholar]
26. He L, Hou X, Kanel G, Zeng N, Galicia V, Wang Y, Yang J, Wu H, Birnbaum MJ, Stiles BLThe critical role of AKT2 in hepatic steatosis induced by PTEN loss. Am J Pathol. 2010;176:2302–2308.[Google Scholar]
27. Leavens KF, Easton RM, Shulman GI, Previs SF, Birnbaum MJAkt2 is required for hepatic lipid accumulation in models of insulin resistance. Cell Metab. 2009;10:405–418.[Google Scholar]
28. Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y, Greenberg MEAkt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell. 1997;91:231–241.[PubMed][Google Scholar]
29. Pastorino JG, Hoek JBEthanol potentiates tumor necrosis factor-alpha cytotoxicity in hepatoma cells and primary rat hepatocytes by promoting induction of the mitochondrial permeability transition. Hepatology. 2000;31:1141–1152.[PubMed][Google Scholar]
30. Antico Arciuch VG, Galli S, Franco MC, Lam PY, Cadenas E, Carreras MC, Poderoso JJAkt1 intramitochondrial cycling is a crucial step in the redox modulation of cell cycle progression. PLoS One. 2009;4:e7523.[Google Scholar]
31. Bijur GN, Jope RSRapid accumulation of Akt in mitochondria following phosphatidylinositol 3-kinase activation. J Neurochem. 2003;87:1427–1435.[Google Scholar]
32. Das AMRegulation of mitochondrial ATP synthase activity in human myocardium. Clin Sci (Lond) 1998;94:499–504.[PubMed][Google Scholar]
33. Rosing J, Harris DA, Kemp AJ, Slater ECNucleotide-binding properties of native and cold-treated mitochondrial ATPase. Biochim Biophys Acta. 1975;376:13–26.[PubMed][Google Scholar]
34. Gohil K, Jones DAA sensitive spectrophotometric assay for pyruvate dehydrogenase and oxoglutarate dehydrogenase complexes. Biosci Rep. 1983;3:1–9.[PubMed][Google Scholar]
35. Wang LC, Okitsu CY, Kochounian CH, Rodriguez A, Hsieh CL, Zandi EA simple and inexpensive on-column frit fabrication method for fused-silica capillaries for increased capacity and versatility in LC-MS/MS applications. Proteomics. 2008;8:1758–1761.[PubMed][Google Scholar]
36. Garcia J, Han D, Sancheti H, Yap LP, Kaplowitz N, Cadenas ERegulation of mitochondrial glutathione redox status and protein glutathionylation by respiratory substrates. J Biol Chem. 2010;285:39646–39654.[Google Scholar]
37. Dranka BP, Benavides GA, Diers AR, Giordano S, Zelickson BR, Reily C, Zou L, Chatham JC, Hill BG, Zhang J, Landar A, Darley-Usmar VMAssessing bioenergetic function in response to oxidative stress by metabolic profiling. Free Radic Biol Med. 2011;51:1621–1635.[Google Scholar]
38. Sansbury BE, Jones SP, Riggs DW, Darley-Usmar VM, Hill BGBioenergetic function in cardiovascular cells: the importance of the reserve capacity and its biological regulation. Chem Biol Interact. 2011;191:288–295.[Google Scholar]
39. Hoshi M, Takashima A, Noguchi K, Murayama M, Sato M, Kondo S, Saitoh Y, Ishiguro K, Hoshino T, Imahori K. Regulation of mitochondrial pyruvate dehydrogenase activity by tau protein kinase I/glycogen synthase. kinase 3beta in brain. Proc Natl Acad Sci USA. 1996;93:2719–2723.
40. Jope RSLithium and GSK-3: one inhibitor, two inhibitory actions, multiple outcomes. Trends Pharmacol Sci. 2003;24:441–443.[PubMed][Google Scholar]
41. Zhou Q, Lam PY, Han D, Cadenas Ec-Jun N-terminal kinase regulates mitochondrial bioenergetics by modulating pyruvate dehydrogenase activity in primary cortical neurons. J Neurochem. 2008;104:325–335.[PubMed][Google Scholar]
42. Zhou Q, Lam PY, Han D, Cadenas EActivation of c-Jun-N-terminal kinase and decline of mitochondrial pyruvate dehydrogenase activity during brain aging. FEBS Lett. 2009;583:1132–1140.[Google Scholar]
43. Vander Heiden MG, Cantley LC, Thompson CBUnderstanding the Warburg effect: the metabolic requirements of cell proliferation. Science. 2009;324:1029–1033.[Google Scholar]
44. Gottlob K, Majewski N, Kennedy S, Kandel E, Robey RB, Hay NInhibition of early apoptotic events by Akt/PKB is dependent on the first committed step of glycolysis and mitochondrial hexokinase. Genes Dev. 2001;15:1406–1418.[Google Scholar]
45. Majewski N, Nogueira V, Bhaskar P, Coy PE, Skeen JE, Gottlob K, Chandel NS, Thompson CB, Robey RB, Hay NHexokinase-mitochondria interaction mediated by Akt is required to inhibit apoptosis in the presence or absence of Bax and Bak. Mol Cell. 2004;16:819–830.[PubMed][Google Scholar]

PI3K/AKT Signaling Regulates Bioenergetics in Immortalized Hepatocytes

Introduction

Materials and Methods

Cell lines

Chemicals

XF24 extracellular metabolic flux analysis

Mitochondria isolation

Proteinase K treatment of mitochondria

Western blot analysis

Enzyme activity assay

Immunoprecipitation

LC/MS/MS analysis

Statistical analysis

Cell lines

Chemicals

XF24 extracellular metabolic flux analysis

Mitochondria isolation

Proteinase K treatment of mitochondria

Western blot analysis

Enzyme activity assay

Immunoprecipitation

LC/MS/MS analysis

Statistical analysis

Results

PI3K/AKT signaling activation induces mitochondrial respiration

Inhibition of PI3K/AKT signaling attenuates mitochondrial respiration

Mitochondrial bioenergetics in Pten immortalized hepatocytes

Activated AKT translocates to mitochondria

AKT may regulate mitochondrial respiration through GSK3β and ATP synthase

Table 1

Table 2

PI3K/AKT signaling activation induces mitochondrial respiration

Inhibition of PI3K/AKT signaling attenuates mitochondrial respiration

Mitochondrial bioenergetics in Pten immortalized hepatocytes

Activated AKT translocates to mitochondria

AKT may regulate mitochondrial respiration through GSK3β and ATP synthase

Table 1

Table 2

Discussion

Acknowledgments

Abstract

Footnotes

References

Mitochondrial bioenergetics in Pten^{immortalized hepatocytes}

Mitochondrial bioenergetics in Pten^{immortalized hepatocytes}