|Home | About | Journals | Submit | Contact Us | Français|
Mitochondrial dysfunction is often associated with aging and neurodegeneration. c-Jun N-terminal kinase (JNK) phosphorylation and its translocation to mitochondria increased as a function of age in rat brain. This was associated with a decrease of pyruvate dehydrogenase (PDH) activity upon phosphorylation of the E1α subunit of PDH. Phosphorylation of PDH is likely mediated by PDH kinase, the protein levels and activity of which increased with age. ATP levels were diminished, whereas lactic acid levels increased, thus indicating a shift toward anaerobic glycolysis. The energy transduction deficit due to impairment of PDH activity during aging may be associated with JNK signaling.
Aging is marked by a general decline of physiological functions, including a pronounced effect on brain activities, such as neuromuscular coordination, cognitive performance, and environmental awareness . The decrease in these neurological activities during normal brain aging has been found to be related to oxidative stress , mitochondrial dysfunction  (mitochondria have become a focal point of the free radical theory of aging ), and dysregulation of cell redox signaling [4–6].
Within these notions, c-Jun N-terminal kinase (JNK) –a stress-activated protein kinase (SAPK) [7–10] and a member of the mitogen-activated protein kinase (MAPK) subfamily– is considered to be a central signal transducer in neuronal death in the mammalian brain  and, among others, functions as a signal transducer that conveys cytosolic oxidative stress signals to mitochondria . Oxidative stress-induced JNK activation entails the phosphorylation of its threonine and tyrosine residues at specific positions by upstream JNK kinases (MAP kinase kinases, MKK). Three major isoforms of JNK have been identified: JNK1 and JNK2 are expressed ubiquitously, whereas the expression of JNK3 appears to be limited to the brain, heart, and testis [13,14]. These three JNK isoforms exhibit differences in specificity toward substrates and binding proteins and in their regulation by upstream kinases and scaffold proteins [13,15]. The activation (phosphorylation) of JNK leads to its translocation to the outer mitochondrial membrane, from where it triggers a phosphorylation cascade affecting different mitochondrial targets [12,16]. Activation of JNK pathways was also shown to enhance neuronal cell death in cultured primary neurons and, conversely, JNK knockout mouse models show protection against excitotoxicity, MPTP, and hypoxia [17–20]. The activity of JNK is significantly increased in the brains of patients with Parkinson’s or Alzheimer’s disease [21,22].
Mitochondria play a key role in brain aging, as these organelles are (a) the sites of energy transduction, (b) major cellular sources of oxidants, (c) targets for radical damaging effects, and (d) sources of ‘redox’ signaling molecules and pro-apoptotic factors . Pyruvate dehydrogenase plays a fundamental role in mitochondrial bioenergetics, for this enzyme complex bridges the anaerobic and aerobic brain energy metabolism, and it is the entry point of carbohydrates into the tricarboxylic acid cycle in the form of acetyl-CoA units. The activity of the pyruvate dehydrogenase complex is regulated at different levels, one of them being phosphorylation/dephosphorylation: phosphorylation by specific pyruvate dehydrogenase kinases (PDK) leads to inactivation of the complex, where dephosphorylation (catalyzed by a specific phosphatase) to its reactivation. Previous work from this laboratory showed that activation of JNK by H2O2 or anisomycin in primary cortical neurons led to its translocation to mitochondria  and that the mitochondrion-associated active JNK induced phosphorylation of the pyruvate dehydrogenase and, thereby, its inhibition.
The goals of this study work are to assess (i) JNK activation in brain as a function of age, (ii) the ensuing translocation to mitochondria, (iii) JNK-mediated modulation of mitochondrial bioenergetics, and (iv) the physiological consequences inherent in these processes.
Antibodies against JNK1, JNK2, PDK-2 and COX were purchased from Santa Cruz Biotech (Santa Cruz, CA). Antibodies against JNK3 and pJNK were from Upstate Biotechnology (Waltham, MA). Antibody against the PDH-E1α was from Mitoscience (Eugene, OR). All other chemicals or reagents were obtained from Sigma-Aldrich (St. Louis, MO).
Male Fisher 344 rats of different ages (6, 14, and 24 months) were from the National Institute on Aging (Baltimore, MD). Each rat was individually caged in the animal facility under standard conditions (12 h light/12 h dark cycle, humidity at 50 ± 15%, t22 ± 2°C, and 12 air changes/h) for 3 days to recover from the shipment stress.
Whole brain mitochondria were isolated from adult male Fisher rats by differential centrifugation followed by discontinuous Percoll density-gradient centrifugation . Brains were excised, rinsed, and homogenized using a Dounce homogenizer in isolation buffer (250 mM sucrose, 20 mM HEPES, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, protease inhibitor (100 μl per brain), 0.5% bovine serum albumin (BSA), pH 7.4). The homogenate was centrifuged at 1,330 g (5 min) to remove nuclei and cell debris and the resulting supernatant was centrifuged at 21,200 g (10 min). The pellet was resuspended in 15% Percoll and was centrifuged 21,000 g for 10 min. The resulting loose pellet was layered onto a pre-formed discontinuous Percoll gradient and centrifuged at 31,000 g for 10 min. Mitochondrial fractions were collected and washed twice with isolation buffer followed by washing in BSA-free isolation buffer.
The purity of the mitochondrial fraction was assessed as previously described  by measuring markers of microsomal (NADPH-cytochrome P450 reductase) and cytosolic (lactic dehydrogenase, β-actin) contamination. Activities of NADPH-cytochrome P450 reductase and lactic dehydrogenase were negligible when compared to those in the crude homogenate. β-actin was absent in the mitochondrial fraction when assessed by immunoblot analysis (shown in Fig. 2).
Mitochondria or total brain homogenate were lysed in RIPA buffer containing Tris-HCl (50 mM), NP-40 (1%), sodium deoxycholate (0.25%), NaCl (150 mM), EDTA (1 mM), pH 7.4. Samples (50 μg/well) were denatured at 95°C for 5 min; separated on 12% SDS-PAGE gels and electro-transferred to a PVDF membrane (Millipore, Billerica, MA). Membranes were blocked with casein (Pierce, Rockford, IL) and then incubated with specific antibodies at concentrations indicated by the manufacturers. Chemiluminescence detection was used to visualize protein bands. The bands of interest were quantified by Scion Image beta 4.0.2.
For PDH activity measurements, mitochondria were sonicated (30 s, setting of 3.0, 100% pulse rate) in a buffer containing 35 mM KH2PO4, 5.0 mM MgCl2, 2.0 mM NaCN, 0.5 mM EDTA, 0.25% Triton X-100, and phosphatase inhibitor at pH 7.25. PDH activity was assayed at 37°C by measuring the reduction of NAD+ at 340 nm upon supplementation of 50 μg mitochondrial protein/ml with 0.5 mM NAD+ in the presence of 200 μM TPP, 40 μM coenzyme A, and 4.0 mM pyruvate. The assay was carried out in the presence of 2.5 μM rotenone to prevent NADH consumption by complex I.
PDK activity was measured by a two-step immunocapture plus spectrophotometric assay. (a) Immnuocapture of PDK-2. α-PDK-2 (5 μg) was attached onto each well of Protein-G coated 96-well immunoprecipitation plate (Pierce, Rockford, IL) by 1 h at room temperature. Brain mitochondrial lysate (50 μg) from rats of different ages (6, 14, and 24 months) was then incubated in the well for 2 h at room temperature to facilitate PDK-2 immunocapture onto the well. (b) In-well PDK-induced PDH phosphorylation. Mitochondrial lysate (100 μg) from brain of 6-month old rats was diluted into phosphorylation buffer (100 μl) containing 30 mM HEPES, 1.5 mM MgCl2, 0.05% Triton X-100, 0.1 mM EDTA, 5 mM DTT, 0.5 mM ATP, proteases inhibitor and phosphotases inhibitor. The lysate was then incubated in the α-PDK-coated well for 5 min at 30°C. Aliquots (25 μl) were removed from the plate and PDH activity was measured as mentioned above. PDK activity was expressed as percentage of inhibition of PDH activity.
(a) 2D gel. Mitochondrial protein was separated by isoelectric point (pI) on precast gel strips (17 cm) with a pH gradient of 3–10 (Bio-Rad, Hercules, CA) by using the Bio-Rad Protean IEF System . The gels were fixed overnight and stained with Pro-Q® Diamond phosphoprotein gel stain (Molecular Probes, Eugene, OR) then imaged using a VersaDoc™ imaging system (Bio-Rad, Hercules, CA). The same gels were then stained with SYPRO™ Ruby protein gel stain (Molecular Probes, Eugene, OR) and imaged again using a VersaDoc™ imaging system. Density of protein spots were quantified by Scion Image beta 4.0.2. Proteins of interest (candidate spots of E1α-PDH: based on the molecular weight, pI value, and phosphorylation signals) were excised from the 2D gel using biopsy punches (Acuderm, Fort Lauderdale, FL) and subjected to LC/MS/MS (USC Proteomics Core Facility). (b) LC/MS/MS. Samples were reduced with DTT and then alkylated with iodoacetamide prior to in-gel tryptic digestion using trypsin that was reductively methylated to reduce autolysis (Promega, Madison, WI). The digestion product were extracted twice from the gel with 5% formic acid/5% acetonitrile solution and once with acetonitrile followed by evaporation using an ADP SpeedVac (Thermo Savant, Watham, MA). Tryptic peptides were analyzed by tandem mass spectrometry. Protein identification was carried out with the MS/MS search software Mascot 1.9 (Matrix Science).
Total brain homogenates were lysed in an equal volume of perchloric acid (2 M) and centrifuged for 10 min at 12,000 g. Supernatants were neutralized with KHCO3 (3 M) and recentrifuged at 12,000 g. Extracts (50 μl) were added to 500 μl of reaction buffer and the concentration of lactic acid was measured using a lactic acid assay kit (r-Biopharm, Germany). Samples for ATP measurements were prepared as described above; ATP levels were determined using an ATP determination kit (Molecular Probes, Eugene, OR).
Data are expressed as the means ± SE of at least three independent experiments. Statistical comparisons were performed by one way ANOVA. Differences were considered significant when P < 0.05.
The basal levels of JNK1, JNK2, JNK3, and phosphorylated JNK (pJNK, reflecting activation) in brain homogenate from rats of with different ages (6, 14, and 24 months) are shown in Fig. 1 (n ≥ 5). The expression level of JNK1, JNK2, and JNK3 in the brain homogenates did not change significantly with age (Fig. 1A–C). JNK activation, determined by antibodies against pJNK (dual phosphorylation of JNK is essential for kinase activity), increased with age (Fig. 1D). The increased pJNK levels reflects an increase in the activation of all three JNK isoforms and is an effect consistent with previous findings showing that JNK activity was constitutively high and significantly increased in older rats .
Activation (phosphorylation) of JNK under defined stress conditions results in its partial translocation to mitochondria . Immunoblot analyses showed that protein levels of JNK2 and JNK3 associated with rat brain mitochondria did not change significantly as a function of age (6-, 14-, and 24-month old rats) (Fig. 2B–C; (n ≥ 5)), whereas those of JNK1 and phosphorylated JNK (pJNK) associated with mitochondria increased as a function of age (Fig. 2A,D; (P < 0.05, n ≥ 5)).
As previously reported , pJNK associated with the outer mitochondrial membrane and triggered a phosphorylation cascade that resulted in the inhibition (phosphorylation) of mitochondrial matrix pyruvate dehydrogenase (PDH) activity. Accordingly, PDH activity in brain mitochondria decreased significantly as a function of age: decreases in activity of ~25% and ~45% were found in mitochondria from 14- and 24-month old rats, respectively (Fig. 3A, (P < 0.01, n ≥ 5)).
PDH activity is, in part, controlled by phosphorylation/dephosphorylation, where a specific pyruvate dehydrogenase kinase (PDK) phosphorylates three serine residues of PDH-E1α, thereby decreasing the oxidative decarboxylation of pyruvate to acetyl-CoA. Because pJNK associates with the outer mitochondrial membrane (without crossing into the mitochondrial matrix), it may be surmised that PDK is the effector that ultimately conveys the inhibitory signal from pJNK to PDH. Protein levels of PDK-2, the most abundant PDK isoenzyme in brain, increased during aging with significantly higher levels at 14- and 26 months as compared to 6 months (Fig. 3B). Furthermore, PDK activity (Fig. 3C) correlated with its protein levels: the PDK-2-dependent inhibition of PDH activity amounting to ~20-, 43-, and 49% at 6-, 14-, and 24 months of age, respectively. Taken together, it may be inferred that increased phosphorylated JNK association to mitochondria may up-regulate PDK activity, thus causing increased phosphorylation (and inhibition) of PDH.
Inhibition of PDH activity upon translocation of pJNK to mitochondria was reported to be a consequence of phosphorylation of the E1 subunit of the pyruvate dehydrogenase complex . The decrease of PDH activity as a function of age is likely to involve a similar inactivation by phosphorylation; to validate this notion, the extent of PDH-E1α was determined: phosphorylated and non-phosphorylated forms of PDH-E1α were identified on 2D gels by staining mitochondrial proteins with Syproruby stain (Fig. 4A) and Pro-Q® Diamond stain (Fig. 4B), respectively. Spots a, b, and c (Fig. 4A,B) were subjected to LC/MS/MS analyses in order to identify the proteins present. In each case the major species identified was pyruvate dehydrogenase subunit E1α. In the three analyses, 8-, 5-, and 6 peptides were identified with high confidence (Mascot score with p < 0.5) and corresponded to 25, 18, and 12% coverage of the protein, respectively. A representative tryptic peptide that was observed in each LC/MS/MS analysis was LPCIFICENNR, corresponding to amino acids 218–228 (Fig. 4C). The theoretical mass for the double charged peptide with two carbamidomethyl cysteine modifications (i.e., alkylation by iodoacetamide) was 718.3. Masses of 718.2, 718.0, and 718.9 were observed in the three runs for pepdides identified as LPCIFICENNR. The collision induced dissociation (m/z) spectrum is presented for the mass of 718.9 in Fig. 4D. This result was consistent with our previous report  in which we found that proteins spots b and c were the two major phosphorylated mitochondrial proteins and were identified as PDH-E1α.
The percentage of phosphorylated PDH-E1α was quantified based on the protein amount on 2D gels and was increased significantly during aging from 14- to 24 months; the percentage of the non-phosphorylated-form –the actual functional form– decreased during aging (Fig. 5, (n ≥ 3, P < 0.05)). The total amount of PDH-E1α did not change during aging (data not shown). Therefore, it can be concluded that the amount of the non-phosphorylated form of PDH-E1α (the active form) decreases during aging (Fig. 5), thus accounting for the decrease in PDH activity as a function of age (Fig. 3A).
PDH activity links glycolysis to the tricarboxylic acid cycle where reducing equivalents in the form of NADH and FADH2 are generated. Accordingly, an adequate PDH activity is particularly important for tissues to maintain a reducing environment and high ATP production. Inhibition or a decrease activity of PDH may lead to an increase in the anaerobic reduction of pyruvate to lactate via lactic dehydrogenase. To assess these metabolic effects related to PDH inhibition, levels of ATP and lactic acid were measured in brain homogenates from rats of different ages. A decrease in ATP levels along with a significant increase of lactic acid concentration was observed in rat brain homogenates as a function of age (Fig. 6; n ≥ 4, P < 0.05). These metabolic patterns may be ascribed to the aforementioned inhibition of the PDH complex during aging.
The age-dependent decrease and increase in ATP production and lactate accumulation, respectively, in brain tissue (Fig. 6) appear to represent a shift from aerobic glycolysis (mitochondrial pyruvate dehydrogenase-dependent) to anaerobic glycolysis (cytosolic lactate dehydrogenase dependent). The mechanistic implications of this shift are primarily based on the inactivation by phosphorylation of the E1α subunit of mitochondrial matrix pyruvate dehydrogenase with the consequent diminished metabolism of acetyl-CoA. Within the context of this study, the impairment of pyruvate dehydrogenase activity may be associated with JNK signaling: endogenous JNK activation, especially at a low level, may reflect a chronic and cumulative stress process that contributes to mitochondrial dysfunction during brain aging. These data showed that JNK1 activation and its translocation to- and association with mitochondria were significantly enhanced during aging, thus suggesting that the regulation of mitochondrial function by JNK is more potent in the aging brain. Of note, the ex vivo approach in this study (mitochondria isolated from brains of rats of different ages) shows a clear correlation –but not causality– between increase active JNK and PDK activities and decrease PDH activity over age; other experimental approaches, such as specific radioactive signals (γ-32[P]ATP) in the mitochondrial proteome (2D IEF/SDS PAGE) upon incubation of brain mitochondria with active recombinant JNK  strengthen the notion that JNK mediates PDH phosphorylation  as well as mitochondrial Bcl-2 and Bcl-xL .
Mitochondrial oxidative stress, JNK activation, and alterations of mitochondrial bioenergetics seem to be intimately linked: (a) mitochondrial H2O2 production increases with aging [26,27], (b) mitochondrially-released H2O2 leads to the activation of JNK  (by mechanisms likely entailing dissociation of JNK from glutathione transferase  or suppression of phosphatases involved in JNK activation [29,30]), (c) activation (phosphorylation) of JNK results in its translocation to the outer mitochondrial membrane , and (d) association of JNK with the outer mitochondrial membrane triggers phosphorylation cascades that affect mitochondrial bioenergetics and mitochondrion-driven apoptotic pathways [12,16] as shown by specific radioactive signals (γ-32[P]ATP) on the mitochondrial proteome upon incubation of brain mitochondria with active recombinant JNK  and identification of Bcl-xL and Bcl-2  and pyruvate dehydrogenase  as phosphorylation targets.
The above-mentioned effects represent a feedback loop that entails coordination of cytosolic and mitochondrial responses and a delicate balance that may be impaired during brain aging. It could be argued that JNK activation, especially at a low level, may reflect a chronic and cumulative stress that contributes to mitochondrial dysfunction during brain aging. Of note, the decline in pyruvate dehydrogenase activity in brain during aging is expected to be linked to decreased mitochondrial NADH levels and –via nucleotide transhydrogenase– to decreased NADPH reducing power, thereby diminishing H2O2 removal by the glutathione system.
The precise mechanism by which JNK leads to inactivation of PDH is not known: the pyruvate dehydrogenase complex is regulated by reversible phosphorylation of the E1α subunit; pyruvate dehydrogenase kinase (PDK consists of four isozymes PDK1, PDK2, PDK3, and PDK4; among the four isoenzymes, PDK2 is the predominant form in brain ) is responsible for the phosphorylation (inactivation) , whereas a Ca2+-sensitive phosphatase is responsible for the dephosphorylation (activation) [33,34]. Thus, at least two mechanism by which JNK may be involved in enhanced PDH phosphorylation can be envisaged: first, JNK upregulates PDK protein and enzymic activity and, second, JNK interacts with the Ca2+ channels on mitochondria causing a reduction of Ca2+ influx and deactivation of the Ca2+-sensitive phosphatase. Although the former mechanism is somewhat suggested by the data in the present study, the underlying mechanism on how JNK may be modulating PDK remains unclear. Other signaling pathways may also contribute to increase PDK2 expression as a function of age.
Alterations of mitochondrial bioenergetics by MAPKs has been reported: in renal cells, the oxidant-induced activation of ERK1/2 decreases complex I activity, thereby diminishing mitochondrial respiration and energy transduction (ATP synthesis); pyruvate dehydrogenase activity was not affected . Of note, ERK1/2 is present in mitochondria and exposure of renal cells to a tertiary hydroperoxide leads to activation (phosphorylation) of ERK1/2, which was not due to translocation of active ERK1/2 from cytosol to mitochondria . ERK1/2, JNK, and p38 are present in mitochondria from different cell types: activation of mitochondrial ERK1/2 during rat brain development is apparently under the control of mitochondrial H2O2 levels ; mitochondrial protein kinase C ε (PKCε) forms signaling modules with ERK1/2, JNK, and p38 in murine heart with implications for cardioprotection ; studies of Lewy body disease neurons provided evidence for active ERK1/2 in mitochondria (colocalizing with Mn-superoxide dismutase) . MAPKs are also involved in the regulation of the mitochondrion-driven apoptotic pathway by mechanisms entailing phosphorylation of Bcl-2 family members either cytosolic (and further translocation to mitochondria) or constitutive of the outer mitochondrial membrane (Bcl-xL and Bcl-2) [39,40].
The shift from aerobic glycolysis (mitochondrial pyruvate dehydrogenase-dependent) to anaerobic glycolysis (cytosolic lactic dehydrogenase-dependent) in brain during aging acquires further significance on two accounts: first, glucose is a primary energy source for mammalian brain and anaerobic glycolysis is an inefficient energy source and, second, pyruvate dehydrogenase was reported to be deficient in the brain of Alzheimer’s disease patients [41–44]. This deficiency occurs not only in regions of brain that are neuropathologically damaged in Alzheimer’s disease, but also in regions that are histopathologically normal, which suggests that the decreased PDH activity occurs in early stages of the disease [41–44]. Therefore, rescue of PDH activity in an early stage of Alzheimer’s disease can be a therapeutic strategy: in this context, supplementation with lipoic acid –a naturally-occurring disulfide compound– may rescue the decreased PDH activity caused by the phosphorylation of E1α-PDH. Acetyl-L-carnitine has also been purported as an agent for the treatment of early stages of Alzheimer’s disease [45–47]: the presence of an acetylcarnitine-CoA transferase in the brain allows the entry of acetyl units from acetyl-L-carnitine into the TCA cycle . This strategy should be viewed as an energy source other than that from the oxidative decarboxylation of pyruvate.
This work was supported by NIH grant 2RO1 AG016718.
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.
Qiongqiong Zhou, Department of Cell Biology, School of Medicine, Johns Hopkins University, Baltimore, MD 21205-2186, USA.
Philip Y. Lam, Department of Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of Southern California, 1985 Zonal Avenue, Los Angeles, CA 90089-9121, USA.
Derick Han, Research Center for Liver Diseases, Keck School of Medicine, University of Southern California, Los Angeles, CA 90089-9121, USA.
Enrique Cadenas, Department of Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of Southern California, 1985 Zonal Avenue, Los Angeles, CA 90089-9121, USA.