Pyruvate Dehydrogenase Regulation

From Cancer to Diabetes, and Beyond

October 15, 2009

Edited by Petr Hajek, Ph. D.

Past issues are available for review in the archives.

Table of Contents








If you are a regular reader of MitoNews, our love and admiration for the Pyruvate Dehydrogenase Complex will not shock you, and recent progress in understanding dysregulation of this complex in many diseases, as reported here, both inform and excite. Although it is now ancient history, feel free to read our previous review of recent studies on PDH in the August 2006 issue in our archives. Here we provide an update on PDH with a focus on PDH regulation and its role in disease.

PDH is the rate-limiting link between cytosolic glycolysis and mitochondrial Krebs cycle, and is tightly regulated by continuous cycles of phosphorylation and dephosphorylation reactions that determine the proportion of PDH in its active (dephosphorylated) state. When blood glucose is increased with intake of dietary carbohydrates, PDH is dephosphorylated and active, and this promotes glucose oxidation and fatty acid synthesis. When the glucose levels are low, PDH is phosphorylated and inactive in most tissues to preserve an energy source for the brain, which relies on glucose.

PDH is phosphorylated at three specific serine residues of the E1 α subunit; phosphorylation at any of the three sites inhibits PDH activity in vitro. In mammals, the phosphorylation is catalyzed by four PDH kinases (PDK1-4), which show variation in site preference and kinetic activity. The dephosphorylation reaction is catalyzed by two PDH phosphatases (PDP1 and PDP2). The expression of PDH kinases and phosphatases is differentially regulated in a tissue-specific manner by a variety of factors, including glucose to fatty acid ratio in diet, absence of nutrients in starvation, and oxygen levels during hypoxia and anoxia. Dysregulation of specific PDKs and/or PDPs is a sentinal feature in obesity, diabetes, and other metabolic diseases, as well as in cancer.


As reviewed by Roche and Hiromasa (Roche & Hiromasa. Cell Mol Life Sci. 2007 Apr;64(7-8):830-49.), with starvation, PDK expression is increased to prevent hypoglycemia. Of the enzymes that modulate PDH activity, PDK4 is of special interest because it is responsible for maintenance of fasting blood glucose levels. PDK4 is also up-regulated by a high fat diet and by extended exercise. Aberrant elevated expression of PDK4 is characteristic of, and detrimental in diabetes. High PDK4 levels downregulate glucose oxidation and conserve gluconeogenic substrates, and this exacerbates hyperglycemia. As reviewed by Kwon and Harris (Kwon & Harris. Adv Enz Regul. 2004;44:109-21.), under basal conditions, PDK4 expression may be repressed by maintaining relevant histones in a deacetylated state. Retinoids may recruit histone acetyl transferase (HAT), which leads to acetylation of relevant histones and thus increases PDK4 expression. During starvation, particularly in heart, skeletal, and other muscle tissues, kidney and liver glucocorticoids, along with FoxO1 recruits HAT which, in turn, promotes histone hyperacetylation and thus massive transcriptional activation of PDK4. Insulin inhibits PDK4 expression via series of events that sequentially lead to PI3K activation, FoxO1 phosphorylation, and its consequent re-localization to the cytosol, and thus lower histone acetylation. As reviewed by Degenhardt and associates, PDK4 (along with PDK2 and PDK3) are also regulated by PPAR transcription factors (Degenhardt, et al. J Mol Biol. 2007 Sep 14;372(2):341-55. Epub 2007 Jul 19.).

In a recent report by Park and colleagues analyzing the promoter region or PDK4 which is already known to include glucocorticoid receptor, FoxO1 and estrogen related receptor α (EER α) sites, yet another regulatory site, that for the estrogen-related receptor binding site was identified, contributing an effect on insulin inhibition, and adding even more complexity to the hormonal regulation of PDK4 expression (Connaughton, et al. Mol Cell Endocrinol. 2009 Aug 22. [Epub ahead of print].).

In another report, Nahle and coworkers uncovered regulation of PDK4 expression by Rb-E2F1 retinoblastoma-E2F1 tumor suppressor complex which suggests intimate coupling between cell cycle regulators and mitochondrial glucose oxidation (Hsieh, et al. J Biol Chem. 2008 Oct 10;283(41):27410-7. Epub 2008 Jul 30.). Loss of transcription factor E2F1 in vivo lowered PDK4 expression and improved myocardial glucose oxidation. Inactivation of Rb, the repressor of E2F1, markedly increased PDK4 expression. These findings may be of interest in elucidation of many cancers characterized by inactivated Rb tumor suppressor.

Importantly, in diabetes due to insulin deficiency or insulin insensitivity, the uninhibited over-expression of PDK4 leads to PDH phosphorylation which prevents glucose oxidation. In addition, the inhibition of PDH promotes gluconeogenesis in liver and worsens the hyperglycemia. When mice were fed a high fat diabetologic diet, the consequent PDK4 deficiency lowered blood glucose and improved glucose tolerance (Jeoung & Harris. Am J Physiol Endocrinol Metab. 2008 Jul;295(1):E46-54. Epub 2008 Apr 22.). In a continuation of these studies Harris and colleagues made an interesting finding (Hwang, et al. Biochem J. 2009 Sep 25;423(2):243-52.). If the high fat diet was rich in saturated fatty acids, PDK4 knockout mouse maintained lower blood glucose, but surprisingly, they gained less weight and accumulated less fat when compared to the control mice. An analysis of levels of transcription factors and regulatory enzymes suggested that PDK4 deficiency protects mice from the negative effects of high saturated fat diet by lowering amounts of enzymes of fatty acid synthesis. Regardless of these new findings, the study confirmed the therapeutic potential of PDK inhibitors to treat type 2 diabetes induced by high saturated fat diet.

In their recently published paper, Wanders and coworkers discovered a crosstalk between fatty acid and energy status signaling in the regulation of PDK4 expression which ensures a cell dependence on fatty acids as an energy source (Houten, et al. Cell Mol Life Sci. 2009 Apr;66(7):1283-94.). With low energy status, AMPK, as the key sensor and regulator of cellular energy balance, is activated, initiating increased energy production by fatty acid oxidation. Further, these authors show that AMPK activation by hypoxia, together with the presence of fatty acids, induces PDK4 expression and decrease glucose oxidation via activation of PPAR. They also argue that HIF-1α activation does not play a role in PDK4 regulation under these conditions.


As discussed in a previous issue of MitoNews, hypoxia induces expression of transcription factor HIF-1 which leads to increased expression of PDK1, along with upregulation of glycolytic enzymes. These transcriptional changes increase glucose flux to pyruvate and glycolytic ATP production, and suppress both Krebs cycle and respiration by the PDK1-mediated phosphorylation of PDH, thus causing a metabolic switch from oxidative to glycolytic glucose utilization. Importantly, the HIF-1-dependent block of respiration promotes cell survival.

In most solid tumors, cells rely on glycolysis for the energy production even in the presence of sufficient amounts of oxygen, phenomena, (known as “aerobic glycolysis”), or the Wartburg effect. Verma and colleagues have recently shown that inhibition of PDH activity via enhanced expression of PDK1, contributes to aerobic glycolysis and malignant phenotype in squamous cell carcinoma (McFate T. J Biol Chem. 2008 Aug 15;283(33):22700-8. Epub 2008 Jun 9.). The authors also show that PDH inhibition in cancer cells is linked to stabilization of HIF-1 by glycolytic metabolites. Pharmacological reversal of the Wartburg effect was found to cause selective apoptosis of tumor cells. To explore this mechanism further, the authors used two approaches to activate PDH and thereby, reverse the metabolic switch in favor of oxidative glucose utilization. This proved to increase the sensitivity of cancer cell to apoptosis. Results utilizing the PDK inhibitor dichloroacetate are summarized in a recent review by Mackey and associates (Michelakis, et al. Br J Cancer. 2008 Oct 7;99(7):989-94. Epub 2008 Sep 2.). Another approach was chosen by Stacpoole and colleagues, who reported that adeno-associated, virus-based delivery and expression of PDH E1α subunit caused metabolic remodeling and apoptosis in HepG2 cells (Glushakova, et al. Mol Genet Metab. 2009 Jun 23. [Epub ahead of print]).

The reduced mitochondrial respiration of cancer cells, and thus reduced production of reactive oxygen species (ROS), may be one of the reasons for decreased cell death. Komatsu and colleagues found recently that thrombopoietin, which induces ROS in hematopoietic cells, also induces HIF-1 and PDK1 (Kirito, et al. Cell Cycle. 2009 Sep 1;8(17):2844-9. Epub 2009 Sep 16.). The authors suggest that the cytokine-induced production of ROS leads, as a feedback mechanism, to activation of HIF-1 which, in turn, leads to increased transcription of PDK1, PDH inhibition and consequently, decrease of toxic ROS.

In a continuation about PDK1 regulation, Appenzeller and colleagues report that in chronic mountain sickness (CMS), a condition caused by lack of adaptation to low ambient oxygen tension in high altitude, the children of CMS patients showed impaired adaptation to hypoxia, including lower levels of PDK1 (Huicho, et al. BMC Pediatr. 2008 Oct 27;8:47.).


The brain is the only organ that relies almost exclusively on glucose metabolism to cover its energy needs. Disrupted glucose metabolism and lactate accumulation, as well as oxidative stress are hallmarks following traumatic brain injury (TBI). Immediate hyperglycemia, enhanced glucose utilization, but depletion of ATP are immediate characteristics of TBI that may be critical for brain damage. These detrimental effects may be further exacerbated by chronic hypoglycemia and inhibited glucose utilization in the delayed phase of TBI. In their recent paper, Verma and colleagues report TBI-induced decrease of protein levels and phosphorylation of the E1 α subunit of PDH using a rat model (Xing, et al. Neurosci Lett. 2009 Apr 17;454(1):38-42. Epub 2009 Jan 23.). A more recent report by Ling and colleagues shows that TBI induces lower levels of PDH E1 α protein in blood and brain (as measured by MitoSciences dipsticks [Sharma, et al. J Emerg Trauma Shock. 2009 May;2(2):67-72.]). The authors also show that pyruvate treatment can prevent the reduction of PDH E1 α and suggest that this reduction is due to a global oxidative stress. Thus a combination of oxidative stress and PDH dysregulation may be involved in TBI.


Alterations in PDH (possibly via oxidative modifications) were also reported in Alzheimer’s disease. In their recent report, the Brinton laboratory demonstrated that mitochondrial dysfunction precedes the development of Alzheimer’s disease pathology (Yao, et al. Proc Natl Acad Sci USA. 2009 Aug 25;106(34):14670-5. Epub 2009 Aug 10.). Using a triple-transgenic AD mouse model, the authors show a decrease of respiration and a decrease of both expression and activity of PDH and COX as early as three months of age, while mitochondrial amyloid beta levels increased at nine months. The good news is that with advances in research, therapeutics strategies are emerging, including estrogen therapy to upregulate key regulatory enzymes of brain metabolism or lipoic acid (the cofactor of PDH and alpha-ketoglutarate dehydrogenase), as reviewed by Munch and colleagues (Maczurek, et al. Adv Drug Deliv Rev. 2008 Oct-Nov;60(13-14):1463-70. Epub 2008 Jul 4.).


The majority of PDH deficiencies identified so far are caused by mutation of PDH E1 α subunit with significantly fewer cases due to mutations of E2, E3 and E3 binding protein. Recently, a role of pyruvate dehydrogenase phosphatases in PDH related disease has been reported. In a seminal paper, Cameron and associates identified in two brothers disease-causing homozygous three base pair deletion, which removes Leu 213 (Maj, et al. J Clin Endocrinol Metab. 2005 Jul;90(7):4101-7. Epub 2005 Apr 26.). Also the Robinson group has reported identification of a first human patient with PDP1 null mutation. (Cameron, et al. Hum Genet. 2009 Apr;125(3):319-26. Epub 2009 Jan 30.).

If you would like to opt-out of future MitoSciences announcements, please send an e-mail to Mandy Cochran at

Copyright (C) 2009 MitoSciences Inc. All rights reserved.