INVESTIGATIONS OF ANAPLEROSIS FROM PROPIONYL-COA

PRECURSORS AND OXIDATION IN THE BRAIN OF VLCAD AND

CONTROL MICE

by

XIAO WANG

Submitted in partial fulfillment of the requirements

for the Degree of Doctor of Philosophy

Thesis Advisor: Henri Brunengraber, M.D., Ph.D.

Department of Nutrition

CASE WESTERN RESERVE UNIVERISITY

May, 2009

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

______

candidate for the ______degree *.

(signed)______(chair of the committee)

______

______

______

______

______

(date) ______

*We also certify that written approval has been obtained for any proprietary material contained therein. Table of Contents

Table of Contents.……………………………………………………………….………i

List of Table...……………………………………………………………………………v

List of Figures.……………………………………………………………………..….…vi

Acknowledgements …………………………………..…………………………..….…xi

List of Abbreviations……………………………………………………………………xii

Abstract………………………………………………………………………………....xvi

CHAPTER 1: Substrate Utilization In The Brain

1.1 Overview of brain energy ..…….………………………………….....1

1.2 The blood-brain barrier ……..……….…….………………………………….…... 2

1.3 Utilization of glucose in the brain………………………………………………..…4

1.3.1 Regional metabolism of glucose..……………………………………….…...5

1.3.2 Overview of astrocytes correlated with the brain energy metabolism……6

1.3.3 Metabolic trafficking between astrocytes and neurons...…………….……9

1.4 Utilization of fatty acids in the brain.………………………………………….….10

1.4.1 Overview of fatty acids in the brain ……………………………………...... 10

1.4.2 The mechanisms of fatty acid transport across the blood brain

barrier ………………………………………………………….……………..…11

1.4.3 The uptake of polyunsaturated fatty acids ………..………………………12

1.4.4 The uptake of saturated fatty acids…………………………………….…..12

1.5 Utilization of ketone bodies in the brain ……..………………………………....14

1.5.1 Metabolic roles……….……………………………………………………....14

1.5.2 Metabolism and regulation…………….…………………………………....16

i 1.5.3 Therapeutic implications of cerebral ketone

bodies ….…………………………………………………………………..…....20

CHAPTER 2: Anaplerosis And Cataplerosis

2.1 Overview……………………………………………………………………..……23

2.2 Functions of anaplerosis in the , muscle, and heart………………..…….26

2.3 Anaplerosis in the brain…………………………………………………..……...29

2.3.1 Overview……………………………………………………………...………30

2.3.2 The significance of the brain anaplerosis…………………………..……..30

2.4 Major anaplerotic pathways…..……………………………………...... …….….37

2.4.1 From pyruvate……………………………………………………..………....37

2.4.2 From propionyl-CoA…………….……………………………..…………….40

CHAPTER 3: Fatty Acid Oxidation And Its Disorders

3.1 The definition and function of fatty acids..………...…..……………….………46

3.2 Fatty acid transport system and cellular uptake..…..………………...…..…..48

3.3 Fatty acid oxidation (β-oxidation)……………..………………………………..49

3.4 Malonyl-CoA metabolism……………………………………………………..…54

3.5 Regulation of fatty acid oxidation..……………………………………………..54

3.5.1 Overview……………………………………………..………………………56

3.5.2 Regulation by malonyl-CoA via the palmitoyl transferase

system ..….……………………………………………………………………..58

3.5.3 Regulation by acetyl-CoA carboxylase .…………………………….……59

3.5.4 Regulation of PPARα……………..………………………………………...62

3.5.5 Regulation by malonyl-CoA decarboxylase ..……………………………64

ii 3.5.6 Regulation by L-carnitine and CoA availability…………………………..65

3.6 Fatty acid oxidation disorders (FOD) ….…………………………………...….65

3.6.1 Pathophysiology of FOD ..…...…………………………………………….66

3.6.2 Very-long-chain acyl-CoA dehydrogenase (VLCAD) deficiency the

VLCAD knock-out mouse mode .…….………………………………………66

3.6.3 Traditional treatment of long-chain FOD …………………………………68

3.6.4 Limitations of traditional therapy call out for new available

treatments …...………..……………………..………………………………...69

3.6.5 Triheptanoin: a novel treatment of FOD …...………………………….…71

3.6.6 Triheptanoin: Treatment for deficiency………….74

CHAPTER 4: Research Proposal

4.1 Project 1. Fatty acid oxidation and anaplerosis from propionyl-CoA

precursors ………………………………………………………………………...76

4.2 Project 2. Assay of the activity of malonyl-CoA decarboxylase …………….84

4.3 Publications ………………………………………………..……………………..86

4.3.1 Wang X, Zhang GF, Puchowicz MA, Kasumov T, Allen F Jr, Rubin A,

Lu M, Yu X, Roe CR, Brunengraber H. Fatty acid oxidation and

anaplerosis from propionyl-CoA precursors in the brain of mice deficient in

very-long-chain acyl-CoA dehydrogenase and control mice. To be

submitted to the Journal of Biological chemistry.

4.3.2 Wang X, Stanley WC, Darrow CJ, Brunengraber H, Kasumov T. Assay

of the activity of malonyl- decarboxylase by gas

iii chromatography-mass spectrometry. Anal Biochem 2007 Apr 15; 363(2):

169-74.

CHAPTER 5: Discussion And Future Directions

5.1 Fatty acid oxidation and anaplerosis from propionyl-CoA precursors in the

brain of VLCAD mice and control mice ………………………………………178

5.1.1 Summary and implications ..………..…………………………………..178

5.1.2 Discussion and conclusions ………………..…………………………..182

5.1.3 Future directions ………..…………………………………………….…191

5.2 The assay of the activity of malonyl-CoA decarboxylase by GC-MS .…...195

5.2.1 Discussion and conclusions .…………………………………………...195

5.2.2 Future directions ...……..……………………………………….….…....196

LITERATURE CITED………………………………………………………………..198

iv List of Table

Table 4.1 Acyl-CoA content of whole brain of control and VLCAD mice ……….118

v List of Figures

Figure 1.1 Simplified diagram showing pathways of cerebral C4-ketone body metabolism ……………………………………………………….……………………18

Figure 2.1 Scheme of the main anaplerotic pathways ..….……………………... 28

Figure 2.2 Brain and the pathways of anaplerosis and cataplerosis ………………………..…………………………………………………..36

Figure 4.1 Concentrations of octanoate and C4 ketone bodies in the blood of control and VLCAD mice from increasing infused doses of [1-13C]octanoate

…………………………………………………………………………………………..130

Figure 4.2 Labeling of ketone bodies from increasing infused doses of

[1-13C]octanoate or [1-13C]heptanoate ………..…………………………………….131

Figure 4.3 Concentrations of selected metabolites in the blood of control and

13 VLCAD mice infused with increasing amounts of [5,6,7- C3]heptanoate …….. 132

Figure 4.4 Mass isotopomer distribution of blood glucose in VLCAD mice infused

13 13 with increasing amounts of [3,4,5- C3]pentanoate or [3,4,5- C3]BKP………….133

Figure 4.5 Mole percent enrichment of the parent acyl-CoA and acetyl-CoA in brains of control and VLCAD mice infused with increasing amounts of

[1-13C]octanoate , [1-13C]heptanoate, or [1-13C]pentanoate …………………..….134

Figure 4.6 Concentration of acetyl-CoA in the brain of control and VLCAD mice infused with increasing amounts of octanoate, heptanoate, pentanoate or

β-ketopentanoate …………………………………………………………………..... 136

vi Figure 4.7 Mol percent enrichment of M3 isotopomers of pentanoyl-CoA, propionyl-CoA, methylmalonyl-CoA and succinyl-CoA in mice infused with

13 [3,4,5- C3]pentanoate …...………………………………………………………...138

Figure 4.8 Relative anaplerosis in brain of control and VLCAD mice infused with

13 13 increasing amounts of [5,6,7- C3]heptanoate, [3,4,5- C3]pentanoate,

13 13 [3,4,5- C3]BKP, or [ C3]propionate …………………………..………..…………139

Figure 4.9 Relative brain concentrations of CAC intermediates and related compounds in VLCAD mice vs control mice ………….…………………………..141

Figure 4.10 Concentrations of selected metabolites in the blood of control and

13 VLCAD mice infused with increasing amounts of [5,6,7- C3]pentanoate ……..142

Figure 4.11 Concentrations of total M3 C5-ketone bodies and of total C4-ketone bodies in the blood of mice infused with increasing amounts of

13 β-keto-[3,4,5- C3]pentanoate ……………………………………………………....143

Figure 4.12 Concentrations of M3 propionate, total M3 C5-ketone bodies and of total C4-ketone bodies in the blood of mice infused with increasing amounts of

13 [ C3]propionate ………………………………………………….…………………...144

Figure 4.13 Concentrations of pentanoyl-CoA and propionyl-CoA in the brains of control and VLCAD mice infused with increasing amounts of

13 [3,4,5- C3]pentanoate ……………………………………………………………….145

vii Figure 4.14 Concentrations of heptanoyl-CoA, pentanoyl-CoA and propionyl-CoA in the brains of control and VLCAD mice infused with increasing amounts of

13 [5,6,7- C3]heptanoate ...... 146

Figure 4.15 Concentrations of octanoyl-CoA, hexanoyl-CoA and butyryl-CoA in the brains of control and VLCAD mice infused with increasing amounts of

[1-13C]Octanoate…………………………………...………………………...……..…147

Figure 4.16 MPE of brain acetyl-CoA, average MPE of acetyls of brain octanoyl-CoA, and average MPE of acetyls of blood β-hydroxybutyrate in mice infused with increasing amounts of [1-13C]octanoate …………...... 148

Figure 4.17 MPE of brain acetyl-CoA, average MPE of acetyls of brain heptanoyl-

CoA, average MPE of acetyls of blood β-hydroxybutyrate, and MPE of acetyl of blood β-hydroxypentanoate in mice infused with increasing amounts of

[1-13C]heptanoate ………………………………..…………………..….…………....149

Figure 4.18 Ion chromatograph of acetyl-CoA (analyzed as the thiophenol derivative) formed from [13C]malonyl-CoA by MCD present in an extract of rat liver. ………………………………………………………………...…………………. 171

13 Figure 4.19 Calibration curve of [1,2- C2]acetyl-CoA concentration by GC-MS

……………………………………………………………………………...……...... 172

13 Figure 4.20 Inhibition by 2 mM ADP of [1,2- C2]acetyl-CoA hydrolysis .……...173

viii Figure 4.21 Kinetics of MCD in the absence and presence of 10 μM inhibitor

CBM-301940 ………………………………………………………………..……….. 174

Figure 4.22 Linearity of MCD assay with the amount of liver tissue ….………. 175

Figure 4.23 MCD activities in from fed, one day starved, and two day starved rats ………………………………….………………………………………………….. 176

Figure 4.24 Lineweaver-Burk plot of MCD kinetics , with calculated Km of 202 μM and Vmax of 3.3 μmol•min-1• g tissue-1 ………………………………………..……..177

x Acknowledgements

This dissertation marks the end of a long journey with a wealth of knowledge and rewarding experiences I had the opportunity to learn and enjoy in the past several years. It would be impossible for me to complete this endeavor without the support of my teachers, co-workers, friends, family and many others.

I would like to express my sincere thanks to my mentor and thesis advisor, Dr.

Henri Brunengraber, for the guidance and constructive criticism he provided throughout the duration of my work. I would also give credit to my former and current co-workers, especially Dr. Guofang Zhang and Dr. Michelle Puchowicz who have contributed to the development of key methodologies and data measurements. I thank members of my thesis committee, Dr. Colleen Croniger,

Dr. Janos Kerner, Dr. Joseph LaManna, Dr. Jonathan Whittaker, and Dr. Stephen

Previs, for critically evaluating my work and for their valuable suggestions to improve it.

I would like to dedicate this work to my parents and my grandmother for their love and support. They are everything to me. Unfortunately, my dear grandmother, who has always been the greatest source of strength in my life, is no longer here to share this joy with me. But she will always live in my memories with me.

Last but not the least, I am grateful to my boyfriend Jialin Tao for his extraordinary support. Without him, I would not have made this far.

This work was supported by the National Institutes of Health (RoadMap grant

3R33DK070291 and grant RO1DK069752 to Dr Brunengraber) and by the Mt.

Sinai Health Care Foundation.

xi

List of Abbreviations

ACC Acetyl-CoA carboxylase

AMPK 5’-AMP-activted kinase

CACT Carnitine: acylcarnitine

CPT Carnitine palmitoyltransferase

FAO Fatty acid oxidation

FOD Fatty acid oxidation disorders

GC-MS Gas chromatography-mass spectrometry

LC-MS Liquid chromatography-mass spectrometry

MCD Malonyl-CoA decarboxylase

MCFA Medium-chain fatty acid

MCAD Medium-chain fatty acyl-CoA dehydrogenase deficiency

MID Mass Isotopomer Distribution

NAD Nicotinamide adenine dinucleotide (oxidized)

NADH Nicotinamide adenine dinucleotide (reduced)

NMR Nuclear magnetic resonance

PC Pyruvate carboxylase

PPAR Peroxisome proliferator-activated receptor

SCFA Short-chain fatty acid

VLCFA Very long-chain fatty acid

VLCAD Very long-chain fatty acyl-CoA dehydrogenase deficiency

xii Investigations of Anaplerosis from Propionyl-CoA Precursors and Fatty Acid

Oxidation in the Brain of VLCAD and Control Mice

Abstract

by

XIAO WANG

Anaplerotic therapy with triheptanoin is currently investigated for the treatment of long-chain fatty acid oxidation disorders. The recently developed mouse model deficient in very long-chain acyl-CoA dehydrogenase (VLCAD mouse) provides a good tool to study anaplerotic therapy in vivo. The goal of this research was to characterize brain fatty acid oxidation and to test the effect of anaplerotic therapy on brain metabolism in VLCAD mice and their controls.

First, concentrations of major acyl-CoA esters (C2-C20) were profiled in the brain of

VLCAD and control mice. The data indicated significant reductions of acetyl-CoA, methylmalonyl-CoA, propionyl-CoA, butyryl-CoA, hexanoyl-CoA and octanoyl-CoA in VLCAD brains vs. controls. The data also revealed clear evidence of long-chain fatty acid oxidation defect suggested by the accumulations of stearoyl-CoA (C18:0-

CoA) and linoleoyl-CoA (C18:2-CoA) in VLCAD brains vs. controls.

Second, concentrations of major citric acid cycle intermediates and related neurotransmitters (glutamate, glutamine, GABA) in the whole brain were found lower in VLCAD mice vs. control mice, with three significant decreases identified

(α-ketoglutarate, glutamate, and GABA). These reductions implicated a

xvi compromised pool size of the citric acid cycle intermediates which possibly lead to abnormalities in the brain of VLCAD mice.

Last, VLCAD and control mice were infused intravenously with increasing amounts of odd-chain fatty acids or C5 ketone body (heptanoate, pentanoate, β- ketopentanoate or propionate). Other mice of the two genotypes were infused with non-anaplerotic octanoate. The concentration and labeling pattern of medium-chain acyl-CoAs demonstrated that the C8, C7, and C5 fatty acids were taken up by the brain and activated to form CoA esters. The odd-chain fatty acids infused were metabolized via the β-oxidation cascade and contributed to a substantial fraction of acetyl-CoA in mice brains of both genotypes. Ketone bodies derived from partial oxidation of infused fatty acids in the liver also contributed to the brain acetyl-CoA. All odd-chain fatty acids or C5 ketone body infused (heptanoate, pentanoate, and propionate) were strongly anaplerotic at low blood concentrations, and more so in VLCAD brains than in control brains. A fraction of anaplerosis from the odd-chain fatty acids involved their conversions in the liver to C5-ketone bodies which were also anaplerotic in the brain.

In conclusion, data of the present research revealed that medium-chain fatty acids enter the brain as such, where they are metabolized by the fatty acid oxidation cascade. Odd-chain fatty acids and C5-ketone bodies of propionyl-CoA precursors contribute to brain anaplerosis rapidly in normal mouse brains, and even more rapidly in VLCAD brains. In VLCAD brains the long-chain fatty acid oxidation occurs by a vicariant enzyme system to be identified.

xvii Chapter 1

Substrate Utilization in the Brain

1.1 Overview of brain energy metabolism

The brain is one of the most vital, complex, yet fragile systems among human organs. It represents only 2 to 3% of one’s body weight, but requires a constant, substantial fraction of one’s total energy supply, including 15% of cardiac output,

20% of total oxygen consumption and 25% of total glucose utilization (1).

Glucose has long been recognized as the most important energy source for the brain. Glucose utilization by the brain quantified from arterial-venous difference is 31 mmol/100 g·min (2). In addition to direct energy production through oxidation to CO2 ( + citric acid oxidation (CAC) + (ETC)), glucose flows down other metabolic fates as well. If needed, glucose can be converted to metabolic intermediates, such as lactate and pyruvate, for use by pathways other than the CAC; or glucose can be released and removed through the circulation. Glucose is anabolic when incorporated into glycogen, , , or converted to neurotransmitters (e.g. g-aminobutyric acid (GABA), glutamate, and acetylcholine) (3; 4).

In prolonged fasting when the glucose level is insufficient to generate enough energy, the brain switches to utilize C4 ketone bodies (acetoacetate (AcAc) and

β-hydroxybutyrate (BHB)). In other conditions that involve insulin resistance (e.g. type-2 diabetes), high ketone (e.g. Atkins) or high-fat diet (e.g. suckling), the

1 brain generally relies on energy input from C4 ketone bodies (KBs) as their plasma concentrations increase markedly (5). Detailed cerebral utilization of C4

KBs will be discussed later in this chapter.

Although lactate and pyruvate can be used by the brain during neuron activation and sustain synaptic activity in vitro (6; 7), they cannot substitute for glucose to maintain brain functions (8). One possible reason is their limited permeability to the BBB, similar to all other carboxylic acids. However, lactate and pyruvate are utilized when originated from glucose inside of the brain (9). For example, they participate in the transfer of nutrients between the astrocytes and the neurons

(10). This transfer largely contributes to the complex heterogeneity of the brain energy metabolism.

The utilization of lactate as an oxidative energy source was further confirmed in vivo during exercise from arterial-venous difference studies (11). Human positron emission tomography (PET) studies observed significantly reduced utilization of glucose when plasma lactate concentration increased (12).

Therefore, enough findings support the utilization of lactate in the brain, at least in certain circumstances. What needs to be defined further is under what condition or to what extent lactate is taking up in addition to glucose.

1.2 The blood-brain barrier (BBB)

The blood-brain barrier (BBB) is composed of a layer of endothelial cells connected by tight junctions (13). These endothelial cells form a permeable interface that allows the exchange of nutrients between the blood in capillaries

2 and the neural cells in a selective manner. The physical structure of the BBB maintains the brain homeostasis by i) providing a diffusible passage to small molecular weight hydrophilic or -soluble substances, while ii) restricting the bidirectional transfer of large molecules. These large molecules are selectively restricted by various criteria, such as size, charge, hydrogen binding potential and lipid solubility (14).

Investigation methodology of the BBB

A number of in vitro, in vivo and in situ techniques have been well developed to study the uptake of nutrients or drugs, and their subsequent metabolism in the brain (15). In vitro techniques are mainly set up in isolated cerebral endothelial cell cultures. The obvious advantage of in vitro cell culture is that this simple controllable preparation avoids anesthesia or surgery. The most recent preparation (16) co-cultures endothelial cells with astrocytes simultaneously, to mimic the integrated structural and functional characteristics of the BBB.

In vivo techniques maintain structured cells and matrix in physiological states and anatomical positions. Widely applied in vivo techniques include: (i) the traditional brain uptake index (BUI) (17) and (ii) the indicator diffusion methods (18), both of which are restricted to single-passage studies of rapidly uptaking substances, (iii) microdialysis (19; 20) which measures regional extracellular concentrations, (iv)

IV administration (21; 22) which allows the determination of kinetics and metabolic profiles when combined with radioactive or stable isotope techniques, and (v) positron emission tomography (PET) (23) which detects radionuclide

3 distribution.

The procedure carried out in the present study involves IV infusion of stable isotope 13C labeled fatty acids to the jugular vein of the VLCAD mouse. This administration is non-invasive, and applies to metabolic tracing of both rapid and slow uptaking substances.

The in-situ perfusion describes “the maintenance of the brain in a viable state, isolated from the animal by means of the mechanically assisted circulation of an artificial medium through its vascular bed” (24). However, this technique requires complicated surgery (see (18) and (24)).

1.3 Utilization of glucose in the brain

Glucose is the predominant and obligatory energy source for the brain in normal physiological state. Glucose is taken up by the brain primarily via the glucose transporter 1 (GLUT1). GLUT1 (25) is mainly distributed on the membranes of red blood cells and endothelial cells of barrier tissues including the BBB (26).

GLUT1 expression is capable of up- or down-regulation induced by reduced or elevated circulating glucose levels (26). Other major glucose transporters are

(26): i) GLUT2 mainly expressed in the small intestinal epithelial cells, liver and the pancreatic β cells; ii) GLUT3 expressed in neurons; and iii) GLUT4 mostly distributed in adipose tissue and skeletal/cardiac muscle. Note that GLUT4 functions in the insulin-regulated glucose uptake.

Glucose provides energy (ATP) via oxidative . This oxidative catabolism is composed of cytosolic glycolysis coupled with mitochondrial

4 oxidative phosphorylation cascade (citric acid cycle (CAC) + oxidative electron transport chain (ETC)). Glycolysis converts 1 mol glucose (6-carbon molecule) to 2 mols pyruvate (3-carbon molecule), producing 6 mols net ATPs (2 mols from direct production, 4 mols from generated NADH). Although only generating a relatively small amount of the ATP, glycolysis is important to produce pyruvate, as well as other 6- or 3-carbon intermediates. Pyruvate or other 6- or 3-carbon intermediates then either enter the mitochondria to be oxidized via the CAC, or exit the glycolysis to be used in anabolic processes. Additionally, glycolysis is

“capable of activation-induced increase in energy demands” (27). Subsequently,

2 mols pyruvate generated from 1 mol glucose enter the mitochondria where 30 mols additional net ATP can be generated via the oxidative phosphorylation cascade which is composed of coupled processes of the CAC and the ETC.

Note that the mitochondrial oxidative phosphorylation cascade is coupled with the ETC for the ATP synthesis, thus requiring the supply of O2. Therefore, the constantly active energy production from glucose in the brain explains its substantial consumption of 20% of the total body O2 requirement.

1.3.1 Regional metabolism of glucose

The brain energy consumption is not limited to reflect neuron metabolism.

Human anatomy has revealed in the brain the presence of other neuron supporting cells: the glia and vascular endothelial cells. Glia cells dominate the brain quantitatively, while neuron cells only contribute to less than half of the total brain volume (28). In addition, there is clear evidence indicating that “the

5 astrocyte-to-neuron ratio increases with increasing brain size” (29). From a functional perspective, glia cells play a significant role in the transfer of nutrients and shuttle of neurotransmitters in and out of neuron cells. Based on these findings, it is now clear that multiple cell types in the brain other than neurons require energy input as well.

The complexity of mixed cell types has created a multi-compartmentation structure that affects the brain metabolism. This multi-compartmentation structure requires a complex system in the uptake and transfer of energy substrates and neurotransmitters. Therefore, understanding the regional metabolism and the shuttle between various cell types and/or compartments is critical in the investigation of integrated brain energy metabolism.

1.3.2 Overview of astrocytes in correlated with brain energy metabolism

Astrocytes are star-shaped glia cells that offer special support of the vascular endothelial cells to maintain the integrity of the BBB. Astrocytes are generally categorized by location into type I and II (30). Type I astrocytes usually encircle endothelial cells with their end-feet (30), thus they are in direct contact with blood capillaries. This special anatomical position enables type I astrocytes to transport energy substrates from the blood to neuron cells. Type II astrocytes usually surround neurons and the synaptic gap, thus they are responsible to restore the presynaptic membrane potential (30).

In addition of being a structural and a physical scaffold, astrocytes serve two other functions of particular importance in correlation with brain energy

6 metabolism: the maintenance of extracellular K+ homeostasis (30; 31) and the clearance of neurotransmitters from the synaptic cleft (30).

The mechanism of neurotransmitter release has been well characterized.

Basically, a neuron cell is triggered by the arrival of an impulse to achieve action potential and rapidly releases neurotransmitters via exocytosis.

Neurotransmitters are classified as excitatory or inhibitory based on type of action. The most abundant excitatory neurotransmitter is glutamate; the most abundant inhibitory neurotransmitter is GABA. Upon releasing from the presynaptic cell membranes, excitatory neurotransmitter opens channels in the membrane of postsynaptic neuron cell, resulting in the efflux of K+ from intracellular storage and influx of Na+ from extracellular milieu. The decreased

K+ concentration and the increased Na+ concentration in the postsynaptic neuron result in the postsynaptic cell “depolarization” which further propagates along the cell membrane and serves as the nerve impulse for the next synaptic transmission event. Excitatory neurotransmitters generally stimulate postsynaptic depolarization; inhibitory neurotransmitters on the other hand, encourage the “hyperpolarization”, thus inhibiting action potential generation.

Whether a neurotransmitter is excitatory or inhibitory is determined by the type of an interaction between the transmitter and its receptor.

To maintain the next transmission cascade, the re-uptake/clearance of K+ and release of neurotransmitters during the last event is crucial. Any interference with this clearance that leads to sustained neuronal depolarization will result in

7 abnormal or complete loss of neuronal functions.

Astrocytes regulate transmitter re-uptake and K+ homeostasis in the synaptic cleft

Astrocytes express abundance potassium channel on the cell membrane, primarily the high density Na+,K+-ATPase (31). The action of Na+,K+-ATPase requires the ATP (32). By hydrolyzing 1ATP to 1ADP, Na+,K+-ATPase extrudes

3Na+ against 2K+ (33). The Na+ extruded by Na+,K+-ATPase then facilitates the uptake of neurotransmitters. These neurotransmitters include glutamate and

GABA. They are taken up by the astrocyte simultaneously with the re-uptake of

Na+ (33). The advantage of this Na+ co-transport is to further activate Na+,K+-

ATPase by the increased intracellular Na+ concentration upon its re-uptake (33).

Therefore, the uptake of K+ and neurotransmitters by the astrocyte is coupled together very efficiently by a positive feedback loop.

The astrocyte: the location of glucose uptake

The astrocyte which encircles endothelial cells and provides direct contact with blood capillaries connects the blood with the neuronal system. Evidence from in vitro experiments shows that astrocytes utilize glucose (32; 34), at a basal rate similar to that determined in whole-animal studies (35). Neuronal uptake of 2-DG is found to occurs “primarily in the area enriched in axon terminals, dendrites and synapses ensheathed by astrocytic processes and not where neuronal perikarya are located” (36). In addition, glucose itself is insufficient to maintain the survival of neurons in vitro (37; 38). Neuronal survival is achieved when other substrates

8 such as pyruvate, glutamine, and lactate are present in addition to glucose (6; 37;

38). Therefore, it is now clear that astrocyte is responsible for the uptake of glucose from the blood. As a result, energy supply for the neuron relies on substrates trafficking through the astrocytes. These substrates involve not only glucose, but also pyruvate, glutamine, and lactate etc. in order to sustain neuronal survival.

1.3.3 Metabolic trafficking between astrocytes and neurons

Although glia cells represent a larger portion of the brain volume, they generally consume only 5% of the brain energy budget (39). Yet strikingly, local distribution studies (40) found that the major utilization of glucose occurs in glia.

It was thus hypothesized that astrocytes themselves take up glucose rather than neurons, but only partially oxidize it for energy (39). Consequently there must be a specialized system present to transport partially oxidized glucose from astrocytes to neurons. This transport system was later found to depend on lactate. Glucose taken up by astrocytes gets catabolized glycolytically and exported as lactate which is transported to the neurons as the direct energy substrate.

This proposed hypothesis is further supported by the distributions of enzymes involved in the oxidation of glucose and lactate. The enzymes for the glycolytical oxidation of glucose, such as the glucose transporter (GT) protein, is primarily found near “axon terminals or dendrites” (41; 42). However, enzymes for the

CAC oxidation of lactate/pyruvate, such as the (PDH)

9 complex, is localized predominantly in the “neuronal perikaryon” (41; 42).

The presence of the astrocyte-neuron lactate shuttle has been widely agreed upon in recent years, although it is still unclear why lactate is directly used and possibly more preferable used (43) in neurons rather than glucose itself. In the basal physiological state, plasma lactate concentration is pretty close to glucose concentration. There must be an explanation for why lactate is used much less quantitatively than glucose across the BBB, but is much more preferable used in neurons when present inside of the brain.

1.4 Utilization of fatty acids in the brain

1.4.1 Overview of fatty acids in the brain

The brain is rich in fatty acids (FAs), primarily in the form of complex lipids which account for as much as half of the brain’s total dry weight. In addition to the glycerophospholipids and glycerolether lipids that belong to normal constituents of cell membranes, the brain contains other forms of FAs, such as cerebrosides, sulfatides, globosides, and gangliosides (44). The FA composition of the brain extremely varies, reflecting the wide range of structural and metabolic functions of these complex lipids (44). Unlike many other organs, these complex lipids function beyond membrane composition and fat metabolism. They constitute the basic structure of the glia that support the integrity and isolation of neurons.

These lipids also participate in the transduction of nervous impulses by forming the specialized structure, the myelin. Although many FA components of the brain lipids can by synthesized de novo, some of them (e.g., essential FAs) must be

10 transported into the brain from the circulation (45; 46). These FAs can only be originated from one source, the diet, and cannot be synthesized in the human body.

1.4.2 The mechanisms of fatty acid transport across the BBB

The semi-permeability of the BBB makes it difficult for large metabolic molecules

(e.g. long-chain (LC) FAs) to enter the brain. Since the brain uses glucose as its primary energy source, FA is not generally considered a major contributor to the brain energy consumption. The consensus among most scientists is that the brain uses FAs mostly to fulfill structural and functional requirements but not metabolic requirements.

The mechanism by which the brain takes up FAs is believed to be similar to that of most other organs, but the precise mechanism still remains controversial.

Several distinct hypotheses were proposed. Two major models of these hypotheses have evolved and are supported by experimental data. They are: i) the passive diffusion model which supports the easy entry of unionized FAs into the outer leaflet of the cell-membrane lipid bilayer, across the bilayer to the inner leaflet, and into the cytoplasm (47; 48). In this passive diffusion model, transport is largely regulated by the physical properties of the membrane and the structure

(chain length and degree of unsaturation) of the FA. This model implies that all dietary fatty acids are in principle available to the brain. This is inconsistent with the data suggesting the selectivity of polyunsaturated fatty acids (PUFA) uptake. ii) the opposing protein-mediated (48; 49) facilitated transport across cell

11 membranes. Several proteins are believed to involve in this process, including fatty acid transport protein (FATP), plasma membrane fatty acid binding protein

(FABPpm) , and fatty acid translocase (FAT) (48).

The debate between above two mechanisms is currently ongoing, with confirmative data from both models in many tissues. In the brain, the uncertainty of this debate is exaggerated because of the more complex structure of the BBB.

It is highly possible that the brain may utilize specialized transport systems that differ from those in most other tissues.

1.4.3 The uptake of polyunsaturated fatty acids (PUFA)

After nearly 30 years of research, it is now agreed that essential polyunsaturated fatty acids (PUFA) provided in the diet are taken up by the brain (48; 50). More importantly, the utilization of PUFA is vital in the early development of the brain functions. The primary PUFA used in the brain are ω6 linoleic (18:2, LA), ω3 α- linolenic (18:3, ALA) acids, long-chain ω6 PUFA, such as the arachidonic acid

(20:4, AA), and ω3 PUFA, such as the eicosapentaenoic acid (20:5, EPA), or the docosahexaenoic acid (22:6, DHA).

Confirmative data of PUFA uptake in the brain was obtained from both in vitro and in vivo studies, supported by observations of labeled products derived from

PUFA chain elongation or de-saturation after the labeled precursor was directly injected into the brain or supplemented in the diet (46; 51).

1.4.4 The uptake of saturated fatty acids

The cerebral uptake of saturated FAs has long been debated. If PUFA are taken

12 up by the brain, there is no reason why saturated FAs cannot be taken up, unless the brain has a specialized transporter system targeting PUFA only.

Evidence in support of saturated FA transport across the BBB were obtained from in vitro (52-54), in vivo (21; 55), and in situ (56) studies, most of which were determined with medium-chain FAs. In particular, Edmond et al. (52) cultured neurons, astrocytes and oligodendrocytes, suggesting that all three cell types readily utilize ketone bodies (KBs) and glucose for oxidative metabolism. This study also reported that these three cell types utilize KBs at a rate 7 to 9 times greater than glucose (52). The utilization of saturated FAs for oxidation only occurred in astrocytes, and at a more preferable rate (52).

However, dietary intervention in rat pups failed to recover in their brains the labeled fatty acids (e.g. palmitic, oleic, stearic acids) and cholesterol supplemented in milk formula (45; 46; 57). Possible explanations for the discrepancy between dietary intervention and local infusion studies could be: (i) the active isotope clearance by the liver before transported to peripheral tissues in dietary trials, (ii) encouraged uptake due to non-physiological, direct isotope infusions (57) and (iii) artificial brain uptake, due to radioactivity contaminations contributed from retained isotopes in the extracellular fluid, or from labeled substrates synthesized in other organs and transported to the brain via the circulation. To avoid possible artifacts, I designed my present study with the following simple but powerful methodology. As indicated in my orientation study,

13 the cerebral uptake of [5, 6, 7- C3]heptanoate is demonstrated by the

13 accumulated M3 heptanoyl-CoA and its direct metabolites, M3 pentanoyl-CoA in the brain. M3 heptanoyl-CoA and M3 pentanoyl-CoA detected can only form in

13 the brain tissue from infused [5, 6, 7- C3]heptanoate, thus excluding contaminations from other possible sources.

Note that the brain uptake of PUFA occurs in both astrocytes and neurons (48), while the uptake of saturated fatty acids seems to be restricted to astrocytes (52;

53).

The uptake of saturated FAs in the brain is further supported by the presence of key regulatory enzymes involved in the FA oxidation pathway, such as MCD,

ACC and CPT-1 (58). These enzymes indicate the presence of normal FA catabolism in the brain.

1.5 Utilization of ketone bodies (KBs) in the brain

1.5.1 Metabolic roles

A source of energy

Ketone bodies (KBs) are a source of energy for the brain during times of glucose shortage, like prolonged fasting or ketonemia resulting from conditions associated with high-fat diet, ketogenic diet, diabetes, or neonatal suckling etc.

The importance of KBs (generally refer to as C4 KBs) as an energy source to the brain was not fully recognized, until one study found that cerebral KB uptake provided 60% of total brain energy in three obese subjects who were starved for

5-6 weeks (5). Subsequent studies have indicated that the proportion of the brain’s energy requirement that can be supplied by KBs differs among species

14 (59-61). For example, KBs can provide only a small portion (3.2%) of energy to the brain in 2-day fasted rats (61), but a much higher portion (30%) in sucking rats (60). This fraction of energy supply is still much less than in the case of humans (5). The reason for this difference is because more KBs are present in the in obese humans (5) (62) than in suckling rats (63-65).

A source of

In addition to supplying energy to the brain, KBs play a vital role in the brain anabolism, especially during the early stage of the brain development. KBs are more favorable substrates than glucose for lipid synthesis in the brain of 18-day old rats (66). KBs are also important in cholesterol biosynthesis (67) particularly in the formation of myelin which is required in normal and rapid events of neuronal transduction. So why are KBs more preferable than glucose in ? Two possible mechanisms were proposed to explain this. The fist mechanism is adaptation to neonatal suckling. Preferable utilization of KBs for lipogenesis helps maintain plasma glucose concentration for other purposes.

This is particularly important in neonatal suckling because of high-fat (HF) low (CHO) contained in milk composition (68). Another explanation is fact that KBs can be directly converted to cytosolic acetoacetyl-CoA by acetoacetyl-CoA synthetase (Figure 2.1). Cytosolic acetoacetyl-CoA thiolase can then be used in hydroxymethylglutaryl-CoA (HMG-CoA) cycle for cholesterol biosynthesis.

KBs are also incorporated into amino acids, including important antioxidant

15 glutathione, neurotransmitters glutamate, glutamine and GABA. Although KBs seem capable of fulfilling almost all functions of glucose in the brain, they do not appear to be essential. Rats fed a hypoketogenic diet develop normally (69).

Moreover, KBs are beneficial for pathological conditions associated with various

CNS malfunctions. This will be discussed later in this chapter.

1.5.2 Metabolism and regulation

KBs are primarily synthesized in the liver, via a pathway defined as ketogenesis.

C4 KBs (acetoacetate (AcAc) and β-hydroxybutyrate (BHB)) are produced from even-chain FAs present in the physiological state from natural dietary sources.

Even-chain FAs taken up and oxidized by the liver lead to the production of both acetyl-CoA and acetoacetyl-CoA in the cytosol. The combination of acetyl-CoA and acetoacetyl-CoA generates HMG-CoA. Subsequent cleavage of an acetyl-

CoA from HMG-CoA yields acetoacetate (AcAc). BHB is generated from reversible reducing of AcAc (see Fig 2.3 in Chapter 2 for details).

KBs synthesized in the liver must be exported and utilized in peripheral tissues, because liver does not have the enzyme required to activate them to CoA esters.

Note that mitochondrial acetyl-CoA itself cannot be used for ketosynthesis, because it is not available for the cytosolic enzymes required in ketogenesis.

This explains why glucose is not ketogenic, although it can generate plenty of mitochondrial acetyl-CoAs.

The amount of circulating KBs is the most significant factor affecting their uptake by the brain (5) (62). For example, in a normal state when the glucose level is

16 high and the KB level is low, the brain utilizes glucose almost exclusively.

However, when glucose is scarce or is induced, cerebral KB uptake is up- regulated. The main stimulus of ketosis is prolonged fasting or high-fat diet.

Ketosis can lead to a small increase of circulating KBs in rats (63-65), but a large increase in humans (5) (62). The amount of circulating KBs affects their contributions to the brain energy (5) (62). Peak KB concentration defined in obese humans (7.8 mmol/L (5)) is even higher than glucose (5-6 mmol/L).

The elevation of cerebral KB uptake during ketosis is achieved by up-regulating the expression of KB transporter on the BBB (64; 65; 70; 71). The transporter responsible for cerebral KB uptake was identified in the 1970s as the monocarboxylic acid transporter (MCT) (63). A family of MCTs was later found that corresponds to at least 9 genes in mammals (72). They are able to transport all monocarboxylic acids across the BBB, including KBs, pyruvate and lactate

(65). MCTs are up-regulated when plasma levels of KBs are high, based on data in both rodents (64; 65; 70; 71) and humans (73; 74). The elevation of MCT expression increases the permeability of the BBB to all monocarboxylic acids.

This explains how the brain is able to benefit from lactate and KBs during prolonged starvation but not during physiological conditions.

17

Figure 1.1 Simplified diagram showing pathways of cerebral C4 kB metabolism. Numbered enzymes are: 1=3-hydroxybutyrate dehydrogenase; 2=succinyl-CoA 3-oxoacid CoA transferase; 3=mitochondrial acetoacetyl-CoA (AcAc-CoA) thiolase; 4=AcAc-CoA synthetase; 5=cytoplasmic AcAc-CoA thiolase; 6=cytoplasmic hydroxymethylglutaryl-CoA synthase; 7=ATP-citrate lyase. A: part of pyruvate/malate shuttle for transport of acetyl-CoA out of mitochondria (from Morris, 2005)

18 Numerous studies in rodents provide evidence of the up-regulation of MCT expression (64; 65). The BBB permeability can be elevated to as much as 7 times more than basal levels in rats during suckling (64; 65), but returns to baseline post-weaning (64; 65). Prolonged fasting can also increase the BBB permeability to KBs (64; 65). Studies in humans show some increase in starving- induced KB uptake across the BBB, but not as high as in rats (75).

The detailed mechanism of the up-regulation of MCT is not full understood.

Typically, up-regulating a membrane transporter can be achieved at the cellular or genetic level. Up-regulation at the cellular level can result from the release from cytoplasmic storage, sequester from an inhibitor, or conformational change upon activation. Up-regulation at the genetic level can result from increased gene expression (transcription and/or translation). It is still not clear at what stage the dynamics of MCTs are regulated when plasma KBs are elevated.

As discussed in Section 1.3.3, astrocytes function as the glucose-metabolizing cells to make lactate. Lactate is then exported and picked up by neurons as the direct energy source. This fact is supported by the distribution of MCTs which primarily locate in neurons rather in astrocytes under normal conditions (76).

MCTs on the membranes of neuronal cells suggest the possible transfer of KBs from astrocytes to neurons, similar to the transfer of lactate.

The possible ketogenic capability of astrocytes was reported in vitro, from sources of either FAs (53) or leucine (77). It is possible that astrocytes make

KBs from FAs or other possible sources in vivo as well. KBs are then

19 transported to neurons and oxidized as an additional energy source. In fact, data from my research have demonstrated in the fasting state an elevated utilization of medium-chain (MC) FAs. These MC FAs could be used for ketogenesis in astrocytes and transported to neurons as KBs.

The expression of MCTs changes with age in rodents as discussed earlier.

Consistently, the activities of enzymes required in KB metabolism increase in suckling rats, and gradually fall back to 20-50% of their initial activities as rats grow older (78-80). These enzymes include (78-80) the cytoplasmic acetoacetyl-

CoA synthetase, the cytoplasmic acetoacetyl-CoA thiolase, the mitochondrial 3- hydroxybutyrate dehydrogenase, the mitochondrial succinyl-CoA 3-oxoacid-CoA transferase and mitochondrial acetoacetyl-CoA thiolase.

1.5.3 Therapeutic implications of cerebral KBs

Ketosis is proposed of potential therapeutic benefits in clinical controls of seizure and refractory epilepsy (81-83). The distinctive mechanism underlying the anticonvulsant activity of KBs is not completely understood, but clinical benefit of a ketogenic diet is recognized in the treatment of epilepsy (81).

Glutamate, the major excitatory neurotransmitter, is responsible for over- activation induced excitotoxicity which is presented as seizures (84; 85). Rapid, efficient neurotransmission requires that the release of a transmitter from an axonal terminal is able to induce depolarization at the post-dendritic membrane.

For this to occur, the concentration of the neurotransmitter in the synaptic cleft prior to release must be low enough. In the case of glutamate, its concentration

20 in the synaptic cleft must be more strictly controlled in order to reduce potential excitotoxicity. As discussed in Section 1.3.2, astrocytes are responsible for the clearance of glutamate from the synaptic cleft originally released from neurons after each neurotransmission event. To maintain homeostasis, astrocytes need to complete a cycle by returning glutamate back to neurons. However, this return of glutamate could further elevate its extracellular concentration. This is why the transfer of glutamate is achieved in the form of glutamine which is only synthesized in astrocytes. Glutamine is then transferred to neurons where glutamate can be re-generated. This process is called the glutamate-glutamine cycle.

Elevated extracellular glutamate can be dangerous and is the major cause of seizures. The cerebral metabolism of KBs influences the dynamics of glutamate

(83) (86) which can contribute to the reduction of excitotoxicity. KBs were reported to reduce the intracellular transamination of glutamate to aspartate in cultured mouse astrocytes (87). The transamination of glutamate to aspartate is coupled with the simultaneous conversion from oxaloacetate (OAA) to α- ketoglutarate (α-KG). The mechanism of reduced transamination is probably due to the depletion of OAA from rapid synthesis of citrate via citrate synthetase (87), induced by elevated flux from KBs through acetyl-CoA into the CAC. As a consequence, elevated glutamate can be used for GABA synthesis in GABAergic neurons and therefore increases the inhibitory neuronal signals that prevent seizures.

21 The ketogenic diet is also useful in several inherited disorders. Some patients with pyruvate dehydrogenase (PC) deficiency (see Section 3.6.6) improved with the treatment of triheptanoin which generates C5 KBs (88). PC deficiency impairs carbohydrate oxidation but does not interfere with C5 utilization by the brain. The beneficial effect of C5 KBs in patients of PC deficiency suggests their potential to improve neurological defects. As a further development, my research proposes the possible therapeutic therapy of odd-chain C5 KBs (β- ketopentanoate (BKP) and β-hydroxypentanoate (BHP)) in the treatment of neurological symptoms (developmental delay, cognitive impairment), induced by various fatty acid oxidation disorders (FOD).

The ketogenic diet is more useful in GLUT1 deficiency (68). GLUT1 deficiency syndrome, also known as De Vivo disease, is a rare genetic condition cased by inadequate transport of glucose across endothelial cells of barrier tissues. It can lead to severe mental retardation, seizures, and a series of other neurological problems (68).

GLUT1 belongs to a family of glucose transporters that are distributed differently in various organs; in adults, it is expressed at highest levels in red blood cells and endothelial cells of barrier tissues, such as the BBB. When glucose uptake cannot meet the needs of the brain energy consumption, an alternative energy source is needed. In this case, KBs are a good substitute for glucose (68). KBs can provide energy, as well as anticonvulsant actions, both of which contribute to the improvement of neurological symptoms in patients of GLUT1 deficiency.

22

Chapter 2

Anaplerosis and Cataplerosis

2.1 Overview

The processes of anaplerosis and cataplerosis refer to a series of reciprocal and/or inter-correlative reactions involved in the balance and biologic function of the citric acid cycle (CAC) (also known as the TCA or Krebs cycle). Anaplerosis describes processes that replenish the CAC by reactions other than (89). Cataplerosis describes processes that deplete the CAC by exporting the CAC intermediates.

The function of the CAC is to oxidize acetyl-CoA generated from energy substrates into useable energy, in the form of the ATP or reducing equivalents

NADH/FADH2 that are directly linked to the ATP synthesis via the electron transport chain (ETC). Proper functioning of the CAC involves the following key steps: (i) the conversion of two acetyl carbons into CO2, and (ii) the regeneration of OAA, the acceptor of the acetyl group. The CAC is composed of a series of reversible (except citrate synthase and α-ketoglutarate dehydrogenase) chemical transformation reactions. The only carbon that enters the CAC is the two-carbon unit of acetyl-CoA which combines with OAA to form citrate; the only carbon that exits the CAC is eliminated as two CO2 molecules. The four-carbon backbone

(e.g. OAA) remains unchanged for the cycle to continue constantly.

23 The CAC relies on the presence of O2 coupled with the ETC in order to generate the ATP from NADH or FADH2. The CAC turns around very rapidly. It can process as much as 1-2 μmol acetyl units per gram tissue per minute (90).

However, the concentrations of the catalytic intermediates present in the CAC are extremely low. For example, the total pool size of the eight CAC intermediates is only 1-2 μmol/g in the liver (91). The rapid turnaround rate and high throughput in the CAC requires a strict control of all catalytic intermediates to maintain their optimal concentrations. This is why anaplerosis and cataplerosis work cooperatively to maintain efficient CAC operations.

The CAC also functions in multiple biosynthetic pathways. The most commonly known biosynthetic pathways requiring the CAC are (92): (in the liver and kidney cortex), glyceroneogenesis (in adipose tissue and mammary gland), synthesis (in the liver, muscle and the kidney), and glutathione synthesis (in the liver and the brain) etc. In these pathways, CAC intermediates are required for processing (e.g. gluconeogenesis), and/or as a direct source of substrates (e.g. gluconeogenesis or ). The

CAC also regulates the balance of 3-carbon and/or 4-carbon intermediates between the cytosol and the mitochondria via the citrate/malate shuttle. This shuttle of 3-carbon and/or 4-carbon intermediates is important in glucose-induced insulin secretion from pancreatic β-cells (93). The citrate/malate shuttle also leads to the production of NADPH which is required for lipogenesis and glutathione biosynthesis (94). All of these biosynthetic pathways can contribute

24 to excessive cataplerosis that may result in the depletion of some CAC intermediates. The main cataplerotic enzymes are (92) phosphoenolpyruvate carboxykinase (PEPCK), , and aspartate aminotransferase etc.

All cataplerotic fluxes could potentially diminish the pool size of the CAC intermediates. Under most conditions, malate, citrate, and aspartate are exported in higher amounts than isocitrate and α-KG (91). During certain circumstances when biosynthetic pathways are very active (e.g. in the liver during prolonged fasting or in the brain during neurotransmission), sufficient supply of anaplerotic substrates is extremely important to maintain the pool of the

CAC intermediates and the efficient energy generation (95-98). In addition, the constant, spontaneous efflux of some CAC intermediates from the mitochondria to the cytosol or to circulation occurs during the physiological state and contributes to cataplerosis as well (91).

In addition to generating NADPH, the citrate/malate shuttle plays a role in recycling of the anaplerotic OAA to the mitochondria. Cytosolic OAA cannot re- enter the mitochondria directly. First, it must be reduced to malate by . Malate can enter the mitochondria in exchange for citrate which is in turn cleaved by cytosolic ATP-citrate lyase to OAA and acetyl-CoA. This shuttle is also required to export acetyl-CoA from the mitochondria for cytosolic lipogenesis. Note that the citrate/malate shuttle does not result in net anaplerosis, because one influx of OAA requires one efflux of citrate.

25 2.2 Functions of anaplerosis in the liver, muscle and the heart

The important role of anaplerosis in hepatic gluconeogenesis is self-evident in the liver. Gluconeogenesis is activated in the fasting state when the insulin level is low and the glucagon level is high. It works actively to maintain fasting glucose level. At the same time, the net anaplerosis is up-regulated to refill the CAC intermediates that are lost through substantial glucose production. In fact,

PEPCK flux from OAA to phospho(enol)pyruvate (PEP) can be 4.6-fold higher than the CAC flux in 24-h-fasted rats (99).

In the heart and myocardium, numerous investigations have demonstrated the coupling of stimulated anaplerosis by exogenous substrates with the improvement of myocardial reperfusion injury and other cardiomyopathies (97;

100; 101).

These cardiomyopathies, especially reperfusion injuries, damage the integrity of the membrane system, leading to the leakage of some CAC intermediates. This

“leakage” can induce excessive cataplerosis beyond the compensating ability of anaplerosis. This could potentially contribute to the progression of various pathological conditions. Among various anaplerotic substrates, propionate (102;

103), lactate (98), pyruvate (104-106), succinate (100), fumarate (107), and glutamate (108) are reported helpful in post-ischemic recovery, reduced reperfusion injury, and/or improved metabolic/contractile functions.

The most important function of the CAC is to produce energy. In conditions when elevated energy output is required, e.g. during vigorous physical activity,

26 the body accumulates AMP, because more ATP is being oxidized. At the same time, increased Ca2+ is released from intracellular storage in skeletal muscle trigged by contraction. Both increased AMP and Ca2+ activate energy-producing pathways including glycolysis (109; 110) and the CAC (111-113). The activation of the CAC can be achieved via the stimulations of pyruvate dehydrogenase, , α-KG dehydrogenase and the production of α- ketoglutarate (αKG) through alanine aminotransferase reaction (111) etc.

Consequently, the post-exercise concentration of total CAC intermediates is rapidly enhanced within the initial few minutes of muscle contraction (111-114)

Although the initial elevation of the CAC intermediates declines (112; 113) with time, Walton et al. observed a marked and sustained anaplerotic flux which persisted along with muscle contraction for at least 90 minutes (115; 116).

Fig 2.1 summarizes the main anaplerotic reactions in mammalian cells. Pyruvate is anaplerotic via pyruvate carboxylase (PC) and/or the malic enzyme (ME).

Glutamate and its precursor glutamine are converted to α-ketoglutarate via glutamate dehydrogenase and/or aminotransferases. Propionyl-CoA originating from odd-chain fatty acids, propionylcarnitine, and C5 KBs can be used to form succinyl-CoA via a three-step reaction chain catalyzed by propionyl-CoA carboxylase, D-methylmalonyl-CoA racemase, and L-methylmalonyl-CoA mutase.

Last but not the least, aspartate derived from protein degradation can be converted to oxaloacetate via transamination.

In the brain, anaplerotic reactions contribute to the regulation of the metabolism

27 + ATP 2

Propionate CO

FA

Odd-chain

KB

5

C Propionyl-CoA -Methylmalonyl-CoA -Methylmalonyl-CoA

(S) (R) Cholesterol

Ile Val Met Glutamate Glutamine -Ketoglutarate α Citrate Lactate Succinyl-CoA Acid Citric Cycle Succinate Pyruvate Acetyl-CoA Fig. 2.2 Main anaplerotic processes feeding into the citric acid cycle Fig. 2.1 2.1 Fig. Fum Oxaloacetate Malate 2 CO Aspartate

28 of neurotransmitters. Anaplerotic glutamine synthesis is also coupled with nitrogen removal from the brain in hyperammonemia (117). The detailed information of the brain anaplerosis is described in Section 2.3.

There is good evidence that a number of pathological conditions could benefit from anaplerotic therapy. Conditions associated with reperfusion injury

(myocardial infarction, stroke, and organ transplantation), PC deficiency and fatty acid oxidation disorders (FOD) were all reported of improvements after various anaplerotic therapies (see Chapter 3 for details). However, it is extremely difficult to demonstrate whether the concentrations of mitochondrial CAC intermediates are decreased or increased for several reasons: i) these CAC intermediates are present at very low concentrations, making it nearly impossible to trace each of them; and ii) the concentration analyses of the CAC intermediates from most tissue homogenates do not necessarily reflect their mitochondrial levels. This is why the beneficial effects of anaplerotic therapies were mostly demonstrated after substrates were given for treatments, although some evidence of the leakage of large molecules, such as during myocardial infarction or rhabdomyolysis, implies the leakage of the CAC intermediates as well. The fact that rhabdomyolysis (leakage of large molecules from cells) is one of the major symptoms of FOD further supports the beneficial function of anaplerotic therapy in the mouse of very-long-chain FOD (VLCAD) proposed in the present study.

2.3 Anaplerosis in the brain

29 2.3.1 Overview

The significance of anaplerosis in the brain was not recognized until the discovery of highly active pyruvate carboxylase (PC) which catalyzes the carboxylation of pyruvate to OAA (118). PC is the most widely distributed anaplerotic enzyme that is actively involved in gluconeogenesis in the liver and in the citrate/malate shuttle present in pancreatic β-cells. High activity of PC was later discovered in astrocytes (119), whereas activities of other gluconeogenic enzymes remain very low, indicating non-gluconeogenic anaplerosis (119). The activity of PC was later demonstrated to be required in the labeling of aspartate and glutamate from H14CO3-(120). These findings support the presence of anaplerosis via PC catalyzed carboxylation. Meanwhile, some CO2 fixation is also reported independent of PC, but requires NADPH rather than ATP, consistent with reaction from pyruvate to malate catalyzed by malic enzyme (ME)

(121).

2.3.2 The significance of the brain anaplerosis

The most important cerebral anaplerotic pathway is the carboxylation of pyruvate to OAA (catalyzed by PC) and to malate (catalyzed by ME). PC is a mitochondrial, biotin-dependent enzyme whose action depends on Mg2+ and ATP.

PC catalyzes the irreversible carboxylation of pyruvate to form OAA as follows:

Pyruvate + HCO3- + ATP → oxaloacetate + ADP + Pi (122)

PC is expressed in a tissue-specific manner. It is most active in the liver

(gluconeogenesis), kidney (gluconeogenesis and amino acid synthesis),

30 pancreatic β-cells (glucose-induced insulin secretion), adipose tissue

(glyceroneogenesis) and lactating mammary gland (glyceroneogenesis) (123).

PC activity is moderate in the brain, heart, adrenal gland, white blood cells and skin fibroblasts (123).

ME is present as two isoforms, one cytosolic, and the other mitochondrial. ME catalyzes the following reversible reaction (121) which depends on NADPH-,

Mg2+ or Mn2+;

+ + Pyruvate + CO2 + NADPH + H ↔ malate + NADP (124)

In addition to anaplerosis, ME also contributes to biosynthetic pathways, including the syntheses of FAs and glutathione (94). These biosynthetic pathways require NADPH which is formed during the reversible decarboxylation from malate to pyruvate.

There is good evidence indicating that the astrocyte is the anaplerotic compartment for several reasons: i) the activities of PC (119) and ME (121) were identified only in astrocytes; ii) when radioactively labeled bicarbonate was injected intravenously to the brain, the radioactivity in glutamine was found to be higher than that of glutamate (125). This is consistent with the role of astrocyte in the glutamine/glutamate cycle. As previously mentioned, the astrocyte plays an important role beyond being a supporting tissue to the neuron. It functions in the shuttle of energy substrates (glucose/lactate cycle) and neurotransmitters

(glutamate/glutamine cycle), that are vital for the functional efficiency and anatomical integrity of the central nervous system (CNS).

31 The glutamate/glutamine cycle contributes to the compartmentation effect of the glutamate/glutamine between the astrocytes and neurons. Glutamate is synthesized in the nervous system by the astrocytes and converted to glutamine by the glia-specific enzyme glutamine synthetase (126; 127). Glutamine is then secreted to the neuronal synapses, where it is taken up by the neurons and subsequently converted to glutamate by glutaminase. During neurotransmission, glutamate is released to synaptic cleft by glutamatergic neurons, where it is rapidly cleared by astrocytes to restore pre-neurotransmission conditions. In

GABAergic neurons, glutamine transferred from astrocytes is also converted to glutamate which is used for GABA synthesis via decarboxylation. GABA released by GABAergic neurons during neurotransmission is also cleared by astrocytes. When the brain is very active, constant neurotransmission results in busy trafficking of the glutamate/glutamine shuttle which requires an elevated production of glutamine from α-KG. The continuous consumption of α-KG

(cataplerosis) eventually results in the depletion of the CAC intermediates in astrocytes. Therefore, the operation of the glutamine-glutamate cycle requires a continuous supply of the CAC intermediates, especially α-KG to be made available through anaplerotic pathways.

Nearly 60% of the astrocytic CAC intermediates are continuously lost per turn of the CAC, mostly in the form of glutamine (128; 129). Because active neurotransmission requires an elevated energy supply in both astrocytes and neurons, a sufficient supply of the CAC intermediates is required for high energy

32 output, in addition to the synthesis and cycling of glutamate and glutamine. This is why anaplerosis is needed to refill the pool of the CAC intermediates in astrocytes. Elevated energy requirements for neurons can lead to further consumption of glutamine for catabolism, rather than for the synthesis of glutamate (129-131). This could result in further cataplerosis from increased consumption of α-KG in astrocytes. An insufficient supply of the CAC intermediates from anaplerosis can result in severe compromises of the CNS functions, similar to the brain abnormalities seen in PC-deficient patients who carry one or more forms of mutations on the PC gene.

Glutamine can be used for other purposes other than neurotransmission.

Glutamine formation is a main pathway of cerebral ammonia detoxification in astrocytes (132; 133). Ammonium is formed in various de-amination reactions and from the purine nucleotide cycle. Ammonium toxicity in mammals involves perturbation of cellular pH and activation of certain ion channels (such as the antagonist of N-methyl d-aspartate type glutamate (NMDA) receptor) (134) that can result in increased permeability of the BBB to allow the entry of large molecules nonselectively. Ammonium toxicity can lead to elevated glutamine synthesis and can result in astrocyte swelling, cellular dysfunction, brain edema and death (134). The elimination of ammonium is achieved by the exportation of glutamine, at about 10 nmol/g · min-1, from the brain to the circulation (135).

Therefore ammonium detoxification via glutamine eventually contributes to the loss of α-KG.

33 Like all other amino acids, glutamate and glutamine are also consumed in peptide and/or protein synthesis, such as the synthesis of glutathione (136).

Glutathione (136) is the most important antioxidant that protects cells from toxin of free radicals; and can be synthesized from cysteine, glutamate and glycine, via the following two ATP-dependent steps (136): i) First, glutamate and cysteine are used to form γ-glutamylcysteine, catalyzed by the rate-limiting enzyme γ- glutamylcysteine synthetase (also known as glutamate cysteine ligase, GCL); ii)

Second, glycine is added to the C-terminal of γ-glutamylcysteine to form glutathione, catalyzed by glutathione synthetase. Glutathione exists in both reduced (GSH) and oxidized (GSSG) forms. The reduced form GSH has a thiol group in its cysteine moiety and can donate a reducing equivalent (H+ + e-) to other unstable molecules, such as reactive oxygen species (137). After donating an electron, glutathione itself combines with another reactive glutathione to form oxidized GSSG (136). GSH is then regenerated from GSSG by the enzyme glutathione reductase, which catalyzes a NADPH-dependent reaction (136). In physiological conditions, more than 90% of the total glutathione pool is in the reduced form GSH (137). The ratio GSSG/GSH is a good reflection of oxidative stress in human (136; 137). Since the brain consumes about 20% of the total body oxygen consumption (1), reactive oxygen species are continuously generated at high rates (136). This is why GSH maintains at a high concentration in the brain in order to defend against oxidative stress (136).

Recent studies indicate the supply of glutathione by astrocytes to neurons (136).

34 This suggests active biosynthesis of glutathione in astrocytes, thus leading to further loss of the glutamate and α-KG.

In addition to α-KG, other CAC intermediates in astrocytes can leave the CAC for alternative metabolic functions. For example, malate leaves the mitochondria and enters the cytoplasm, where it becomes decarboxylated into OAA via decarboxylation of to pyruvate catalyzed by ME. This reaction leads to the formation of NADPH which is consumed in a number of biosynthetic pathways, such as the lipogenesis and for the reduction of oxidized glutathione (94) (136).

Glial cells, and possibly neurons as well, may also lose the CAC intermediates through the export of citrate to the extracellular fluid as well (138-140).

Note that glutamate and GABA can also be anaplerotic, after they are re-taken up by the astrocyte from the synaptic cleft after neurotransmission events.

Glutamate can be used to form glutamine for the next turn of glutamate/glutamine shuttle, or feed into the CAC. On the other hand, GABA is anaplerotic when it is degraded to succinate via the GABA shunt, after converting to succinic semialdehyde (see Fig. 2.2).

35 Neuron CAC KG α GABA Glutamate Glutamine neurotransmission neurotransmission -ketoglutarate inherent in the release of of release the in inherent -ketoglutarate GABA α Glutamate plasma plasma Synaptic cleft Synaptic GABA

Ammonium detoxification

e e e

Glutamine Glutathion Glutamate

d d d

Other proteins

c c c

hy hy hy

ni ni ni

i i i

de de de

l l l

c c c

a a a

c c c

i i i

u u u

m m m

S S S

e e e

-KG

s s s α Citrate Propionyl-CoA Methylmalonyl-CoA Succinyl-CoA CAC Succinate Glucose Pyruvate Acetyl-CoA KBs, 5 Oxaloacetate Malate odd-chain FAs odd-chain AAs, C AAs, transmitters glutamate, glutamine and GABA and glutamine glutamate, transmitters Fig. 3.3 Simplified scheme of the CAC and the anaplerosis for the loss of of loss the for anaplerosis the and CAC the of scheme Simplified 3.3 Fig. Fig. 2.2 Fig. 2.2 Astrocyte Aspartate 36 Figure 2.2 summarizes the main cataplerotic and anaplerotic pathways in astrocytes, as well as the coupled glutamate/glutamine shuttle. A net loss of the

CAC intermediates induced from cataplerosis via α-KG, malate, or citrate will reduce the oxidative capacity of the cell via an insufficient supply of OAA, the recipient of actyl-CoA. This could lead to rapid depletion of the ATP and subsequent loss of membrane potentials and systematic neurological functions.

For a long time, anaplerosis was considered present only in astrocytes, based on the detection of PC in astrocytes instead of neurons (119). It was later found in neurons as well (118; 141; 142). The carboxylation of pyruvate to malate was also confirmed by malic enzyme (ME) activity detected in both astrocytes and neurons (143-145). Therefore, neurons themselves may also have anaplerotic pathways, from pyruvate and/or glutamate, in order to maintain an optimal pool of the CAC intermediates for efficient and sufficient energy production.

2.4 Major anaplerotic pathways

2.4.1 From pyruvate

Anaplerosis from pyruvate via PC is most important in physiological conditions.

Carboxylation via PC generates OAA, the direct recipient of acetyl-CoA that propels the continuation of the CAC. PC is widely distributed in a tissue-specific manner, and plays an important role in nearly all metabolic functions that requires anaplerosis, including gluconeogenesis, glyceroneogenesis, amino acid synthesis, lipogenesis and β-cell functions.

37 PC contributes to gluconeogenesis in both the liver and the kidney. The highly active PC together with other gluconeogenic enzymes, such as PEPCK, fructose-

1,6-bisphosphatase and glucose-6-phosphatase, function in glucose output in the fasting state. This output could account for up to 96% of total glucose production in humans (96). The activities of these gluconeogenic enzymes are elevated during prolonged fasting, through the hormonal effect of glucagon or the up- regulation of gene expression (96).

In lipogenic tissues, such as the liver, the adipose tissue and the lactating mammary gland, PC contributes to lipogenesis via glyceroneogenesis (146-150) which refers to the generation of glycerol moiety of triacylglycerol from pyruvate, and is up-regulated by peroxisome proliferators-activated receptor γ (PPAR γ) during the fasting state (151). Glyceroneogenesis continuously forms glycerol which is used in the esterification of FAs to form (TGs). Note that lipogenesis also requires NADPH which can be produced in the citrate/malate shuttle (94). Therefore, PC can contribute to NADPH production by generating more mitochondrial malate.

In pancreatic β-cells, anaplerosis through PC provides essential signals for glucose-stimulated insulin secretion (GSIS) (152; 153). The underlying mechanism could be the pyruvate/malate cycle that shuttles 3C and 4C between the mitochondria and the cytosol (93). Pancreatic β-cells sense extracellular glucose concentration and secret insulin when glucose is high (154). This glucose sensing and insulin secretion processes require (154): i) the rapid influx

38 in β-cells of glucose which is catabolized via aerobic glycolysis plus the CAC; and ii) the carboxylation of pyruvate to OAA catalyzed by active PC which stimulates the shuttle of pyruvate/malate between the mitochondria and the cytosol. The pyruvate/malate shuttles across the mitochondrial membrane through the following steps (93): First, PC catalyzes the formation of OAA from pyruvate, which is subsequently converted to malate; Then, malate crosses the membrane to the cytosol, where it is decarboxylated to pyruvate, producing

NADPH; Next, a rise in the ATP/ADP ratio induced by glucose catabolism leads to elevated intracellular K+ via closure of ATP-regulated K+ channels; Finally, K+ induces subsequent membrane depolarization, Ca2+ influx and exocytosis of insulin (155). High PC activity discovered in pancreatic β-cells was once thought to contribute only to the supply of OAA for more efficient ATP production via the

CAC, until a link between GSIS and the 3C/4C (pyruvate/malate) shuttle was identified (93), with an emphasis on NADPH production. In contrast, an alternative citrate/malate shuttle was later proposed (156). The citrate/malate shuttle (156) starts with the synthesis of the mitochondrial OAA from pyruvate via

PC. OAA accepts acetyl-CoA to form citrate by citrate synthase. Citrate then crosses the mitochondrial membrane to the cytosol, where it undergoes ATP- dependent cleavage to OAA and acetyl-CoA. OAA is subsequently reduced to malate by MDH with NADPH formed simultaneously.

Anaplerosis from pyruvate can also occur via the ME which catalyzes a reversible NADPH-dependent reaction to convert pyruvate into malate (94). The

39 expression of ME is regulated by complex hormonal system and nutritional status, including elevated thyroid hormones or higher dietary intake of CHO, for example

(157).

2.4.2 From propionyl-CoA

Anaplerosis from propionyl-CoA is the primary focus of this research study.

Propionic acid is an odd-3-carbon carboxylic acid. Although different from naturally present even-chain FAs, some propionic acid can be found in foodstuffs, because it is a commonly used preservative. Bacteria of the genus

Propionibacterium class produce propionic acid as the end product of their anaerobic metabolism (158). This class of bacteria is commonly seen in the stomachs of ruminants and the sweat glands of humans. In non-ruminant mammals, some propionate can be formed in the gut by intestinal .

About 99% of the portal vein propionate is taken up by the liver, resulting in arterial propionate concentrations of about 5 μM (91).

Propionyl-CoA can be formed from i) the activation of free propionate; ii) the catabolism of odd-chain fatty acids which are generally present at very low concentrations; iii) the catabolism of four amino acids: isoleucine, valine, threonine, and methionine; and iv) the conversion of cholesterol to bile salts in the liver. In physiological state, since the catabolism of amino acids produces little of propionyl-CoA, and since plasma concentration of propionate remains very low, only small amounts of propionyl-CoA is available for anaplerosis in peripheral tissues.

40 Propionyl-CoA contributes to anaplerosis through succinyl-CoA which is a CAC intermediate. It is first carboxylated by the ATP and biotin-dependent propionyl-

CoA carboxylase to S-methylmalonyl-CoA. The latter is then epimerized to R- methylmalonyl-CoA via a B12-dependent mutase to succinyl-CoA.

Precursors of propionyl-CoA

Propionate

Propionate is proved to be an excellent anaplerotic substrate, even at a low concentration of 0.1-1 mM (159), however, propionate does not provide an energy source for of the ATP or acetyl-CoA.

The direct use of propionate as an anaplerotic substrate could be dangerous because of its toxicity when overloaded. Propionic acidemia (160; 161) induced by propionyl overload is characterized by abnormal propionyl-CoA carboxylation or biotin deficiency that may lead to trapping of free CoAs. Propionyl overload is characterized by accumulations of free propionate, propionylcarnitine, 3- hydroxypropionate (160), methyl citrate, and C5 KBs (BHP and BKP), as well as odd long-chain FAs. The clinical representations of propionic acidemia are largely neurological, including neuron-degeneration, mental retardation and motor impairment (161). This is possibly due to the rapid utilization of propionic acid by glia, as indicated in cell culture studies (162). In contrast, the oxidation of propionic acid was not found in cultured neurons in the same preparation (162).

The oxidation of propionic acid in glia was found to increase histone H4 acetylation (162) which is linked to various events of translational activation and

41 is proposed to contribute to neurotoxicity of propionic acidemia (162). Histone acetylation alters nucleosome and chromatin structure, exposing nucleosomal

DNA buried in chromatin fibers accessible to transcription factors (163).

To summarize, propionate is a potential anaplerotic substrate in peripheral tissues. Propionate’s contribution to the anaplerosis required in the shuttle of neurotransmitters makes it potentially beneficial to the brain; but its alteration of

DNA makes it potentially toxic. Therefore, anaplerotic therapy directly using propionic acid could be dangerous. Neurological toxicity tests of propionate are required before treatments in humans.

Triheptanoin

Triheptanoin is a based on 7-carbon FA heptanoate. After hydrolysis, one glycerol and three heptanoate are produced. With oral administration, most of the heptanoate reaching the liver via the portal vein is absorbed and oxidized into 2 acetyl-CoAs + 1 anaplerotic propionyl-CoA. Propionyl-CoA ether enters the CAC via succinyl-CoA for gluconeogenesis or used for C5 KB synthesis of

BKP (β-ketopentanoate) and BHP (R-β- hydroxypentanoate) via reactions of the

HMG-CoA cycle:

AcAc-CoA thiolase

Propionyl-CoA + acetyl-CoA β-ketopentanoyl-CoA + CoA

HMG-CoA synthase

β-ketopentanoyl-CoA + acetyl-CoA hydroxyethylglutaryl-CoA + CoA

42 HMG-CoA lyase

hydroxylethylglutaryl-CoA R-3- hydroxypentanoate + acetyl-CoA

Currently triheptanoin has been widely applied clinically as a standard anaplerotic therapy in the treatment of various fatty acid oxidation disorders

(FOD). For example, patients with long-chain FOD have benefited from replacing their trioctanoin (medium even-chain triglyceride) diets with triheptanoin.

This diet has improved their clinical conditions and quality of life (164) (see

Section 3.6.5). The triheptanoin diet has also been effective in a patient with inborn PC deficiency (88) (see section 3.6.6 in chapter 3).

However, the triheptanoin therapy is obviously limited to the management of medium- or short-chain FOD that are result from defective enzymes to degrade medium- or short-chain FAs.

Tripentanoin

Pentanoate, a 5-carbon FA, is also an anaplerotic substrate since its initial degradation forms propionyl-CoA and acetyl-CoA. Unlike heptanoate, it can be metabolized by patients with medium-chain FAO disorders. Hydrolysis of tripentanoin forms 5-carbon based pentanoyl-CoA by isovaleryl-CoA dehydrogenase, an enzyme of branched-chain amino acid catabolism.

C5 Ketone bodies

In peripheral tissues, the C5 KBs (BHP, BKP) are metabolized similarly to C4 KBs

(BHB, AcAc). R-β-hydroxybutyrate dehydrogenase catalyzes the equilibrium between β-ketopentanoate (BKP) and R-β-hydroxypentanoate (BHP):

43 BKP + NADH + H+ ↔ R-BHP + NAD+

BKP is activated and catabolized via the following two mitochondrial reactions:

3-oxoacid-CoA transferase

BKP + succinyl-CoA BKP-CoA + succinate

acetoacyl-CoA thiolase

BKP-CoA propionyl-CoA + acetyl-CoA

BKP is also activated by cytosolic acetoacyl-CoA synthetase:

BKP + CoA + ATP → BKP-CoA + AMP + PPi

The C5 KBs BKP and BHP are analogs of the C4 KBs acetoacetate and R-β- hydroxybutyrate, respectively. They are exported from the liver and utilized in the peripheral tissues, generating both acetyl-CoA and anaplerotic propionyl-CoA

(see Fig. 2.3).

C5 KBs can be very effective anaplerotic substrates in peripheral tissues of the heart, brain and kidney, because these organs have highly active KB utilizing enzymes (165-167). This is supported by improvements in the cardiac, muscular or neurological symptoms of patients with inborn errors of LC FOD (164), CPTII deficiency (168), or PC deficiency (88), after dietary treatment with anaplerotic triheptanoin.

44 Ac-CoA Energy 2AcCoA Suc AcAcCoA Anaplerosis Prop-CoA SucCoA -CoA BKP BKP BHP BHB BHP BKP BKP BHP Ac-CoA Ac-CoA modifications) HEG -CoA BKP-CoA Ac-CoA Prop-CoA CoA Odd-chain Fig. 3.4 Mitochondrial ketonebody in hepatic and non-hepatic tissues(from Fukao et al.2004with Fig. 2.3

45

Chapter 3

Fatty Acid Oxidation and Its Disorders

3.1 The definition and function of fatty acids

A fatty acid (FA) is a carboxylic acid with a straight alphatic hydrocarbon chain.

Naturally occurring FAs commonly have a chain of 4 to 28 carbons which are usually even-numbered and unbranched. Odd-chain FAs are unusual, but can be derived from odd-chain triglycerides (TG), branched fats or certain amino acids.

By definition, FAs are categorized into saturated and unsaturated FAs, depending on whether a double bond is present. FAs also differ in carbon length at the same time. Saturated FAs do not have double bonds, or other functional groups along the chain. Saturated FAs have a condense structure which only contains carbons and hydrogens in addition to the carboxylic group. The most commonly occurring FAs are C4:0 (butyric), C6:0 (caproic), C8:0 (caprylic),

C10:0 (capric), C12:0 (lauric), C14:0 (myristic), C16:0 (palmitic), C18:0 (stearic),

C20:0 (arachidic), C22:0 (behenic) and C24:0 (lignoceric) acid.

In contrast, unsaturated FAs form multiple geometrical structures because they have a bendable carbon bone with one or more double bounds. For example, two carbon atoms that are bound to a double bound can exist in a cis- or trans- configuration. A cis configuration causes the chain of the FA to bend because

46 the hydrogen atoms bound to the adjacent carbons are on the same side of the double bond. The bendy configuration of cis- FAs can avoid closely packed FAs especially when they are part of a bilayered membrane or TG in a lipid droplet.

The degree to which FAs are packed together largely affects the mobility of a lipid-composed structure, e.g. the cell membrane. In contrast, a trans configuration maintains a straight structure because its hydrogen atoms bound to the adjacent carbons are on the opposite side of the double bond. The most naturally occurring unsaturated FAs have a cis- configuration. Most trans- FAs result from food processing in industry.

FAs have three major physiological roles. First, they are building blocks of phospholipids and glycolipids that constitute cellular structures like the cell membrane. Second, they are stored as triglycerides (TGs) which are esters of fatty acids with glycerol. The oxidation of TGs and FAs directly is the most important energy source in the fasting state. Third, their derivatives can function as hormones and intracellular messengers that affect metabolic fluxes, immune system, inflammation, and/or gene expressions, e.g. via association with PPAR pathways.

The human body can synthesize most of FAs needed from other sources, except for polyunsaturated FAs (PUFA) linoleic acid (LA) and α-linolenic acid (ALA). In addition, the human body requires supplies of longer chain omega-3 fatty acids eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), because of its limited ability to synthesize them from LA and ALA. These PUFA are essential,

47 and must originate from diet. Food sources like plant oils and marine oils (e.g. from seal or fish) are rich in parent compounds of essential PUFAs of omega-6 and omega-3 FAs. Omega-6 and omega-3 FAs contain double bonds at the omega-6 or omega-3 position which are beyond the synthetic ability in mammals.

Essential PUFAs are important for the regulation of blood pressure and inflammation, because they are precursors of important signaling lipids, such as prostaglandins that regulate vascular functions (44; 169). PUFAs are also vital to the development of the CNS (44; 169). This is why PUFAs are recommended for newborns and lactating mothers.

For reasons that are not fully understood, trans fat and trans FAs are considered harmful, because of their strong correlation with cardiovascular disease (CVDs)

(170), such as atherosclerosis and coronary heart disease. Elevated trans fat intake has been shown to raise the bad LDL cholesterol but reduce the good

HDL cholesterol levels, both of which are risk factors in the induction and progression of various CVDs.

FAs can also be categorized by length: short-chain FAs (SCFA) generally contain

8 carbons or less; medium-chain FAs (MCFA) generally contain 8-14 carbons; long-chain FAs (LCFA) contain more than 16 carbons, and very-long-chain FAs

(VLC FA) contain more than 20 carbons.

3.2 Fatty acid transport system and cellular uptake

Non-esterified FAs (NEFA) are free FAs that can travel in the blood when complexed with albumin (a soluble 66-kD protein). Association of NEFAs with

48 albumin (4 FAs: 1 albumin) can largely improve the solubility of long-chain FAs, most of which are normally insoluble in water. Esterified FAs (TGs) and cholesterol are transported through a complex system composed of a family of lipoproteins that are differentiated by function and size. A lipoprotein contains both proteins (apolipoproteins) and lipids. Major circulating lipoproteins include chylomicrons, high density lipoprotein (HDL), low density lipoprotein (LDL), intermediate density lipoprotein (IDL), and very low density lipoprotein (VLDL).

The cellular uptake of NEFAs starts with their dissociation from the albumin-fatty acid complex, facilitated by the albumin-receptor (ALB-R). Cellular transport of fatty acids involves (i) passive diffusion (flip-flop) of FAs through the lipid bilayer,

(ii) fatty acid translocase (FAT/CD36) (171), and (iii) fatty acid-binding protein

(FABPpm) (172), and (iv) FA transport protein (FATP) to facilitate the transport of

FA into plasma membrane (173).

3.3 Fatty acid oxidation (β-oxidation)

The purpose of fatty acid degradation is to convert an aliphatic compound into acetyl-CoA that can be processed into usable energy through CAC. FA degredation occurs both in the cytosol and the mitochondria. FAs are first activated in the cytosol to generated fatty acyl-CoA esters that enter the mitochondria via the carnitine cycle. Subsequent processes of fatty acyl-CoAs occur inside of the mitochondria and involve the beta-oxidation cycle, the CAC, and the electron-transfer chain (ETC).

Activation of FAs into acyl-CoAs

49 In the cytosol, VLC and LC FAs have to be activated into their coenzyme A (CoA) esters before they enter the mitochondria for oxidation or are trans-esterified to other lipids. The global reaction of FA activation is:

RCOO- + CoASH + ATP4-↔ RCOSCoA + AMP2- + PPi3- (122)

This reaction is catalyzed by a number of different acyl-CoA synthetases called thiokinases. These actyl-CoA synthetases are present as different isozymes that catalyze the activation of FAs with different structures, e.g. straight-chain saturated and unsaturated FAS, branched-chain FAs, and dicarboxylic FAs. The cellular localization of acyl-CoA synthetases also varies by the type of isozymes.

Acyl-CoA synthetases for VLC (C>20) and LC (C16-C20) FAs are present in the outer membrane of the mitochondria and in the membrane of the peroxisomes;

Acyl-CoA synthetase for MC (C4-C14) and SC (C2-C4) FAs is located in the inner mitochondrial membrane.

The carnitine palmitoyltransferase (CPT) system

LC acyl-CoAs enter the mitochondria via the CPT system which contains three enzymes: carnitine palmitoyl transferase I (CPT I) in the outer mitochondrial membrane; the carnitine: acylcarnitine translocase (CACT), an integral inner mitochondrial membrane protein; and carnitine palmitoyl transferase II (CPT II), in the inner mitochondrial membrane (174).

CPT I catalyzes the formation of LC acylcarnitine from LC acyl-CoA and L- carnitine, releasing free CoA into the . This reaction is:

LCFA-CoA + L-carnitine ↔ LCFA-carnitine + CoA (122)

50 The newly formed acylcarnitine is then rapidly transported into the mitochondrial matrix via a carnitine: acyl carnitine translocase (CACT). In the last step, acyl-

CoA is regenerated in the mitochondrial matrix by CPT II:

LCFA-carnitine + CoA ↔ LCFA-CoA + L-carnitine (122)

There are three human isoforms of CPT I that are defined according to their primary distributions: a liver isoform (L-CPT I), a muscle isoform (M-CPT I) and a brain isoform (C-CPT I). CPT I is the rate limiting enzyme. Unlike CPT I, there is only one isoform of CPT II in humans.

The CPT system is the rate-limiting step of FA degredation which makes it an obvious regulatory target. All three CPT enzymes appear to be associated with the outer mitochondrial membrane and connected to malonyl-CoA which is the major inhibitor of FA oxidation. The regulational mechanism of malonyl-CoA will be discussed latter in this chapter. Note that the brain isoform of CPT I (C-CPT-I) which is not correlated with FA translocation into the mitochondria or beta- oxidation (175) in vitro, possibly suggests its distinct function.

In contrast to LC FAs, MC and SC FAs can traverse the mitochondrial membrane independent of the CPT system. They are activated within the mitochondrial matrix as MC or SC acyl-CoAs. This is especially beneficial for the clinical treatments of patients having genetic disorders of LC or VLC FOD which are resulted from the absence or dysfunction of one or more enzymes involved in any step of LC/VLC FA degredation. For example, genetic deficiency of CPT I or

CPT II cannot translocate LC FA into the mitochondria for degredation, thus it

51 can be treated therapeutically with MC or SC FAs which do not require the CPT system for degradation.

Mitochondrial β-oxidation

Once inside the mitochondria, the β-oxidation of FAs occurs via four recurring steps:

Oxidation by acyl-CoA dehydrogenase:

Fatty acyl-CoA is converted by acyl-CoA dehydrogenase to form trans-Δ2-enoyl-

CoA ester by creating a double bond between C2 and C3:

2 Cn Acyl-CoA + E-FAD → trans-Δ -enoyl-CoA + E-FADH2 (176)

There are four different acyl-CoA dehydrogenases present in the mitochondria, with specificities for various acyl-CoAs of different chain length, such as VLC, LC,

MC or SC acyl-CoAs.

Hydration by enoyl-CoA hydratase:

Enoyl-CoA hydratase converts trans-Δ2-enoyl-CoA to L-3-hydroxyacyl-CoA, eliminating the double bond between C2 and C3:

2 Trans-Δ -enoyl-CoA + H2O ↔ L-3-hydroxyacyl-CoA (176)

This reaction is stereospecific, generating only the L isomer.

Oxidation by NAD+:

NAD+ dependent L-3-hydroxyacyl-CoA dehydrogenase catalyzes the formation of

3-ketoacyl-CoA and NADH:

L-3-hydroxyacyl-CoA + NAD+ ↔ 3-ketoacyl-CoA + NADH + H+ (176)

Thiolysis by β-ketothiolase:

52 3-ketoacyl-CoA thiolase inserts a thiol group between α and β carbons, forming acetyl-CoA and a new acyl-CoA containing two C atoms fewer than the original one:

3-keto- Cn-acyl-CoA + CoA ↔ acetyl-CoA + Cn-2 acyl-CoA (176)

The overall equation of each beta-oxidation cycle is:

+ - Cn Acyl-CoA + FAD + NAD + H2O + CoAS → Cn-2 Acyl-CoA + FADH2

+ NADH + Acetyl-CoA (176)

The enzymes of beta-oxidation are organized into a multienzyme complex (177-

179). This metabolic channeling results in high efficiency of the sequence of reactions with little accumulation of intermediates.

The final product acetyl-CoA enters the CAC to produce 3 NADH, 1 FADH2, 1

GTP, and 2 CO2. The process of acetyl-CoA by the CAC is introduced in Section

2.1 of Chapter 2.

To sum up, each step of the beta-oxidation cycle shortens the fatty acyl-CoA by two carbons until it is completely converted into acetyl-CoA. The reducing equivalents (NADH, FADH2) generated in the CAC enter the ETC with other equivalents produced in the CAC, resulting in the net synthesis of the ATP.

In the liver, most of the acetyl groups derived from β-oxidation are used for KB synthesis of BHB and AcAc which are exported to other tissues (e.g. the heart and the brain) for terminal oxidation.

Oxidation of unsaturated FAs and odd-chain FAs

The presence of a cis double bond in some cis unsaturated FAs can prevent the

53 formation of a trans-Δ2-enoyl-CoA. The cis double bond can be converted to trans-Δ2 bond that can be processed by normal β-oxidation. This conversion is handled by one of the two other enzymes: cis-Δ3-Enoyl CoA isomerase which converts a cis-Δ3 (odd numbered) double bond into a trans-Δ2 bond; or NADPH- dependent 2,4-dienoyl CoA reductase which reduces a cis-Δ4 double bond (even numbered double bond) to a trans-Δ3-enoyl CoA.

The oxidation of odd-chain FAs, as introduced in Chapter 2, generates a propionyl-CoA as an end product in addition to acetyl-CoA. Propionyl-CoA can be utilized for C5 KB (BKP, BHP) synthesis in the liver or used for anaplerosis via succinyl-CoA into the CAC in peripheral tissues (e.g. the muscle, adipose tissue, brain, pancreatic β-cells etc.).

Oxidation in peroxisomes

When the chain is too long to be handled by the mitochondria, FA oxidation occurs partially in peroxisomes. Peroxisomal oxidation ends at octanoyl-CoA, and responsible for initial degradations of VLC FAs, branched-chain FAs, and LC dicarboxylic acids.

3.4 Malonyl-CoA metabolism

Malonyl-CoA is important for the regulation of FA oxidation, because it inhibits

CPT I (180). Malonyl-CoA, an intermediate in FA synthesis, is generated from the carboxylation of acetyl-CoA with CO2. This carboxylating reaction is catalyzed by acetyl-CoA carboxylase (ACC), a multifunctional enzyme with a biotin carboxyl carrier protein, biotin carboxylase, and carboxyl transferase

54 domains (181):

¯ ¯ ATP + HCO3 + ACC·biotin → ACC·biotin·CO2 + ADP + Pi (176)

¯ ACC·biotin·CO2 + acetyl-CoA → ACC·biotin + malonyl-CoA (176)

Malonyl-CoA is predominately present in the cytosol, where it has two major metabolic fates: i) the cytosolic FA synthesis in lipogenic tissues of the liver, white adipose tissue and the lactating mammary gland, during which the production of malonyl-CoA by ACC1 is the rate-limiting step; ii) the degradation back to acetyl-CoA in non-lipogenic tissues of the heart and skeletal muscle, via decarboxylation catalyzed by malonyl-CoA decarboxylase (MCD). In vitro, a side-reaction of the mitochondrial propionyl-CoA carboxylase can also synthesize malonyl-CoA. Whether or not this reaction occurs in vivo is unclear (182).

In mammals, ACC is present as two distinct isoforms defined by their major distributions: ACC1 (the liver form, Mr = 265 kDa) and ACC2 (the muscle form,

Mr = 280 kDa). These two isoforms differ in 15% of their amino acid sequences, but still have the same domain structure (183).

In addition, ACC2 has an extra 200-residue sequence at the N-terminus (183).

This unique structure of the N-terminal peptide suggests that ACC2 is a membrane-targeted protein that produces malonyl-CoA in close proximity to the active site of CPT I (184; 185). This explains why a different isoform of ACC

(ACC2) is needed in non-lipogenic tissues such as the heart and skeletal muscle.

The association of ACC2 with the mitochondria is consistent with the hypothesis that ACC2 is involved in the regulation of mitochondrial FAO through the

55 inhibition of CPT I by malonyl-CoA.

Additional studies with affinity-purified anti-ACC2-specific antibodies have targeted that ACC2 as a membrane binding protein, and ACC1 as a purely cytosolic protein (186). This is consistent with the fact that ACC1, predominantly distributed in lipogenic organs, controls the rate of , while

ACC2, predominantly distributed in oxidative organs, regulates energy metabolism by regulating FA oxidation.

The decarboxylation by MCD is the only known fate of malonyl-CoA in non- lipogenic organs (heart and skeletal muscle). MCD activity is found in the mitochondria, peroxisomes, and the cytosol (187-190), with the peroxisomal enzyme showing the highest specific activity. Active cytosolic MCD is directly involved in the regulation of malonyl-CoA. Peroxisomal MCD may interact with a peroxisomal-specific CPT I (191) which may control the LC FA oxidation occurring locally.

To sum up, ACC and MCD catalyze a substrate cycle that controls the concentration of malonyl-CoA, thus regulating FA oxidation via the inhibition of

CPT I.

3.5 Regulations of fatty acid oxidation

3.5.1 Overview

The regulation of FA oxidation, primarily by hormones is a component of net energy balance which remains stable despite a constant switch between the feeding and fasting state. The network of hormonal system is rather complex,

56 involving many important players, among which the insulin and glucagon are the most important.

The regulation of FA oxidation is always negatively balanced with FA synthesis.

The mechanisms of regulation vary from short-term factors (substrate availability, allosteric effects or enzymatic modification) to long-term factors (enzymatic activity, gene expression and metabolic turnover).

Although regulation via substrate availability only plays a small role in FA mobilization, studies in perfused hearts and livers did indicate that fatty acid oxidation increases with substrate concentrations (192-195). Hormonal regulations affect multiple downstream mediators (e.g. ACC, peroxisome proliferator-activated receptors (PPARs), AMP-activated protein kinase (AMPK), and MCD etc.) and play a big role in metabolic switch, including FA oxidation.

The metabolic switch favors nutrient absorption and storage during feeding, shutting down endogenous energy production fluxes including FA oxidation. In contrast, endogenous energy production is activated in the fasting state. This switch is controlled by two major hormones: insulin and glucagon. In the feeding state, elevated insulin stimulates the activity of (FAS), thus increasing lipogenesis. At the meantime, insulin activates ACC which elevates intracellular malonyl-CoA formation, thus inhibiting FA oxidation via CPT I. The primary mechanism underlying the action of insulin occurs via phosphorylation and the activation of phosphodiesterase, the decreased level of cAMP, and the subsequent decreases of protein kinase A activity (196).

57 In the fasting state, when energy requirement is elevated, the insulin level is low and the glucagon level is high. Glucagon is capable of activating AMPK activity which results in the activation of hormone-sensitive lipase, thereby stimulating . Simultaneously, activated AMPK phosphorylates certain serine residues in ACC and inhibits ACC activity, leading to decreased concentration of malonyl-CoA and decreased inhibition of FA oxidation.

The regulatory mechanisms through malonyl-CoA, ACC, MCD, AMPK and

PPARs will be introduced in the following sections.

3.5.2 Regulation by malonyl-CoA via CPT system

Malonyl-CoA potently inhibits FA oxidation via the inhibition of CPT I which is involved in the transfer of activated FA as acyl-CoAs from the cytosol to the mitochondria, where the enzymes of β-oxidation are located.

The inhibition of CPT I requires its association with malonyl-CoA. Distinct from all other carnitine acyltransferases, CPT I protein contains an additional N- terminal domain of about 160 residues and allows the attachment and regulation by malonyl-CoA (197). CPT I activity in oxidative tissues (heart and skeleton muscle) (M-CPT I isoform) is sensitive to cytosolic malonyl-CoA concentration, consistent with high energy demands in these tissues. Although it is a membrane protein, CPT I has both the catalytic site and malonyl-CoA binding domain exposed to the cytosol (197; 198). The proximity and sensitive interrelation between CPT I and malonyl-CoA mediates the acute regulation of

FA oxidation, when malonyl-CoA concentration changes in response to hormonal

58 fluctuations.

In addition to malonyl-CoA, CPT I activity is also regulated by and PPARα (199).

This will be discussed in Section 3.5.4.

3.5.3 Regulation by ACC

The regulation of FA oxidation by ACC is achieved through modulating intracellular malonyl-CoA. ACC catalyzes the formation of malonyl-CoA from acetyl-CoA. Activated ACC activity can potentiate the formation of malonyl-CoA, thus inhibiting FA oxidation. Reduced ACC activity on the contrary, decreases malonyl-CoA synthesis which leads to reduced inhibition of FA oxidation. For example, tissues from ACC2 null mice have lower malonyl-CoA concentration than controls (200). This explains why ACC2 null mice have higher FAO rates and less fat accumulation in the heart and soleus muscle compared to controls

(200).

The regulatory mechanisms of ACC can be either short-term or long-term. Short- term regulation involves i) allosteric activation by cytosolic citrate, ii) allosteric inhibition by LC acyl-CoAs and iii) covalent inactivation upon phosphorylation by activated AMPK. Long-term regulation of ACC involves alterations of gene expression induced by hormones or changes in nutritional states.

Allosteric regulations

Citrate is an allosteric activator of ACC and a precursor of ACC’s substrate: cytosolic acetyl-CoA. Cytosolic citrate concentration is elevated when exported out of the mitochondria via the citrate/malate shuttle. This can be a result of

59 metabolic signals indicating excess energy supplies such as high glucose concentration (201). Citrate acutely activates ACC activity and increases malonyl-CoA concentration, thus decreasing FA oxidation.

AMPK regulations

AMPK functions as the metabolic switch that controls energy balance. AMPK is a heterotrimeric kinase with subunits of catalytic α (Mr = 63 kDa), regulatory β

(Mr = 38 kDa), and γ (Mr = 36 kDa) (202). The α subunit is the catalytic site, and the γ subunit is the activator of the α subunit (203).

The energy-sensing capability of AMPK is attributed to its sensitivity to the ratio of [AMP]/[ATP]. It can be activated up to over 200 fold (204; 205) by elevated

[AMP] or reduced [ATP]. The activation of AMPK itself involves: i) the allosteric activation by AMP, or ii) phosphorylation (activation) by upstream AMPK kinase

(AMPKK), or iii) inhibition of the dephosphorylation (inactivation) of AMPK via protein phosphatase 2C. The activation of AMPK can only be achieved when the following two conditions are met: i) the γ subunit of AMPK must undergo a conformational change which exposes the active site (Thr-172) of the α subunit; and ii) the active site (Thr-172) of the α subunit must be phosphorylated and consequently activated by AMPKK. The complex formed by LKB1 (a serine/threonine kinase), mouse protein 25 (MO25), and the pseudokinase STE- related adaptor protein (STRAD) is identified as the major upstream AMPKK on its active site Thr-172 (206). AMPK can also be regulated by allosteric modulators (e.g. 3-phosphoglycerate) that can increase its affinity to AMPKK, or

60 decrease its affinity to phosphatases (207).

The ratio of [AMP]/[ATP] changes in different nutritional states. For instance, it increases in the fasting state or under metabolic stresses (e.g. electrical or exercise-induced muscle contraction, starvation, ischemia and ischemia reperfusion), when most ATP is hydrolyzed to AMP for energy production. In this case, activation of AMPK slows down the ATP depletion and facilitates the ATP production. Activated AMPK functions in a variety of metabolic fluxes, including

(208) i) stimulating glucose uptake and glycolysis via the translocation of GLUT 4 to the plasma membrane, (ii) enhancing FAO by phosphorylating ACC, thus decreasing malonyl-CoA concentration and reducing the inhibition of FA oxidation, (iii) activating CPT I activity and increasing fatty acyl-CoA availability in the mitochondria, and (iv) up-regulating biogenesis of GLUT4 and mitochondrial quantity, etc. AMPK activation also inhibits anabolic energy consuming pathways, such as lipogenesis, and protein synthesis.

AMPK regulates ACC through phosphorylation/dephosphorylation. ACC is inactivated when phosphorylated by AMPK. This generally occurs in the fasting state when [AMP]/[ATP] ratio is high and AMPK itself is activated. Inactivation of

ACC in the fasting state decreases the formation of malonyl-CoA, leading to the elevation of FA oxidation. Conversely, ACC is activated in feeding state upon de-phosphorylation induced by high insulin or relieved phosphorylation when

AMPK activity is low. In this case, activated ACC increases malonyl-CoA formation, thereby reducing FA oxidation. For example, an AMPK activator, 5-

61 aminoimidazole-4-carboxamide-1-β-D-ribofuranoside (AICAR), inhibits the activity of ACC in the rat skeletal muscle (209).

Regulations of gene expression

Alterations of ACC gene expression via the regulation of transcription rate and mRNA levels can be induced by hormones and nutritional states. For example,

ACC expression is increased by (185) insulin or high CHO low fat. Up-regulation of ACC expression is also seen in various nutritional states to meet different metabolic requirements. For example, ACC1 expression is positively regulated by high demand of FA synthesis, such as in lactating rat mammary gland where

ACC1 activity is more than 30 fold greater than in glands from pregnant rats

(185). Conversely, ACC1 expression is negatively regulated by high dietary fat, starvation and diabetes, when a high level of FA oxidation is needed

3.5.4 Regulation by PPARα

Peroxisome proliferator-activated receptors (PPARs) are a group of nuclear receptor proteins that function as transcription factors to regulate the expression of genes widely involved in cellular differentiation, development, and metabolisms of CHO, lipids, and proteins.

There are three types of PPARs identified in mammals (151; 210): PPARα,

PPARβ/δ, and PPARγ. PPARα is widely distributed in the liver, kidney, heart, muscle, adipose tissue and many others; PPARβ/δ is primarily expressed in the brain, adipose tissue and skin; PPARγ is distributed in virtually all tissues, but as

3 isoforms that are encoded from the same gene but modified differently through

62 alternative splicing.

All PPARs need to be activated through association with endogenous ligands, including free FAs and eicosanoids. Activated PPARs heterodimerize with the retinoid X receptor (RXR) and then enters nucleus, where they bind with specific regions on the promoter of target genes, called PPREs (peroxisome proliferator hormone response elements) (151). The association of PPARs with DNA promoters induces alterations (increase or decrease) of the expression of target genes.

All three PPARs possess similar structures, and working mechanisms, but have different metabolic significance (151; 210). The functions of PPARα and γ have been thoroughly studied and documented. PPARα is involved primarily in . It activates i) FA uptake via up-regulating the expression of FA transport protein (FATP) and FA translocase (FAT) (211), ii) FA oxidation via up- regulating oxidation enzymes (212) and iii) lipolysis via up-regulating LPL expression (151). PPARγ sensitizes insulin, accelerates glucose metabolism, differentiates adipocyte, and reduces inflammatory and atherosclerosis etc. The function of PPARβ/δ is not completely understood, but it is believed to be beneficial in hyperlipidemia, atherosclerosis, obesity, and cholesterol efflux (213;

214).

Enzymes involved in FA oxidation, such as acyl-CoA dehydrogenase in the mitochondria and acyl-CoA oxidase in the peroxisome, are among the regulatory targets for PPARs. For example, PPARα is highly expressed in oxidative tissues

63 (210), where it activates FA oxidation by up-regulating transcriptions of FA oxidation enzymes. This is why in PPARα null mice, the expression of genes and the rate of FA oxidation are low (210).

PPARα increases transcription of M-CPT I, during starvation or when long chain acyl-CoAs accumulate. However, the L-CPT I gene cannot be regulated by

PPARα, because it does not have PPREs in the promoter region (215).

3.5.5 Regulation by MCD

Similar to ACC, MCD contributes to the regulation of FA oxidation by modulating the intracellular concentration of malonyl-CoA. MCD can be activated via AMPK- dependent phosphorylation in isolated skeletal muscles (113), when intracellular

[AMP]/[ATP] ratio is high, e.g. during muscle contraction. AICAR, an AMP agonist, was found to induce a 2-3 fold increases in the Vmax and a 40% decrease in the Km of muscular MCD (113). However, other studies have reported that neither contraction-induced nor AICAR-induced AMPK had any effect on the activity or the phosphorylation status of MCD (189; 216). Thus, the regulatory mechanisms of MCD activity via AMPK remain unclear.

MCD can be regulated at the transcription level through the action of PPARα

(217). MCD expression is activated by prolonged fasting (218), high fat diet, and streptozotoein-induced diabetes, when plasma NEFA level increases and PPARα is subsequently activated via ligand binding (219). Activated PPARα stimulates

FA oxidation to meet the elevated demand of energy supply by activating MCD gene expression, with other FA oxidation enzymes. My previous study in rat liver

64 (218) reports elevated MCD activities with prolonged fasting time.

3.5.6 Regulation by L-carnitine and CoA availability

The supply of free L-carnitine or CoA can affect mitochondrial FA oxidation as well. Plasma carnitine can be generated from diet (beef, chicken and cow milk but not in grains, fruits, and vegetables) or through the liver synthesis of lysine and methionine coupled with cofactors of micronutrients, such as vitamin C, iron, pyridoxine and niacin (220; 221).

The supply of carnitine is required by the transport of fatty acyl-CoA into the mitochondria. The limited pool of free carnitine inhibits FA oxidation, leading to the accumulation of LC acyl-CoAs which may subsequently trap free CoAs and prohibits activation of other free FAs.

The pool of CoA is primarily located in the mitochondria (222). Due to substantial intramitochondrial CoA acelytion, only a small amount of free CoA is available to sustain mitochondrial oxidation (223). Since the mitochondrial CoA pool is limited, depletion of free CoA can potentially inhibit CoA-requiring enzymes such as CPT II and 3-ketoacyl-CoA thiolase which slows down the flux of entire FA oxidation. This inhibition can lead to the accumulation of LC acyl-CoAs involved in every step of the oxidation pathway, resulting in more depletion of free CoAs.

Bian et al (224) reported that heart perfusion of propionate of higher than 1 mM inhibits FA oxidation, possibly due to the trapping of CoAs induced by accumulated load of propionyl-CoA.

3.6 Fatty acid oxidation disorders (FOD)

65 3.6.1 Pathophysiology of FOD

Fatty acid oxidation disorders (FOD) are autosomal recessive diseases caused by alterations of genes that lead to abnormalities of plasma membrane carnitine uptake, of the carnitine palmitoyltransferase cycle (CPT I, translocase, CPT II) and of defective enzymes involved in the pathway of mitochondrial β-oxidation spiral (225). The most common disorders of β-oxidation affect very-long-chain acyl-CoA dehydrogenase (VLCAD), mitochondrial trifunctional protein (MTP), long-chain acyl-CoA dehydrogenase (LCAD), as well as medium- and short- chain acyl-CoA dehydrogenase (MCAD, SCAD). All FOD have a combined prevalence of 1: 15,000 and affect both males and females (226).

Patients of FOD commonly present recurrent hypoketotic hypoglycemia, cardiomyopathy, cardiac arrhythmias, rhabdomyolysis, muscle weakness, developmental delay, cognitive impairment, neuropathy and retinopathy (225).

These symptoms may appear at any age from birth to adulthood for unknown mechanisms.

3.6.2 Very-long-chain acyl-CoA dehydrogenase (VLCAD) deficiency and

the VLCAD knock-out (KO) mouse model

Very long-chain acyl-CoA dehydrogenase (VLCAD) catalyzes the first reaction in the mitochondrial β-oxidation spiral of LC fatty acids. In humans, VLCAD deficiency (VLCADD) is characterized by the absence of the VLCAD protein (227) and by symptom variations of either a mild or a severe phenotype (228).

Patients with mild VLCADD usually present hypoketotic hypoglycemia with no

66 cardiac symptoms, resembling MCAD deficiency (229), whereas patients with severe VLCADD usually suffer from severe cardiomyopathy which can appear at early or late stage of one’s life-span (228). Other additional clinical representations seen in individuals with VLCADD mimic symptoms of LCAD, including hypoglycemia associated with fasting, hepatomegaly, hypotonia and the absence of ketones in the urine (230-232). Fasting or exercise often triggers in patients of VLCAD episodes of metabolic derangement, which manifest as the accumulated C14–C18 acylcarnitines in the plasma or urine (C14:1 in particular)

(233), due to deficient oxidation of LC acyl-CoAs.

The VLCAD knockout mouse is viable and does not exhibit any specific clinical symptoms under non-stressed conditions (234). However, deficient animals are sensitive to starvation, stress, or cold, all of which induce hypoglycemia, skeletal myopathy and elevated plasma C16–C18 acylcarnitines, similar to symptoms in humans (233; 235). After eight hours of fasting at 4 ˚C, mean glucose concentrations in the blood decreased progressively in homozygous VLCAD- deficient mice (VLCAD -/-), leading to severe death events (233), whereas in wild-type (WT) mice, mean glucose concentrations remained normal (233).

Spiekerkoetter et. al. (236) found stress-induced decreases in the formation of

G6P, concentrations of hepatic G6P and F6P, but normal activity of glucose-6- phosphatase in VLCAD-/- mice. This suggests that stress-induced severe hypoglycemia is due at least in part to impaired gluconeogenesis. In summary, the VLCAD KO mouse is a good animal model to study human FOD.

67 3.6.3 Traditional treatment of long-chain (LC) FOD

The general aim of FOD treatments is to minimize the utilization of fat via dietary fat restriction coupled with CHO surplus, while removing the accumulated intermediates of impaired FA oxidation.

The diets of FOD patients are adjusted to lower daily fat intake from 30-35% to about 15% total Kcal, in which essential PUFAs are still included (225; 237). A surplus of CHO is provided to compensate for the lowered calories intake. A supplementation of MC FA, usually the medium even-chain trioctanoin, is given to patients with LC FOD, because it can be used via mitochondrial β-oxidation independent of CPT I, translocase, CPT II, VLCAD, MTP, or LCAD. MC FA provides an alternative fuel that can bypass the defective pathways manifested in

LC FOD. However, it is clearly not applicable for patients of MC or SC FOD.

Stress challenges, such as fasting, fever, infection or trauma, can induce massive lipolysis and are strictly prohibited via frequent feeding with slowly- digestible cornstarch at bed time (225). In acute conditions of metabolic crisis, an intravenous infusion of 10% glucose is given to maintain normal plasma glucose level (238). Massive lipolysis can be detrimental due to the accumulation of LC acy-CoAs in addition to insufficient fuel supplies. These accumulated LC acyl-CoAs can damage the structure of membrane system and inhibit enzymatic activities due to their detergent effects. This is found in hearts of reperfusion-injury. Moreover, the accumulation of acyl-CoAs may largely reduce the pool size of free CoAs, especially inside of the mitochondria. The

68 trapping of free CoAs can inhibit oxidation of other FAs, including MC FAs supplemented in dietary therapy, because an optimum CoA pool is required for

FA activation and functions of many enzymes involved in β-oxidation pathway.

One way to export these accumulated acyl-CoAs and release the trapped CoAs is to increase carnitine levels, thereby binding FAs to form acyl-carnitines that can be secreted via the urine. This is why carnitine is routinely supplemented to patients of FOD.

3.6.4 Limitations of traditional therapy call out for new available

treatments

Traditional therapy for LC FOD: The limitations

Traditional therapy for LC FOD introduced in previous section is pretty straight forward. However, it is limited solely to the improvement of an acute metabolic crisis. In other words, it fails to prevent, control or improve almost all related cardiac, muscular, neurological, retinal symptoms, and unexpected sudden death

(238-240). More specifically, limited efficacy was achieved during the treatment of cardiomyopathy with associated arrhythmias and conduction defects, rhabdomyolysis, muscle weakness, progressive peripheral neuropathy, retinopathy and sudden infant death (164). A review of 107 patients that represented nearly all the pathological symptoms caused by FOD reveals the overall severity of FOD (238). Of these patients, 47 died prior to the identification of the disorder and 50 died within the first two years of life. Despite the wide applied treatment, patients of FOD are still suffering from poorly controlled

69 symptoms and a poor quality of life.

Limitations of traditional treatments: Why?

Based on the symptoms of rhabdomyolysis and muscle weakness, it is evident that FOD patients suffer from an insufficient energy supply, despite the ample supply of glucose from traditional dietary therapy. This suggests impaired conversion from glucose to usable energy of the ATP. The generation of the

ATP can be divided into three major steps: i) the production of the mitochondrial acetyl-CoA either from glycolysis or from FA β-oxidation; ii) the production of reducing equivalents of NADH or FADH2; and ii) the process of reducing equivalents by the electron transport chain to produce the ATP. When the production of the mitochondrial acetyl-CoA is identical, FAs and glucose should contribute equally to the energy production via the CAC. In the case of FOD patients, clearly the amply glucose provided from traditional treatment cannot compensate for the energy deficiency that results from impaired FA oxidation.

Energy production from mitochondrial acetyl-CoA requires the processes of the

CAC and the electron transport chain. Since there is no evidence suggesting any defects in the electron transport chain associated with FOD, the problem most possibly occurs in the running of the CAC. As discussed in Chapter 2, proper operation of the CAC is dependent on the pool size of all catalytic intermediates; this requires proper balance between anaplerosis and cataplerosis.

Any disturbance in anaplerosis or cataplerosis may deplete the CAC intermediates, leading to improper processing of acetyl-CoA to generated NADH

70 and FADH2 normally.

Testing disturbance in anaplerosis or cataplerosis is extremely difficult, due to the low concentrations of all CAC intermediates and the challenges of measuring inside of the mitochondria. For these reasons, the efficacy of anaplerotic therapies is usually demonstrated by improved cardiac function (164; 241), muscle strength (164), or neurological status (88).

3.6.5 Triheptanoin: a novel treatment of FOD

Due to the limitations of traditional therapy, an alternative treatment is needed to improve clinical managements of FOD. Brunengraber and Roe have developed a new treatment that uses anaplerotic triheptanoin, an odd-chain TG based on 7- carbon FA heptanoate. Six years of clinical trials on patients with severe FOD along with animal studies have demonstrated improved outcomes compared with conventional therapy (164; 168; 225; 242).

The rationale underlying the new anaplerotic therapy is primarily based on: i) anaplerosis from odd-chain FA/TG can refill the pool of the CAC intermediates by producing propionyl-CoA which enters the CAC via succinyl-CoA; and ii) odd- chain FA/TG can also provide energy via the ATP production, which re- establishes the equilibrium of energy balance.

The odd-chain 7-carbon FA, heptanoate, hydrolyzed from triheptanoin, can be utilized for energy in patients of VLC and LC FOD, because it enters the mitochondria independent of the CPT system. Simultaneously, heptanoate oxidation generates the 3-carbon propionyl-CoA which is anaplerotic to the CAC.

71 Triheptanoin is also utilized in the liver for C5 ketogenesis. The C5 KBs produced

(BKP and BHP) enter the circulation and can be used in peripheral tissues (88).

The oxidation of C5 KBs can also produce propionyl-CoA that contributes to anaplerosis.

Note that an energy crisis is characterized by severe hypoglycemia and is induced by various stimuli (starvation, infection, heat shock, hypoxia ischemia, glucose deprivation, exercise etc.). It is dangerous in FOD patients, because it induces excessive consumption of the ATP, leading to activation of AMPK caused by the elevated [AMP]/[ATP] ratio. The activation of AMPK deteriorates symptoms of FOD by elevating lipolysis while shutting down the clearance of accumulated acyl-CoAs via lipogenesis. This is why providing an energy source is crucial in order to maintain a relatively low [AMP]/[ATP] ratio.

An initial clinical trial was conducted at Baylor University Medical Center by Roe et al., where the dietary treatment of trioctanoin was replaced by triheptanoin in

LC FOD patients (164). This new treatment rapidly improved the patients’ major clinical symptoms and quality of life (164).

In this trial, three patients included were two, six and nine years of age and their

VLCAD deficiency was diagnosed from blood acylcarnitine analysis. Before the study was initiated, two of these patients were on Portagen (30% of Kcals as

MCT oil, Mead Johnson Nutritionals), and the third was receiving medium chain triglycerides, all primarily by gastrostomy. These traditional treatments with even-chain triglycerides (MCT) failed to improve symptoms of cardiomyopathy

72 and muscle weakness constantly experienced by these three patients (164).

The switch to triheptanoin treatment was performed with a protocol involving a nine-day admission to Baylor Medical Center in Dallas, Texas, with follow-up visits at 2, 6, 12, 18 months. Triheptanoin was substituted isocalorically for trioctanoin. Blood samples for metabolite profile measurements were taken during and after breakfast before and after their diets were switched from even- chain MCT to triheptanoin. The even-chain MCT diet lowered the cis-5-C14:1 acylcarnitine (derived from oleate) which is the most specific acylcarnitine that accumulates in VLCAD deficiency. The levels of BHB and AcAc increased substantially 60 to 120 minutes after a meal containing MCT. The odd-chain triheptanoin diet however, considerably elevated plasma BHB and AcAc, but also

BHP and BKP, the total of which peaked between 90 and 150 min. After starting the triheptanoin diet therapy, the patients experienced immediate improvements in muscle strength, endurance and activity within 5 – 12 hours, resolution of cardiomyopathy and only minor episodes of rhabdomyolysis (164).

The success of this trial has promoted the application of triheptanoin diets in patients of various types of FOD, including CPTI, CACT, CPTII, VLCAD, LCAD,

TFP, and SCAD etc. The effects of dietary treatment of triheptanoin given at 30-

35% total kcal/day were summarized in a recent review (242) with the following major beneficial outcomes: i) a reduced mortality from 50% to 6% in 77 patients of various types of FOD, ii) resolved cardiomyopathy, iii) attenuated rhabdomyolysis, iv) normalized glucose homeostasis, v) eliminated

73 hepatomegaly, and vi) improved muscle strength and endurance. However, peripheral neuropathy and retinopathy were unchanged with this new therapy

(242).

In addition to diet, IV infusions of odd-chain MCT fatty acids (heptanoate, pentanoate, and propionate) and C5 KBs (BKP, BHP) should also be considered in the acute treatment of LC FOD. However, the possible complications of propionyl overload must be carefully considered in acute infusions, although the indices of propionyl overload in three VLCAD patients who received triheptanoin diets were significantly lower than in samples from propionic academia patients

(164).

3.6.6 Triheptanoin: Treatment for pyruvate carboxylase deficiency

Triheptanoin has been applied for clinical treatment of other disorders of anaplerosis, for example, the pyruvate carboxylase deficiency. Pyruvate carboxylase (PC) catalyzes the ATP-dependent carboxylation of pyruvate to oxaloacetate (OAA), thus contributing to anaplerosis. PC catalyzes the most important anaplerotic pathway and plays an important role in many tissues where anaplerosis is required for normal functions. For example, PC is involved in hepatic and renal gluconeogenesis, adipose glyceroneogenesis, glucose-induced insulin secretion in β-cells and de novo synthesis of neurotransmitters in astrocytes. PC gene expression is regulated by PPAR γ, suggesting that PC is involved in the metabolic switch controlling fuel partitioning toward lipogenesis.

PC deficiency is a very rare autosomal recessive disease, characterized by

74 impairment of gluconeogenesis and lactate metabolism, producing severe lactic acidosis (243). The primary result of the defect is the deficiency of OAA for the

CAC (244). This leads to profound energy deficiency due to the compromised

CAC to convert acetyl-CoA into a sufficient amount of the ATP.

Mochel et al. used triheptanoin to treat a newborn girl with PC deficiency (88).

This patient was born at 35 weeks, with normal Apgar scores, but later developed severe hepatic failure, dehydration, axial hypotonia, lactic acidosis and ketoacidosis. An enzyme assay in cultured fibroblasts confirmed PC deficiency.

The anaplerotic treatment of triheptanoin was immediately initiated and the following important observations indicating beneficial outcomes were documented during the following 6 months of treatment with triheptanoin: i) the immediate reversal (less than 48 h) of major hepatic failure with full correction of all biochemical abnormalities from lactic acidosis and ketoacidosis; ii) enhanced levels of anaplerotic C5 KBs synthesized in the liver are available for peripheral use; (iii) the utilization by the brain of C5 KBs that were transported across the

BBB, associated with improved neurological status indicated from increased levels of glutamine and free γ-aminobutyric acid (f-GABA) in the cerebrospinal fluid; (iv) no indication of toxicity or propionyl overdose.

75 Chapter 4

Research Proposal

4.1 Project 1. Fatty acid oxidation and anaplerosis from propionyl-CoA

precursors

The ineffective traditional therapy of LC FOD suggests that an ample supply of acetyl-CoA from glucose and medium even-chain FAs cannot resolve the progressive cardiac, muscular or neurological defects commonly associated with

FOD. In contrast, a novel triheptanoin diet remarkably improves many of the major clinical and biochemical conditions. Therefore, in addition to an inadequate supply of acetyl-CoA, FOD patients may suffer from a sustained leakage of the catalytic CAC intermediates. This leakage of the catalytic CAC intermediates obstacles the conversion of acetyl-CoA fuel into the usable energy, thereby resulting in poor outcomes of the traditional therapy. The advantage of anaplerotic therapy is that it supplies not only acetyl-CoA, but also propionyl-CoA to replenish the CAC intermediates pool, enabling efficient production of the ATP.

In chapter 2, I discussed the significance of anaplerosis to sustain normal functions of the liver, heart, muscle and the brain, by compensating for the physiological cataplerosis. This cataplerosis is elevated in certain cell injuries that damage cell membranes, such as the reperfusion injury associated with myocardial infarction, stroke, or organ transplantation. Under such circumstances, anaplerosis from endogenous substrates can no longer sustain

76 enough CAC intermediates for the proper operation of the CAC. Many anaplerotic substrates were reported helpful when given exogenously in the post- ischemic recovery, decreased reperfusion injury, and/or improved metabolic/contractile functions. These anaplerotic substrates reported were propionate (102; 103), lactate (98), pyruvate (104-106), succinate (100), fumarate (107), and glutamate (108) etc.

In the case of the LC FOD, the presentation of rhabdomyolysis (creatine kinase release in plasma) indicates the loss of enzymes and metabolites from cells.

This presentation of rhabdomyolysis is most likely accompanied with the leakage of the catalytic CAC intermediates as well. When this occurs, the operation of the CAC could be compromised, despite a sufficient supply of acetyl-CoA from glucose and medium-even chain FAs.

The loss of the CAC intermediates is also identified in the brain. In vivo study of

NMR in the brain has suggested a 60% loss of the pool of CAC intermediates

(especially α-KG) per turn of the CAC under physiological conditions (90), primarily in the form of glutamine which is used for protein synthesis, ammonium clearance and the glutamine-glutamate cycle (90). This loss of the CAC intermediates in the brain is well balanced by physiological anaplerosis, primarily via the pyruvate carboxylase (PC) pathway (88; 245). PC is a key anaplerotic enzyme that functions in the liver, heart, skeletal muscle, β-cells and the brain

(see Section 2.4.1). PC-deficiency, a rare genetic disease, will lead to defective liver and brain functions (88; 245). Although treatments with OAA precursors

77 (aspartate, citrate) (245) are helpful in PC-deficient patients, they fail to prevent neurological outcomes, such as the progressive brain atrophy and delayed myelination (245). An alternative anaplerotic therapy with triheptanoin has resulted in the rapid improvements of liver functions, the elevations of the Gln and f-GABA levels in the cerebrospinal fluid (CSF), and the achievement of normal myelination with no signs of CNS abnormalities or degeneration post treatments (88).

The anaplerotic therapy of triheptanoin is currently applied to patients of various types of FOD, including CPTI, CACT, CPTII, VLCAD, LCAD, TFP, and SCAD etc.

The primary beneficial outcomes documented are mostly associated with cardiac and muscular improvements (242). The effect of anaplerotic therapy in the neurological functions is poorly investigated. During the therapy of triheptanoin in the treatment of PC-deficiency (88), the production of anaplerotic C5 KBs was observed; and some improvements of CNS functions were demonstrated. Since the leakage of the CAC intermediates proposed in the FOD most likely occurs in the brain, this could contribute to the induction and progression of neurological symptoms manifesting in FOD patients, including the developmental delay and cognitive impairment etc. The successful anaplerotic therapy-induced improvement of CNS functions observed in PC-deficient patients (88) has shed light on the present proposed study, which is aimed to investigate potential benefits of anaplerotic therapy in the brain of mice deficient in VLC acyl-CoA dehydrogenase (VLCAD). This study will aid the development of anaplerotic

78 therapy that targets neurological symptoms in the clinical treatment of FOD.

On the basis of above considerations, I’ve designed the present research study to investigate the uptake and metabolism of various anaplerotic substrates, in the brains of control and VLCAD-deficient mice. The VLCAD-deficient mouse in general provides a good animal model to study human FOD. They are viable and their mild phenotype allows the surgical implantation of arterial and venous catheters under anesthesia. However, they are sensitive to starvation, stress, or cold, all of which induce hypoglycemia, skeletal myopathy and elevated plasma

C14-C18 acylcarnitines, similar to symptoms observed in humans (233; 235). The various anaplerotic substrates studied are medium-odd-chain FAs (heptanoate, pentanoate, propionate) or C5 KB (BKP). The medium-even chain octanoate is also tested for control purposes. These substrates are labeled with stable isotopes and infused in vivo to the mouse jugular vein. Their contributions to acetyl-CoA and/or anaplerosis in the brain are calculated and compared. In the meantime, the concentrations of the CAC intermediates and major neurotransmitters (glutamate, glutamine, GABA) are analyzed to evaluate the effects of anaplerotic substrates infused. A separate study is set up to investigate the profile of acyl-CoA esters from C2 to C20, which sheds light on the catabolism of endogenous FAs in the brain.

My primary working hypotheses are formulated as follows: i) LC FOD possibly leads to a depletion of the pool of the CAC intermediates, or an insufficient refill of this pool from anaplerosis; ii) A larger than normal supply of anaplerosis is

79 required to maintain the proper functioning of the CAC; iii) Odd-medium-chain

FAs (heptanoate, pentanoate, propionate), as well as C5 KBs from their initial hepatic metabolism, can be utilized in the brain to provide acetyl-CoA, as well as propionyl-CoA. The advantage is that propionyl-CoA replenishes the CAC intermediate pool while acetyl-CoA is used to generated ATP; iv) As a consequence to the deletion of very-long-chain fatty acyl-CoA dehydrogenase

(vlcad), VLCAD brain is characterized by a distinctive acyl-CoA profile and property of FA oxidation; v) VLCAD brains may possess defects in the pool of the

CAC intermediates and the concentrations of major neurotransmitters, such as the glutamate, glutamine and GABA.

My specific research aims are:

1 To test the entry of 13C labeled octanoate, heptanoate, pentanoate, BKP and

propionate into the brain using the profile of labeled acyl-CoA esters.

2 To characterize and compare in the brain the mass isotopomer distribution

(MID) of acyl-CoA esters metabolized from infused MC FAs as well as

acetyl-CoA. Anaplerosis can be evaluated from the MID ratio of M3

13 succinyl-CoA versus M3 propionyl-CoA derived from [5,6,7- C3]heptanoate,

13 13 13 [3,4,5- C3]pentanoate, [3,4,5- C3]BKP or [ C3]propionate. The contribution

to the brain acetyl-CoA can be assessed by comparing the M1 enrichment of

acetyl-CoA and the average enrichment of the acetyl moieties derived from

infused [1-13C]octanoate, [1-13C]heptanoate, or [1-13C]pentanoate. The

average enrichment of the acetyl moieties derived from [1-13C]octanoate, [1-

80 13C]heptanoate, or [1-13C]pentanoate is calculated as (M1 octanoyl-CoA)/4,

(M1 heptanoyl-CoA)/2, or (M1 pentanoyl-CoA) respectively. These

calculations are based on the maximum number of acetyl-CoA produced

from each of the MC FAs infused.

3 To assay, in the brain of mice used in Aims 1&2, the concentrations and

labeling patterns of the CAC intermediates and related neurotransmitters Gln,

Glu and GABA. These assays may reveal a defect in the pool of the CAC

intermediates and/or the concentrations of neurotransmitters which can be

caused by insufficient brain anaplerosis.

4 To compare the full profile of acyl-CoA esters from C2 to C20 in the brain of

control and VLCAD-deficient mice.

We are currently in collaboration with Dr. Xin Yu, the associate professor at the

Department of Biochemical Engineering at Case. All the analyses needed in my project can be completed in our lab by GC-MS, LC-MS or in Dr. Yu’s lab by NMR.

We are fortunate to have obtained breeders to produce VLCAD-deficient mice and their controls provided by our collaborator Dr Arnold W. Strauss. All of the studies described in Project 1 will be conducted in homozygous VLCAD-/- mice and their controls. Since VLCAD-/- mice are sensitive to stress, fasting and cold challenges, they will be raised with the help of our collaborator Colleen Croniger, expert in genetic knockout mice.

Research plan

The specific questions investigated in this project are:

81 1) Do medium-chain fatty acids (MC-FAs) enter brains as such or as ketone

bodies (KBs) made in the liver?

2) Are odd-chain MC-FAs and C5 KBs anaplerotic in the brain?

3) What is the contribution of MC-FA to acetyl-CoA in the brain?

4) Do VLCAD brains present any defect in LC FA breakdown and the pool of

acyl-CoA esters?

5) Do VLCAD brains present any defect in the pool of CAC intermediates?

The primary research strategies used are:

2-hour fasted mice were anesthetized and fitted with carotid and jugular catheters. Rectal temperature was kept at 37-37.5 ºС with a water mattress.

After 10 min of post-surgical equilibration, increasing dosage of sodium [1-

13 13 13 13 C]octanoate, [5,6,7- C3]heptanoate, [1- C]heptanoate, [3,4,5- C3]pentanoate,

13 13 13 [1- C]pentanoate, β-keto-[3,4,5- C3]pentanoate, or [ C3]propionate was infused in the jugular vein of each mouse for 30 min, at various rates of 0 (control), 10%,

20%, 30%, 40% or 50% of the caloric requirements (CR). Arterial blood was collected and the whole brain was removed and freeze-clamped with liquid N2 at the end of each infusion.

For the assays of acyl-CoA esters, major CAC intermediates and brain neurotransmitters (glutamate, glutamine, GABA), the frozen brain tissues were extracted, with acyl-CoA esters separated from CAC intermediates and neurotransmitters. Acyl-CoA esters were analyzed by LC-MS. Major CAC intermediates and neurotransmitters were analyzed by GC-MS.

82 In support of my hypothesis, we developed isotopic techniques to quantify substrate uptake, anaplerotic flux and the contribution to acetyl-CoA originated from infused substrates labeled in various positions. Briefly, the elevation of MC fatty acyl-CoAs and its direct downstream metabolites must be synthesized inside of the brain cells and serve as the direct proof for the entry of infused MC-

FAs into the brain. The anaplerotic property of an odd MC-FA was quantified by the kinetics of M3 labeling of propionyl-CoA and succinyl-CoA (246). This was achieved by infusions of the odd MC-FAs labeled in their propionyl moieties, for

13 13 example [5,6,7- C3]heptanoate, [3,4,5- C3]pentanoate, β-keto-[3,4,5-

13 13 C3]pentanoate or [ C3]propionate. The relative anaplerosis was calculated from the mass isotopomer distribution (MID) ratio of M3 succinyl-CoA versus M3 propionyl-CoA, indicating the contribution to catalytic intermediates of the CAC which carry the acetyl groups as they are oxidized (246).

To assess the contribution of MC-FAs to brain acetyl-CoA production, the concentration and M1 labeling of acetyl-CoA were analyzed with infusions of

MC-FAs labeled in their acetyl moieties. The relative contribution to brain acetyl-CoA was assessed by comparing M1 labeling of brain acetyl-CoA with M1 labelings of infused [1-13C]MC-FAs and [1-13C]KBs which is derived from initial metabolisms in the liver of infused [1-13C]MC-FAs.

For the assay of acyl-CoA profile (C2 - C20), brain tissues were harvested from

VLCAD and control mice without fasting, vascular catheterization nor infusion of any substrate. The acyl-CoA esters extracted from brain tissues were analyzed

83 by LC-MS.

The concentrations and labeling patterns of infused MC-FAs and their metabolites (shorter chain MC-FAs, C4 KBs and/or C5 KBs) from peripheral tissues (e.g. the liver and kidney (166; 167)) were assayed by GC-MS. The labeling of circulating glucose was also analyzed, which was utilized in the brain as a secondary tracer.

4.2 Project 2. Assay of the activity of malonyl-CoA decarboxylase (MCD)

MCD catalyzes the decarboxylation of malonyl-CoA to acetyl-CoA (see section

3.4 of chapter 3). Together with acetyl-CoA carboxylase (ACC), MCD governs a substrate cycle that determines the cytoplasmic level of malonyl-CoA. Malonyl-

CoA is an intermediate of FA synthesis; it also contributes to the regulation of mitochondrial FA oxidation by inhibiting CPTI (180), the enzyme that controls the transfer of LC fatty acyl-CoAs from the cytosol to the mitochondrial matrix.

Very little is known about the regulatory mechanisms of MCD. MCD is activated at the transcription level through the action of PPARα (217). MCD is also activated via AMPK-dependent phosphorylation (113), but data discovered were quite controversial (189; 216).

Further investigations in the action of MCD require a simple but sensitive assay to measure the MCD enzymatic activity. There are several assays currently available. The spectrophotometric assay (247) involves (i) forming acetyl–CoA from the decarboxylation of malonyl–CoA by MCD, (ii) generating oxaloacetate from malate + NAD+ + malate dehydrogenase, and (iii) forming citrate from the

84 condensation of oxaloacetate with acetyl-CoA via citrate synthase. The assay is monitored by the production of NADH. This technique has low sensitivity and is not applicable to crude preparations. The first described radiochemical assay

(189; 248) is based on the same reactions except that [14C]oxaloacetate is generated by transamination of [14C]aspartate with glutamate. The assay is monitored by the accumulation of [14C]citrate. A second, more direct radiochemical assay (249) measures the formation of 14CO2 from [3-14C]malonyl-

CoA. A third radiochemical assay, developed by Kerner and Hoppel (187), involves (i) the production of [2-14C]acetyl–CoA from [2-14C]malonyl–CoA, (ii) the conversion of [2-14C]acetyl–CoA to [2-14C]acetylcarnitine in the presence of excess L-carnitine and carnitine acetyltransferase, and (iii) the isolation of [2-

14C]acetylcarnitine on ion exchange resin and the counting of its radioactivity.

The above three radiochemical assays are sensitive and have been used extensively.

To avoid using radioactive substrates, I developed a stable isotope-based assay of the MCD activity that can be analyzed by more commonly available GC-MS.

13 My assay involves (i) incubating [ C3]malonyl-CoA with the tissue extract

13 containing MCD to form [1,2- C2]acetyl-CoA, (ii) spiking the reaction mixture with

2 13 [ H3,1- C]acetyl-CoA internal standard, (iii) reacting the acetyl-CoA with thiophenol to form the corresponding labeled acetylthiophenols (250), and (iv) assaying the concentration of acetylthiophenol by gas chromatography-mass spectrometry (GC-MS).

85 This new assay was applied to measure the liver MCD activities in three groups of rats with different nutritional status (fed, one-day fasted or two-day fasted).

The results confirmed that fasting activates MCD activity in rat liver, consistent with the activation of FA catabolism in the fasting state (210), when the liver glycogen store is depleted.

4.3 Publications

4.3.1 Wang X, Zhang GF, Puchowicz MA, Kasumov T, Allen Jr F, Rubin AJ, Lu

M, Yu X, Roe CR, and Brunengraber H. Fatty acid oxidation and

anaplerosis from propionyl-CoA precursors in the brain of mice deficient in

very-long-chain acyl-CoA dehydrogenase and control mice (to be

submitted to the Journal of Biological Chemistry, 2008).

4.3.2 Wang X, Stanley WC, Darrow CJ, Brunengraber H, and Kasumov T.

Assay of the activity of malonyl-coenzyme A decarboxylase by gas

chromatography-mass spectrometry. Anal Biochem 2007 Apr 15; 363(2):

169-74.

86 4.3.1

Fatty acid oxidation and anaplerosis from propionyl-CoA precursors in the

brain of mice deficient in very-long-chain acyl-CoA dehydrogenase and

control mice

Xiao Wang1, Guofang Zhang1, Michelle A. Puchowicz1 ,Takhar Kasumov1,

Frederick Allen Jr1, Adam J. Rubin1, Ming Lu2, Xin Yu2 , Charles R. Roe3,

and Henri Brunengraber1

From the Departments of Nutrition1 and Biomedical Engineering2, Case Western

Reserve University, Cleveland OH 44106, and the Institute for Metabolic

Disease3, Baylor University Medical Center, Dallas TX 75226

87 Running title: Anaplerosis from propionyl-CoA in mouse brain

Key words: fatty acid oxidation disorders, mass isotopomer analysis, C5-ketone bodies, odd-chain fatty acids, citric acid cycle.

*: to whom correspondence should be addressed at Department of Nutrition,

Case Western Reserve University, 10,900 Euclid Avenue, Cleveland OH 44106.

E-mail: [email protected]

88 Abbreviations

BHB, β-hydroxybutyrate; BHP, β-hydroxypentanoate (3-hydroxyvalerate); BKP,

β-ketopentanoate (3-ketovalerate); CAC, citric acid cycle; FOD, fatty acid oxidation disorders; LC-FOD, long-chain fatty acid oxidation disorders; VLCAD mouse, mouse deficient in very-long-chain acyl-CoA dehydrogenase;

Footnotes

1. Mass isotopomers are designated as Mn, where n is the number of heavy atoms in the molecules. The molar percent enrichment of a mass isotopomer is designated as mn.

Acknowledgments

This work was supported by the NIH (Road Map grant 5R33DK070291 and grant

5R01DK069752) and the Cleveland Mt. Sinai Health Care Foundation. We thank

Dr Arnold W. Strauss for providing us with breeders of VLCAD-deficient and control mice. We thank the Case Mouse Metabolic Phenotyping Center

(mmpc.org) for running the in vivo experiments.

89 Abstract

Anaplerotic therapy with glycerol triheptanoate (triheptanoin) is currently investigated for the treatment of long-chain fatty acid oxidation disorders. The goal of this study was to test the effect of anaplerotic therapy on brain metabolism in mice deficient in very-long-chain acyl-CoA dehydrogenase

(VLCAD) and in control mice. Mice were infused intravenously with increasing amounts of precursors of propionyl-CoA: heptanoate, pentanoate, β- ketopentanoate or propionate. The compounds were either triply 13C-labeled in their propionyl moiety, or singly 13C-labeled in one acetyl moiety. Other mice were infused with non anaplerotic [1-13C]octanoate. The concentrations and mass isotopomer distributions of short-chain acyl-CoAs and citric acid cycle intermediates were assayed. Our data show that the C8, C7, and C5 fatty acids are taken up by the brain and activated to CoA esters. β-Oxidation of these fatty acids contributes a substantial fraction of the acetyl-CoA of the whole brain.

Ketone bodies derived from partial oxidation of the fatty acids in liver also contribute to brain acetyl-CoA. The concentration and labeling pattern of medium-chain acyl-CoAs show that long-chain fatty acid oxidation occurs in the brain of normal and VLCAD mice. The odd-chain compounds are strongly anaplerotic at low blood concentrations, and more so in VLCAD brains than in control brains. A fraction of anaplerosis from the odd-chain fatty acids involves their conversion in liver to C5-ketone bodies which are used by the brain.

90 Introduction

Proper operation of the citric acid cycle (CAC) requires the maintenance of the concentrations of eight catalytic intermediates which carry acetyl units as they are oxidized to CO2. The physiological “leakage” of intermediates through the mitochondrial and cell membrane, i.e., cataplerosis, must be balanced by an equivalent re-filling of intermediates, i.e., anaplerosis. The main anaplerotic processes are (i) pyruvate carboxylation to oxaloacetate, (ii) the conversion of glutamate or glutamine to α-ketoglutarate, and (iii) the conversion of propionyl-

CoA to methylmalonyl-CoA and succinyl-CoA (1). Propionyl-CoA is formed from the metabolism of some amino acids, odd-chain fatty acids, and C5-ketone bodies (β-hydroxypentanoate (BHP) and β-ketopentanoate (BKP)(2)). C5-Ketone bodies are formed from the β-oxidation of odd-chain fatty acids in the liver (3), by the same reactions of the HMG-CoA cycle which form C4-ketone bodies, β- hydroxybutyrate (BHB) and acetoacetate. C5-ketone bodies are rapidly used by peripheral tissues (2). In addition to maintaining the pool of CAC intermediates, anaplerotic and cataplerotic reactions play a central role in gluconeogenesis from amino acids, lactate and pyruvate, as well as in the .

Excessive cataplerosis which occurs in cardiac reperfusion injury can lead to decreased energy production in spite of ample supply of acetyl-CoA to the CAC

(4). Under these conditions, the physiological anaplerotic reactions cannot fully compensate for the loss of CAC intermediates. Pharmacological concentrations

91 of pyruvate (5), lactate (6), propionate (7), or propionylcarnitine (8) improve the mechanical performance of ischemia-reperfused hearts and of hearts pre- perfused without anaplerotic substrates.

Anaplerotic therapy with dietary triheptanoin (glycerol triheptanoate, a propionyl-

CoA precursor) has recently been proposed for the treatment of patients with long-chain fatty acid oxidation disorders (LC-FOD) (9,10). During acute episodes of LC-FOD decompensation, rhabdomyolysis leads to the release of large molecules (creatine kinase), and presumably of small molecules (CAC intermediates) from muscle cells. Between acute episodes when the creatine kinase activity in plasma is normal, there is no evidence of excessive cataplerosis in these patients. However, the condition of the patients (especially muscle and heart strength) is much more improved by the dietary treatment with anaplerotic triheptanoin compared with the classical treatment with non- anaplerotic trioctanoin (9). Thus, in LC-FOD patients, whether or not the pools of

CAC intermediates are chronically depleted, boosting anaplerosis improves the metabolic perturbations of their condition.

In the brain, anaplerotic reactions contribute to the regulation of the metabolism of the neurotransmitters glutamate, glutamine and GABA (11-14). The export of

+ NH4 from the brain as glutamine is a cataplerotic process which must be compensated by constant anaplerosis.

92 The goals of the present study were (i) to quantitate brain anaplerosis from propionyl-CoA precursors in normal mice and in mice with a LC-FOD, and (ii) to gain insight on the metabolism of medium-chain fatty acids in the brain. This was achieved using anaplerotic substrates labeled in their acetyl or propionyl moieties,

13 13 for example [1- C]heptanoate or [5,6,7- C3]heptanoate. We used mice deficient in very long-chain acyl-CoA dehydrogenase (VLCAD) (15) because their mild phenotype allows the surgical implantation of arterial and venous catheters under anesthesia. Our data show faster rates of anaplerosis from four propionyl-CoA precursors in VLCAD-deficient compared to normal mice. Our data also demonstrate the entry of medium-chain fatty acids into the brain and their contribution to acetyl-CoA.

93 Experimental Procedures

13 Materials. General chemicals were from Sigma-Aldrich. [U- C3]propionic,

2 13 13 2 [ H5]propionic, [1- C]pentanoic, [3,4,5- C3]pentanoic, [ H9]pentanoic, [1-

13 2 2 13 C]octanoic, [ H15]octanoic, [ H13]heptanoic, [5,6,7- C3]heptanoic, γ-amino-

2 13 [ H6]butyric acids, as well as β-keto-[3,4,5- C3]pentanoate ethyl ester,

2 2 13 2 2 13 [ H5]glycerol, [ H5]glutamine, [ C6]glucose, H2O, NaO H, and CO2 gas were obtained from Isotec. [1-13C]Heptanoic acid was prepared by reacting

13 hexylmagnesium bromide with CO2. The purity of the vacuum-distilled product was assessed by GC-MS and NMR. Internal standards of R,S-ain β-hydroxy-

2 2 [2,2,3,4,4,4- H6]butyrate, R,S-β-hydroxy-[2,2,3,4,4- H5]pentanoate and R,S-3-

2 hydroxy-[2,2,3,4,4- H5]glutarate were prepared by (i) incubating ethyl

2 acetoacetate, ethyl β-ketopentanoate or dimethyl-3-ketoglutarate in H2O +

2 2 NaO H overnight, (ii) reacting with NaB H4, (iii) acidification to destroy excess

2 NaB H4, (iv) neutralization and lyophilization (2,16). Solutions of β-keto-[3,4,5-

13 C3]pentanoate were prepared by overnight hydrolysis of the ethyl esters with

1.1 equivalents of NaOH, followed by neutralization and lyophilization.

2 2 [ H5]Propionyl-CoA, pentanoyl-CoA, [ H13]heptanoyl-CoA, and [5,6,7-

13 C3]heptanoyl-CoA were synthesized from the acids by the procedure of Weaire and Kelwick (17), and purified by preparative HPLC (18). Reagents for trimethylsilyl, tert-butyldimethylsilyl, and pentafluorobenzyl derivatization were purchased from Pierce.

94 Mice. Breeders of homozygous VLCAD -/- (called hereunder VLCAD mice) and of control mice were kindly provided by Dr Arnold W. Strauss (15). The mice were fed regular Teklad mouse chow. Male mice were used for infusion experiments at 11-12 weeks of age.

Protocols. After a short 2 hr fasting period to avoid hypoglycemia, mice were anesthetized with 2% isoflurane, placed on a heated water mattress, and fitted with jugular and carotid catheters. Body temperature, monitored with a rectal probe, was kept at 37˚C. After 10 min equilibration under saline infusion, a 150 mM solution, of either [1-13C]octanoate, [1-13C]heptanoate, [5,6,7-

13 13 13 13 C3]heptanoate, [1- C]pentanoate, [3,4,5- C3]pentanoate, [3,4,5- C3]BKP, or

13 [ C3]propionate was infused at rates ranging from 0 to 50% of caloric requirement (1 rate/mouse, 6-9 mice/group) for 30 min. Rates of infusions were calculated based on the following Kcal/g: octanoic (8.0), heptanoic (7.7), pentanoic (6.7), β-ketopentanoic (5.1), and propionic (4.8). The rate of normal saline infusion was adjusted for each mouse so that the total volume of isotonic solutions infused was 26 μl/min. At 24 and 29 min, arterial blood samples (100 μl) were taken. At 30 min, while the infusion was still running, the skull was opened with scissors; the brain was scooped out with a round spatula and immersed in liquid N2.

Analytical Procedures. Brains were assayed for the concentration and labeling pattern of acyl-CoA esters, CAC intermediates, glutamate, glutamine and GABA.

The whole frozen brains (≈ 200 mg), powdered under liquid N2 and spiked with

95 2 2 internal standards (5 nmol [ H5]propionyl-CoA, 5 nmol [ H13]heptanoyl-CoA, 1

2 2 μmol of [ H6]GABA, 30 nmol [ H5]glutamine, 60 nmol R,S-3-hydroxy-

2 [ H5]glutarate) were extracted at -25˚C for 1 min with 0.5 ml of acetonitrile:methanol (2:1) using a Polytron homogenizer. After addition of 3 ml of ice-cold hexane, 30 sec re-extraction with the Polytron, addition of 3 ml of ice- cold ammonium acetate 50 mM, pH 5.8, and 1 min extraction, the slurry was centrifuged at 13,000 g for 30 min at 4˚C. The lower aqueous layer, containing the CoA esters, was removed. The remaining material was re-extracted with 3 ml ammonium acetate buffer. The two aqueous phases were combined, adjusted to pH 1.5 with HCl, and run on an 3 ml ion exchange cartridge packed with 300 mg of 2-2(pyridyl)ethyl silica gel (Sigma) (19). The cartridge had been pre-activated with 3 ml methanol, then with 3 ml of HCl pH 1.5. The effluent of the cartridge, after loading the extract, contains CAC intermediates, glutamate, glutamine and

GABA. The acyl-CoAs trapped on the silical gel cartridge were released with (i) 3 ml of a 3:1 mixture of ammonium formate 100 mM pH 6.3, and methanol (to release the short- and medium-chain acyl-CoAs), then (ii) 3 ml of a 1:3 mixture of ammonium formate 100 mM pH 6.3, and methanol (to release the medium- and long-chain acyl-CoAs (19). The combined effluent was lyophilized and stored at -80˚C until LC-MS analysis.

After dissolving the acyl-CoAs in 20 μl of buffer A (5% acetonitrile in ammonium formate 50 mM, pH 5.5), 10 μl were injected on a Thermo Electron Hypersil

GOLD column (150 x 2.1 mm) protected by a guard column (Hypersil Gold 5 μm,

96 10 x 2.1 mm, Drop in Guard), in an Agilent 1100 liquid chromatograph. The column was developed at 0.25 ml/min (i) for 2 min with buffer A, (ii) from 2 to 47 min with a 0 to 90% gradient of buffer B (95% acetonitrile in ammonium formate

5 mM, pH 6.3) in buffer A, (iii) from 47 to 50 min with 10% buffer A/90% buffer B,

(iv) from 50 to 53 min with a 90% to 0% gradient buffer B in buffer A, and (v) 10 min stabilization with buffer A before the next injection. The order of acyl-CoA elution (min) was methylmalonyl-CoA (12.5), succinyl-CoA (14.5), acetyl-CoA

(16.2), propionyl-CoA (17.9), pentanoyl-CoA (22.0), heptanoyl-CoA (26.4), octanoyl-CoA (28.7).

The liquid chromatograph was coupled to an API 4000 QTrap mass spectrometer

(Applied Biosystems, Foster City, CA) operated under positive ionization mode with the following source settings: turbo-ion-spray source at 600 °C under N2 nebulization at 65 psi, N2 heater gas at 55 psi, curtain gas at 30 psi, collision- activated dissociation gas pressure was held at high, turbo ion-spray voltage at

4,500 V, declustering potential at 70 V, entrance potential at 10 V, collision cell exit potential at50 V. The Analyst software (version 1.4.1; Applied Biology) was used for data registration.

For the assay of CAC intermediates, glutamate, glutamine and GABA, the initial effluent of the silica gel cartridge was treated with 10 μl of 10 M methoxylamine-

HCl to protect ketoacids. After raising the pH to 7-8 with NaOH, and 40 min incubation at 45˚C, the solution was divided into two tubes and lyophilized. The tubes were treated to prepare the trimethylsilyl (TMS) and tert-butyldimethylsilyl

97 (TBDMS) derivatives of (i) the CAC intermediate derivatives, and (ii) glutamate, glutamine, GABA, respectively. The derivatives were assayed by GC-MS using previously published conditions (20).

Blood samples were assayed for the concentration and labeling pattern of the infused labeled substrates, as well as C4-ketone bodies (β-hydroxybutyrate + acetoacetate) and C5-ketone bodies (β-hydroxypentanoate + β-ketopentanoate).

Blood samples (100 μl) were immediately pipetted into glass tubes containing 1

2 ml aqueous solution of internal standards: R,S-β-hydroxy-[ H6]butyrate (20 nmol),

2 13 2 R,S-β-hydroxy-[ H5]pentanoate (20nmol), [ C6]glucose (600 nmol) , [ H5]glycerol

(30 nmol) and depending on the substrate infused to the mice, 60 nmol of

2 2 2 2 [ H15]octanoate, [ H13]heptanoate, [ H9]pentanoate, and/or [ H5]propionate. After quick mixing, the solution of hemolyzed blood was treated with 40 μl of 1 M

2 2 NaB H4 in 0.1 M NaOH. The treatment with NaB H4 converts unstable acetoacetate and BKP to the stable BHB and BHP with one amu greater than the

BHB and BHP present in blood (2). It also converts glucose to monodeuterated sorbitol. The samples were then frozen until analysis. Upon thawing, the samples were divided into 2 aliquots: 200 μl for glucose and glycerol assays, and the rest for fatty acids and ketone bodies. For the latter assays, the samples were first

2 acidified with HCl to destroy excess NaB H4. Then, after adjusting pH to 9.0, the samples were subjected to extractive alkylation by adding 1 ml of 0.1 M tetrabutyammonium sulfate pH 9.0, followed by 1 ml of 0.13 M pentafluorobenzyl bromide in methylene chloride (21). After 5 min vortexing, 1 hr sonicating, and

98 overnight shaking at room temperature, the samples were extracted 3 times with

4 ml hexane. After centrifugation, the upper hexane phase (containig the pentafluorobenzyl derivatives of fatty acid and ketone bodies) was separated and evaporated. The residue was treated overnight with 330 μl of a mixture acetic anhydride: pyridine:chloroform (15:30:285). This procedure converts the pentafluorobenzyl derivatives of BHB and BHP to acetyl- pentafluorobenzyl derivatives. After extraction with diethyl ether, drying over Na2SO4, and evaporating the solvent, the residue was dissolved in 50 μl chloroform. The GC-

MS analyses of the fatty acids, ketone bodies, glucose (as sorbitol) and glycerol are described in (3). In some series, the mass isotopomer1 distribution of plasma glucose was assayed on the permethyl derivative (22) to minimize the natural enrichment of the derivative. To assay the mass isotopomer distribution of the

2 C4- and C5-ketone bodies, plasma samples (not treated with NaB H4) were deproteinized with acetonitrile, and the residue of extract evaporation treated to prepare the tert-butyldimethylsilyl derivatives of BHB and BHP. The mass isotopomer distributions of the latter were assayed as described previously (2).

13C-NMR analyses. The initial effluent of the silica gel cartridge was lyophilized

2 and the residue dissolved in 0.5 mL of H2O. NMR spectra were acquired with a

5 mm 13C/1H probe inside a Bruker 900MHz Spectrometer at room temperature from 45000 scans (51° pulse, 0.18 s acquisition and 1.1 s delay) with broadband decoupling. The FID was acquired with an 8192 data set and zero filled to 16000 to improve digital resolution. Before Fourier transformation, signals were

99 processed with 2Hz exponential filtering to improve resolution of J coupled 13C-

13C multiplets.

Calculations. Correction of measured mass isotopomer distributions for natural enrichment was performed using the CORMAT software (23). Relative rates of

13 anaplerosis from [ C3]propionyl-CoA precursors were calculated as the ratio 100 x (m3 succinyl-CoA)/(m3 propionyl-CoA) (24). These percentages refer to the contribution of the anaplerotic substrates to the catalytic intermediates of the

CAC which carry acetyl units as they are oxidized. Note that, when label enters the CAC only via propionyl-CoA, M3 succinyl-CoA is only formed from the sequence: M3 propionyl-CoA ➔ M3 methylmalonyl-CoA ➔ M3 succinyl-CoA.

Recycling of label in the CAC cannot form M3 succinyl-CoA. However, if part of the M3 propionyl-CoA precursor is converted to glucose in the liver, some of the new glucose molecules released in plasma may be labeled on carbons 1,2,5, or

6. This could result in the labeling of acetyl-CoA in peripheral tissues (see

Discussion).

The calculation of relative anaplerosis described above requires total chromatographic separation of the isomeric succinyl-CoA and methylmalonyl-

CoA. The enrichment of methylmalonyl-CoA is intermediate between those of propionyl-CoA and succinyl-CoA because of the partial reversibility of the methylmalonyl-CoA mutase reaction (25). Thus, incomplete separation of

100 methylmalonyl-CoA and succinyl-CoA would lead to overestimation of relative anaplerosis.

Statistics. All data are reported as mean ± SD calculated by Excel on a personal computer. Groups were compared using a two-tailed unpaired (Table 4.1& Fig

4.9) or paired (Figs 4.5&4.8) Student’s t test. Differences between two groups were considered significant at p < 0.05.

101 Results

Blood concentrations of infused substrates and their metabolites; secondary tracers.

The blood concentrations of the infused substrates were identical at 24 and 29 min of infusion, as shown by the blood concentration ratios (24 min/29 min) which was 0.99 on average (n = 98) with a global coefficient of variation of 5%

(ranging from 2.2% to 7.6% between groups). Thus, the brains were sampled while the concentrations of each infused substrate were in steady-state.

The labeling pattern of brain metabolites can result not only from the metabolism of the infused labeled substrate, but also from the metabolism in the brain of labeled compounds derived from the metabolism of the infused substrate in other tissues, mostly in the liver. Such secondary tracers include (i) glucose labeled in liver via net gluconeogenesis (from odd-chain fatty acids labeled on their terminal

14 3 carbons) , isotopic exchanges ( CO2 incorporation (26)) , or the uptake of

13 [ C]glutamine released from muscle (27,28) , (ii) C4-ketone bodies formed from the acetyl groups of even-chain and odd-chain fatty acids, (iii) C5-ketone bodies formed from odd-chain fatty acids (and labeled in their propionyl (3) or acetyl moiety), and (iv) short-chain fatty acids derived from the hydrolysis of β-oxidation intermediates. This why we assayed in arterial blood (i) the concentrations of the infused substrates, and (ii) the concentrations and mass isotopomer distributions of their metabolites most likely to contribute to the labeling of brain metabolites.

102 In mice infused with [1-13C]octanoate, the arterial blood concentration of the substrate and of total C4-ketone bodies increased with the dose of [1-

13C]octanoate infused (Figs 4.1A, 4.1B). Fig 4.1 shows a clear precursor-to- product relationship between octanoate and C4-ketone bodies, as originally shown by McGarry and Foster (29). Fig 4.2A shows the mass isotopomer distribution of BHB at the end of the [1-13C]octanoate infusions. The M1 and M2 enrichments of BHB plateaued at about 27% and 3%, respectively.

13 In the blood of mice infused with [5,6,7- C3]heptanoate, we observed (Fig 4.3) the accumulation of the infused substrate and small amounts of M3 C5-ketone bodies, M3 pentanoate, and M3 propionate. The concentration of C4-ketone bodies increased in the blood of control, but not of VLCAD mice. In mice infused

13 with [1- C]heptanoate, both the C4- and C5-ketone bodies became labeled via incorporation of [1-13C]acetyl-CoA (Fig 4.2 B). The M1 enrichment of BHP, which has only one acetyl, is one-half that of BHB which has two acetyl groups in its molecule.

13 In mice infused with [3,4,5- C3]pentanoate, we also observed the accumulation of M3 C5-ketone bodies and a very small amount of M3 propionate

(Supplementary Fig 4.10). Also, we observed the presence of M1 to M5 mass isotopomers of glucose (Fig 4.4A). Similar profiles were observed in mice infused

13 13 with [5,6,7- C3]heptanoate or [ C3]propionate (not shown). In mice infused with

13 non-gluconeogenic [3,4,5- C3]BKP, we also observed substantial glucose

103 labeling (Fig 4.4B). The labeling profiles of glucose in the VLCAD mice were similar to those assayed in the control mice (not shown).

13 Lastly, in mice infused with [3,4,5- C3]BKP, the accumulation of C5-ketone bodies (Supplementary Fig 4.11) did not affect the concentration of C4-ketone bodies which remained unlabeled as expected (not shown).

Acyl-CoA profiles in mouse brain.

Table 4.1 shows the concentrations of acyl-CoA esters in the brains of ad libitum fed control and VLCAD mice whose brains were sampled early afternoon without vascular catheterization or infusion of any substrate. In VLCAD mice brains compared to control, the profile of acyl-CoA concentrations is compatible with an impairment of long-chain fatty acid oxidation: five-fold increase in stearoyl-CoA concentration, and decrease in most β-oxidation intermediates shorter than 14 carbons. In addition, the concentrations of propionyl-CoA and methylmalonyl-

CoA were lower in the brains of VLCAD mice compared to control mice. This suggests a stimulation of anaplerosis from endogenous propionyl-CoA in VLCAD brains compared to control brains.

In the brains of 2 hr-fasted mice infused with increasing amounts of [1-

13C]octanoate, [1-13C]heptanoate, or [1-13C]pentanoate, we observed the accumulation of the corresponding M1 acyl-CoAs (Figs 4.5A, 4.5B, 4.5C). In

13 mice infused with increasing amounts of [5,6,7- C3]heptanoate, [3,4,5-

104 13 13 C3]pentanoate, or[ C3]propionate, we observed the accumulation of the corresponding M3 acyl-CoA (not shown).

Infusion of octanoate, heptanoate, pentanoate or BKP increased the concentration of acetyl-CoA in the brain of control and VLCAD mice (Fig 4.6).

Similar increases in acetyl-CoA were observed when triheptanoin was infused

(not shown). Infusion of propionate did not increase acetyl-CoA concentration, as expected (not shown). Also, during the infusions of [1-13C]octanoate, [1-

13C]heptanoate or [1-13C]pentanoate, brain acetyl-CoA became labeled (Fig 4.5).

Anaplerosis from propionyl-CoA precursors

Relative anaplerosis, ie the contribution of propionyl-CoA precursors to citric acid cycle intermediates which carry acetyl groups as they are oxidized, was calculated from the m3 enrichment ratio (succinyl-CoA)/(propionyl-CoA). As an example of raw data, Fig 4.7 shows the dependence of the m3 enrichments of pentanoyl-CoA, propionyl-CoA, methylmalonyl-CoA and succinyl-CoA on the rate

13 of [3,4,5- C3]pentanoate infusion. The four m3 enrichments show clear precursor-to-product relationships. The mass isotopomer distribution of succinyl-

CoA included M1 and M2 isotopomers resulting from the recycling of M3 succinyl-CoA in the CAC, but no M4 isotopomer (not shown). The two panels of

Fig 4.7 show that, in brains of VLCAD mice, maximal M3 labeling of propionyl-

13 CoA and succinyl-CoA was achieved at lower rates of [3,4,5- C3]pentanoate infusion than in brains of control mice.

105 The four panels of Fig 4.8 show the profiles of brain relative anaplerosis

13 corresponding to increasing infusion rates of [5,6,7- C3]heptanoate, [3,4,5-

13 13 13 C3]pentanoate, [3,4,5- C3]BKP, or [ C3]propionate. Since relative anaplerosis is expressed as contribution of propionyl-CoA precursors to the catalytic intermediates of the CAC, it is clear that these substrates are strongly anaplerotic, even at low concentrations. Also, in all cases, relative anaplerosis was higher in brains of VLCAD mice compared to controls.

Relative anaplerosis from M3 propionyl-CoA, calculated as above, assumes that label enters the CAC only by the conversion of M3 propionyl-CoA to M3 succinyl-

CoA. Then, recycling of label in the CAC cannot yield M3 succinyl-CoA, only M2 and M1 succinyl-CoA. This avoids the artifact of overestimation of anaplerosis from precursors of less-than-triply labeled propionyl-CoA (24). However, if label entered the CAC via M1 or M2 acetyl-CoA, in addition to the propionyl-CoA pathway, some M3 or M4 succinyl-CoA could be formed from α-ketoglutarate.

This could happen if glucose, labeled in the liver from M3 propionyl-CoA precursors (Fig 4.4), were metabolized in the brain at a rate that would substantially label acetyl-CoA. This would lead to some overestimation of relative anaplerosis. To test for this possibility, we analyzed by 13C-NMR brain extracts of

13 mice infused with [3,4,5- C3]pentanoate. We tested whether glutamate was labeled only on C1, C2, and C3 as should be the case if label entered the CAC only via the propionyl-CoA to succinyl-CoA reactions. The first three glutamate carbons were strongly labeled as expected (not shown). Barely detectable traces

106 of 13C were identified on C4 and C5 of glutamate after long data acquisition (12 hr). These traces of 13C probably result from the metabolism in the brain of

13 glucose labeled from [3,4,5- C3]pentanoate in the liver. However, since our GC-

MS assay of glutamate and our LC-MS-MS assay of succinyl-CoA did not detect any M4 isotopomers, we concluded that our calculations of relative anaplerosis were not overestimated by the formation in the brain of traces of [13C]acetyl-CoA derived from plasma glucose.

Concentrations of CAC intermediates and related metabolites.

In each series of control and VLCAD mice, one mouse was infused only with saline (zero substrate infusion in Figs 4.1-4.8). We assayed the basal brain concentrations of CAC intermediates in 6 control and 6 VLCAD mice infused with saline. Fig 4.9 shows the concentrations of CAC intermediates in the brains of

VLCAD mice, expressed relative to the concentrations in the brains of control mice. In the brains of VLCAD mice, six intermediates of the CAC or related compounds were present at lower concentrations in VLCAD brains vs control, with three statistically significant decreases (α-ketoglutarate, glutamate and γ- aminobutyrate). In the brains of mice infused with the anaplerotic substrates, the relative concentrations of CAC intermediates (1 dose per mouse) did not show a clear trend compared to baseline levels (not shown). This contrasts with the stimulation of anaplerosis clearly demonstrated by the labeling data (Figs

4.8&4.9).

107 Discussion

To the best of our review of the literature, brain anaplerosis from propionyl-CoA precursors has not been investigated so far. This probably results from the very little availability of propionyl-CoA precursors to brain cells. In non-ruminant mammals, propionate derived from intestinal is almost completely removed from portal vein blood in a single passage through the liver (30). Since normal diets contain virtually no odd-chain fatty acids, the concentration of propionate in arterial plasma of non-ruminant mammals is very low (5 - 10 μM,

(30-32)). Thus peripheral tissues, including the brain, derive propionyl-CoA only from low plasma propionate concentrations, and from the local catabolism of isoleucine, methionine, threonine and valine. In contrast, when the diet contains a precursor of propionyl-CoA, as is the case when LC-FOD patients are treated with triheptanoin (9,10), peripheral tissues receive substantial concentrations of anaplerotic heptanoate and C5-ketone bodies (3). Since these compounds also provide acetyl groups, they are not only anaplerotic, but they also provide fuel for energy production. Our study was designed to assess, in normal and VLCAD mice, the fates of the propionyl and acetyl moieties of odd-medium-chain fatty acids in the brain. Analyses of the labeling data provided evidence that long- chain fatty acid oxidation occurs in the brain of control and VLCAD mice.

Secondary labeled substrates and tracers.

108 When labeled even-chain or odd-chain fatty acids are infused in vivo, the formation of ketone bodies represents a major source of secondary labeled substrates for peripheral tissues, especially for the brain which has a high capacity to utilize ketone bodies (33-35). When the fatty acids are labeled in one of their acetyl moiety, the acetyl groups of ketone bodies (C4 and C5) are substantially labeled (Figs 4.2A, 4.2B). When the odd-chain fatty acids are labeled in their propionyl-CoA moiety, the C5-ketone bodies are maximally M3 labeled because there is no dilution of the propionyl moiety in the liver. Thus the

M3 C5-ketone bodies can contribute to the anaplerotic labeling of brain CAC intermediates.

Comparison between Figs 4.1 and 4.3 shows that and the conversion of octanoate to C4-ketone bodies in the liver is much more efficient that the conversion of heptanoate to C5-ketone bodies. This presumably results from the pulling of propionyl-CoA away from the HMG-CoA cycle by anaplerosis followed by gluconeogenesis. The accumulation of M3 pentanoate and M3 propionate

13 during the infusion of [5,6,7- C3]heptanoate shows that M3 pentanoyl-CoA and

M3 propionyl-CoA are hydrolyzed in some tissues. Thus, when heptanoate is infused, substrate cycling occurs between pentanoyl-CoA and pentanoate, as well as between propionyl-CoA and propionate. Similar cycling between acetyl-

CoA and acetate was reported (36-38).

Another secondary tracer is glucose which becomes labeled from the in vivo metabolism of many [13C]substrates even if they are not gluconeogenic

109 (26,28,39). The β-oxidation in the liver of an odd-chain fatty acid e.g. [3,4,5-

13 C3]pentanoate forms M3 propionyl-CoA. The latter has two fates relevant to the present discussion: gluconeogenesis and C5-ketogenesis. First, anaplerosis from

M3 propionyl-CoA, followed by multiple turns of the CAC, and cataplerosis via

PEPCK, fuels gluconeogenesis with PEP enriched in M3, M2, and M1 mass isotopomers. Second, C5-ketogenesis forms M3 BKP and M3 BHP which are

13 released from the liver. When we infused [3,4,5- C3]BKP in other mice, we also observed glucose labeling, but with a different distribution of mass isotopomers

(Fig 4.4B). The labeling of glucose from M3 BKP must involve: (i) uptake by muscle and labeling of CAC intermediates via 3-oxoacid-CoA transferase, 3- ketoacyl-CoA thiolase, and anaplerosis from M3 propionyl-CoA, (ii) cataplerosis at α-ketoglutarate forming labeled glutamine which is released by muscle, (iii) uptake of labeled glutamine by liver and labeling of CAC and gluconeogenic intermediates. We previously showed that the labeling of glucose differs whether a labeled substrate is metabolized mostly in the liver or in muscle. In the latter case, abeling of glucose results from the muscle-to-liver transfer of labeled

13 glutamine (27). In the case of [3,4,5- C3]pentanoate which is metabolized in liver and in muscle, the labeling pattern of glucose (Fig 4.4A) results most likely from the combination of (i) net gluconeogenesis in the liver, (ii) metabolism of the substrate in muscle with transfer of labeled glutamine to the liver, and (ii) the metabolism of M3 BKP in muscle followed by transfer of labeled glutamine to the liver. The net gluconeogenic component of the metabolism of the heptanoate

110 moiety of triheptanoin helps in the maintenance of normoglycemia in LC-FOD patients which are prone to hypoglycemic episodes.

Entry of medium-chain fatty acids into the brain.

After systemic or enteral administration of a labeled medium-chain fatty acid such as octanoate, label from the substrate is found in brain metabolites derived from acetyl-CoA. This can occur either via entry of octanoate as such into the brain, followed by β-oxidation, or indirectly via entry into the brain of C4-ketone bodies formed from octanoate in the liver. To the best of our review of the literature, the entry of octanoate as such into the brain has not been unequivocally demonstrated. The time scale of accumulation of 11C radioactivity in the brain after intravenous injection of [1-11C]octanoate is compatible with indirect labeling via ketone bodies (40,41). Our study sheds light on this question since, in the

13 brains of mice infused with [1- C]octanoate, [1-C3]heptanoate, [1-C3]pentanoate,

13 or [3,4,5- C3]pentanoate, we observed the accumulation of the corresponding

M1or M3 CoAs esters (Figs 4.5A, 4.5B, 4.5C, Supplementary Figs 4.13-4.15).

Thus, the C8, C7, and the C5 n-fatty acids are taken up as such by the brain before CoA activation and β-oxidation. This does not exclude simultaneous uptake into the brain of ketone bodies formed from the medium-chain fatty acids in the liver (see below).

Evidence of long-chain fatty acid oxidation in mouse brain

111 In addition to demonstrating the oxidation of medium-chain fatty acids in mouse brains, our data provide indirect evidence that the brain oxidizes long-chain fatty acids. This evidence is based on a comparison of the profiles of enrichment of

M1 octanoyl-CoA, M1 heptanoyl-CoA and M1 pentanoyl-CoA in the brain when increasing amounts of [1-13C]octanoate, [1-13C]heptanoate or [1-13C]pentanoate were infused (Fig 4.5). When [1-13C]octanoate was infused the m1 enrichment of octanoyl-CoA plateaued at about 45% and 60% in control and VLCAD mice, respectively (Fig 4.5A). This implies that, in control mice, slightly more than one- half of whole-brain octanoyl-CoA was derived from an endogenous source, most likely long-even-chain fatty acids. Also, the 60% enrichment of octanoyl-CoA in

VLCAD mice shows that some long-chain fatty acid oxidation occurred in their brain in spite of the total absence of VLCAD (15). This strongly suggest that, in the brain of VLCAD mice, a vicariant β-oxidation enzyme allows to degrade long- chain acyl-CoA esters, albeit at a lower rate than in the brain of control mice.

Indeed, as shown in Table 4.1, octanoyl-CoA is present in the brain of VLCAD mice, but at a 3 times lower content than in the brains of control mice.

When [1-13C]heptanoate or [1-13C]pentanoate was infused, the m1 enrichment of heptanoyl-CoA and of pentanoyl-CoA increased to about 90% (Figs 4.5B,

4.5C). Thus, only about 10% of heptanoyl-CoA or pentanoyl-CoA was of endogenous origin, presumably derived from the small proportion of long-odd- chain fatty acids present in the brain. In support of this interpretation is the

112 presence of heptanoyl-CoA and pentanoyl-CoA in the brains of control and

VLCAD mice which were not infused with any substrate (Table 4.I).

The occurrence of long-chain fatty acid oxidation in the brain has been questioned because of the very low activity of 3-ketoacyl-CoA thiolase compared to other β-oxidation enzymes (42). Although our data on acyl-CoA concentration and enrichment (Fig 4.5, Supplementary Figs 4.13-4.15) reveal a β-oxidation process in the brain, they do not provide actual rates of fatty acid oxidation. Also, the labeling of brain acetyl-CoA during the intravenous infusion of a [1-

13C]medium-chain fatty acid can result either from the β-oxidation of this fatty acid in the brain, or/and the oxidation in the brain of ketone bodies derived from the partial oxidation of the [1-13C]fatty acid in liver. The presence of substantial activity of acetoacetyl-CoA thiolase in brain (42) allows the metabolism of ketone bodies following their activation by the very active 3-oxoacid-CoA transferase

(43). To shed some light on this question, Supplementary Figs 4.7 and 4.8 show, for mice infused with [1-13C]octanoate or [1-13C]heptanoate, (i) the enrichment of brain acetyl-CoA, as well as the enrichments of precursors of acetyl-CoA, i.e., (ii) the average enrichment of the acetyl groups of the [1-13C]acyl-CoA derived from the activation of the infused [1-13C]fatty acid ((m1 of acyl-CoA)/(number of acetyl in acyl-CoA)) , (iii) the average enrichment of the acetyl groups of C4-ketone bodies ([m1 + 2·m2]/2), and (iv) the enrichment of the acetyl group of C5-ketone bodies (= m1 of BHP, when [1-13C]heptanoate was infused). On Supplementary

Figs 4.7A and 4.7B, consider rates of [1-13C]octanoate infusion up to 60

113 μmol·min−1·kg−1 (corresponding to a plasma octanoate concentration of 0.5 mM

(Fig 4.1). In that range, the enrichment of acetyl-CoA is equal to the average enrichment the acetyls of octanoyl-CoA (1/4 that octanoyl-CoA) and is greater than the average enrichment of the acetyls of BHB. This suggests that, in that range of plasma [1-13C]octanoate concentration, [1-13C]acetyl-CoA is derives mostly from the oxidation of [1-13C]octanoyl-CoA. At higher rates of [1-

13C]octanoate infusion, [1-13C]acetyl-CoA is formed from the oxidation of both [1-

13 13 C]octanoyl-CoA and [ C]C4-ketone bodies. A similar situation occurs

(Supplementary Figs 4.8A and 4.8B) when mice were infused with [1-

13C]heptanoate at rates up to 50 μmol·min−1·kg−1 (corresponding to a plasma heptanoate concentration of 0.3 mM (Fig 4.2). At high [1-13C]heptanoate concentrations, both C4- and C5-ketone bodies could be precursors of [1-

13C]acetyl-CoA along with [1-13C]heptanoyl-CoA.

What is the metabolic significance of long-chain fatty acid oxidation in mouse brain given that the balance of long-chain fatty acids across the brain is zero

(44,45)? It is likely that fatty acid oxidation is part of a slow process of turnover and remodeling of brain lipids. It is also likely that the rate of oxidation of exogenous medium-chain fatty acids is faster than that of endogenous long-chain fatty acids. This is because the oxidation of long-chain fatty acids requires seven passages through the very-low activity 3-ketoacyl-CoA thiolase (42), versus two passages for octanoate or heptanoate.

114 Anaplerosis from propionyl-CoA precursors

In patients with long-chain fatty acid oxidation disorders successfully treated with dietary triheptanoin, the plasma concentrations of heptanoate and C5-ketone bodies peaked at 1.5 and 0.15 mM, respectively 1.5 hr after a meal containing

35% of the calories as triheptanoin (see Table 2 of (9)). To the extent that our mouse data can be extrapolated to humans, it is likely that the dietary treatment of patients with long-chain fatty acid oxidation disorders stimulated anaplerosis in their brains. In one patient born with a deficit in pyruvate carboxylase, an anaplerotic enzyme in brain (11), the improvement in the patient’s neurological status upon dietary treatment with triheptanoin was ascribed to the anaplerotic effect of heptanoate and C5-ketone bodies (46). This observation inspired the present study.

The high efficiency of anaplerosis from propionyl-CoA precursors (Fig 4.8) was also observed in rat hearts perfused with increasing concentrations of [U-

13 C3]propionate (24). This high efficiency was ascribed to the activation of succinyl-CoA thiolase by the decrease in free CoA resulting from the trapping of

CoA in propionyl-CoA. In the present study, the infusion of heptanoate, pentanoate or propionate increased the concentrations of the corresponding CoA esters (not shown) presumably resulting in a decrease in free CoA. This probably also explains the efficiency of anaplerosis from propionyl-CoA precursors in the brain.

115 In patients with LC-FOD, the rapid improvement of cardiac and muscular strength upon treatment with triheptanoin led to the hypothesis that, in these patients, the

CAC did not operate optimally. This occurs in spite of ample supply of acetyl-CoA by the high carbohydrate diet, and of a normal respiratory chain. It was proposed without direct evidence that the detergent effect of long-chain acyl-CoAs (47) would deplete the pools of some CAC intermediates which carry acetyl groups as they are oxidized. This would explain the release of large molecules such as creatine kinase from muscle during episodes of rhabdomyolysis. If large protein molecules are released, it is likely that small molecules like the catalytic intermediates of the CAC are also released. Our observation that the concentration of stearoy-CoA in VLCAD brain is four times that in control brains supports this hypothesis (Table 4.1). The depletion of some CAC intermediates would result in a less than optimal operation of the CAC. In the present study, it appears that the pools of some CAC intermediates and related compounds are depleted in the brains of VLCAD mice (Fig 4.9). However, the unclear effect of the infusion of propionyl-CoA precursors on the pools of CAC intermediates, in spite of evidence of stimulated anaplerosis (Fig 4.8) does not allow conclusions to be made on the relationship between rates of anaplerosis and poll sizes of

CAC intermediates in the brain. Still, rates of anaplerosis from propionyl-CoA precursors are faster in the brains of VLCAD mice compared to control mice (Fig

4.8). Also, the basal levels of propionyl-CoA and methylmalonyl-CoA are lower in

VLCAD brains than in control brains (Table 4.I). This must reflect a greater pull

116 on anaplerotic propionyl-CoA in VLCAD brains than in control brains, possibly as a means to optimize CAC operation and energy production.

Conclusions.

In the present in vivo study, metabolite concentrations and labeling patterns were assayed on total brain tissue. There is extensive evidence of compartmentation of brain metabolism between different cell types, in particular between neurons and astrocytes. We do not imply that our global measurements reflect the metabolism of all brain cells or of one type of brain cell. Global measurements such as those presented here complement measurements conducted on single cell types by others. However, our data clearly demonstrate the following conclusions. First, medium-chain fatty acids enter the brain as such before they are activated to CoA esters. Second, the brain is the site of long-chain and medium-chain fatty acid oxidation. Third, long-chain fatty acid oxidation occurs in the brain of VLCAD mice by a vicariant enzyme system to be identified. Fourth, anaplerosis from propionyl-CoA precursors is rapid in normal mouse brain, and even more rapid in VLCAD brain. We hope that the present study will stimulate measurements of fatty acid oxidation and anaplerosis from propionyl-CoA precursors in different types of brain cells from mice with metabolic disorders that could be alleviated by anaplerotic therapy.

117 Table 4.1

Acyl-CoA content of whole brain of control and VLCAD mice. Data, expressed as nmol/g, are presented as mean ± SD ( n = 7). Stars indicate statistically significant differences (p < 0.05).

Acyl-CoA ester Control VLCAD

Acetyl 43.3 ± 4.5 25.6 ± 6.3*

Propionyl 0.79 ± 0.11 0.48 ± 0.098*

Methylmalonyl 0.014 ± 0.007 0.004 ± 0.0006*

Succinyl 0.029 ± 0.007 0.025 ± 0.008

4:0 0.29 ± 0.15 0.065 ± 0.04*

5:0 2.1± 0.65 2.1 ± 1.3

6:0 0.17 ± 0.08 0.051 ± 0.03*

7:0 0.14 ± 0.06 0.10 ± 0.07

8:0 4.86 ± 1.00 1.72 ± 0.45*

10:0 0.34 ± 0.16 0.12 ± 0.08*

10:1 0.34 ± 0.28 0.23 ± 0.14

12:0 0.36 ± 0.17 0.17 ± 0.08*

12:1 0.19 ± 0.12 0.12 ± 0.08

118 14:0 3.7 ± 0.35 2.1 ± 1.1

14:1 0.78 ± 0.35 0.56 ± 0.25

14:2 0.50 ± 0.21 0.59 ± 0.30

15:0 0.11 ± 0.04 0.062 ± 0.03

16:0 6.80 ± 1.4 6.05 ± 2.3

16:1 3.91 ± 2.9 3.22 ± 1.2

16:2 0.29 ± 0.28 0.23 ± 0.13

18:0 8.37 ± 2.9 44.8 ± 10.2*

18:1 11.3 ± 2.8 10.1 ± 3.6

18:2 0.15 ± 0.1 0.72 ± 0.29*

20:4 6.98 ± 5.3 6.48 ± 2.9

119 References

1. Brunengraber, H. and Roe, C. R. (2006) J Inherit. Metab Dis. 29, 327-331

2. Leclerc, J., Des Rosiers C., Montgomery, J. A., Brunet, J., Ste-Marie, L.,

Reider, M. W., Fernandez, C. A., Powers, L., David, F., and Brunengraber, H.

(1995) Am. J Physiol 268, E446-E452

3. Kinman, R. P., Kasumov, T., Jobbins, K. A., Thomas, K. R., Adams, J. E.,

Brunengraber, L. N., Kutz, G., Brewer, W. U., Roe, C. R., and Brunengraber, H.

(2006) Am. J Physiol Endocrinol. Metab 291, E860-E866

4. Peuhkurinen, K. J., Nuutinen, E. M., Pietilainen, E. P., Hiltunen, J. K., and

Hassinen, I. E. (1982) Biochem J 208, 577-581

5. Bunger, R., Swindall, B., Brodie, D., Zdunek, D., Stiegler, H., and Walter, G.

(1986) J Mol Cell Cardiol. 18, 423-438

6. Kline, J. A., Thornton, L. R., Lopaschuk, G. D., Barbee, R. W., and Watts, J.

A. (2000) Shock 14, 215-221

7. Russell, R. R., III and Taegtmeyer, H. (1991) J. Clin. Invest 87, 384-390

8. Russell, R. R., III, Mommessin, J. I., and Taegtmeyer, H. (1995) Am. J

Physiol 268, H441-H447

120 9. Roe, C. R., Sweetman, L., Roe, D. S., David, F., and Brunengraber, H.

(2002) J Clin Invest 110, 259-269

10. Roe, C. R., Yang, B. Z., Brunengraber, H., Roe, D. S., Wallace, M., and

Garritson, B. K. (2008) Neurology 71, 260-264

11. Lapidot, A. and Gopher, A. (1994) J Biol. Chem 269, 27198-27208

12. Shen, J., Sibson, N. R., Cline, G., Behar, K. L., Rothman, D. L., and

Shulman, R. G. (1998) Dev. Neurosci 20, 434-443

13. Lebon, V., Petersen, K. F., Cline, G. W., Shen, J., Mason, G. F., Dufour, S.,

Behar, K. L., Shulman, G. I., and Rothman, D. L. (2002) J Neurosci 22, 1523-

1531

14. Xu, Y., Oz, G., LaNoue, K. F., Keiger, C. J., Berkich, D. A., Gruetter, R., and

Hutson, S. H. (2004) J Neurochem. 90, 1104-1116

15. Exil, V. J., Roberts, R. L., Sims, H., McLaughlin, J. E., Malkin, R. A.,

Gardner, C. D., Ni, G., Rottman, J. N., and Strauss, A. W. (2003) Circ. Res 93,

448-455

16. Des Rosiers C., Montgomery, J. A., Desrochers, S., Garneau, M., David, F.,

Mamer, O. A., and Brunengraber, H. (1988) Anal Biochem 173, 96-105

17. Weaire, P. J. and Kekwick, R. G. (1975) Biochem J 146, 425-437

121 18. Kasumov, T., Martini, W. Z., Reszko, A. E., Bian, F., Pierce, B. A., David, F.,

Roe, C. R., and Brunengraber, H. (2002) Anal Biochem 305, 90-96

19. Minkler, P. E., Kerner, J., Ingalls, S. T., and Hoppel, C. L. (2008) Anal

Biochem 376, 275-276

20. Yang, L., Kasumov, T., Kombu, R. S., Zhu, S. H., Cendrowski, A. V., David,

F., Anderson, V. E., Kelleher, J. K., and Brunengraber, H. (2008) J Biol. Chem

21. Hachey, D. L., Patterson, B. W., Reeds, P. J., and Elsas, L. J. (1991) Anal

Chem 63, 919-923

22. Beylot, M., Previs, S. F., David, F., and Brunengraber, H. (1993) Anal

Biochem 212, 526-531

23. Fernandez, C. A., Des Rosiers C., Previs, S. F., David, F., and

Brunengraber, H. (1996) J Mass Spectrom. 31, 255-262

24. Kasumov, T., Cendrowski, A. V., David, F., Jobbins, K. A., Anderson, V. E., and Brunengraber, H. (2007) Arch. Biochem Biophys 463, 110-117

25. Reszko, A. E., Kasumov, T., Pierce, B. A., David, F., Hoppel, C. L., Stanley,

W. C., Des Rosiers C., and Brunengraber, H. (2003) J Biol. Chem 278, 34959-

34965

26. Esenmo, E., Chandramouli, V., Schumann, W. C., Kumaran, K., Wahren, J., and Landau, B. R. (1992) Am. J Physiol 263, E36-E41

122 27. Beylot, M., Soloviev, M. V., David, F., Landau, B. R., and Brunengraber, H.

(1995) J Biol. Chem 270, 1509-1514

28. Yang, D., Previs, S. F., Fernandez, C. A., Dugelay, S., Soloviev, M. V.,

Hazey, J. W., Agarwal, K. C., Levine, W. C., David, F., Rinaldo, P., Beylot, M., and Brunengraber, H. (1996) Am. J Physiol 270, E882-E889

29. McGarry, J. D. and Foster, D. W. (1971) J Biol. Chem 246, 1149-1159

30. Puchowicz, M. A., Bederman, I. R., Comte, B., Yang, D., David, F., Stone,

E., Jabbour, K., Wasserman, D. H., and Brunengraber, H. (1999) Am. J Physiol

277, E1022-E1027

31. Powers, L., Osborn, M. K., Kien, C. L., Murray, R. D., and Brunengraber, H.

(1995) J. Mass Spectrom. 30, 747-754

32. Martini, W. Z., Stanley, W. C., Huang, H., Des Rosiers C., Hoppel, C. L., and Brunengraber, H. (2003) Am. J Physiol Endocrinol. Metab 284, E351-E356

33. Williamson, D. H., Bates, M. W., Page, M. A., and Krebs, H. A. (1971)

Biochem J 121, 41-47

34. Hawkins, R. A., Williamson, D. H., and Krebs, H. A. (1971) Biochem J 122,

13-18

35. Ruderman, N. B., Ross, P. S., Berger, M., and Goodman, M. N. (1974)

Biochem J 138, 1-10

123 36. Grigat, K. P., Koppe, K., Seufert, C. D., and Soling, H. D. (1979) Biochem J

177, 71-79

37. Knowles, S. E., Jarrett, I. G., Filsell, O. H., and Ballard, F. J. (1974)

Biochem J 142, 401-411

38. Crabtree, B., Gordon, M. J., and Christie, S. L. (1990) Biochem J 270, 219-

225

39. Schumann, W. C., Magnusson, I., Chandramouli, V., Kumaran, K., Wahren,

J., and Landau, B. R. (1991) J Biol. Chem 266, 6985-6990

40. Kuge, Y., Yajima, K., Kawashima, H., Yamazaki, H., Hashimoto, N., and

Miyake, Y. (1995) Ann. Nucl. Med. 9, 137-142

41. Vannucci, S. and Hawkins, R. (1983) J Neurochem. 41, 1718-1725

42. Yang, S. Y., He, X. Y., and Schulz, H. (1987) J. Biol. Chem. 262, 13027-

13032

43. Sugden, P. H. and Newsholme, E. A. (1973) Biochem J 134, 97-101

44. Spitzer, J. J. (1973) Physiologist 16, 55-68

45. Owen, O. E., Morgan, A. P., Kemp, H. G., Sullivan, J. M., Herrera, M. G., and Cahill, G. F., Jr. (1967) J Clin Invest 46, 1589-1595

124 46. Mochel, F., DeLonlay, P., Touati, G., Brunengraber, H., Kinman, R. P.,

Rabier, D., Roe, C. R., and Saudubray, J. M. (2005) Mol Genet. Metab 84, 305-

312

47. Jones, P. M., Butt, Y., and Bennett, M. J. (2003) Pediatr. Res 53, 783-787

125 Legends for figures

Fig 4.1. Concentrations of octanoate and C4-ketone bodies in the blood of control

(panel A) and VLCAD mice (panel B) infused with increasing amounts of [1-

13C]octanoate. One mouse per infusion rate.

Fig 4.2. Labeling of ketone bodies from increasing infused doses of [1-

13C]octanoate (panel A) or [1-13C]heptanoate (panel B). BHB: β-hydroxybutyrate

(C4-ketone body); BHP: β-hydroxypentanoate (C5-ketone body). In panel B, the low M2 labeling of BHB (similar to that shown in panel A) has been omitted for clarity.

Fig 4.3. Concentrations of selected metabolites in the blood of control (panel A) and VLCAD mice (panel B) infused with increasing amounts of [5,6,7-

13 C3]heptanoate. Heptanoate, pentanoate, C5-ketone bodies (C5-KB) and propionate were M3 labeled. C4-ketone bodies (C4-KB) remained unlabeled.

Fig 4.4. Mass isotopomer distribution of blood glucose in VLCAD mice infused

13 13 with increasing amounts of [3,4,5- C3]pentanoate (panel A) or [3,4,5- C3]BKP

(panel B).

Fig 4.5. Mole percent enrichment (MPE) of the parent acyl-CoA and acetyl-CoA in brains of control and VLCAD mice infused with increasing amounts of [1-

13C]octanoate (panel A), [1-13C]heptanoate (panel B) or [1-13C]pentanoate (panel

C).

126 Fig 4.6. Concentration of acetyl-CoA in the brain of control and VLCAD mice infused with increasing amounts of octanoate (panel A), heptanoate (panel B), pentanoate (panel C) or β-ketopentanoate (BKP) (panel D).

Fig 4.7. Mol percent enrichment (MPE) of M3 isotopomers of pentanoyl-CoA, propionyl-CoA, methylmalonyl-CoA and succinyl-CoA inmice infused with [3,4,5-

13 C3]pentanoate. This figure shows the measurements used to calculate rates of relative anaplerosis (Fig 8).

Fig 4.8. Relative anaplerosis in brain of control and VLCAD mice infused with

13 13 increasing amounts of [5,6,7- C3]heptanoate (A) , [3,4,5- C3]pentanoate (B) ,

13 13 [3,4,5- C3]BKP (C), or [ C3]propionate (D).

Fig 4.9. Relative brain concentrations of CAC intermediates and related compounds in VLCAD mice vs control mice (mean ± SD, n = 7 in each group).

Legends for supplementaty figures

Fig 4.10. Concentrations of selected metabolites in the blood of control (panel A) and VLCAD mice (panel B) infused with increasing amounts of [5,6,7-

13 C3]pentanoate. Pentanoate, C5-ketone bodies (C5-KB) and propionate were M3 labeled. C4-ketone bodies (C4-KB) remained unlabeled.

127 Fig 4.11. Concentrations of total M3 C5-ketone bodies and of total C4-ketone bodies in the blood of mice infused with increasing amounts of β-keto-[3,4,5-

13 C3]pentanoate.

Fig 4.12. Concentrations of M3 propionate, total M3 C5-ketone bodies and of total C4-ketone bodies in the blood of mice infused with increasing amounts of

13 [ C3]propionate.

Fig 4.13. Concentrations of pentanoyl-CoA and propionyl-CoA in the brains of control and VLCAD mice infused with increasing amounts of [3,4,5-

13 C3]pentanoate.

Fig 4.14. Concentrations of heptanoyl-CoA, pentanoyl-CoA and propionyl-CoA in the brains of control and VLCAD mice infused with increasing amounts of [5,6,7-

13 C3]heptanoate.

Fig 4.15. Concentrations of octanoyl-CoA, hexanoyl-CoA and butyryl-CoA in the brains of control and VLCAD mice infused with increasing amounts of [1-

13C]octanoate.

Fig 4.16. MPE of brain acetyl-CoA, average MPE of acetyls of brain octanoyl-

CoA, and average MPE of acetyls of blood β-hydroxybutyrate (BHB) in mice infused with increasing amounts of [1-13C]octanoate.

Fig 4.17. MPE of brain acetyl-CoA, average MPE of acetyls of brain heptanoyl-

CoA, average MPE of acetyls of blood β-hydroxybutyrate (BHB), and MPE of

128 acetyl of blood β-hydroxypentanoate (BHP) in mice infused with increasing amounts of [1-13C]heptanoate.

129 Fig. 4.1

130 Fig. 4.2

131 Fig. 4.3

132 Fig. 4.4

133 Fig. 4.5

134 Fig. 4.5 (continued)

135 Fig. 4.6

136 Fig. 4.6 (continued)

137 Fig. 4.7

138 Fig. 4.8

139 Fig. 4.8 (continued)

140 Fig. 4.9

141 Fig 4.10

142 Fig 4.11

143 Fig. 4.12

144 Fig. 4.13

145 Fig. 4.14

146 Fig. 4.15

147 Fig. 4.16

148 Fig. 4.17

149 4.3.2

Assay Of The Activity Of Malonyl-CoA Decarboxylase

By Gas Chromatography-Mass Spectrometry

Xiao Wang1, William C. Stanley2, Henri Brunengraber1*

and Takhar Kasumov1

Departments of Nutrition1, and Physiology & Biophysics2, Case Western Reserve

University, Cleveland, OH 44106

Running title: GC-MS assay of malonyl-CoA decarboxylase

* To whom correspondence should be addressed: Department of Nutrition, Case

Western Reserve University, School of Medicine – WG 48, 10900 Euclid Avenue

Cleveland, OH 44106-4954. Email: [email protected]

150 Abbreviations

GC-MS: gas chromatography-mass spectrometry; MPE: molar percent enrichment;

TMS: trimethylsilyl; PCI: positive chemical ionization.

Footnotes

1. Mass isotopomers are designated as Mn, where n is the number of atomic mass units above the molecular weight of the unlabeled isotopomer M.

Acknowledgments

This work was supported by NIH Roadmap grant 5R33DK070291，PO1

HL074237, and by the Cleveland Mt. Sinai Health Care Foundation.

151 Abstract

We developed a GC-MS assay to measure the activity of malonyl-CoA decarboxylase (MCD) in crude tissue homogenates. Liver extracts are incubated

13 13 with [U- C3]malonyl-CoA to form [U- C2]acetyl-CoA by the action of MCD. The

13 reaction mixture contains 2 mM ADP to prevent the hydrolysis of [1,2- C2]acetyl-

CoA by acetyl-CoA hydrolase present in the extracts. Newly-formed [U-

13 2 13 C2]acetyl-CoA and internal standard of [ H3, 1- C]acetyl-CoA are analyzed as thiophenol derivatives by GC-MS. This assay was applied to a study of the kinetics of MCD in rat liver. Using Lineweaver-Burke plot of MCD kinetics, Km of

−1 −1 202 μM and Vmax of 3.3 μmol·min ·(g liver) were calculated. The liver MCD activities (μmol·min−1·g−1 ± S.D.) in three groups of rats with different nutritional status (fed, one-day fasted or two-day fasted) were 1.80 ± 0.41, 2.59 ± 0.37 (p <

0.05), and 3.07 ± 0.70 (p < 0.05), respectively. In conclusion, we report a practical, non-radioactive, sensitive assay of MCD in crude tissue extract.

152 Introduction

Malonyl-CoA is an intermediate of fatty acid synthesis. In lipogenic and non- lipogenic organs, malonyl-CoA contributes to the regulation of mitochondrial fatty acid oxidation by inhibiting carnitine palmitoyltransferase I, the enzyme that controls the transfer of long chain fatty acyl-CoAs from the cytosol to the mitochondrial matrix [1-3]. Malonyl-CoA is formed by cytosolic acetyl-CoA carboxylase (ACC), and by a side-reaction of mitochondrial propionyl-CoA carboxylase [4]. The fates of malonyl-CoA are (i) fatty acid synthesis, and (ii) decarboxylation via malonyl-CoA decarboxylase (MCD). Thus, ACC and MCD catalyze a substrate cycle, which contributes to the regulation of .

MCD occurs in a wide array of organisms, from prokaryotes to mammals [5]. In rat liver, MCD was originally identified only in peroxisomes and cytosol [6].

However, a recent study reported that MCD is present in rat liver mitochondria, peroxisomes and cytosol, with the peroxisomal enzyme showing the highest specific activity [7]. MCD participates to the regulation of fatty acid oxidation in heart and skeletal muscle [8,9]. In muscle, MCD is activated by phosphorylation

[9]. Also, conditions associated with elevated plasma free fatty acid levels (high- fat feeding, starvation, and streptozotocin-induced diabetes) increase MCD mRNA in rat heart and skeletal muscle [8]. MCD expression is activated by the peroxisome-proliferator-activated receptor α (PPARα) [10], a member of the nuclear hormone receptor superfamily and a fatty acid-activated transcription

153 factor that regulates lipid metabolism [10]. Indeed, in the heart of PPARα- deficient mice, the expression of MCD is decreased, malonyl-CoA concentration is increased, and the rate of fatty acid oxidation is decreased [11].

There are several assays currently available to measure MCD activity. The spectrophotometric assay [12] involves (i) forming acetyl-CoA from the decarboxylation of malonyl-CoA by MCD, (ii) generating oxaloacetate from malate + NAD+ + malate dehydrogenase, and (iii) forming citrate from the condensation of oxaloacetate with acetyl-CoA via citrate synthase. The assay is monitored by the production of NADH. This technique has low sensitivity and is not applicable to crude preparations. The first described radiochemical assay

[13,14] is based on the same reactions, except that [14C]oxaloacetate is generated by transamination of [14C]aspartate with glutamate. The assay is monitored by the accumulation of [14C]citrate. A second, more direct

14 14 radiochemical assay [15] measures the formation of CO2 from [3- C]malonyl-

CoA. A third radiochemical assay, developed by Kener at al [16] involves (i) the production of [2-14C]acetyl-CoA from [2-14C]malonyl-CoA, (ii) the conversion of

[2-14C]acetyl-CoA to [2-14C]acetylcarnitine in the presence of excess L-carnitine and carnitine acetyltransferase, and (iii) the isolation of [2-14C]acetylcarnitine on ion exchange resin, and the counting of its radioactivity. The above three radiochemical assays are sensitive and have been used extensively. To avoid using radioactive substrates, we developed a stable isotope-based assay of

13 MCD activity. Our assay involves (i) incubating [U- C3]malonyl-CoA with the

13 tissue extract containing MCD, to form [1,2- C2]acetyl-CoA, (ii) spiking the

154 2 13 reaction mixture with [ H3, 1- C]acetyl-CoA internal standard, (iii) reacting the acetyl-CoA with thiophenol to form the corresponding labeled acetylthiophenols

[17], and assaying the concentration of acetylthiophenol by gas chromatography-mass spectrometry.

Methods

Materials: Chemicals, biochemicals, and enzymes were obtained from Sigma-

Aldrich. A MCD inhibitor (MCDi), CBM-301940 (3-{[5-(2,2,2-trifluoro-1-hydroxy-1- trifluoromethylethyl)-4,5-dihydro-isoxazole-3-carbonyl]-amino}-butyric acid tert- butyl ester), was generously provided by Chugai Pharma USA [13,18]. N-methyl-

N-trimethylsilyl-trifluoroacetamide was supplied by Regis Technologies Inc. [U-

13 2 13 C3]Malonic acid and [ H3, 1- C]acetic anhydride (99%) were purchased from

13 Isotec (Miamisburg, OH). [U- C3]Malonyl-CoA, was prepared from [U-

13 2 13 C3]malonic acid by the method of Trams and Brady [19]. [ H3, 1- C]Acetyl-

2 13 CoA internal standard was prepared by reacting [ H6, 1,1- C2]acetic anhydride

13 2 13 with CoA [20]. Both [U- C3]malonyl-CoA and [ H3, 1- C]acetyl-CoA were purified by HPLC on a Hewlett-Packard 1090 liquid chromatograph, equipped with a diode array UV detector monitoring absorbance at 260 and 280 nm, autosampler, and a 250-mm C18 semi-preparative column (30 cm, 10 mm id,

Phenomenex, C18(2) Luna) with a Security Guard column (50 x 10 mm,

Phenomenex, C18(2) Luna). Solvent reservoir A contained acetonitrile, and reservoir B contained 75 mM KH2PO4. The gradient profile was as follows: from

155 0 to 15 min, increase linearly the acetonitrile concentration from 5.8 to 7% and the flow rate from 3.25 to 3.4 mL/min; from 15 to 20 min, wash the column with

60% acetonitrile. Malonyl-CoA and acetyl-CoA elute at 7.54 and 17.74 min, respectively.

Sample preparation

Rats were killed by skull crushing just before laparotomy and excision of a liver lobe which was quick-frozen between aluminum blocks pre-cooled in liquid N2.

The frozen liver samples were powdered under liquid N2 and the powder kept in plastic tubes immersed in liquid N2 until analysis.

Liver powder was extracted with a Polytron homogenizer in ice-cold 130 mM KCl/

20 mM Hepes buffer pH 7.4. To the slurry were added 0.1% Triton X-100 and

0.1% (v/v) Sigma protease inhibitor cocktail. The extract was kept for 10 min on ice with occasional vortexing before being centrifuged at 12,000 g for 20 min.

The supernatant was frozen until analysis.

Assay of MCD activity

Various volumes of tissue extract (undiluted or 10-fold diluted) and water were added to 450 μl of buffer (45 mM potassium phosphate pH 7.7, 0.1 mM EDTA,

0.1 mM dithiothreitol, 0.2 mM KCN) for a total volume of 600 μl. When the MCD inhibitor CBM-301940 was tested, it was added in DMSO, resulting in a final

DMSO concentration of 1% [13]. The same DMSO concentration was added to

13 control incubations. After equilibration at 37°C, 0.4 mM [U- C3]malonyl-CoA was

156 added to start the reaction. After 0 to 5 min incubation at 37°C, 100 μl of medium was removed and transferred to a glass tube containing 200 μl of ice-cold

2 13 methanol and 7.6 nmol of [ H3, 1- C]acetyl-CoA internal standard. This was followed by adding 0.5 ml of 100 mM potassium phosphate buffer, pH 8.5 and

0.15 ml of 1 mM thiophenol solution in tetrahydrofurane (dried on sodium, freshly distilled and tested for the absence of acetylthiophenol). The reaction between acetyl-CoA and thiophenol was allowed to proceed for 4 hr at 60 ºC, under agitation. Excess thiophenol was precipitated with 0.1 ml of 5 mM silver nitrate.

Acetylthiophenol was extracted with 3 x 4 ml of diethyl ether. The extract was dried with Na2SO4 and the solvent evaporated down to about 0.1 ml to avoid evaporation of volatile acetylthiophenol. The acetylthiophenol solution in ether was analyzed by ammonia positive chemical ionization gas chromatography-

+ + mass spectrometry. Ions at m/z 170 (M + NH4 ) to 174 (M + NH4 + 4) were monitored.

TMS derivatization of malonylthiophenol

13 [U- C3]malonyl-CoA was incubated with thiophenol using the same protocol as for acetyl-CoA. After extraction with diethyl ether, malonylthiophenol was dried under N2. The dry residue was incubated with N-methyl-N-trimethylsilyl- trifluoroacetamide for 30 min at 60°C to form the trimethylsilyl (TMS) derivative.

The TMS ester of malonylthiophenol was analyzed by ammonia positive chemical ionization gas chromatography-mass spectrometry. Ions at m/z 286 (M

+ + + NH4 ) to 289 (M + NH4 + 3) were monitored.

157 GC-MS conditions

Analyses of acetylthiophenol and the TMS ester of malonylthiophenol were carried out on an Agilent 5973 mass spectrometer, equipped with a model 6890 gas chromatograph, autosampler and a 007 series OV-225 bonded phase fused silica capillary column (30 m, 320 μm i.d., 1 μm film thickness). The carrier gas was helium (2 μl injection is in splitless). The injector temperature was set at

240ºС. GC temperature program was as follows: start at 80ºС, hold for 1 min, increase by 8.0ºС/min to 180ºС, increase by 45ºС/min to 240ºС, and hold for 15 min. Both the ion source and the quadrupole were set at 150ºС.

Acetylthiophenol and the TMS ester of malonylthiophenol eluted at 12.9 and 15 min, respectively. Under ammonia-positive chemical ionization (pressure adjusted to optimized peak areas), ions monitored were 170 to 174 for

+ + acetylthiophenol (M + NH4 to M + NH4 + 4) and 286 to 289 for the TMS ester of

+ + malonylthiophenol (M + NH4 to M + NH4 + 3).

Results and Discussion

For the assay of MCD activity by GC-MS, we adapted a procedure, which we had previously developed for the assay of acetyl-CoA concentration and enrichment, namely transesterification of acetyl-CoA with thiophenol, followed by GC-MS assay of acetylthiophenol [17]. We had previously demonstrated that, under our experimental conditions, thiophenol does not react with ubiquitous traces of unlabeled acetate present in many reagent solutions [17]. However, blank assays occasionally revealed a small but variable peak at m/z 170, the mass of

158 unlabeled acetylthiophenol. We observed no blank at m/z 172. This is why we

13 decided to conduct the MCD assay using [U- C3]malonyl-CoA, which is

13 converted by MCD to [1,2- C2]acetyl-CoA. The latter is transesterified chemically

13 to M2 [1,2- C2]acetylthiophenol, the signal of which is monitored at m/z 172.

Fig 4.18 shows the GC-MS chromatograph of thiophenol derivatives of [1,2-

13 2 C2]acetyl-CoA generated by MCD and of internal standard [ H3, 1-

13C]acetylthiophenol. We described previously an assay of the enrichment of acetyl-CoA by LC-MS of the whole molecule [21] The acetylthiophenol technique allows assaying the concentration and enrichment of acetyl-CoA by GC-MS, which is more generally available than LC-MS. Also, the basal M+1 enrichment of acetylthiophenol (10.3%) is much lower than that of the intact acetyl-CoA molecule (26.6%) measured by LC-MS. This is useful for measuring low 13C- einrichments of acetyl-CoA [17]. However, the LC-MS assay of acetyl-CoA does not require any derivatization step.

13 The assay of the concentration of standard solutions of [1,2- C2]acetyl-CoA in

2 13 the presence of an internal standard of [ H3, 1- C]acetyl-CoA is linear in the range 0 to 22 nmol (r2 = 0.999, Fig 4.19).

In setting up the assay for MCD activity, we were aware that tissue extracts

13 contain enzymes which might hydrolyze [1,2- C2]acetyl-CoA formed in the MCD reaction, and lead to an underestimation of the MCD activity. Prass et al [22] had characterized liver acetyl-CoA hydrolase and shown that it is completely inhibited by 2 mM ADP. Therefore, we tested whether 2 mM ADP would inhibit [1,2-

159 13 C2]acetyl-CoA hydrolysis under the conditions of the MCD assay. As shown

13 below, the MCD assay generates about 2-3 nmol [1,2- C2]acetyl-CoA over 5 min.

13 Accordingly, we simulated the production of [1,2- C2]acetyl-CoA by infusing over

13 5 min, using a syringe pump, 2.2 nmol [1,2- C2]acetyl-CoA into three assay

13 mixtures devoid of [U- C3]malonyl-CoA, but containing: (i) no tissue extract, no

ADP (blank), (ii) tissue extract, no ADP, and (iii) tissue extract + 2 mM ADP. At 5

13 min, we stopped the infusion of [1,2- C2]acetyl-CoA, and added methanol +

2 13 [ H3, 1- C]acetyl-CoA internal standard + thiophenol to the solution (n = 5 for

13 each condition). The recovery of [1,2- C2]acetyl-CoA (Fig 4.20) were 2.03 ±

0.05 nmol in the blanks, 1.84 ± 0.04 nmol (91.4 ± 1.9% of blanks, p < 0.05) in the presence of tissue extract but no ADP, and 2.12 ± 0.16 nmol (104 ± 7.7% of the blanks, p > 0.05, not significantly different) in the presence of tissue extract and

ADP (n = 5). The data indicate that 2 mM ADP inhibits the hydrolysis of [1,2-

13 C2]acetyl-CoA formed in the MCD assay. As a result, we added 2 mM ADP to all assays of MCD activity.

We have not determined the yield of the transesterification reaction with

13 thiophenol. However, since the [1,2- C2]acetyl-CoA formed by MCD and the

2 13 internal standard [ H3, 1- C]acetyl-CoA were derivatized simultaneously, the yield of the transesterification step does not affect the calculated acetyl-CoA production.

13 The initial concentration of [U- C3]malonyl-CoA in the assay (0.4 mM) was at

13 least 20 times higher than the concentration of [1,2- C2]acetyl-CoA formed during the assay (0.02-0.03 mM maximum at 5 min). Consequently, the excess

160 13 of [U- C3]malonyl-CoA not used by the MCD-catalyzed decarboxylation may

13 spontaneously decarboxylate to [1,2- C2]acetyl-CoA, which would also be

13 derivatized with thiophenol in parallel with MCD-produced [1,2- C2]acetyl-CoA.

13 In addition, the excess of [U- C3]malonyl-CoA may also react with thiophenol and form the M3 thiophenol ester of malonic acid, which would lose the acidic carboxylic group to produce M2 acetylthiophenol. Such chemical decarboxylation, a common process for β-keto acids and their derivatives, can be stimulated by heat and acidic pH. Since the derivatization reaction with thiophenol requires a four-hour incubation at 60°C, some decarboxylation of M3 malonylthiophenol intermediate to M2 acetylthiophenol might occur. The decarboxylation of M3 malonyl-CoA and/or M3 malonylthiophenol would lead to an overestimated production of M2 acetyl-CoA in the MCD activity assay. In fact, we identified malonylthiophenol, produced from the incubation of malonyl-CoA with thiophenol (after TMS derivatization). To correct for artifactual M2 acetyl-

CoA production, blank assays without tissue extracts were run in parallel with the

MCD assays. Therefore, all measured acetyl-CoA values shown in Figs 4.21-

4.24 were corrected by subtracting the values of acetyl-CoA produced as a result of non-enzymatic decarboxylation yielded in blank samples.

Our assay of MCD activity is linear with time (Fig 4.21) and with protein concentration (up to 1 mg of tissue equivalent, Fig 4.22). The sensitivity of the assay allows analyzing MCD activity in the equivalent of 0.1 mg of liver tissue.

Fig 4.21 also shows that addition of the MCD inhibitor CBM-301940 (10 μM) resulted in an 85% inhibition of MCD activity. CBM-301940 is a recently

161 developed MCD inhibitor which has been used to increase malonyl-CoA concentration in heart [13,18].

To assess the effect of nutritional status on MCD activity, we assayed MCD activity in livers of fed, one-day fasted and two-day fasted rat. The activities

(μmol⋅min−1⋅g−1 ± S.D.) were calculated as 1.80 ± 0.41, 2.59 ± 0.37 (p < 0.05), and 3.07 ± 0.70 (p < 0.05), respectively (Fig 4.23). Our data confirm that fasting activates MCD activity in rat liver [8]. This is consistent with the activation of fatty acid catabolism in the fasting state, when the liver glycogen store is depleted.

The activated MCD decreases malonyl-CoA content, resulting in the activation of fatty acid oxidation. Also, the MCD activities (μmol⋅min−1⋅g−1 ± S.D.) in fed (1.80

± 0.41) and two-day fasted rat livers (3.07 ± 0.70) are in good agreement with the data presented by Dyck et al [23], based on the ratio between wet and dry weight of 3.3 in rat liver [24].

We used our new GC-MS assay of the MCD activity, to determine the Km and

Vmax of MCD in rat liver. The Lineweaver-Burke plot of the kinetics of MCD (Fig

−1 4.24) yields a Km for malonyl-CoA of 202 μM and a Vmax of 3.3 μmol⋅min ⋅(g

−1 liver) . The Km of MCD in rat liver is close to the Km measured in rat heart (250

μM) [15]. Since the concentration of malonyl-CoA in these tissues is less than 5 nmol/g, the flux through MCD is directly proportional to malonyl-CoA concentration. For reference, the Km of MCD we measured in rat liver is close to that measured in Mycobacterium tuberculosis H37Ra (200 μM) [5] and the recombinant human MCD (200 μM) [6]. It is lower than the Km of a partially

162 purified MCD from the mammary gland of the rat (330 μM) [25]. It is higher than the Km measured in swine heart MCD (47 μM) [26], in MCD cloned from rat brain and expressed in E. coli (68 μM) [27] and in MCD of rat liver mitochondria

(54 μM) [28].

In conclusion, we report a practical, non-radioactive, sensitive assay of MCD in crude tissue extract. The M2 acetyl-CoA generated can be assayed as such by

LC-MS [21], or as acetylthiophenol by more commonly available GC-MS. This assay will help investigating the regulation of fatty acid oxidation in various tissues.

163 Reference List

[1] J.D. McGarry, S.E. Mills, C.S. Long, and D.W. Foster, Observations on the

affinity for carnitine, and malonyl-CoA sensitivity, of carnitine

palmitoyltransferase I in animal and human tissues. Demonstration of the

presence of malonyl-CoA in non-hepatic tissues of the rat, Biochem. J.

214 (1983) 21-28.

[2] J.D. McGarry, G.F. Leatherman, and D.W. Foster, Carnitine

palmitoyltransferase I. The site of inhibition of hepatic fatty acid oxidation

by malonyl-CoA, J. Biol. Chem. 253 (1978) 4128-4136.

[3] D. Chien, D. Dean, A.K. Saha, J.P. Flatt, and N.B. Ruderman, Malonyl-

CoA content and fatty acid oxidation in rat muscle and liver in vivo, Am. J.

Physiol Endocrinol. Metab 279 (2000) E259-E265.

[4] H.R. Scholte, I.E. Luyt-Houwen, M.L. Dubelaar, and W.C. Hulsmann, The

source of malonyl-CoA in rat heart. The calcium paradox releases acetyl-

CoA carboxylase and not propionyl-CoA carboxylase, FEBS Lett. 198

(1986) 47-50.

[5] P.E. Kolattukudy, A.J. Poulose, and Y.S. Kim, Malonyl-CoA decarboxylase

from avian, mammalian, and microbial sources, Methods Enzymol. 71 Pt

C (1981) 150-163.

164 [6] K.A. Sacksteder, J.C. Morrell, R.J. Wanders, R. Matalon, and S.J. Gould,

MCD encodes peroxisomal and cytoplasmic forms of malonyl-CoA

decarboxylase and is mutated in malonyl-CoA decarboxylase deficiency, J.

Biol. Chem. 274 (1999) 24461-24468.

[7] E. Joly, M. Bendayan, R. Roduit, A.K. Saha, N.B. Ruderman, and M.

Prentki, Malonyl-CoA decarboxylase is present in the cytosolic,

mitochondrial and peroxisomal compartments of rat hepatocytes, FEBS

Lett. 579 (2005) 6581-6586.

[8] M.E. Young, G.W. Goodwin, J. Ying, P. Guthrie, C.R. Wilson, F.A. Laws,

and H. Taegtmeyer, Regulation of cardiac and skeletal muscle malonyl-

CoA decarboxylase by fatty acids, Am. J. Physiol Endocrinol. Metab 280

(2001) E471-E479.

[9] A.K. Saha, A.J. Schwarsin, R. Roduit, F. Masse, V. Kaushik, K. Tornheim,

M. Prentki, and N.B. Ruderman, Activation of malonyl-CoA decarboxylase

in rat skeletal muscle by contraction and the AMP-activated protein kinase

activator 5-aminoimidazole-4-carboxamide-1-beta -D-ribofuranoside, J.

Biol. Chem. 275 (2000) 24279-24283.

[10] G.Y. Lee, N.H. Kim, Z.S. Zhao, B.S. Cha, and Y.S. Kim, Peroxisomal-

proliferator-activated receptor alpha activates transcription of the rat

hepatic malonyl-CoA decarboxylase gene: a key regulation of malonyl-

CoA level, Biochem. J. 378 (2004) 983-990.

165 [11] F.M. Campbell, R. Kozak, A. Wagner, J.Y. Altarejos, J.R. Dyck, D.D.

Belke, D.L. Severson, D.P. Kelly, and G.D. Lopaschuk, A role for

peroxisome proliferator-activated receptor alpha (PPARalpha ) in the

control of cardiac malonyl-CoA levels: reduced fatty acid oxidation rates

and increased glucose oxidation rates in the hearts of mice lacking

PPARalpha are associated with higher concentrations of malonyl-CoA and

reduced expression of malonyl-CoA decarboxylase, J. Biol. Chem. 277

(2002) 4098-4103.

[12] C. Hamilton and E.D. Saggerson, Malonyl-CoA metabolism in cardiac

myocytes, Biochem. J. 350 Pt 1 (2000) 61-67.

[13] J.R. Dyck, J.F. Cheng, W.C. Stanley, R. Barr, M.P. Chandler, S. Brown, D.

Wallace, T. Arrhenius, C. Harmon, G. Yang, A.M. Nadzan, and G.D.

Lopaschuk, Malonyl coenzyme a decarboxylase inhibition protects the

ischemic heart by inhibiting fatty acid oxidation and stimulating glucose

oxidation, Circ. Res. 94 (2004) e78-e84.

[14] J.R. Dyck, A.J. Barr, R.L. Barr, P.E. Kolattukudy, and G.D. Lopaschuk,

Characterization of cardiac malonyl-CoA decarboxylase and its putative

role in regulating fatty acid oxidation, Am. J. Physiol 275 (1998) H2122-

H2129.

[15] G.W. Goodwin and H. Taegtmeyer, Regulation of fatty acid oxidation of

the heart by MCD and ACC during contractile stimulation, Am. J. Physiol

277 (1999) E772-E777.

166 [16] J. Kerner and C.L. Hoppel, Radiochemical malonyl-CoA decarboxylase

assay: activity and subcellular distribution in heart and skeletal muscle,

Anal. Biochem. 306 (2002) 283-289.

[17] T. Kasumov, J.E. Adams, F. Bian, F. David, K.R. Thomas, K.A. Jobbins,

P.E. Minkler, C.L. Hoppel, and H. Brunengraber, Probing peroxisomal

beta-oxidation and the labelling of acetyl-CoA proxies with [1-

(13C)]octanoate and [3-(13C)]octanoate in the perfused rat liver, Biochem.

J. 389 (2005) 397-401.

[18] A.E. Reszko, T. Kasumov, F. David, K.R. Thomas, K.A. Jobbins, J.F.

Cheng, G.D. Lopaschuk, J.R. Dyck, M. Diaz, C. Des. R, W.C. Stanley, and

H. Brunengraber, Regulation of malonyl-CoA concentration and turnover

in the normal heart, J. Biol. Chem. 279 (2004) 34298-34301.

[19] P.J. Weaire and R.G. Kekwick, The synthesis of fatty acids in avocado

mesocarp and cauliflower bud tissue, Biochem. J. 146 (1975) 425-437.

[20] A.E. Reszko, T. Kasumov, F. David, K.A. Jobbins, K.R. Thomas, C.L.

Hoppel, H. Brunengraber, and C. Des. R., Peroxisomal fatty acid oxidation

is a substantial source of the acetyl moiety of malonyl-CoA in rat heart, J.

Biol. Chem. 279 (2004) 19574-19579.

[21] R.L. Prass, F. Isohashi, and M.F. Utter, Purification and characterization of

an extramitochondrial acetyl coenzyme A hydrolase from rat liver, J. Biol.

Chem. 255 (1980) 5215-5223.

167 [22] J.R. Dyck, L.G. Berthiaume, P.D. Thomas, P.F. Kantor, A.J. Barr, R. Barr,

D. Singh, T.A. Hopkins, N. Voilley, M. Prentki, and G.D. Lopaschuk,

Characterization of rat liver malonyl-CoA decarboxylase and the study of

its role in regulating fatty acid metabolism, Biochem. J. 350 Pt 2 (2000)

599-608.

[23] M. Wimmer, B. Wilmering, and D. Sasse, The relation of rat liver wet

weight to dry weight, Histochemistry 83 (1985) 571-572.

[24] Y.S. Kim and P.E. Kolattukudy, Malonyl-CoA decarboxylase from the

mammary gland of lactating rat. Purification, properties and subcellular

localization, Biochim. Biophys. Acta 531 (1978) 187-196.

[25] K.L. King, I.C. Okere, N. Sharma, J.R. Dyck, A.E. Reszko, T.A. McElfresh,

J. Kerner, M.P. Chandler, G.D. Lopaschuk, and W.C. Stanley, Regulation

of cardiac malonyl-CoA content and fatty acid oxidation during increased

cardiac power, Am. J. Physiol Heart Circ. Physiol 289 (2005) H1033-

H1037.

[26] G.Y. Lee, Y.Y. Bahk, and Y.S. Kim, Rat malonyl-CoA decarboxylase;

cloning, expression in E. coli and its biochemical characterization, J.

Biochem. Mol. Biol. 35 (2002) 213-219.

[27] Y.S. Kim and P.E. Kolattukudy, Purification and properties of malonyl-CoA

decarboxylase from rat liver mitochondria and its immunological

168 comparison with the enzymes from rat brain, heart, and mammary gland,

Arch. Biochem. Biophys. 190 (1978) 234-246.

Legends for figures

Fig. 4.18. Ion chromatograph of acetyl-CoA (analyzed as the thiophenol derivative) formed from [U-13C]malonyl-CoA by MCD present in an extract of rat

2 liver. After stopping the reaction with methanol, the mixture was spiked with [ H3,

1-13C]acetyl-CoA internal standard and treated with thiophenol. The deuterated internal standard of acetyl-CoA migrated slightly ahead of the 13C -labeled acetyl-

CoA.

13 Fig. 4.19. Calibration curve of [1,2- C2]acetyl-CoA concentration by GC-MS. A neutralized perchloric acid extract from rat liver was spiked with increasing

13 2 amounts of [1,2- C2]acetyl-CoA and a constant amount (5.8 nmol) of [ H3, 1-

13C]acetyl-CoA.

13 Fig. 4.20. Inhibition by 2 mM ADP of [1,2- C2]acetyl-CoA hydrolysis. The

13 vertical scale is the amount of [1,2- C2]acetyl-CoA recovered at the end of a 5 min infusion of the substrate in blanks, or in model assay mixtures ± 2 mM ADP.

169 Fig. 4.21. Kinetics of MCD in the absence and presence of 10 μM inhibitor CBM-

301940

Fig. 4.22. Linearity of MCD assay with the amount of liver tissue

Fig. 4.23. MCD activities in livers from fed, one day starved, and two day starved rats.

Fig. 4.24. Lineweaver-Burk plot of MCD kinetics, with calculated Km of 202 μM and

Vmax of 3.3μmol•min-1• g tissue-1.

170 Fig. 4.18

171 Fig. 4.19

172 Fig. 4.20

173 Fig. 4.21

174 Fig. 4.22

175 Fig. 4.23

176 Fig. 4.24

177 Chapter 5

Discussion and Future Directions

5.1 Fatty acid oxidation and anaplerosis from propionyl-CoA precursors in

the brain of VLCAD mice and controls

5.1.1 Summary and implications

The major hypotheses of my investigation in the brain of VLCAD mice are: LC

FOD possibly leads to depletion of the pool of the CAC intermediates; and a larger supply of anaplerosis is required to maintain the proper functioning of the

CAC. The hypotheses are mainly based on the LC FOD-induced rhabdomyolysis which suggests the leakage of the CAC intermediates accompanied with the leakage of large molecules. Normal function of the brain requires an optimum pool size of the CAC intermediates which is maintained by physiological anaplerosis, primarily via the pyruvate carboxylase pathway. This loss of the CAC intermediates can compromise the operation of the CAC which could potentially lead to reduced energy production, defective turnover of neurotransmitters, and ultimately the progression of neurological abnormalities.

The present investigation is proposed in brains of VLCAD mice an elevated loss of the CAC intermediates induced by massive cataplerosis and/or defective anaplerosis. As a result, acute infusions of odd-chain compounds (MC-FAs or C5

KBs) can be beneficial to VLCAD brains due to their anaplerotic properties. To demonstrate these hypotheses, the basal concentrations of major CAC intermediates and related metabolites were assayed in the brains of control and

178 VLCAD mice. The data reveals an excessive cataplerosis and/or insufficient anaplerosis in VLCAD brains vs. controls, based on the observations that six

CAC intermediates are present at lower concentrations in VLCAD brains, with three statistically significant decreases. Second, acyl-CoA profile data indicates significantly reductions in the basal concentrations of propionyl-CoA and methylmalonyl-CoA in VLCAD brains, suggesting a larger requirement of anaplerosis that fuels the CAC and depletes the metabolites in the pathway of propionyl-CoA to succinyl-CoA. Reduced pool of the CAC intermediates can obstacle the efficient production of the ATP, as well as the maintenance of α-KG level which is the primary precursor of neurotransmitters glutamate, glutamine, and GABA.

The efficacy of anaplerotic therapy has been demonstrated in patients of various types of FOD (88; 164; 168; 242) and PC-deficiency (88). In particular, anaplerotic therapy of triheptanoin rapidly improves major neurological symptoms in PC-deficient patients, including the elevations of CSF Gln and f-GABA levels and the achievement of normal myelination (88). Therefore, the present study is proposed to investigate the potential benefits of various anaplerotic substrates in the brains of control and VLCAD mice.

Since odd MC-FAs heptanoate, pentanoate, propionate and C5 KBs are precursors of propionyl-CoA, they can contribute to the brain anaplerosis when infused intravenously, as used in the acute treatment of FOD. This contribution requires the uptake of infused compounds by the brain. This uptake was demonstrated by stable isotope technology coupled with liquid chromatography-

179 mass spectrometry. The research approach to assess the brain anaplerosis was

13 designed as follows: i) the MC-FAs and C5 KB infused were labeled with C on

13 their propionyl moieties. For example, the [5,6,7- C3]heptanoate, [3,4,5-

13 C3]pentanote etc.; ii) the labeled MC-FAs and C5 KB were infused to the jugular vein of control or VLCAD mice under anesthesia; iii) the labeling patterns of M3 heptanoyl-CoA, M3 pentanoyl-CoA and M3 propionyl-CoA in the harvested brain tissues were analyzed. The novelty of this strategy is that these labeled odd- chain acyl-CoA esters can only form in the brain cells, from labeled compounds infused that are taken up by the brain from the circulation. Particularly, the M3 labeled acyl-CoA corresponding to the infused FA precursor cannot be derived from other secondary tracers labeled in peripheral tissues. This stable isotope- based strategy eliminates any contamination commonly accompanied with radioactive isotopes that retains in extracellular fluid or derived from a secondary tracer (21; 45; 46; 52-55; 57). In the present study, the accumulation of labeled acyl-CoA corresponding to the MC-FA infused was observed, for example M3

13 heptanoyl-CoA accumulated in the brain when [5,6,7- C3]heptanoate was infused. Similar accumulations of shorter chain acyl-CoAs were also observed, for example M3 pentanoyl-CoA and M3 propionyl-CoA accumulated in the brain

13 when [5,6,7- C3]heptanoate was infused, which were metabolized from the oxidation of infused MC-FA in the brain. Therefore, the concentration and labeling pattern of these odd-chain acyl-CoAs provide strong evidence for the direct uptake of the C8, C7 and C5 FAs as such by the brain and their utilization via β-oxidation.

180 In addition to demonstrating the oxidation of MC-FA, my data clearly proves the oxidation of endogenous LC FAs in mouse brains. This is concluded from comparisons of the enrichments of M1 octanoyl-CoA, M1 heptanoyl-CoA and M1 pentanoyl-CoA in the brain when increasing amounts of [1-3C]octanoate, [1-

13C]heptanoate or [1-13C]pentanoate were infused. When [1-13C]octanoate was infused, the M1 enrichment of octanoyl-CoA plateaued at about 45% and 60% in control and VLCAD mice, respectively. This implies that, in control mice, slightly more than one-half of whole-brain octanoyl-CoA was derived from an endogenous source, most likely LC FAs. Also, the 60% enrichment of octanoyl-

CoA in VLCAD mice shows that some LC FA oxidation occurred in their brain in spite of the total absence of VLCAD. This strongly suggests that, in the brain of

VLCAD mice, a vicariant β-oxidation enzyme enables some degradation of LC acyl-CoA esters, albeit at a lower rate than in the brain of control mice.

Although some oxidation of LC FA was identified in mouse brains from the labeling pattern discussed above, the concentration profile of acyl-CoAs in

VLCAD revealed a clear blockage in LC FA oxidation. This is based on the finding that octanoyl-CoA concentration is 3 times lower in VLCAD brains vs. control brains. Also, VLCAD brains have significantly reduced levels of short- to medium-chain acyl-CoAs (C2-C8), but accumulations of C18:0-CoA and C18:2-CoA.

Since this acyl-CoA profile is consistent with acylcarnitine profile (C14-C18) found in the plasma or urine of VLCAD patients, it clearly demonstrates in VLCAD brains the presentation of VLC FOD. The accumulated C18:0-CoA and C18:2-CoA are harmful amphiphiles that damage membrane structure, due to their detergent

181 effects (251). This suggests some blockage of LC FA oxidation in VLCAD brains and possible elevation of the cataplerosis with excessive loss of the CAC intermediates when the mitochondrial membrane is damaged.

The contribution to the brain anaplerosis was evaluated by the MID ratio of M3 succinyl-CoA vs. M3 propionyl-CoA. This calculation reflects relative anaplerosis which represents the contribution of propionyl-CoA precursors to the catalytic intermediates of the CAC that carry the acetyl group as it is oxidized. Data of my investigation reveal that all odd MC-FAs and C5 KB tested increase the concentrations of their corresponding CoA esters; and they are strongly anaplerotic even at low plasma concentrations of 0.1 mM. Importantly, the calculated anaplerosis is higher in VLCAD brains than in control brains, suggesting larger demands of anaplerosis in VLCAD brains. This is consistent with reductions of CAC intermediates observed in VLCAD brains vs. controls, possibly due to the elevated cataplerosis or defective anaplerosis.

Among all substrates tested, the contributions to the brain anaplerosis are in the following order: pentanoate > BKP > heptanoate > propionate.

Infusions of octanoate, heptanoate, pentanoate and BKP contribute to the brain acetyl-CoA production, based on the findings of (i) the elevation of acetyl-CoA concentration and (ii) the elevation of M1 enrichment of acetyl-CoA, when increasing amounts of [1-13C]MC-FAs were infused.

5.1.2 Discussion and conclusions

The VLCAD deficiency commonly leads to abnormal accumulations of C14–C18 acylcarnitines during episodes of metabolic derangement, due to defective

182 oxidations of LC acyl-CoAs. The analysis of plasma or serum, especially when collected during a crisis, reveals accumulated C14-C18 (especially C14:1) acylcarnitine derived from the catabolism of oleic acid (233). Recently, Zhang

DY et. al. (252)profiled acyl-CoA esters in the livers of LCAD mice fasted for 6 hours without or with insulin clamp. Consistently, their data indicated in fasting livers the accumulations of C14–C18 acyl-CoA which were attenuated by insulin clamp. Whereas in the present study, only excessive C18:0 and C18:2 were identified accumulating in VLCAD brains. This may be attributed to different substrate preference for the enzymes of long-chain acyl-CoA dehydrogenase

(lcad) vs. vlcad. Also, VLCAD mice used in the present acyl-CoA profile study were not fasted. They probably have higher levels of insulin compared to LCAD mice used in Zhang DY’s paper. Insulin reduces LC FA oxidation cascade, thus decreasing the concentrations of C14:0-CoA and C14:1-CoA while increasing the concentrations of C18:0-CoA and C18:1-CoA. Additionally, food composition may also affect acyl-CoA profile in the brain. It is possible that mouse chow used in the present study is rich in C18:0 FA, resulting in the greater level of C18:0-CoA observed in VLCAD brains. However, similar level of C18:0-CoA was not observed in control brains fed on the same chow diet, thereby eliminating the possibility of dietary intervention. Another possibility that can alter diet composition in VLCAD mice but not in control mice is through weaning. Suckling

VLCAD mouse pups used in the present study consumed milk from VLCAD mothers during weaning. Deficiency of VLC acyl-CoA dehydrogenase could possibly affect fat composition in the mammary gland, thus altering the fat

183 composition of milk produced during lactation. Future studies are required to test milk composition in VLCAD mothers.

In mice infused with heptanoate (0-50% caloric requirement (CR)), plasma concentration of total C5 KBs only reached 0.2 mM; whereas in mice infused with similar amounts of octanoate (0-50% CR), plasma concentration of total C4 KBs reached as high as 1.25 mM. This suggests that hepatic C4 KB production from octanoate is much more efficient than C5 KB production from heptanoate. This presumably results from the pulling of propionyl-CoA away from the HMG-CoA cycle by anaplerosis followed by gluconeogenesis. The accumulation of M3

13 pentanoate and M3 propionate during the infusions of [5,6,7- C3]heptanoate indicated that M3 pentanoyl-CoA and M3 propionyl-CoA were hydrolyzed in some tissues. Thus, when heptanoate was infused, substrate cycling occurs between pentanoyl-CoA and pentanoate, as well as between propionyl-CoA and propionate. Hydrolyses of M3 pentanoyl-CoA and M3 propionyl-CoA released

M3 pentanoate and M3 propionate to the circulation, both of which can contribute to brain anaplerosis as secondary substrates in addition to the odd-chain MC-FA infused.

Labeled even-chain or odd-chain MC-FAs infused were also used for hepatic ketogenesis to form C4 or C5 KBs, which represented a major source of secondary labeled substrates for peripheral tissues, especially for the brain which has a high capacity to utilize KBs (59-61). When MC-FAs infused were labeled in one of their acetyl moiety, the acetyl groups of KBs (C4 and C5) were substantially labeled. When odd-chain FAs infused were labeled in their

184 propionyl-CoA moiety, the C5-ketone bodies were maximally M3 labeled because there is no dilution of the propionyl moiety in the liver. Thus the M3 C5-ketone bodies can also contribute to brain anaplerosis.

Since propionyl-CoA precursors are also gluconeogenic, labeled odd-chain FAs infused can also label hepatic glucose which when exported can serve as a secondary tracer for all peripheral tissues. The labeling patters of glucose were measured to evaluate the possible effect of this secondary tracer. In mice

13 infused with increasing amounts of [3,4,5- C3]pentanoate, glucose labeling pattern from M1 to M5 was observed, with MID in the order of M2 > M3 > M1 >

13 M4 > M5. The oxidation of [3,4,5- C3]pentanoate in the liver formed M3 propionyl-CoA via β-oxidation. Anaplerosis from M3 propionyl-CoA, followed by cataplerosis via PEPCK, fuel the hepatic gluconeogenesis, thereby generating

M2 and M3 PEP. In addition, multiple turns of the CAC followed by PEPCK activity can generate additional M and M1 PEP. The combination of all mass isotopomers of PEP results in the observed glucose MID profile of M2 > M3 > M1

13 > M4 > M5. When [3,4,5- C3]BKP was infused, glucose labeling was observed as well, but with a different MID profile in the order of M1 > M2 > M3 > M4 > M5.

This is because C5 KBs are not gluconeogenic substrates in the liver. KBs cannot be activated in the liver, but they can be taken up and activated in the proximal tubule of the rat kidney (166; 167). Thus M3 BKP when infused as such, or generated from the hepatic metabolism of M3 heptanoate or M3 pentanoate, could contribute to renal gluconeogenesis and thereby labeling glucose. The distinct labeling profile of glucose derived from M3 BKP infusions must be

185 generated from a separate mechanism. This mechanism must involve: (i) the uptake by muscle of M3 BKP and subsequent labeling of the CAC intermediates via anaplerosis, (ii) the cataplerosis as labeled glutamine which is then released by the muscle, (iii) the uptake of labeled glutamine by the liver and the subsequent labeling of hepatic CAC intermediates, and (iv) the gluconeogenic cataplerosis that generates labeled glucose. A previous study in my lab (253) proved that glucose labeling differs substantially whether a labeled substrate is metabolized mostly in the liver or in the muscle. In the latter case, the labeling of glucose results from the muscle to liver transfer of labeled glutamine (253). Additionally, the uptake in kidney of M3 BKP (166; 167) could also contribute to renal gluconeogenesis, thereby labeling glucose. In the case

13 of [3,4,5- C3]pentanoate which is utilized by the liver and the muscle, the labeling pattern of glucose resulted most likely from the combination of its metabolism in both organs. The formation of labeled glucose from precursors of

13 [ C3]propionyl-CoA makes it complicated to interpretate labeling patterns of metabolites and the calculations of anaplerosis. This will be discussed later.

The uptake and oxidation of infused MC-FAs by the brain was demonstrated by the concentration and labeling profiles of the odd-chain acyl-CoA esters. In mice infused with increasing amounts of heptanoate, the elevations of heptanoyl-CoA, pentanoyl-CoA, and propionyl-CoA in the brain were observed, ranging from 0.05 to 0.25, 0.5 to16, 0.15 to 0.3 nmol/g, respectively. Among three acyl-CoAs that accumulated, the concentration of pentanoyl-CoA increased most rapidly and plateaued at the highest level, whereas the concentration of propionyl-CoA

186 presented a delayed increase, probably being dragged into the direction of anaplerosis through succinyl-CoA.

13 The four propionyl-CoA precursors infused, [5,6,7- C3]heptanoate, [3,4,5-

13 13 13 C3]pentanoate, [3,4,5- C3]BKP, or [ C3]propionate, all contributed to a substantial fraction of the brain anaplerosis, even at very low blood concentrations. The high efficiency of anaplerosis from propionyl-CoA precursors was also observed in rat hearts perfused with increasing

13 concentrations of [U- C3]propionate (159). This high efficiency was attributed to the activation of succinyl-CoA thiolase by the decrease in free CoA resulted from the trapping of CoA in propionyl-CoA (159). In my present study, the infusions of heptanoate, pentanoate and propionate increased the concentrations of their corresponding CoA esters, presumably resulting in a decrease in free CoA as well. This probably also explains the efficiency of anaplerosis from propionyl-

CoA precursors in the brain.

Relative anaplerosis from M3 propionyl-CoA precursors is calculated from the enrichment ratio (m3 succinyl-CoA)/(m3 propionyl-CoA). When 13C labeling enters the CAC only through M3 succinyl-CoA (originates from M3 propionyl-

CoA), the recycling of labeling in the continued turns of the CAC cannot yield additional M3 succinyl-CoA, but only M2 and M1 succinyl-CoA. This avoids the artifact of overestimating the anaplerosis from secondary precursors of less-than- triply labeled propionyl-CoA. However, if other 13C labeled metabolites enter the

CAC via M1 or M2 acetyl-CoA, some M3 or M4 succinyl-CoA could be formed, when all labelings from propionyl-CoA, the CAC cycling and acetyl-CoA

187 eventually combine. Since labeled glucose was found in the circulation when labeled heptanoate, pentanoate or propionate was infused, and since labeled glucose generates labeled acetyl-CoA, calculated anaplerosis could be overestimated under these circumstances. To test this possible overestimation, the labeling patterns of succinyl-CoA from M to M4 were compared, with no M4 isotopomer detectable. This suggests that labeling from acetyl-CoA (via glycolysis from glucose) only contributed to a trivial amount to succinyl-CoA, if there was any. Therefore, it is unlikely that any substantial M3 succinyl-CoA was formed due to the labeled glucose that can result in an overestimation of brain anaplerosis. This conclusion was further supported by the kinetics of anaplerosis which rapidly plateaued at fairly low blood concentration of infused [3,4,5-

13 C3]pentanoate. This means that although the amount of labeled glucose

13 continued to grow as the infusion rate of [3,4,5- C3]pentanoate increased, it did not result in continued elevations of brain anaplerosis.

Additional finding that eliminated the possible overestimation of anaplerosis was obtained from NMR analysis of the brain glutamate. The positional isotopomers on carbons 4 and 5 of glutamate correspond to carbons 2 and 1 of acetyl-CoA, respectively. The 13C-NMR spectra of the brain extracts revealed some 13C enrichments in carbons 4 and 5 of glutamate, but at very low levels (about 5%).

This suggests a small contribution of labeled acetyl-CoA to the CAC, eliminating a substantial overestimation of anaplerosis.

As expected, infusions of all FAs tested contributed to the brain acetyl-CoA production. In the brain of control mice, [1-13C]heptanoate and [1-13C]pentanoate

188 only accounted for 33% and 20% of the brain acetyl-CoA pool, respectively, while

67% and 80% must be derived from unlabeled sources, most likely from glucose.

In VLCAD brains, the contribution of [1-13C]pentanoate to acetyl-CoA (40%) is about twice of that in control brains. This higher contribution in VLCAD brains must be resulted from less dilution from unlabeled acetyl-CoA, consistent with reduced acetyl-CoA concentration in VLCAD brains demonstrated in the acyl-

CoA profile study.

Also, the labeling of brain acetyl-CoA during the intravenous infusion of a [1-

13C]MC-FA can result either from the β-oxidation of this FA in the brain, or/and the oxidation in the brain of KBs derived from the partial oxidation in liver of the

[1-13C]FA infused. The presence of the substantial activity of acetoacetyl-CoA thiolase in brain (254) allows the metabolism of KBs following their activation by the very active 3-oxoacid-CoA transferase (255). For example, when [1-

13C]octanoate was infused, the enrichment of brain acetyl-CoA was equal to the average enrichment of the acetyls of octanoyl-CoA (1/4 of the enrichment of octanoyl-CoA because octanoyl-CoA generates 4 acetyl-CoA and only 1 acetyl moiety is labeled). However, the enrichment of brain acetyl-CoA should be less than the average enrichment of the acetyls of octanoyl-CoA, because of substantial dilution from unlabeled acetyl-CoA generated from other energy substrates, such as glucose and endogenous LC FAs. Therefore, some of this labeled acetyl-CoA must derive from secondary substrates that were labeled elsewhere. This was confirmed by the finding of [13C]BHB in the circulation as infusion rates of [1-13C]octanoate increased. This suggests that, [1-13C]acetyl-

189 13 13 CoA was formed from the oxidation of both [1- C]octanoyl-CoA and [ C]C4-KBs, especially at higher rates of [1-13C]octanoate infusion. Similar situation occurred when mice were infused with higher rates of [1-13C]heptanoate. In this case,

13 13 13 both [1- C]C4- and [1- C]C5-KBs could be precursors of [1- C]acetyl-CoA along with [1-13C]heptanoyl-CoA.

The basal concentrations of major CAC intermediates and closely related metabolites, glutamate, glutamine and GABA, were largely reduced in VLCAD brains vs. controls. Three significant reductions identified in VLCAD brains were

αKG, glutamate and GABA. This observation suggests that, in VLCAD brains, the operation of the CAC could be severely compromised that resulted in the reduction of neurotransmitters downstream. This explained larger demands of anaplerosis observed in VLCAD brains.

Glutamate and GABA are the most widely distributed excitatory and inhibitory neurotransmitters in the brain, respectively. Glutamate functions in various brain processes, including neuronal development, neurotoxicity, and various functions of neuronal plasticity, such as learning and memory (256). The deficiency of neurotransmitters glutamate, glutamine and GABA found in VLCAD brains could contribute to the commonly observed neurological defects that FOD patients suffer from. Importantly, infusions of all anaplerotic substrates tested

(heptanoate, pentanoate, BKP and propionate) contributed to brain anaplerosis which refills the intermediates of the CAC and possibly the downstream neurotransmitters.

190 To summarize, the present study was designed to measure global metabolism of the whole brain and does not reflect the metabolism of all brain cells or one particular cell type. Based on the data presented above, the present study clearly demonstrated the following conclusions. First, MC-FAs enter the brain as such before they are activated to CoA esters. Second, LC and MC FA oxidation occurs in the brain of adult mouse. Third, LC FA oxidation occurs in the brain of

VLCAD mice by a vicariant enzyme system to be identified. Fourth, anaplerosis from propionyl-CoA precursors is rapid in normal mouse brain, and even more rapid in VLCAD brain. These results suggest the potential benefit of anaplerosis from odd MC-FAs in the brain of VLCAD mice. This could help with a better understanding of the VLCAD mouse model and further clinical development of anaplerotic therapy in patients of LC and VLC FOD.

5.1.3 Future directions

A. To assay the labeling pattern of glutamine in the blood of VLCAD and control

13 13 mice infused with [3,4,5- C3]pentanoate or [3,4,5- C3]BKP

As discussed earlier in this chapter, labeled glucose in the blood when non-

13 anaplerotic [3,4,5- C3]BKP is infused may be derived from (i) the uptake by muscle and labeling of the CAC intermediates via anaplerosis, (ii) the cataplerosis and release of labeled glutamine by the muscle, (iii) the uptake of labeled glutamine by the liver and subsequent labeling of the CAC intermediates, and (iv) the hepatic gluconeogenic cataplerosis. Therefore, glutamine is the major contributor to generate labeled glucose. Analysis of the labeling pattern of glutamine in the blood will help demonstrate this theory.

191 B. To assay the acyl-CoA profile in fasting VLCAD and control brains

The distinctive profile of the accumulated C14-C18 acylcarnitine identified in LC

FOD patients is demonstrated to be more obvious during events of an energy crisis, such as fasting, cold challenge or infection. Similarly, the profile of intracellular acyl-CoAs should change with stress challenges as well. This is because insulin and glucagon levels fluctuate between the feeding and fasting states. They directly regulate FA oxidation and affect acyl-CoA profile. As indicated by Zhang DY et. al. (252), insulin clamp shifted hepatic acyl-CoA content from SC range to LC or VLC ranges.

Although 8-hour fasting with cold challenge is lethal to VLCAD mice (233), it is still feasible to starve them for 6 hours at room temperature, or even up to 12 hours. In my present study of the brain acyl-CoA profile, the mice used were not fasted. Additional study in fasting mice will provide more information on the profile of the brain acyl-CoAs.

C. To assay the acyl-CoA profile in the brain of VLCAD mice weaned with

VLCAD+/- mothers

As pointed out earlier in this chapter, mouse pups weaned with VLCAD-/- mothers could have ingested milk with an altered fat composition. Dietary fat composition can affect the profile of FAs and possibly acyl-CoA esters in the brain, due to substantial fat utilization during weaning as the brain grows rapidly after birth.

Also, VLCAD-/- mouse mothers could produce milk that is deficient in certain essential or non-essential FAs, which could affect normal development of VLCAD pups. The assay of the milk composition in VLCAD-/- mothers will demonstrate

192 this possible intervention. Future research of brain metabolism in VLCAD mice weaned with VLCAD+/- mothers could be a better approach which minimizes possible intervention from milk composition during weaning.

D. To measure the concentrations of the CAC intermediates and

neurotransmitters glutamate, glutamine and GABA with infusions of

anaplerotic substrates

In the current study, reductions in the basal concentrations of the CAC intermediates and neurotransmitters glutamate, glutamine and GABA were observed in VLCAD brains, suggesting insufficient anaplerosis and/or excessive cataplerosis. The concentrations of these metabolites were also assayed with infusions of odd MC-FAs and C5 KB in the present study. The data implied normalized to various extents the concentrations of all CAC intermediates

(except α-KG), glutamine and GABA in VLCAD brains with infusions of all odd- chain compounds infused. However, since only one mouse was used for one infusion rate of each compound tested, statistical analyses cannot be done.

Therefore, the results were difficult for conclusive interpretation. To assess this normalization of the CAC intermediates and neurotransmitters by infusions of anaplerotic substrates with greater certainty and confidence, a group of 5-8 mice should be used for each infusion rate of one substrate, which allows appropriate statistical analyses.

E. To assay the cell mass and distribution of neurotransmitters in VLCAD and

control brains, with or without infusions of anaplerotic substrates

193 Measurements in VLCAD brains revealed lower concentrations of major neurotransmitters (glutamate, glutamine and GABA) based on assays of the whole brain. Since the brain is consisted of at least three different cell-types, endothelial cells, neurons and glia, with multiple compartments channeled together. The complex compartmentation effect leads to massive difficulties in the interpretation of this data. For example, the reductions of neurotransmitters observed could be resulted from their distribution changes from one compartment to another, which may not be functional relevant. Alternatively, reductions of neurotransmitter concentrations could be caused by an increase in brain cell mass. Based on above considerations, studies of cell mass and the distribution of neurotransmitters in the brain of VLCAD mice will further correlate the concentration changes with functional significance.

F. To identify the vicariant enzyme that enables LC FA oxidation in VLCAD

brains

As discussed earlier in this chapter, the enrichment of M1 octanoyl-CoA in

VLCAD brains plateaued at 60% when [1-13C]octanoate was infused, suggesting the oxidation of endogenous LC FAs that contributed to unlabeled octanoyl-CoA in spite of the total absence of VLCAD (234). This strongly suggests that, in the brain of VLCAD mice, a vicariant β-oxidation enzyme allows to degrade LC acyl-

CoA esters. Future studies are needed to identify this enzyme. The most possible candidate could be lcad that partially oxidizes VLC FAs. However, this vicariant enzyme does not completely compensate for VLCAD, as demonstrated by the abnormal acyl-CoA profile observed in VLCAD brains.

194 G. To correlate the neurotransmitter level to the brain functional test, with and

without infusions of anaplerotic substrates

The observed reduction in neurotransmitters glutamate, glutamine and GABA in

VLCAD brains could directly drives the induction and/or progression of neurological abnormalities commonly seen in patients of VLC and LC FOD.

However, no one has actually correlated the neurotransmitter levels to brain functional tests either in animal models of FOD or in humans. As a result, brain functional tests with or without anaplerotic therapy are necessary in the future to better understand the characteristics of FOD and to design clinical treatments.

5.2 The assay of the activity of MCD by GC-MS

5.2.1 Discussion and conclusions

An accurate, relatively simple and sensitive assay of the activity of MCD is required for investigations of the action and regulation of the MCD enzyme. In parallel to investigating FA oxidation, I set up a practical, non-radioactive,

13 sensitive assay of MCD in crude tissue extract. The [1,2- C2]acetyl-CoA

13 generated from [ C3]malonyl-CoA was assayed as acetylthiophenol by more commonly available GC-MS. This assay of MCD activity in rat liver was linear with time and with protein concentration. This assay was applied to assess the effect of nutritional status on MCD activity in livers of fed, one-day fasted and two-day fasted rats. The data confirmed the activation of FA catabolism in the fasting state, consistent with the literature.

In setting up the assay ex vivo, 2mM ADP was used to minimize the action of

13 ample hepatic enzymes which may hydrolyze the [1,2- C2]acetyl-CoA formed in

195 the MCD reaction, and lead to an underestimation of the MCD activity. The hydrolysis of acetyl-CoA was trivial, as demonstrated in vitro by a 104% recovery of acetyl-CoA with the presence of ADP vs. a 91% recovery without ADP.

13 The initial concentration of [ C3]malonyl-CoA in the assay (0.4 mM) was at least

13 20 times higher than the concentration of [1,2- C2]acetyl-CoA formed during the assay (0.02-0.03 mM maximum at 5 min). This was designed to allow the

13 maximum action of the MCD enzyme. Consequently the excess of [ C3]malonyl-

CoA not used by the MCD-catalyzed decarboxylation may spontaneously

13 decarboxylate to [1,2- C2]acetyl-CoA, which would also be derivatized with

13 thiophenol in parallel with MCD-produced [1,2- C2]acetyl-CoA. In addition, the

13 excess of [ C3]malonyl-CoA may also react with thiophenol and form the M3 thiophenol ester of malonic acid, which would lose the acidic carboxylic group to produce M2 acetylthiophenol. The decarboxylation of M3 malonyl-CoA and/or

M3 malonylthiophenol would lead to an overestimated production of M2 acetyl-

CoA in the MCD activity assay. This artifactual M2 acetyl-CoA production was corrected by subtracting the values of acetyl-CoA produced as a result of non- enzymatic decarboxylation yielded in blank samples.

5.2.2 Future directions

13 In the present assay of the activity of MCD, the assay of [1,2- C2]acetyl-CoA formed by MCD requires a transesterification reaction with thiophenol to be detected by the GC-MS. The yield of this transesterification reaction determines the signal of detection. For future studies, the reaction should be optimized to

13 achieve the maximum yield. However, since the [1,2- C2]acetyl-CoA formed

196 2 13 by MCD and the internal standard [ H3, 1- C]acetyl-CoA were derivatized simultaneously, the yield of the transesterification step does not affect the calculated acetyl-CoA production.

Also, for further validation purposes, data from the present assay of the activity of

MCD should be compared with data from traditional methodologies using radioactive isotopes. Those methodologies were introduced in Chapter 4.

13 Additionally, instead of GC-MS, the [1,2- C2]acetyl-CoA formed by MCD from

13 [ C3]malonyl-CoA can be measured by LC-MS for comparison.

197 Reference List

1. Guzman M and Blazquez C. Is there an astrocyte-neuron ketone body

shuttle? Trends Endocrinol Metab 12: 169-173, 2001.

2. Kety SS and Schmidt CF. The nitrous oxide method for the quantitative

determination of cerebral blood flow in man: theory, procedure and normal

values. J Clin Invest 27: 476-483, 1948.

3. Edvinsson L, MacKenzie ET and McCulloch J. Cerebral blood flow and

metabolism. New York: Raven Press, 1993.

4. Sokoloff L. Circulation and energy metabolism of the brain. In: Basic

Neurochemnistry, edited by Siegel G, Agranoff B, Albers RW and Molinoff

P. New York: Raven Press, 1989, p. 565-590.

5. Owen OE, Morgan AP, Kemp HG, Sullivan JM, Herrera MG and Cahill

GF, Jr. Brain metabolism during fasting. J Clin Invest 46: 1589-1595, 1967.

6. McIlwain H and Bachelard HS. Biochemistry and the central nervous

system. Edinburgh: Churchill Livingstone, 1985.

7. Schurr A, West CA and Rigor BM. Lactate-supported synaptic function

in the rat hippocampal slice preparation. Science 240: 1326-1328, 1988.

198 8. Pardridge WM and Oldendorf WH. Transport of metabolic substrates

through the blood-brain barrier. J Neurochem 28: 5-12, 1977.

9. Teller DN, Bana-Schwartz M, Deguzman T and Lajtha A. Energetics of

amino acid transport into brain slices: effects of glucose depletion and

substitution of Krebs' cycle intermediates. Brain Res 131: 321-334, 1977.

10. Schurr A. Lactate: the ultimate cerebral oxidative energy substrate? J

Cereb Blood Flow Metab 26: 142-152, 2006.

11. Ide K, Schmalbruch IK, Quistorff B, Horn A and Secher NH. Lactate,

glucose and O2 uptake in human brain during recovery from maximal

exercise. J Physiol 522 Pt 1: 159-164, 2000.

12. Smith D, Pernet A, Hallett WA, Bingham E, Marsden PK and Amiel SA.

Lactate: a preferred fuel for human brain metabolism in vivo. J Cereb

Blood Flow Metab 23: 658-664, 2003.

13. Brightman MW and Reese TS. Junctions between intimately apposed

cell membranes in the vertebrate brain. J Cell Biol 40: 648-677, 1969.

14. Saunders NR, Habgood MD and Dziegielewska KM. Barrier

mechanisms in the brain, I. Adult brain. Clin Exp Pharmacol Physiol 26:

11-19, 1999.

199 15. Van Bree JB, De Boer AG, Danhof M and Breimer DD. Drug transport

across the blood-brain barrier. II. Experimental techniques to study drug

transport. Pharm Weekbl Sci 14: 338-348, 1992.

16. Dehouck MP, Meresse S, Delorme P, Fruchart JC and Cecchelli R. An

easier, reproducible, and mass-production method to study the blood-

brain barrier in vitro. J Neurochem 54: 1798-1801, 1990.

17. Oldendorf WH. Measurement of brain uptake of radiolabeled substances

using a tritiated water internal standard. Brain Res 24: 372-376, 1970.

18. Foster KA, Mellick GD, Weiss M and Roberts MS. An isolated in-situ rat

head perfusion model for pharmacokinetic studies. Pharm Res 17: 127-

134, 2000.

19. Ungerstedt U. Microdialysis--principles and applications for studies in

animals and man. J Intern Med 230: 365-373, 1991.

20. Meulemans A, Vicart P, Mohler J, Henzel D and Vulpillat M.

Permeability of nitroimidazoles in rat cortex. Chemotherapy 32: 486-493,

1986.

200 21. Ebert D, Haller RG and Walton ME. Energy contribution of octanoate to

intact rat brain metabolism measured by 13C nuclear magnetic resonance

spectroscopy. J Neurosci 23: 5928-5935, 2003.

22. Ando S and Tanaka Y. Mass spectrometric studies on brain metabolism,

using stable isotopes. Mass Spectrom Rev 24: 865-886, 2005.

23. Webb S, Ott RJ and Cherry SR. Quantitation of blood-brain barrier

permeability by positron emission tomography. Phys Med Biol 34: 1767-

1771, 1989.

24. Ross BD. Perfusion techniques in biochemistry. London: Oxford

University Press, 1972.

25. Hruz PW and Mueckler MM. Structural analysis of the GLUT1 facilitative

glucose transporter (review). Mol Membr Biol 18: 183-193, 2001.

26. Bell GI, Kayano T, Buse JB, Burant CF, Takeda J, Lin D, Fukumoto H

and Seino S. Molecular biology of mammalian glucose transporters.

Diabetes Care 13: 198-208, 1990.

27. Van den Berg C. Energetics and Human Information Processing. Boston:

Nijhoff, 1986.

201 28. Kimelberg HK and Norenberg MD. Astrocytes. Sci Am 260: 66-72, 74,

76, 1989.

29. Tower DB and Young OM. The activities of butyrylcholinesterase and

carbonic anhydrase, the rate of anaerobic glycolysis, and the question of a

constant density of glial cells in cerebral cortices of various mammalian

species from mouse to whale. J Neurochem 20: 269-278, 1973.

30. Saunders WB. The fine structure of the nervous system: the neurons and

supporting cells. Philadelphia: 1991.

31. Barres BA. New roles for glia. J Neurosci 11: 3685-3694, 1991.

32. Brookes N and Yarowsky PJ. Determinants of deoxyglucose uptake in

cultured astrocytes: the role of the sodium pump. J Neurochem 44: 473-

479, 1985.

33. Kimelberg HK, Jalonen T and Walz W. Regulation of the brain

microenvironment: transmitters and ions. In: Astrocytes: pharmacology

and function, edited by Murphy S. San Diego: Academic Press, 1993, p.

193.

202 34. Poitry-Yamate CL and Tsacopoulos M. Glucose metabolism in freshly

isolated Muller glial cells from a mammalian retina. J Comp Neurol 320:

257-266, 1992.

35. Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS,

Pettigrew KD, Sakurada O and Shinohara M. The [14C]deoxyglucose

method for the measurement of local cerebral glucose utilization: theory,

procedure, and normal values in the conscious and anesthetized albino rat.

J Neurochem 28: 897-916, 1977.

36. Kadekaro M, Crane AM and Sokoloff L. Differential effects of electrical

stimulation of sciatic nerve on metabolic activity in spinal cord and dorsal

root ganglion in the rat. Proc Natl Acad Sci U S A 82: 6010-6013, 1985.

37. Muller HW, Beckh S and Seifert W. Neurotrophic factor for central

neurons. Proc Natl Acad Sci U S A 81: 1248-1252, 1984.

38. Selak I, Skaper SD and Varon S. Pyruvate participation in the low

molecular weight trophic activity for central nervous system neurons in

glia-conditioned media. J Neurosci 5: 23-28, 1985.

39. Pellerin L and Magistretti PJ. How to balance the brain energy budget

while spending glucose differently. J Physiol 546: 325, 2003.

203 40. Vega C, Martiel JL, Drouhault D, Burckhart MF and Coles JA. Uptake

of locally applied deoxyglucose, glucose and lactate by axons and

Schwann cells of rat vagus nerve. J Physiol 546: 551-564, 2003.

41. Bagley PR, Tucker SP, Nolan C, Lindsay JG, Davies A, Baldwin SA,

Cremer JE and Cunningham VJ. Anatomical mapping of glucose

transporter protein and pyruvate dehydrogenase in rat brain: an

immunogold study. Brain Res 499: 214-224, 1989.

42. Milner TA, Aoki C, Sheu KF, Blass JP and Pickel VM. Light microscopic

immunocytochemical localization of pyruvate dehydrogenase complex in

rat brain: topographical distribution and relation to cholinergic and

catecholaminergic nuclei. J Neurosci 7: 3171-3190, 1987.

43. Bouzier-Sore AK, Voisin P, Canioni P, Magistretti PJ and Pellerin L.

Lactate is a preferential oxidative energy substrate over glucose for

neurons in culture. J Cereb Blood Flow Metab 23: 1298-1306, 2003.

44. Youdim KA, Martin A and Joseph JA. Essential fatty acids and the brain:

possible health implications. Int J Dev Neurosci 18: 383-399, 2000.

45. Edmond J. Essential polyunsaturated fatty acids and the barrier to the

brain: the components of a model for transport. J Mol Neurosci 16: 181-

193, 2001.

204 46. Edmond J, Higa TA, Korsak RA, Bergner EA and Lee WN. Fatty acid

transport and utilization for the developing brain. J Neurochem 70: 1227-

1234, 1998.

47. Hamilton JA and Kamp F. How are free fatty acids transported in

membranes? Is it by proteins or by free diffusion through the lipids?

Diabetes 48: 2255-2269, 1999.

48. Watkins PA, Hamilton JA, Leaf A, Spector AA, Moore SA, Anderson

RE, Moser HW, Noetzel MJ and Katz R. Brain uptake and utilization of

fatty acids: applications to peroxisomal biogenesis diseases. J Mol

Neurosci 16: 87-92, 2001.

49. Hui TY and Bernlohr DA. Fatty acid transporters in animal cells. Front

Biosci 2: d222-d231, 1997.

50. Moore SA, Yoder E and Spector AA. Role of the blood-brain barrier in

the formation of long-chain omega-3 and omega-6 fatty acids from

essential fatty acid precursors. J Neurochem 55: 391-402, 1990.

51. Chang MC, Arai T, Freed LM, Wakabayashi S, Channing MA, Dunn

BB, Der MG, Bell JM, Sasaki T, Herscovitch P, Eckelman WC and

Rapoport SI. Brain incorporation of [1-11C]arachidonate in normocapnic

205 and hypercapnic monkeys, measured with positron emission tomography.

Brain Res 755: 74-83, 1997.

52. Edmond J, Robbins RA, Bergstrom JD, Cole RA and de VJ. Capacity

for substrate utilization in oxidative metabolism by neurons, astrocytes,

and oligodendrocytes from developing brain in primary culture. J Neurosci

Res 18: 551-561, 1987.

53. Auestad N, Korsak RA, Morrow JW and Edmond J. Fatty acid oxidation

and ketogenesis by astrocytes in primary culture. J Neurochem 56: 1376-

1386, 1991.

54. Kuge Y, Hikosaka K, Seki K, Ohkura K, Nishijima K, Tsukamoto E and

Tamaki N. In vitro uptake of [1-14C]Octanoate in brain slices of rats: basic

studies for assessing [1-11C]Octanoate as a PET tracer of glial functions.

Nucl Med Biol 29: 303-306, 2002.

55. Kuge Y, Yajima K, Kawashima H, Yamazaki H, Hashimoto N and

Miyake Y. Brain uptake and metabolism of [1-11C]octanoate in rats:

pharmacokinetic basis for its application as a radiopharmaceutical for

studying brain fatty acid metabolism. Ann Nucl Med 9: 137-142, 1995.

56. Smith QR and Nagura H. Fatty acid uptake and incorporation in brain:

studies with the perfusion model. J Mol Neurosci 16: 167-172, 2001.

206 57. Edmond J, Korsak RA, Morrow JW, Torok-Both G and Catlin DH.

Dietary cholesterol and the origin of cholesterol in the brain of developing

rats. J Nutr 121: 1323-1330, 1991.

58. Lane MD, Wolfgang M, Cha SH and Dai Y. Regulation of food intake and

energy expenditure by hypothalamic malonyl-CoA. Int J Obes (Lond) 32

Suppl 4: S49-S54, 2008.

59. Dahlquist G and Persson B. The rate of cerebral utilization of glucose,

ketone bodies, and oxygen: a comparative in vivo study of infant and adult

rats. Pediatr Res 10: 910-917, 1976.

60. Cremer JE. Substrate utilization and brain development. J Cereb Blood

Flow Metab 2: 394-407, 1982.

61. Hawkins RA, Mans AM and Davis DW. Regional ketone body utilization

by rat brain in starvation and diabetes. Am J Physiol 250: E169-E178,

1986.

62. Haymond MW, Karl IE, Clarke WL, Pagliara AS and Santiago JV.

Differences in circulating gluconeogenic substrates during short-term

fasting in men, women, and children. Metabolism 31: 33-42, 1982.

207 63. Hawkins RA, Williamson DH and Krebs HA. Ketone-body utilization by

adult and suckling rat brain in vivo. Biochem J 122: 13-18, 1971.

64. Moore TJ, Lione AP, Sugden MC and Regen DM. Beta-hydroxybutyrate

transport in rat brain: developmental and dietary modulations. Am J

Physiol 230: 619-630, 1976.

65. Gjedde A and Crone C. Induction processes in blood-brain transfer of

ketone bodies during starvation. Am J Physiol 229: 1165-1169, 1975.

66. Webber RJ and Edmond J. Utilization of L(+)-3-hydroxybutyrate, D(-)-3-

hydroxybutyrate, acetoacetate, and glucose for respiration and lipid

synthesis in the 18-day-old rat. J Biol Chem 252: 5222-5226, 1977.

67. Koper JW, Lopes-Cardozo M and Van Golde LM. Preferential utilization

of ketone bodies for the synthesis of myelin cholesterol in vivo. Biochim

Biophys Acta 666: 411-417, 1981.

68. Morris AA. Cerebral ketone body metabolism. J Inherit Metab Dis 28:

109-121, 2005.

69. Auestad N, Fisher R, Chiappelli F, Korsak RA and Edmond J. Growth

and development of brain of artificially reared hypoketonemic rat pups.

Proc Soc Exp Biol Med 195: 335-344, 1990.

208 70. Leino RL, Gerhart DZ and Drewes LR. Monocarboxylate transporter

(MCT1) abundance in brains of suckling and adult rats: a quantitative

electron microscopic immunogold study. Brain Res Dev Brain Res 113:

47-54, 1999.

71. Leino RL, Gerhart DZ, Duelli R, Enerson BE and Drewes LR. Diet-

induced ketosis increases monocarboxylate transporter (MCT1) levels in

rat brain. Neurochem Int 38: 519-527, 2001.

72. Halestrap AP and Price NT. The proton-linked monocarboxylate

transporter (MCT) family: structure, function and regulation. Biochem J

343 Pt 2: 281-299, 1999.

73. Hasselbalch SG, Knudsen GM, Jakobsen J, Hageman LP, Holm S and

Paulson OB. Blood-brain barrier permeability of glucose and ketone

bodies during short-term starvation in humans. Am J Physiol 268: E1161-

E1166, 1995.

74. Pan JW, Rothman TL, Behar KL, Stein DT and Hetherington HP.

Human brain beta-hydroxybutyrate and lactate increase in fasting-induced

ketosis. J Cereb Blood Flow Metab 20: 1502-1507, 2000.

209 75. Pan JW, Telang FW, Lee JH, de Graaf RA, Rothman DL, Stein DT and

Hetherington HP. Measurement of beta-hydroxybutyrate in acute

hyperketonemia in human brain. J Neurochem 79: 539-544, 2001.

76. Pellerin L, Pellegri G, Martin JL and Magistretti PJ. Expression of

monocarboxylate transporter mRNAs in mouse brain: support for a distinct

role of lactate as an energy substrate for the neonatal vs. adult brain. Proc

Natl Acad Sci U S A 95: 3990-3995, 1998.

77. Bixel MG and Hamprecht B. Generation of ketone bodies from leucine

by cultured astroglial cells. J Neurochem 65: 2450-2461, 1995.

78. Buckley BM and Williamson DH. Acetoacetate and brain lipogenesis:

developmental pattern of acetoacetyl-coenzyme A synthetase in the

soluble fraction of rat brain. Biochem J 132: 653-656, 1973.

79. Middleton B. The acetoacetyl-coenzyme A thiolases of rat brain and their

relative activities during postnatal development. Biochem J 132: 731-737,

1973.

80. Page MA, Krebs HA and Williamson DH. Activities of enzymes of

ketone-body utilization in brain and other tissues of suckling rats. Biochem

J 121: 49-53, 1971.

210 81. Freeman JM, Vining EP, Pillas DJ, Pyzik PL, Casey JC and Kelly LM.

The efficacy of the ketogenic diet-1998: a prospective evaluation of

intervention in 150 children. Pediatrics 102: 1358-1363, 1998.

82. Likhodii SS, Serbanescu I, Cortez MA, Murphy P, Snead OC, III and

Burnham WM. Anticonvulsant properties of acetone, a brain ketone

elevated by the ketogenic diet. Ann Neurol 54: 219-226, 2003.

83. Niesen CE and Ge S. Chronic epilepsy in developing hippocampal

neurons: electrophysiologic and morphologic features. Dev Neurosci 21:

328-338, 1999.

84. Bittigau P and Ikonomidou C. Glutamate in neurologic diseases. J Child

Neurol 12: 471-485, 1997.

85. Olney JW. Excitotoxicity: an overview. Can Dis Wkly Rep 16 Suppl 1E:

47-57, 1990.

86. DeVivo DC, Leckie MP, Ferrendelli JS and McDougal DB, Jr. Chronic

ketosis and cerebral metabolism. Ann Neurol 3: 331-337, 1978.

87. Erecinska M, Nelson D, Daikhin Y and Yudkoff M. Regulation of GABA

level in rat brain synaptosomes: fluxes through enzymes of the GABA

211 shunt and effects of glutamate, calcium, and ketone bodies. J Neurochem

67: 2325-2334, 1996.

88. Mochel F, DeLonlay P, Touati G, Brunengraber H, Kinman RP, Rabier

D, Roe CR and Saudubray JM. Pyruvate carboxylase deficiency: clinical

and biochemical response to anaplerotic diet therapy. Mol Genet Metab

84: 305-312, 2005.

89. Ashworth JM and Kornberg HL. The anaplerotic fixation of carbon

dioxide by Escherichia coli. Proc R Soc Lond B Biol Sci 165: 179-188,

1966.

90. Hassel B. Carboxylation and anaplerosis in neurons and glia. Mol

Neurobiol 22: 21-40, 2000.

91. Brunengraber H and Roe CR. Anaplerotic molecules: current and future.

J Inherit Metab Dis 29: 327-331, 2006.

92. Owen OE, Kalhan SC and Hanson RW. The key role of anaplerosis and

cataplerosis for citric acid cycle function. J Biol Chem 277: 30409-30412,

2002.

212 93. Macdonald MJ. Feasibility of a mitochondrial pyruvate malate shuttle in

pancreatic islets. Further implication of cytosolic NADPH in insulin

secretion. J Biol Chem 270: 20051-20058, 1995.

94. Chen LC, Boissonneault G, Hayek MG and Chow CK. Dietary fat

effects on hepatic lipid peroxidation and enzymes of H2O2 metabolism

and NADPH generation. Lipids 28: 657-662, 1993.

95. Jitrapakdee S, Vidal-Puig A and Wallace JC. Anaplerotic roles of

pyruvate carboxylase in mammalian tissues. Cell Mol Life Sci 63: 843-854,

2006.

96. Rothman DL, Magnusson I, Katz LD, Shulman RG and Shulman GI.

Quantitation of hepatic and gluconeogenesis in fasting

humans with 13C NMR. Science 254: 573-576, 1991.

97. Russell RR, III and Taegtmeyer H. Pyruvate carboxylation prevents the

decline in contractile function of rat hearts oxidizing acetoacetate. Am J

Physiol 261: H1756-H1762, 1991.

98. Russell RR, III and Taegtmeyer H. Changes in citric acid cycle flux and

anaplerosis antedate the functional decline in isolated rat hearts utilizing

acetoacetate. J Clin Invest 87: 384-390, 1991.

213 99. Reszko AE, Kasumov T, Pierce BA, David F, Hoppel CL, Stanley WC,

Des Rosiers C and Brunengraber H. Assessing the reversibility of the

anaplerotic reactions of the propionyl-CoA pathway in heart and liver. J

Biol Chem 278: 34959-34965, 2003.

100. Cairns CB, Ferroggiaro AA, Walther JM, Harken AH and Banerjee A.

Postischemic administration of succinate reverses the impairment of

oxidative phosphorylation after cardiac ischemia and reperfusion injury.

Circulation 96: II-5, 1997.

101. Sumegi B, Podanyi B, Forgo P and Kover KE. Metabolism of [3-

13C]pyruvate and [3-13C]propionate in normal and ischaemic rat heart in

vivo: 1H- and 13C-NMR studies. Biochem J 312 ( Pt 1): 75-81, 1995.

102. Di LF, Menabo R, Barbato R and Siliprandi N. Contrasting effects of

propionate and propionyl-L-carnitine on energy-linked processes in

ischemic hearts. Am J Physiol 267: H455-H461, 1994.

103. Liedtke AJ, Hacker T, Renstrom B and Nellis SH. Anaplerotic effects of

propionate on oxidations of acetate and long-chain fatty acids. Am J

Physiol 270: H2197-H2203, 1996.

214 104. Bunger R, Swindall B, Brodie D, Zdunek D, Stiegler H and Walter G.

Pyruvate attenuation of hypoxia damage in isolated working guinea-pig

heart. J Mol Cell Cardiol 18: 423-438, 1986.

105. Bunger R, Mallet RT and Hartman DA. Pyruvate-enhanced

phosphorylation potential and inotropism in normoxic and postischemic

isolated working heart. Near-complete prevention of reperfusion

contractile failure. Eur J Biochem 180: 221-233, 1989.

106. Tejero-Taldo MI, Caffrey JL, Sun J and Mallet RT. Antioxidant

properties of pyruvate mediate its potentiation of beta-adrenergic

inotropism in stunned myocardium. J Mol Cell Cardiol 31: 1863-1872,

1999.

107. Laplante A, Vincent G, Poirier M and Des Rosiers C. Effects and

metabolism of fumarate in the perfused rat heart. A 13C mass isotopomer

study. Am J Physiol 272: E74-E82, 1997.

108. Bittl JA and Shine KI. Protection of ischemic rabbit myocardium by

glutamic acid. Am J Physiol 245: H406-H412, 1983.

109. Gibala MJ, Young ME and Taegtmeyer H. Anaplerosis of the citric acid

cycle: role in energy metabolism of heart and skeletal muscle

16. Acta Physiol Scand 168: 657-665, 2000.

215 110. Constantin-Teodosiu D, Simpson EJ and Greenhaff PL. The

importance of pyruvate availability to PDC activation and anaplerosis in

human skeletal muscle

21. Am J Physiol 276: E472-E478, 1999.

111. Gibala MJ, MacLean DA, Graham TE and Saltin B. Anaplerotic

processes in human skeletal muscle during brief dynamic exercise. J

Physiol 502 ( Pt 3): 703-713, 1997.

112. Gibala MJ, MacLean DA, Graham TE and Saltin B. Tricarboxylic acid

cycle intermediate pool size and estimated cycle flux in human muscle

during exercise. Am J Physiol 275: E235-E242, 1998.

113. Sahlin K, Katz A and Broberg S. Tricarboxylic acid cycle intermediates

in human muscle during prolonged exercise. Am J Physiol 259: C834-

C841, 1990.

114. Gibala MJ and Saltin B. PDH activation by dichloroacetate reduces TCA

cycle intermediates at rest but not during exercise in humans. Am J

Physiol 277: E33-E38, 1999.

115. Gibala MJ. Anaplerosis of the muscle tricarboxylic acid cycle pool during

contraction: does size matter?

7. J Physiol 548: 334, 2003.

216 116. Walton ME, Ebert D and Haller RG. Relative rates of anaplerotic flux in

rested and contracted rat skeletal muscle measured by 13C NMR

spectroscopy

3. J Physiol 548: 541-548, 2003.

117. Sibson NR, Mason GF, Shen J, Cline GW, Herskovits AZ, Wall JE,

Behar KL, Rothman DL and Shulman RG. In vivo (13)C NMR

measurement of neurotransmitter glutamate cycling, anaplerosis and TCA

cycle flux in rat brain during. J Neurochem 76: 975-989, 2001.

118. Hassel B and Brathe A. Cerebral metabolism of lactate in vivo: evidence

for neuronal pyruvate carboxylation. J Cereb Blood Flow Metab 20: 327-

336, 2000.

119. Yu AC, Drejer J, Hertz L and Schousboe A. Pyruvate carboxylase

activity in primary cultures of astrocytes and neurons. J Neurochem 41:

1484-1487, 1983.

120. Moldave K, Winzler RJ and Pearson HE. The incorporation in vitro of

C14 into amino acids of control and virus-infected mouse brain. J Biol

Chem 200: 357-365, 1953.

121. Cran RK and Ball EG. Relationship of C14o2 fixation to carbohydrate

metabolism in retina. J Biol Chem 189: 269-276, 1951.

217 122. Reactome. Cold Spring Harbor Laboratory, The European Bioinformatics

Institute, and The Gene Ontology Consortium . 9-9-2008.

Ref Type: Internet Communication

123. Jitrapakdee S, Booker GW, Cassady AI and Wallace JC. Cloning,

sequencing and expression of rat liver pyruvate carboxylase. Biochem J

316 ( Pt 2): 631-637, 1996.

124. GenomeNet. Kyoto University Bioinformatics Center . 9-9-2008.

Ref Type: Internet Communication

125. Waelsch H, Berl HW, Rossi CA, Clarke DD and Purpura DP.

Quantitative aspects of CO2 fixation in mamalian brain in vivo. J

Neurochem 11: 717-728, 1964.

126. Martinez-Hernandez A, Bell KP and Norenberg MD. Glutamine

synthetase: glial localization in brain. Science 195: 1356-1358, 1977.

127. Tansey FA, Farooq M and Cammer W. Glutamine synthetase in

oligodendrocytes and astrocytes: new biochemical and

immunocytochemical evidence. J Neurochem 56: 266-272, 1991.

218 128. Hassel B, Sonnewald U, Unsgard G and Fonnum F. NMR spectroscopy

of cultured astrocytes: effects of glutamine and the gliotoxin fluorocitrate. J

Neurochem 62: 2187-2194, 1994.

129. Hassel B, Sonnewald U and Fonnum F. Glial-neuronal interactions as

studied by cerebral metabolism of [2-13C]acetate and [1-13C]glucose: an

ex vivo 13C NMR spectroscopic study. J Neurochem 64: 2773-2782, 1995.

130. Bradford HF, Ward HK and Thomas AJ. Glutamine--a major substrate

for nerve endings. J Neurochem 30: 1453-1459, 1978.

131. Zielke HR, Collins RM, Jr., Baab PJ, Huang Y, Zielke CL and Tildon JT.

Compartmentation of [14C]glutamate and [14C]glutamine oxidative

metabolism in the rat hippocampus as determined by microdialysis. J

Neurochem 71: 1315-1320, 1998.

132. Berl S, Takagaki G, Clarke DD and Waelsch H. Metabolic compartments

in vivo. Ammonia and glutamic acid metabolism in brain and liver. J Biol

Chem 237: 2562-2569, 1962.

133. Lee WJ, Hawkins RA, Vina JR and Peterson DR. Glutamine transport

by the blood-brain barrier: a possible mechanism for nitrogen removal. Am

J Physiol 274: C1101-C1107, 1998.

219 134. Wee NL, Tng YY, Cheng HT, Lee SM, Chew SF and Ip YK. Ammonia

toxicity and tolerance in the brain of the African sharptooth catfish, Clarias

gariepinus. Aquat Toxicol 82: 204-213, 2007.

135. Grill V, Bjorkman O, Gutniak M and Lindqvist M. Brain uptake and

release of amino acids in nondiabetic and insulin-dependent diabetic

subjects: important role of glutamine release for nitrogen balance.

Metabolism 41: 28-32, 1992.

136. Dringen R. Metabolism and functions of glutathione in brain. Prog

Neurobiol 62: 649-671, 2000.

137. Pressman AH. Glutathione: the Ultimate Antioxidant. New York: St.

Martin's Press, 1997.

138. Sonnewald U, Westergaard N, Krane J, Unsgard G, Petersen SB and

Schousboe A. First direct demonstration of preferential release of citrate

from astrocytes using [13C]NMR spectroscopy of cultured neurons and

astrocytes. Neurosci Lett 128: 235-239, 1991.

139. Westergaard N, Sonnewald U, Unsgard G, Peng L, Hertz L and

Schousboe A. Uptake, release, and metabolism of citrate in neurons and

astrocytes in primary cultures. J Neurochem 62: 1727-1733, 1994.

220 140. Hassel B, Westergaard N, Schousboe A and Fonnum F. Metabolic

differences between primary cultures of astrocytes and neurons from

cerebellum and cerebral cortex. Effects of fluorocitrate. Neurochem Res

20: 413-420, 1995.

141. Cheng SC and Nakamura R. Metabolism related to the tricarboxylic acid

cycle in rat brain slices. Observations on CO 2 fixation and metabolic

compartmentation. Brain Res 38: 355-370, 1972.

142. Hassel B and Brathe A. Neuronal pyruvate carboxylation supports

formation of transmitter glutamate. J Neurosci 20: 1342-1347, 2000.

143. Vogel R, Jennemann G, Seitz J, Wiesinger H and Hamprecht B.

Mitochondrial malic enzyme: purification from bovine brain, generation of

an antiserum, and immunocytochemical localization in neurons of rat brain.

J Neurochem 71: 844-852, 1998.

144. Cruz F, Scott SR, Barroso I, Santisteban P and Cerdan S. Ontogeny

and cellular localization of the pyruvate recycling system in rat brain. J

Neurochem 70: 2613-2619, 1998.

145. McKenna MC, Stevenson JH, Huang X, Tildon JT, Zielke CL and

Hopkins IB. Mitochondrial malic enzyme activity is much higher in

mitochondria from cortical synaptic terminals compared with mitochondria

221 from primary cultures of cortical neurons or cerebellar granule cells.

Neurochem Int 36: 451-459, 2000.

146. Nye CK, Hanson RW and Kalhan SC. Glyceroneogenesis is the

dominant pathway for triglyceride glycerol synthesis in vivo in the rat. J

Biol Chem 283: 27565-27574, 2008.

147. Kalhan SC, Bugianesi E, McCullough AJ, Hanson RW and Kelley DE.

Estimates of hepatic glyceroneogenesis in type 2 diabetes mellitus in

humans. Metabolism 57: 305-312, 2008.

148. Reshef L, Olswang Y, Cassuto H, Blum B, Croniger CM, Kalhan SC,

Tilghman SM and Hanson RW. Glyceroneogenesis and the

triglyceride/fatty acid cycle. J Biol Chem 278: 30413-30416, 2003.

149. Kalhan SC, Mahajan S, Burkett E, Reshef L and Hanson RW.

Glyceroneogenesis and the source of glycerol for hepatic triacylglycerol

synthesis in humans. J Biol Chem 276: 12928-12931, 2001.

150. Nye CK, Hanson RW and Kalhan SC. Glyceroneogenesis is the

dominant pathway for triglyceride glycerol synthesis in vivo in the rat. J

Biol Chem 283: 27565-27574, 2008.

222 151. Schoonjans K, Peinado-Onsurbe J, Lefebvre AM, Heyman RA, Briggs

M, Deeb S, Staels B and Auwerx J. PPARalpha and PPARgamma

activators direct a distinct tissue-specific transcriptional response via a

PPRE in the lipoprotein lipase gene. EMBO J 15: 5336-5348, 1996.

152. Cline GW, Lepine RL, Papas KK, Kibbey RG and Shulman GI. 13C

NMR isotopomer analysis of anaplerotic pathways in INS-1 cells. J Biol

Chem 279: 44370-44375, 2004.

153. Yang S, Fransson U, Fagerhus L, Holst LS, Hohmeier HE, Renstrom

E and Mulder H. Enhanced cAMP protein kinase A signaling determines

improved insulin secretion in a clonal insulin-producing beta-cell line (INS-

1 832/13). Mol Endocrinol 18: 2312-2320, 2004.

154. Cerf ME, Williams K, Chapman CS and Louw J. Compromised beta-cell

development and beta-cell dysfunction in weanling offspring from dams

maintained on a high-fat diet during gestation. Pancreas 34: 347-353,

2007.

155. Newgard CB and McGarry JD. Metabolic coupling factors in pancreatic

beta-cell signal transduction. Annu Rev Biochem 64: 689-719, 1995.

156. Farfari S, Schulz V, Corkey B and Prentki M. Glucose-regulated

anaplerosis and cataplerosis in pancreatic beta-cells: possible implication

223 of a pyruvate/citrate shuttle in insulin secretion. Diabetes 49: 718-726,

2000.

157. Povey S, Wilson DE, Jr., Harris H, Gormley IP, Perry P and Buckton

KE. Sub-unit structure of soluble and mitochondrial malic enzyme:

demonstration of human mitochondrial enzyme in human-mouse hybrids.

Ann Hum Genet 39: 203-212, 1975.

158. Turumi R, Takeda K and Tonouchi A. Characteristics and Propionate

Production of Propionibacterium Isolated from a Methane fermentation

Digester. Microbes and Environments 15: 151-159, 2000.

159. Kasumov T, Cendrowski AV, David F, Jobbins KA, Anderson VE and

Brunengraber H. Mass isotopomer study of anaplerosis from propionate

in the perfused rat heart. Arch Biochem Biophys 463: 110-117, 2007.

160. Ando T, Rasmussen K, Nyhan WL and Hull D. 3-hydroxypropionate:

significance of -oxidation of propionate in patients with propionic acidemia

and methylmalonic acidemia. Proc Natl Acad Sci U S A 69: 2807-2811,

1972.

161. Feliz B, Witt DR and Harris BT. Propionic acidemia: a neuropathology

case report and review of prior cases. Arch Pathol Lab Med 127: e325-

e328, 2003.

224 162. Nguyen NH, Morland C, Gonzalez SV, Rise F, Storm-Mathisen J,

Gundersen V and Hassel B. Propionate increases neuronal histone

acetylation, but is metabolized oxidatively by glia. Relevance for propionic

acidemia. J Neurochem 101: 806-814, 2007.

163. Davie JR and Hendzel MJ. Multiple functions of dynamic histone

acetylation. J Cell Biochem 55: 98-105, 1994.

164. Roe CR, Sweetman L, Roe DS, David F and Brunengraber H.

Treatment of cardiomyopathy and rhabdomyolysis in long-chain fat

oxidation disorders using an anaplerotic odd-chain triglyceride. J Clin

Invest 110: 259-269, 2002.

165. Ardawi MS and Newsholme EA. Maximum activities of some enzymes of

glycolysis, the tricarboxylic acid cycle and ketone-body and glutamine

utilization pathways in lymphocytes of the rat. Biochem J 208: 743-748,

1982.

166. Elhamri M, Martin M, Ferrier B and Baverel G. Substrate uptake and

utilization by the kidney of fed and starved rats in vivo. Ren Physiol

Biochem 16: 311-324, 1993.

225 167. Ferrier B, Martin M, Janbon B and Baverel G. Transport of beta-

hydroxybutyrate and acetoacetate along rat nephrons: a micropuncture

study. Am J Physiol 262: F762-F769, 1992.

168. Roe CR, Yang BZ, Brunengraber H, Roe DS, Wallace M and Garritson

BK. Carnitine palmitoyltransferase II deficiency: successful anaplerotic

diet therapy. Neurology 71: 260-264, 2008.

169. Das UN. Hypertension as a low-grade systemic inflammatory condition

that has its origins in the perinatal period. J Assoc Physicians India 54:

133-142, 2006.

170. Booker CS and Mann JI. Trans fatty acids and cardiovascular health:

translation of the evidence base. Nutr Metab Cardiovasc Dis 18: 448-456,

2008.

171. van d, V, van BM, Glatz JF, Hasselbaink DM and Luiken JJ. Critical

steps in cellular fatty acid uptake and utilization. Mol Cell Biochem 239: 9-

15, 2002.

172. Stremmel W, Strohmeyer G, Borchard F, Kochwa S and Berk PD.

Isolation and partial characterization of a fatty acid binding protein in rat

liver plasma membranes. Proc Natl Acad Sci U S A 82: 4-8, 1985.

226 173. Schaffer JE. Fatty acid transport: the roads taken. Am J Physiol

Endocrinol Metab 282: E239-E246, 2002.

174. Kolodziej MP, Crilly PJ, Corstorphine CG and Zammit VA.

Development and characterization of a polyclonal antibody against rat liver

mitochondrial overt carnitine palmitoyltransferase (CPT I). Distinction of

CPT I from CPT II and of isoforms of CPT I in different tissues. Biochem J

282 ( Pt 2): 415-421, 1992.

175. Price N, van der LF, Jackson V, Corstorphine C, Thomson R,

Sorensen A and Zammit V. A novel brain-expressed protein related to

carnitine palmitoyltransferase I. Genomics 80: 433-442, 2002.

176. Jeremy M.Berg, John L.Tymoczko, Lubert Stryer and Neil D.Clarke.

Biochemistry. Michelle Julet, 2002.

177. Forster ME and Staib W. Beta-oxidation as channeled reaction linked to

citric acid cycle: evidence from measurements of mitochondrial pyruvate

oxidation during fatty acid degradation. Int J Biochem 24: 1111-1116,

1992.

178. Kamijo T, Aoyama T, Miyazaki J and Hashimoto T. Molecular cloning of

the cDNAs for the subunits of rat mitochondrial fatty acid beta-oxidation

multienzyme complex. Structural and functional relationships to other

227 mitochondrial and peroxisomal beta-oxidation enzymes. J Biol Chem 268:

26452-26460, 1993.

179. Ishikawa M, Tsuchiya D, Oyama T, Tsunaka Y and Morikawa K.

Structural basis for channelling mechanism of a fatty acid beta-oxidation

multienzyme complex. EMBO J 23: 2745-2754, 2004.

180. McGarry JD, Takabayashi Y and Foster DW. The role of malonyl-coa in

the coordination of fatty acid synthesis and oxidation in isolated rat

hepatocytes. J Biol Chem 253: 8294-8300, 1978.

181. Hardie DG. Regulation of fatty acid synthesis via phosphorylation of

acetyl-CoA carboxylase. Prog Lipid Res 28: 117-146, 1989.

182. Thampy KG. Formation of malonyl coenzyme A in rat heart. Identification

and purification of an isozyme of A carboxylase from rat heart. J Biol

Chem 264: 17631-17634, 1989.

183. bu-Elheiga L, marza-Ortega DB, Baldini A and Wakil SJ. Human

acetyl-CoA carboxylase 2. Molecular cloning, characterization,

chromosomal mapping, and evidence for two isoforms. J Biol Chem 272:

10669-10677, 1997.

228 184. Ha J, Lee JK, Kim KS, Witters LA and Kim KH. Cloning of human

acetyl-CoA carboxylase-beta and its unique features. Proc Natl Acad Sci

U S A 93: 11466-11470, 1996.

185. Kim KH. Regulation of mammalian acetyl-coenzyme A carboxylase. Annu

Rev Nutr 17: 77-99, 1997.

186. bu-Elheiga L, Brinkley WR, Zhong L, Chirala SS, Woldegiorgis G and

Wakil SJ. The subcellular localization of acetyl-CoA carboxylase 2. Proc

Natl Acad Sci U S A 97: 1444-1449, 2000.

187. Kerner J and Hoppel CL. Radiochemical malonyl-CoA decarboxylase

assay: activity and subcellular distribution in heart and skeletal muscle.

Anal Biochem 306: 283-289, 2002.

188. Sacksteder KA, Morrell JC, Wanders RJ, Matalon R and Gould SJ.

MCD encodes peroxisomal and cytoplasmic forms of malonyl-CoA

decarboxylase and is mutated in malonyl-CoA decarboxylase deficiency. J

Biol Chem 274: 24461-24468, 1999.

189. Dyck JR, Barr AJ, Barr RL, Kolattukudy PE and Lopaschuk GD.

Characterization of cardiac malonyl-CoA decarboxylase and its putative

role in regulating fatty acid oxidation. Am J Physiol 275: H2122-H2129,

1998.

229 190. Joly E, Bendayan M, Roduit R, Saha AK, Ruderman NB and Prentki M.

Malonyl-CoA decarboxylase is present in the cytosolic, mitochondrial and

peroxisomal compartments of rat hepatocytes. FEBS Lett 579: 6581-6586,

2005.

191. McGarry JD and Brown NF. The mitochondrial carnitine

palmitoyltransferase system. From concept to molecular analysis. Eur J

Biochem 244: 1-14, 1997.

192. Ontko JA. Metabolism of free fatty acids in isolated liver cells. Factors

affecting the partition between esterification and oxidation. J Biol Chem

247: 1788-1800, 1972.

193. McGarry JD, Meier JM and Foster DW. The effects of starvation and

refeeding on carbohydrate and lipid metabolism in vivo and in the

perfused rat liver. The relationship between fatty acid oxidation and

esterification in the regulation of ketogenesis. J Biol Chem 248: 270-278,

1973.

194. Oram JF, Bennetch SL and Neely JR. Regulation of fatty acid utilization

in isolated perfused rat hearts. J Biol Chem 248: 5299-5309, 1973.

230 195. Longnus SL, Wambolt RB, Barr RL, Lopaschuk GD and Allard MF.

Regulation of myocardial fatty acid oxidation by substrate supply. Am J

Physiol Heart Circ Physiol 281: H1561-H1567, 2001.

196. Degerman E, Belfrage P and Manganiello VC. Structure, localization,

and regulation of cGMP-inhibited phosphodiesterase (PDE3). J Biol Chem

272: 6823-6826, 1997.

197. Eaton S. Control of mitochondrial beta-oxidation flux. Prog Lipid Res 41:

197-239, 2002.

198. Zammit VA, Fraser F and Orstorphine CG. Regulation of mitochondrial

outer-membrane carnitine palmitoyltransferase (CPT I): role of membrane-

topology. Adv Enzyme Regul 37: 295-317, 1997.

199. Park EA, Mynatt RL, Cook GA and Kashfi K. Insulin regulates enzyme

activity, malonyl-CoA sensitivity and mRNA abundance of hepatic

carnitine palmitoyltransferase-I. Biochem J 310 ( Pt 3): 853-858, 1995.

200. bu-Elheiga L, Matzuk MM, bo-Hashema KA and Wakil SJ. Continuous

fatty acid oxidation and reduced fat storage in mice lacking acetyl-CoA

carboxylase 2. Science 291: 2613-2616, 2001.

231 201. Poirier M, Vincent G, Reszko AE, Bouchard B, Kelleher JK,

Brunengraber H and Des Rosiers C. Probing the link between citrate

and malonyl-CoA in perfused rat hearts. Am J Physiol Heart Circ Physiol

283: H1379-H1386, 2002.

202. Stapleton D, Mitchelhill KI, Gao G, Widmer J, Michell BJ, Teh T,

House CM, Fernandez CS, Cox T, Witters LA and Kemp BE.

Mammalian AMP-activated protein kinase subfamily. J Biol Chem 271:

611-614, 1996.

203. Woods A, Cheung PC, Smith FC, Davison MD, Scott J, Beri RK and

Carling D. Characterization of AMP-activated protein kinase beta and

gamma subunits. Assembly of the heterotrimeric complex in vitro. J Biol

Chem 271: 10282-10290, 1996.

204. Salt IP, Johnson G, Ashcroft SJ and Hardie DG. AMP-activated protein

kinase is activated by low glucose in cell lines derived from pancreatic

beta cells, and may regulate insulin release. Biochem J 335 ( Pt 3): 533-

539, 1998.

205. Hardie DG, Salt IP, Hawley SA and Davies SP. AMP-activated protein

kinase: an ultrasensitive system for monitoring cellular energy charge.

Biochem J 338 ( Pt 3): 717-722, 1999.

232 206. Shaw RJ, Kosmatka M, Bardeesy N, Hurley RL, Witters LA, DePinho

RA and Cantley LC. The tumor suppressor LKB1 kinase directly activates

AMP-activated kinase and regulates apoptosis in response to energy

stress. Proc Natl Acad Sci U S A 101: 3329-3335, 2004.

207. Davies SP, Helps NR, Cohen PT and Hardie DG. 5'-AMP inhibits

dephosphorylation, as well as promoting phosphorylation, of the AMP-

activated protein kinase. Studies using bacterially expressed human

protein phosphatase-2C alpha and native bovine protein phosphatase-

2AC. FEBS Lett 377: 421-425, 1995.

208. Viollet B, Mounier R, Leclerc J, Yazigi A, Foretz M and Andreelli F.

Targeting AMP-activated protein kinase as a novel therapeutic approach

for the treatment of metabolic disorders. Diabetes Metab 33: 395-402,

2007.

209. Abe K, Shinohara Y and Terada H. Isolation and characterization of

cDNA encoding rat heart type acetyl-CoA carboxylase. Biochim Biophys

Acta 1398: 347-352, 1998.

210. Djouadi F, Weinheimer CJ and Kelly DP. The role of PPAR alpha as a

"lipostat" transcription factor. Adv Exp Med Biol 466: 211-220, 1999.

233 211. Motojima K, Passilly P, Peters JM, Gonzalez FJ and Latruffe N.

Expression of putative fatty acid transporter genes are regulated by

peroxisome proliferator-activated receptor alpha and gamma activators in

a tissue- and inducer-specific manner. J Biol Chem 273: 16710-16714,

1998.

212. Wahli W, Braissant O and Desvergne B. Peroxisome proliferator

activated receptors: transcriptional regulators of adipogenesis, lipid

metabolism and more... Chem Biol 2: 261-266, 1995.

213. Dressel U, Allen TL, Pippal JB, Rohde PR, Lau P and Muscat GE. The

peroxisome proliferator-activated receptor beta/delta agonist, GW501516,

regulates the expression of genes involved in lipid catabolism and energy

uncoupling in skeletal muscle cells. Mol Endocrinol 17: 2477-2493, 2003.

214. Tanaka T, Yamamoto J, Iwasaki S, Asaba H, Hamura H, Ikeda Y,

Watanabe M, Magoori K, Ioka RX, Tachibana K, Watanabe Y,

Uchiyama Y, Sumi K, Iguchi H, Ito S, Doi T, Hamakubo T, Naito M,

Auwerx J, Yanagisawa M, Kodama T and Sakai J. Activation of

peroxisome proliferator-activated receptor delta induces fatty acid beta-

oxidation in skeletal muscle and attenuates metabolic syndrome. Proc Natl

Acad Sci U S A 100: 15924-15929, 2003.

234 215. Le MC, Cauzac M, Diradourian C, Perdereau D, Girard J, Burnol AF

and Pegorier JP. Fatty acids induce L-CPT I gene expression through a

PPARalpha-independent mechanism in rat hepatoma cells. J Nutr 135:

2313-2319, 2005.

216. Habinowski SA, Hirshman M, Sakamoto K, Kemp BE, Gould SJ,

Goodyear LJ and Witters LA. Malonyl-CoA decarboxylase is not a

substrate of AMP-activated protein kinase in rat fast-twitch skeletal muscle

or an islet cell line. Arch Biochem Biophys 396: 71-79, 2001.

217. Lee GY, Kim NH, Zhao ZS, Cha BS and Kim YS. Peroxisomal-

proliferator-activated receptor alpha activates transcription of the rat

hepatic malonyl-CoA decarboxylase gene: a key regulation of malonyl-

CoA level. Biochem J 378: 983-990, 2004.

218. Wang X, Stanley WC, Darrow CJ, Brunengraber H and Kasumov T.

Assay of the activity of malonyl-coenzyme A decarboxylase by gas

chromatography-mass spectrometry. Anal Biochem 363: 169-174, 2007.

219. Young ME, Goodwin GW, Ying J, Guthrie P, Wilson CR, Laws FA and

Taegtmeyer H. Regulation of cardiac and skeletal muscle malonyl-CoA

decarboxylase by fatty acids. Am J Physiol Endocrinol Metab 280: E471-

E479, 2001.

235 220. Cho YO and Leklem JE. In vivo evidence for a vitamin B-6 requirement in

carnitine synthesis. J Nutr 120: 258-265, 1990.

221. Johnston CS, Solomon RE and Corte C. Vitamin C depletion is

associated with alterations in blood histamine and plasma free carnitine in

adults. J Am Coll Nutr 15: 586-591, 1996.

222. Oram JF, Wenger JI and Neely JR. Regulation of long chain fatty acid

activation in heart muscle. J Biol Chem 250: 73-78, 1975.

223. Shepherd D, Yates DW and Garland PB. The relationship between the

rates of conversion of palmitate into citrate or acetoacetate and the acetyl-

coenzyme A content of rat-liver mitochondria. Biochem J 97: 38C-40C,

1965.

224. Bian F, Kasumov T, Jobbins KA, Minkler PE, Anderson VE, Kerner J,

Hoppel CL and Brunengraber H. Competition between acetate and

oleate for the formation of malonyl-CoA and mitochondrial acetyl-CoA in

the perfused rat heart. J Mol Cell Cardiol 41: 868-875, 2006.

225. Roe CR and Ding JH. Mitochondrial fatty acid oxidation disorders. In: The

Metabolic and Molecular Bases of Inherited Diseases, edited by Scriver

CR, Beaudet AL, Sly WS and Valle D. New York: McGraw-Hill, 2001, p.

2297-2326.

236 226. Coates PM. New developments in the diagnosis and investigation of

mitochondrial fatty acid oxidation disorders. Eur J Pediatr 153: S49-S56,

1994.

227. Aoyama T, Uchida Y, Kelley RI, Marble M, Hofman K, Tonsgard JH,

Rhead WJ and Hashimoto T. A novel disease with deficiency of

mitochondrial very-long-chain acyl-CoA dehydrogenase. Biochem Biophys

Res Commun 191: 1369-1372, 1993.

228. Kompare M and Rizzo WB. Mitochondrial fatty-acid oxidation disorders.

Semin Pediatr Neurol 15: 140-149, 2008.

229. Vianey-Saban C, Divry P, Brivet M, Nada M, Zabot MT, Mathieu M and

Roe C. Mitochondrial very-long-chain acyl-coenzyme A dehydrogenase

deficiency: clinical characteristics and diagnostic considerations in 30

patients. Clin Chim Acta 269: 43-62, 1998.

230. Naylor EW, Mosovich LL, Guthrie R, Evans JE and Tieckelmann H.

Intermittent non-ketotic dicarboxylic aciduria in two siblings with

hypoglycaemia: an apparent defect in beta-oxidation of fatty acids. J

Inherit Metab Dis 3: 19-24, 1980.

231. Brown-Harrison MC, Nada MA, Sprecher H, Vianey-Saban C,

Farquhar J, Jr., Gilladoga AC and Roe CR. Very long chain acyl-CoA

237 dehydrogenase deficiency: successful treatment of acute cardiomyopathy.

Biochem Mol Med 58: 59-65, 1996.

232. Roe DS, Vianey-Saban C, Sharma S, Zabot MT and Roe CR. Oxidation

of unsaturated fatty acids by human fibroblasts with very-long-chain acyl-

CoA dehydrogenase deficiency: aspects of substrate specificity and

correlation with clinical phenotype. Clin Chim Acta 312: 55-67, 2001.

233. Spiekerkoetter U, Tokunaga C, Wendel U, Mayatepek E, Exil V, Duran

M, Wijburg FA, Wanders RJ and Strauss AW. Changes in blood

carnitine and acylcarnitine profiles of very long-chain acyl-CoA

dehydrogenase-deficient mice subjected to stress. Eur J Clin Invest 34:

191-196, 2004.

234. Exil VJ, Roberts RL, Sims H, McLaughlin JE, Malkin RA, Gardner CD,

Ni G, Rottman JN and Strauss AW. Very-long-chain acyl-coenzyme a

dehydrogenase deficiency in mice. Circ Res 93: 448-455, 2003.

235. Cox KB, Hamm DA, Millington DS, Matern D, Vockley J, Rinaldo P,

Pinkert CA, Rhead WJ, Lindsey JR and Wood PA. Gestational,

pathologic and biochemical differences between very long-chain acyl-CoA

dehydrogenase deficiency and long-chain acyl-CoA dehydrogenase

deficiency in the mouse. Hum Mol Genet 10: 2069-2077, 2001.

238 236. Spiekerkoetter U, Ruiter J, Tokunaga C, Wendel U, Mayatepek E,

Wijburg FA, Strauss AW and Wanders RJ. Evidence for impaired

gluconeogenesis in very long-chain acyl-CoA dehydrogenase-deficient

mice. Horm Metab Res 38: 625-630, 2006.

237. Gillingham MB, Connor WE, Matern D, Rinaldo P, Burlingame T,

Meeuws K and Harding CO. Optimal dietary therapy of long-chain 3-

hydroxyacyl-CoA dehydrogenase deficiency. Mol Genet Metab 79: 114-

123, 2003.

238. Saudubray JM, Martin D, de LP, Touati G, Poggi-Travert F, Bonnet D,

Jouvet P, Boutron M, Slama A, Vianey-Saban C, Bonnefont JP,

Rabier D, Kamoun P and Brivet M. Recognition and management of

fatty acid oxidation defects: a series of 107 patients. J Inherit Metab Dis 22:

488-502, 1999.

239. Bonnet D, Martin D, Pascale DL, Villain E, Jouvet P, Rabier D, Brivet

M and Saudubray JM. Arrhythmias and conduction defects as presenting

symptoms of fatty acid oxidation disorders in children. Circulation 100:

2248-2253, 1999.

240. Saudubray JM, Martin D, de LP, Touati G, Poggi-Travert F, Bonnet D,

Jouvet P, Boutron M, Slama A, Vianey-Saban C, Bonnefont JP,

Rabier D, Kamoun P and Brivet M. Recognition and management of

239 fatty acid oxidation defects: a series of 107 patients. J Inherit Metab Dis 22:

488-502, 1999.

241. Hermann HP, Arp J, Pieske B, Kogler H, Baron S, Janssen PM and

Hasenfuss G. Improved systolic and diastolic myocardial function with

intracoronary pyruvate in patients with congestive heart failure. Eur J

Heart Fail 6: 213-218, 2004.

242. Roe CR and Mochel F. Anaplerotic diet therapy in inherited metabolic

disease: therapeutic potential. J Inherit Metab Dis 29: 332-340, 2006.

243. Bartlett K, Ghneim HK, Stirk JH, Dale G and Alberti KG. Pyruvate

carboxylase deficiency. J Inherit Metab Dis 7 Suppl 1: 74-78, 1984.

244. Peuhkurinen KJ and Hassinen IE. Pyruvate carboxylation as an

anaplerotic mechanism in the isolated perfused rat heart. Biochem J 202:

67-76, 1982.

245. Ahmad A, Kahler SG, Kishnani PS, rtigas-Lopez M, Pappu AS,

Steiner R, Millington DS and Van Hove JL. Treatment of pyruvate

carboxylase deficiency with high doses of citrate and aspartate. Am J Med

Genet 87: 331-338, 1999.

240 246. Kasumov T, Martini WZ, Reszko AE, Bian F, Pierce BA, David F, Roe

CR and Brunengraber H. Assay of the concentration and (13)C isotopic

enrichment of propionyl-CoA, methylmalonyl-CoA, and succinyl-CoA by

gas chromatography-mass spectrometry. Anal Biochem 305: 90-96, 2002.

247. Hamilton C and Saggerson ED. Malonyl-CoA metabolism in cardiac

myocytes. Biochem J 350 Pt 1: 61-67, 2000.

248. Dyck JR, Cheng JF, Stanley WC, Barr R, Chandler MP, Brown S,

Wallace D, Arrhenius T, Harmon C, Yang G, Nadzan AM and

Lopaschuk GD. Malonyl coenzyme a decarboxylase inhibition protects

the ischemic heart by inhibiting fatty acid oxidation and stimulating glucose

oxidation. Circ Res 94: e78-e84, 2004.

249. Goodwin GW and Taegtmeyer H. Regulation of fatty acid oxidation of

the heart by MCD and ACC during contractile stimulation. Am J Physiol

277: E772-E777, 1999.

250. Kasumov T, Adams JE, Bian F, David F, Thomas KR, Jobbins KA,

Minkler PE, Hoppel CL and Brunengraber H. Probing peroxisomal beta-

oxidation and the labelling of acetyl-CoA proxies with [1-(13C)]octanoate

and [3-(13C)]octanoate in the perfused rat liver. Biochem J 389: 397-401,

2005.

241 251. Brecher P. The interaction of long-chain acyl CoA with membranes. Mol

Cell Biochem 57: 3-15, 1983.

252. Zhang D, Liu ZX, Choi CS, Tian L, Kibbey R, Dong J, Cline GW, Wood

PA and Shulman GI. Mitochondrial dysfunction due to long-chain Acyl-

CoA dehydrogenase deficiency causes hepatic steatosis and hepatic

insulin resistance. Proc Natl Acad Sci U S A 104: 17075-17080, 2007.

253. Beylot M, Soloviev MV, David F, Landau BR and Brunengraber H.

Tracing hepatic gluconeogenesis relative to citric acid cycle activity in vitro

and in vivo. Comparisons in the use of [3-13C]lactate, [2-13C]acetate, and

alpha-keto[3-13C]isocaproate. J Biol Chem 270: 1509-1514, 1995.

254. Yang SY, He XY and Schulz H. Fatty acid oxidation in rat brain is limited

by the low activity of 3-ketoacyl-coenzyme A thiolase. J Biol Chem 262:

13027-13032, 1987.

255. Sugden PH and Newsholme EA. Activities of hexokinase,

phosphofructokinase, 3-oxo acid coenzyme A-transferase and

acetoacetyl-coenzyme A thiolase in nervous tissue from vertebrates and

invertebrates. Biochem J 134: 97-101, 1973.

256. Riedel G, Platt B and Micheau J. Glutamate receptor function in learning

and memory. Behav Brain Res 140: 1-47, 2003.

242