Upload
phamlien
View
213
Download
0
Embed Size (px)
Citation preview
Mitochondrial adaptations in insulin resistant muscle
van den Broek, N.M.A.
DOI:10.6100/IR684831
Published: 01/01/2010
Document VersionPublisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the author's version of the article upon submission and before peer-review. There can be important differencesbetween the submitted version and the official published version of record. People interested in the research are advised to contact theauthor for the final version of the publication, or visit the DOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and page numbers.
Link to publication
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?
Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Download date: 01. Sep. 2018
Mitochondrial adaptations in insulin resistant muscle
A catalogue record is available from the Eindhoven University of Technology Library ISBN: 978‐90‐386‐2299‐6 Printed by:
Mitochondrial adaptations in insulin resistant muscle
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de rector magnificus, prof.dr.ir. C.J. van Duijn, voor een
commissie aangewezen door het College voor Promoties in het openbaar te verdedigen
op donderdag 16 september 2010 om 16.00 uur
door
Nicole Martina Adriana van den Broek
geboren te Tilburg
Dit proefschrift is goedgekeurd door de promotor: prof.dr. K. Nicolay Copromotor: dr. J.J. Prompers
Contents Chapter 1 General introduction 1
Chapter 2 Comparison of in vivo post‐exercise phosphocreatine recovery and resting ATP synthesis flux for the assessment of skeletal muscle mitochondrial function
31
Chapter 3 Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics in muscle after exercise are due to differences in proton efflux rates
53
Chapter 4 Early or advanced stage type 2 diabetes is not accompanied by in vivo skeletal muscle mitochondrial dysfunction
77
Chapter 5 Increased mitochondrial content rescues in vivo muscle oxidative capacity in long‐term high‐fat diet fed rats
99
Chapter 6 Carnitine insufficiency in high‐fat diet fed rats does not contribute to lipid‐induced impairment of skeletal muscle mitochondrial function in vivo
123
Chapter 7 Summary and future perspectives 147
Nederlandse samenvatting Dankwoord List of publications Curriculum Vitae
157
Chapter Introduction
Chapter 1
2
What was originally termed adult‐onset diabetes is now one of the most dominant diseases
in the world: type 2 diabetes (T2D). The prevalence of diabetes is increasing dramatically.
The total number of people with diabetes is projected to rise from 171 million in 2000 to
366 million in 2030 [4]. The majority of all diabetes patients (90%) suffers from T2D. The
most important reason for the dramatic increase in prevalence of T2D is the expanding
Western lifestyle, which encompasses the combination of an excessive calorie intake and a
decreased physical activity, leading to fat accumulation and increased body weight.
T2D is characterized by insulin resistance (IR), a disorder in which major metabolic tissues,
in particular skeletal muscle, are less sensitive for insulin. A first attempt of the body to
compensate for the IR of the tissues is to increase insulin production by the pancreatic ‐cells. IR is progressive and therefore during the years the need for insulin will keep
increasing. After some time the overproduction of insulin will lead to -cell failure, resulting in a diminished insulin secretion and sustained elevated blood glucose levels, a state which
is called T2D. After several years, T2D can lead to severe life‐threatening complications like
cardiovascular disease, neuropathy, retinopathy and kidney failure.
During the last decade excessive ectopic lipid storage and muscle mitochondrial
dysfunction have been associated with the pathogenesis of IR and T2D [5]. However, there
is still no consensus about the underlying mechanisms leading to IR and consequently T2D.
Therefore further investigation into the etiology of IR and T2D is important to conquer the
T2D epidemic.
The present chapter provides an overview of the background of this thesis. First the most
important cellular events leading to IR and T2D will be explained and the energy‐producing
organelles, i.e. the mitochondria, will be introduced. Subsequently, different theories about
the mechanism of fatty acid‐induced IR will be summarized and the methods applied in this
thesis for measuring muscle lipids and lipid intermediates and muscle mitochondrial
function are described. Finally, the outline of the thesis is presented.
Glucose homeostasis, insulin resistance and type 2 diabetes
In healthy subjects blood glucose levels are well controlled. After a meal, glucose is taken
up in the blood and transported to tissues, where glucose is used for energy production.
Glucose is the most important substrate for the brain. However, skeletal muscle is
responsible for as much as 70‐80% of postprandial glucose uptake [6]. When blood glucose
levels rise, e.g. after a meal, the ‐cells of the pancreas respond by producing more insulin.
Insulin can stimulate the ‘insulin sensitive’ glucose uptake, by binding to cell membrane‐
bound insulin receptors, which activates the insulin signaling cascade. Binding of insulin to
its receptor leads to autophosphorylation of its ‐subunits and the phosphorylation of
Introduction
3
tyrosine residues in insulin receptor substrates (IRS). IRS activates phosphatidylinositol‐3‐
kinase (PI3K) through its SH2 domain, thus increasing the intracellular concentration of PIP2
and PIP. This, in turn, activates phosphatidylinositol phosphate‐dependent kinase‐1 (PDK‐
1), which subsequently activates Akt/PKB. This activation cascade results in the
translocation of the glucose transporter 4 (GLUT4) from cytoplasmic vesicles to the cell
membrane, where GLUT4 facilitates glucose uptake (reviewed in [3]). Besides stimulating
the uptake of glucose in muscle, insulin also inhibits glucose production in the liver [7]. As a
result of those muscle and liver responses, blood glucose levels will return back to basal
values in a relatively short time period. In the muscle, glucose can be used for ATP
production. However during rest, when little energy is needed, glucose is converted to
glycogen, which is stored in the muscle, thus providing an energy reserve that can quickly
be mobilized to meet a sudden need for glucose. In the case of IR, postprandial glucose is
still cleared from the blood, but because of the decreased sensitivity of the peripheral
tissues for insulin, more insulin is needed for a similar blood glucose lowering effect.
Sustained insulin overproduction by pancreatic ‐cells leads to ‐cell failure and decreased insulin production. This results in increased basal glucose levels and increased duration of
postprandial blood glucose clearance, which is a condition called T2D. The high blood
glucose levels can irreversibly damage tissues, finally leading to well known complications
of T2D, e.g. cardiovascular disease, neuropathy, retinopathy and kidney failure. The
dramatic rise in T2D is associated with the increased occurrence of obesity and the
decrease of physical activity. However, the primary cause of IR and T2D is still not known.
During the last decade, skeletal muscle mitochondrial dysfunction has been proposed to
play a key role in the development of IR and T2D [8‐13].
Mitochondrial energy production in muscle
Mitochondria are called the ‘powerhouses’ of the cell. Without them, cells would be unable
to extract significant amounts of energy from the nutrients, and as a consequence,
essentially all cellular functions would fail under high energy demand conditions. The
specific structure of the mitochondrion is essential for synthesizing adenosine triphosphate
(ATP), the energy source for cells. The organelle is surrounded by two lipid bilayer
membranes: an outer and an inner membrane. Many foldings of the inner membrane form
cristae, in which oxidative enzymes (NADH dehydrogenase (complex I), cytochrome c
reductase (complex III), and cytochrome c oxidase (complex IV)), contributing to the
electron transport chain (ETC), and the F1F0‐ATP synthase complex (complex V), responsible
for proton‐driven ATP synthesis, are located (Figure 1). In addition, the inner cavity of the
mitochondrion, termed the matrix, contains large quantities of enzymes, which are
involved in the oxidation of carbohydrates (CHO) and fatty acids (FAs), i.e. enzymes from
the tricarboxylic acid (TCA) cycle and the ‐oxidation pathway, respectively [14].
Chapter 1
4
ATP is used throughout the cell to drive almost all energy‐dependent cellular processes, like
active transport of molecules and ions, the synthesis of macromolecules and the
performance of mechanical work during muscle contractions and other cellular movements
[14]. ATP is therefore also called the energy currency of the cell. ATP is a nucleotide
composed of an adenine ring and a ribose sugar (adenosine) and three phosphate groups
(triphosphate). The phosphate groups, starting with the group closest to the ribose, are
referred to as the alpha (), beta (), and gamma (γ) phosphates. Two phospho‐anhydride
bonds (those that connect adjacent phosphates) in ATP are responsible for the high energy
content of this molecule. In the context of biochemical reactions, these anhydride bonds
are frequently referred to as high‐energy bonds. Energy stored in ATP is released upon
hydrolysis of the terminal anhydride bond, thereby forming adenosine diphosphate (ADP)
and inorganic phosphate (Pi).
Muscle contraction depends on energy supplied by ATP. The concentration of ATP in the
muscle is sufficient to maintain full contraction for only 1 or 2 seconds at most. Therefore,
the formed ADP has to be rephosphorylated back to ATP immediately, allowing the muscle
to continue its contraction. There are several sources of energy for the rephosphorylation
of ADP. The source of energy that is used to reconstitute ATP within the first few seconds of
muscle contraction is phosphocreatine (PCr), which carries a high‐energy phosphate bond,
Figure 1. Schematic overview depicting the different components of oxidative phosphorylation, i.e. the electron transport chain and the F1F0‐ATP synthase in the inner mitochondrial membrane. NADH and FADH2 donate their electrons to complex I and II, respectively. Electron transfer through complexes I, III and IV is accompanied by the pumping of protons from the matrix into the intermembrane space, leading to the generation of an electrochemical proton gradient. This gradient is used to drive the translocation of protons through the F1F0‐ATP synthase, which goes along with the formation of ATP. IMM, inner mitochondrial membrane; complex I, NADH dehydrogenase; complex II, succinate dehydrogenase; complex III, cytochrome bc1 complex; complex IV, cytochrome c oxidase; F1F0 ATP‐ase, F1F0‐ATP synthase; Q, coenzyme Q; cyt c, cytochrome c.
Introduction
5
with a slightly higher free energy content than those in ATP. Therefore, phosphocreatine is
instantly cleaved and the released energy is used to rephosphorylate ADP to ATP by the
creatine kinase (CK) reaction:
MgADP PCr H MgATP Cr Eq. 1
After about five seconds of maximal muscle contraction, PCr is depleted and the next
source of energy to reconstitute ATP is addressed, i.e. glycogen. Glycogen has previously
been stored in large amounts in muscle cells, thereby providing a rapid source of energy for
the muscle in situations of high‐energy demand, i.e. muscle contractions longer than a few
seconds. Cleavage of glycogen results in the formation of glucose‐1‐phosphate which, after
conversion into glucose‐6‐phosphate, can enter the glycolytic pathway. Via glycolysis,
glucose‐6‐phosphate, either formed from glycogen or from glucose directly, is rapidly
broken down to pyruvate, thereby forming 2 molecules of ATP. Besides the advantage of
fast ATP production, glycolysis has the ability to produce ATP in the absence of oxygen by
producing lactate. Unfortunately, however, during anaerobic glycolysis so many end
products accumulate that the muscle will lose its capability to sustain maximum
contraction after about one minute.
The third source of energy is oxidative metabolism. More than 95% of ATP is formed in the
mitochondria via oxidative phosphorylation, by degrading substrates like CHO and FA
aerobically. Both substrates have to be converted to the general degradation product
acetyl coenzyme A (acetyl‐CoA). Pyruvate, formed via glycolysis, is subsequently
transported into the mitochondria and converted to acetyl‐CoA by the pyruvate
dehydrogenase complex. Within the outer mitochondrial membrane, the enzyme fatty acyl‐
CoA synthetase (ACS) catalyses the first step in intracellular FA metabolism by converting
FAs into fatty acyl‐CoA’s. Upon activation by ACS the resulting fatty acyl‐CoA’s can be
transported across the mitochondrial membrane via the carnitine shuttle, which consists of
three components: carnitine palmitoyltransferase (CPT) 1 and 2 and carnitine acylcarnitine
translocase (CACT). CPT1 executes the initial step in this process by catalyzing the reversible
transesterification of long‐chain acyl‐CoA with carnitine. The long‐chain acylcarnitine
products of CPT1 are then transported across the inner mitochondrial membrane by CACT.
Finally, inside the mitochondrion, CPT2 regenerates acyl‐CoA on the matrix side of the
membrane where it can enter the ‐oxidation. Each cycle of the ‐oxidation consists of four reactions, shortening the acyl‐CoA unit by two carbon atoms. Thereby, every cycle one
nicotinamide adenine dinucleotide (NADH), one flavin adenine dinucleotide (FADH2) and an
acetyl‐CoA molecule are released. This process continues until the entire acyl chain is
cleaved into acetyl‐CoA units. Acetyl‐CoA, derived from CHO via pyruvate or from FAs, then
Chapter 1
6
enters the next step in the oxidative process: the TCA cycle. Acetyl‐CoA transfers its two‐
carbon acetyl group to oxaloacetate to form a six‐carbon compound, i.e. citrate. Citrate
then undergoes a series of chemical transformations resulting in oxaloacetate, to which a
new acetyl‐CoA can be conjugated to enter the TCA cycle again. During each cycle one ATP,
two CO2, one FADH2 and three NADH molecules are released. The electron donors, NADH
and FADH2, transfer their electrons to oxygen, which is reduced to water in the ETC. The
enzymes that catalyze these reactions have the ability to simultaneously create a proton
gradient across the inner mitochondrial membrane, which drives the F1F0‐ATP synthase
complex to form ATP from ADP and Pi:
MgATP H O MgADP P H Eq. 2
Figure 2. Schematic overview of FA transport across the mitochondrial membranes by the carnitine shuttle and the metabolic conversions of FAs and pyruvate to yield ATP. ACS exerts the first step in FA transport by converting FAs into fatty acyl‐CoA’s. CPT1 catalyzes the reversible transesterification of long‐chain acyl‐CoA with carnitine. These long‐chain acylcarnitine species are then transported across the inner mitochondrial membrane by CACT. Inside the mitochondrion, CPT2 regenerates acyl‐CoA on
the matrix side of the membrane where it can enter the ‐oxidation, which generates acetyl‐CoA. Pyruvate is transported into the mitochondrial matrix, where it is converted to acetyl‐CoA by PDH. Acetyl‐CoA can enter the TCA cycle. NADH and FADH2, produced in the TCA cycle, enable ATP production by OXPHOS as explained in more detail in Figure 1. OMM, outer mitochondrial membrane; ACS, acyl‐CoA synthetase; CPT, carnitine palmitoyl transferase; CACT, carnitine‐acylcarnitine translocase; PDH, pyruvate dehydrogenase; TCA, tricarboxylic acid cycle; OXPHOS, oxidative phosphorylation.
Introduction
7
Mechanisms of fatty acid‐induced muscle insulin resistance and the role of mitochondrial dysfunction
T2D and IR are often associated with hyperlipidemia (excessively high plasma triglyceride
and cholesterol levels) as well as elevated levels of plasma FFA (reviewed in [15‐16]). It has
been shown that raising plasma FFA levels by intravenously administering IntralipidTM
decreases whole body glucose uptake, oxidation and storage [17] and skeletal muscle
glucose uptake [18‐19]. Furthermore, lowering of plasma FFA has been shown to reduce
the severity of insulin resistance [20]. Plasma FFAs can easily enter cells where they are
either oxidized to generate energy in the form of ATP or re‐esterified for storage as
triglycerides (TG). Not surprisingly, therefore, raising plasma FFA levels results in
intramyocellular accumulation of triglycerides, the so‐called intramyocellular lipids (IMCL)
[21], which are correlated to the severity of IR [22]. It was shown that IMCL content is a
better predictor for IR than plasma FFA levels [23]. However IMCL is probably not
interfering with the insulin signaling pathway directly, but rather represents a surrogate
marker of a systemic lipid oversupply, which is believed to cause IR. In this chapter several
hypotheses about the underlying mechanism of lipid‐induced IR are summarized.
Inhibition of glucose oxidation
Randle et al. introduced the ‘glucose‐fatty acid cycle’ or ‘Randle cycle’ [2]. In isolated heart
and skeletal muscle preparations, Randle et al. demonstrated that the utilization of one
nutrient inhibited the use of the other directly and without hormonal mediation. Most
emphasis was put on the control of glucose oxidation by FAs. Due to an elevated FA
oxidation, the intra‐mitochondrial acetyl‐CoA/CoA ratio increases, resulting in the inhibition
of PDH, a regulatory enzyme in glucose metabolism. It has been shown that these changes
lead to elevated citrate levels in the cytosol, which inhibit PFK‐1 [24], followed by an
increase in glucose‐6‐phosphate, which eventually inhibits hexokinase and therefore
diminishes glucose uptake, resulting in elevated plasma glucose levels (Figure 3). Overall,
Randle concluded that an increased FA availability leads to increased ‐oxidation and a decrease in muscle glucose uptake and oxidation, and can therefore be seen as a possible
cause for elevated plasma glucose levels and skeletal muscle IR [2].
Inhibition of glucose transport and mitochondrial dysfunction
Several years later, the Randle cycle was challenged. A series of MRS studies performed by
the group of Shulman pointed out that glucose transport was impaired directly by elevated
FFA, without an increase in glucose‐6‐phosphate levels [25]. In addition, others have shown
that infusion of lipids induced IR only several hours after it had already decreased glucose
oxidation, suggesting that the glucose‐fatty acid cycle may not be responsible for IR [26]. As
Chapter 1
8
an alternative for the glucose‐fatty acid cycle, Shulman proposed a mechanism in which
intramyocellular lipids impair glucose uptake by inhibiting the insulin signaling cascade [3,
27]. The proposed mechanism involves the increase in intramyocellular lipid metabolites
like diacylglycerol (DAG), long‐chain acyl‐CoA’s and ceramides, due to increased delivery of
lipids to muscle and/or a decreased intracellular FA oxidation. These metabolites can
activate protein kinase C , leading to the phosphorylation of serine/threonine sites on the insulin receptor substrate 1, which in turn interferes with the activation of
phosphatidylinositol 3‐kinase, eventually resulting in a diminished glucose transport activity
(Figure 4). Since then, several studies in humans, animals and in vitro preparations have
confirmed steps of this mechanism [25, 28‐29].
Figure 3. Mechanism of inhibition of glucose utilization by enhanced fatty acid oxidation as described by Randle et al. [2]. An increase in fatty acid availability promotes fatty acid oxidation. This results in elevated acetyl‐CoA levels, which causes inhibition of PDH activity and reduced pyruvate oxidation. This leads to increased intramitochondrial and consequently intracellular citrate levels. Citrate in turn inhibits PFK‐1, leading to accumulation of G‐6‐P, which inhibits HK and thereby glucose uptake via GLUT4. LC‐FA, long‐chain fatty acid; FAT, fatty acid translocase; CD36, cluster of differentiation 36;
LC‐acyl‐CoA, long‐chain acyl‐coenzyme A; ‐ox, ‐oxidation; TCA, tricarboxylicacid cycle; PDH, pyruvate dehydrogenase; PFK‐1, phosphofructokinase‐1; F‐6‐P, fructose‐6‐phosphate; G‐6‐P, glucose‐6‐phosphate; HK, hexokinase; GLUT4, glucose transporter type 4.
Introduction
9
Furthermore, it was shown that IMCL, assessed by 1H MRS, is a better predictor for IR than
plasma FFA, triglyceride and cholesterol levels, both in adults and in children [23, 30‐33].
Further proof for the relation between IR and IMCL came from Pima Indians in whom IMCL
levels were inversely correlated with insulin action [22]. However, it has to be kept in mind
that IMCL is probaby not directly affecting insulin sensitivity, but more likely represents the
accumulation of lipid intermediates, e.g. DAG, ceramides and long‐chain acyl‐CoA’s, which
can on their turn interfere with the insulin signaling cascade [34].
Although many studies strongly suggest that lipid accumulation within muscle tissue is
associated with IR, paradoxical results have been presented [35‐37]. Not only within muscle
of obese, diabetic subjects, but also within muscle of endurance‐trained, insulin‐sensitive
athletes increased IMCL levels have been found. Therefore, the reported correlations
between IMCL content and IR do not represent a direct, functional relationship. In addition
to this finding, it has been shown that insulin sensitivity and IMCL content are dependent
on muscle fiber type. IMCL levels are higher in insulin sensitive oxidative type 1 muscle
fibers, compared to the less insulin sensitive, more glycolytic type 2 fibers [37‐39]. These
findings together point out that increased lipid content within muscle does not always
denote IR and that muscle lipid content should be evaluated within a context of other
markers of metabolic capacity.
One such marker is the capacity for lipid oxidation [40], or the capacity for substrate
oxidation in general [41]. Several human studies have provided evidence for dysfunctional
muscle mitochondria in insulin‐resistant subjects by showing down‐regulation of genes
encoding mitochondrial enzymes [42‐43], decreased mitochondrial content and lower
mitochondrial respiratory chain activity [44]. Kelly et al. reported a 40% decrease in overall
ETC activity (estimated using the activity of rotenone‐sensitive NADH:oxygen
oxidoreductase) and smaller, deformed mitochondria in skeletal muscle from T2D patients
compared to muscle of healthy volunteers [45]. Moreover, studies in which oxygen
consumption was determined in freshly prepared isolated mitochondria or permeabilized
muscle fibers showed decreased ADP‐stimulated respiration in diabetic patients compared
to BMI‐matched controls [46‐47]. In vivo 31P MRS measurements were also applied to
investigate the correlation between mitochondrial function and IR and/or T2D. Resting ATP
synthesis flux has been measured by saturation transfer (ST) experiments by the group of
Shulman [48‐50]. They showed decreased resting ATP synthesis fluxes in muscle of insulin‐
resistant elderly subjects as well as in insulin‐resistant offspring of type 2 diabetic parents
as compared to healthy controls [48‐50]. Another 31P MRS technique to determine
mitochondrial function is the measurement of PCr recovery after muscle exercise. A slower
PCr recovery, indicating impaired mitochondrial function, was found in overweight diabetic
patients as compared to healthy overweight controls [51]. It was hypothesized that
inherited or acquired skeletal muscle mitochondrial dysfunction, associated with a reduced
Chapter 1
10
mitochondrial capacity to oxidize FAs, leads to a lipid overload in muscle cells (Figure 4),
inducing IR as was explained before.
However, recently more and more studies have been reported in which impaired
mitochondrial dysfunction was not observed in insulin‐resistant subjects or patients with
type 2 diabetes. PCr recovery rates of early and advanced stage diabetic patients matched
for the level of physical activity were not significantly different [52]. Additionally, results
from respiration measurements of permeabilized fibers of T2D patients compared to
controls are in agreement with the previous results when normalized for mitochondrial
DNA content or citrate synthase activity [53]. In other words, the type 2 diabetes patients
showed normal intrinsic mitochondrial function, but an impaired oxidative capacity that
was entirely attributed to a lower mitochondrial content. Environmental factors play an
Figure 4. Mechanism of inhibition of glucose utilization by enhanced fatty acid uptake and/or decreased FA oxidation as proposed by Shulman et al. [3]. Increased LC‐FA supply in combination with decreased LC‐FA oxidation leads to an accumulation of lipids and lipid intermediates in muscle.
These lipid intermediates (e.g. DAG and ceramides) activate PKC, leading to the phosphorylation of serine/threonine sites on the insulin receptor substrate 1, which in turn interferes with the activation of phosphatidylinositol 3‐kinase, eventually resulting in a diminished GLUT4 translocation to the cell membrane and consequently diminished glucose transport. IMCL, intramyocellular lipids; DAG, diacylglycerol; PKC, protein kinase C; IRS, insulin receptor substrate; tyr, tyrosine; ser, serine; thr, threonine; PI3K, phosphatidylinositol 3‐kinase (for other abbreviation see legend Figure 3).
Introduction
11
important role in regulating skeletal muscle oxidative capacity, and the lower mitochondrial
content in type 2 diabetes patients might simply be the result of a reduced habitual
physical activity level [8, 53‐54]. Furthermore, mitochondrial dysfunction in type 2 diabetes
might be secondary to impaired insulin signaling [55‐57] and/or abnormal blood glucose,
insulin [58‐59] and FFA [57] levels. Therefore, the debate continues as to whether
mitochondrial dysfunction represents either cause or consequence of IR and/or T2D.
Incomplete ‐oxidation
The hypothesis that mitochondrial dysfunction induces a decreased capacity to oxidize FAs,
which leads to lipid overload and consequently IR, is losing impact. A recent study links IR
to an increased capacity to oxidize FAs rather than the reverse [1] (Figure 5). However, the
high rates of ‐oxidation in insulin‐resistant states are associated with low rates of complete fat oxidation [1, 60]. High rates of incomplete ‐oxidation occur when carbon flux through the ‐oxidation machinery outpaces the entry of acetyl‐CoA into the TCA cycle.
Most of the evidence for the elevated incomplete ‐oxidation in insulin‐resistant states is based on the profiling of acylcarnitines in muscle, blood and urine. The vast majority of
acylcarnitines is produced in the mitochondria and can therefore, in combination with
measurements of substrate oxidation and mitochondrial function, be interpreted as a
measure for (incomplete) ‐oxidation. A growing number of studies reported a negative
correlation between circulating and/or tissue‐associated acylcarnitines and glucose
tolerance in both animals [1,61] and humans [62‐65]. Several even chain, FA‐derived
acylcarnitine intermediates were elevated in muscle of high‐fat diet‐fed rodents, but
decreased after a 2‐wk exercise intervention that restored glucose tolerance [60].
Furthermore, in vitro measurement of [14C]‐oleate catabolism revealed disproportionally
high rates of incomplete relative to complete fat oxidation in isolated muscle mitochondria
from high‐fat diet‐fed compared to lean rodents [61]. From these studies it can be
concluded that an excessive lipid load can lead to an increase in incomplete ‐oxidation, followed by the accumulation of lipid intermediates (e.g. acylcarnitines). The accumulation
of lipid intermediates reflects a failed attempt to deal with the excess of intra‐
mitochondrial acyl‐CoA’s, causing mitochondrial ‘stress’ (Figure 5). The ‘stress’ is suggested
to induce impairments in mitochondrial function and to activate stress kinases, interfering
with insulin action [66], both leading to IR.
Chapter 1
12
Current status
Although a lot of research has been performed to investigate the role of skeletal muscle
mitochondrial dysfunction in the development of IR, still no consensus has been reached
about the exact interplay between mitochondrial function, lipotoxicity and IR and/or T2D.
Aim of the thesis
The aim of this thesis was to study the timing and nature of muscle mitochondrial
adaptations during the development of IR, by combining both in vivo and in vitro
approaches, in order to gain more insight into the etiology of IR and T2D.
Figure 5. Mechanism of inhibition of glucose utilization by incomplete fatty acid oxidation as
described by Koves et al. [1]. High FA availability increases ‐oxidation, resulting in carbon fluxes through the ‐oxidation machinery which outpace the entry of acetyl‐CoA into the TCA cycle. This
results in the accumulation of incompletely oxidized lipid intermediates, which can interfere via PKC with insulin signaling and glucose uptake. CAT, carnitine acyltransferase; CACT, carnitine‐acylcarnitine translocase; the flash represents the generation of reactive oxygen species (for other abbreviation see legend Figures 3 and 4).
Introduction
13
Measurement of muscle lipids and lipid intermediates
In this thesis, the relation between IR and muscle lipid overload was studied by the
measurement of IMCL by in vivo 1H MRS and muscle acylcarnitines by in vitro tandem mass
spectrometry (MS/MS). These measurement procedures are explained in this paragraph.
IMCL
As was explained in one of the previous paragraphs, the amount of lipids inside the muscle
cell, i.e. intramyocellular lipids (IMCL), are strongly correlated with the degree of IR [22].
IMCL is mainly present as liquid droplets in the cytosol of muscle cells. The lipids which are
present as subcutaneous or interstitial adipose tissue are referred to as extramyocellular
lipids, or EMCL. While EMCL is metabolically relatively inert, there is substantial evidence
that IMCL within droplets can be rapidly mobilized and utilized for energy metabolism,
particularly as they are primarily located immediately adjacent to mitochondria.
IMCL levels have been determined using a variety of techniques including biochemical
extraction [22, 67], Oil red O histochemical staining [36, 68] and electron microscopy
morphometry [69] of needle biopsy samples and magnetic resonance imaging (MRI) and
computed tomography (CT) [70]. Disadvantages of these techniques are that they are
invasive or unable to differentiate between IMCL and EMCL or both. Localized 1H MRS has
proven to be a valuable tool for determining IMCL, because 1H MRS is uniquely capable of
separately detecting IMCL and EMCL non‐invasively and without the use of any harmful
radiation [71]. The IMCL/EMCL peak separation depends on the orientation of the muscle
with respect to the magnetic field and amounts to a maximum of circa 0.2 ppm when the
muscle is parallel to the magnetic field. In this case the CH2 protons of IMCL and EMCL
resonate around 1.28 and 1.47 ppm, respectively (Figure 6). This chemical shift difference
has been shown to originate from bulk magnetic susceptibility (BMS) effects, that are due
to the layered ordering of EMCL depots along the main muscle axis, while IMCL is organized
in spherical droplets in the cytosol of the muscle cell [71].
Figure 6. Localized 1H
MR spectrum
measured in tibialis anterior muscle from a Wistar rat fed with a high‐fat diet for 2.5 weeks, showing peaks of total creatine (tCr), extramyocellular lipids (EMCL) and intramyocellular lipids (IMCL).
Chapter 1
14
Thus, 1H MRS based IMCL measurements are very useful for the determination of IMCL
levels in longitudinal studies in both humans and animal models during the development of
IR and T2D.
Acylcarnitines
The first step for the use of FAs as a substrate for ATP production in the mitochondria is
their transport into the mitochondria. A step required for transport across the
mitochondrial membrane is the formation of acylcarnitine esters from long‐chain acyl‐
CoA’s by CPT1. Besides fulfilling an important role in mitochondrial import of FAs, the
binding to carnitine is also essential for the efflux of excess intra‐mitochondrial acyl‐CoA
into the cytosol, and from the cytosol into the bloodstream. Acylcarnitines represent
byproducts of substrate catabolism and are formed from their respective acyl‐CoA
intermediates by a family of carnitine acyltransferases that reside principally in the
mitochondria. Most even chain species reflect incomplete FA oxidation, odd chain species
stem primarily from amino acid catabolism, whereas acetylcarnitine is derived from acetyl‐
CoA, the universal degradation product of all metabolic substrates.
Blood acylcarnitine profile analysis is the current standard for the diagnosis of ‐oxidation disorders at the metabolite level [72‐73]. Furthermore, recently muscle acylcarnitine
accumulation was related to IR [1, 66, 74]. The acylcarnitine accumulation as was observed
in insulin‐resistant skeletal muscle was suggested to reflect a failed attempt to deal with
the excess of intra‐mitochondrial acyl‐CoA’s, causing mitochondrial stress. The
accumulation of metabolic by‐products (e.g. acylcarnitines) would then activate stress
kinases or other signals, interfering with insulin action [66]. However, how and if
acylcarnitines directly affect insulin‐mediated glucose uptake is currently not known. The
finding that carnitine supplementation improved glucose tolerance while increasing
circulating acylcarnitines and leaving muscle acylcarnitines unaffected favors the
interpretation that production and efflux of these metabolites is beneficial rather than
detrimental. However, the exact physiological relevance of changes in muscle, blood and
urine acylcarnitine levels in IR and as a result of carnitine supplementation is still under
debate.
By means of tandem MS/MS it is possible to analyze 36 independent acylcarnitine species
ranging in size from 2 to 22 carbons in muscle, blood and urine [75]. Important for
analyzing acylcarnitine levels is to consider that steady‐state acylcarnitine concentrations in
tissues and blood represent the net balance between production, consumption, import and
export, and therefore do not directly provide information about fluxes through individual
metabolic pathways.
Introduction
15
Measurement of muscle mitochondrial function
Data to support the proposed role of skeletal muscle mitochondrial dysfunction in the
development of IR and T2D have been obtained with various in vitro methods, including
measurements of oxidative enzyme activities [39, 44, 76‐78], mRNA and/or protein
expression of oxidative phosphorylation genes [42‐43, 78‐80] as well as mitochondrial
content, morphology and respiration [44, 47, 77, 80]. Furthermore, in vivo magnetic
resonance spectroscopy (MRS) measurements of basal mitochondrial ATP synthesis flux
[48‐49, 81] and PCr [11] and ADP [82] recovery kinetics also point towards a potential role
for mitochondrial dysfunction in the etiology of insulin resistance and/or type 2 diabetes. In
this thesis in vitro high‐resolution respirometry and in vivo 31P MRS play a dominant role
and therefore these techniques are explained in more detail below.
High‐resolution respirometry
Levels of dissolved oxygen in solution can be measured polarographically with a Clark‐type
oxygen electrode. Clark electrodes have gold or platinum cathodes and silver or silver/silver
chloride anodes, which are connected by a salt bridge and covered by an oxygen‐
permeable membrane. As oxygen diffuses across the membrane, it is reduced by a fixed
voltage between the cathode and anode which generates current in proportion to the
concentration of oxygen in solution. By calibrating the oxygen electrode with known
oxygen concentrations, it is possible to measure the rate of oxygen consumption in a
Figure 7. Typical example of time‐dependent changes in oxygen concentration during a high‐resolution
respiration measurement on a suspension of isolated mitochondria, isolated from tibialis anterior
muscle of a Wistar rat. The dark line represents the oxygen concentration in the respiratory chamber,
the lighter line represents the slope of the dark line. First, isolated mitochondria were injected in the
hermetically closed respiratory chambers (mit). After some time, glucose (gl), hexokinase (hk) and ATP
were added, inducing maximal respiration or state 3 respiration. When the oxygen consumption rate
reached a steady state, state 4 was induced by adding carboxyatractyloside. Finally, after addition of
carbonylcyanide‐3‐chlorophenylhydrazone (CCCP), the maximal oxygen consumption rate in the
uncoupled state (state U) was determined.
Chapter 1
16
medium containing actively respiring mitochondria. Since reduction of oxygen is a critical
step in the process of mitochondrial electron transport and ATP synthesis, measurement of
mitochondrial oxygen consumption provides a convenient way to assess mitochondrial
function [83]. Oxygen consumption can be measured in permeabilized muscle fibers and
isolated mitochondria. The latter approach has been used in the studies described in this
thesis.
Measuring respiration in isolated mitochondria practically means that together with a
respiration medium, an aliquot of suspension with isolated mitochondria is added to the
closed metabolic chamber. The mitochondria are brought into defined “states” by the
sequential addition of substrates or inhibitors. Since the mitochondria consume oxygen,
the oxygen concentration drops. This change of oxygen concentration is recorded by the
oxygen sensor in the chamber. From the rate of decline in the oxygen concentration (taking
into account correction for oxygen diffusion) the respiration rate can be computed for the
mitochondria in different states (see Figure 7). State 3, state 4 and state uncoupled (state
U) are frequently used to define mitochondrial respiration. State 3 respiration is the
maximal respiration, reached when saturating levels of ADP are added to the mitochondria
supplemented with oxidizible substrate and excess Pi. State 4 respiration is induced by
addition of carboxyatractyloside, which inhibits the exchange of mitochondrial ATP for
extra‐mitochondrial ADP by adenine nucleotide translocase and this way effectively blocks
ATP synthesis. The residual oxygen consumption in the absence of ADP phosphorylation is
attributable to proton leak across the inner mitochondrial membrane. Thus,
carboxyatractyloside inhibited respiration serves as an indicator of the degree of uncoupled
respiration or proton leak under these conditions. The respiratory control ratio (RCR),
which is calculated by dividing state 3 by state 4 respiration, indicates the tightness of the
coupling between respiration and ADP phosphorylation. Stepwise titration of the
protonophore carbonylcyanide‐3‐chlorophenylhydrazone (CCCP) induces an uncoupled
state by dissipating the proton gradient across the inner mitochondrial membrane (state
U). Under these conditions, mitochondrial oxidative capacity can be determined in the
absence of the potential control exerted by ATP synthase, adenine nucleotide translocase,
or phosphate transporters.
By using different combinations of oxidizible substrates it is possible to quantify specific
substrate oxidation capacity in isolated mitochondria. Oxidation of TCA cycle intermediates
such as pyruvate plus malate leads to formation of NADH, which can donate electrons to
ETC complex I. In turn oxidation of another TCA cycle intermediate succinate leads to
formation of FADH2, which bypasses complex I and donates electrons to complex II. By
using fatty acid substrates such as palmitoyl‐L‐carnitine or palmitoyl‐CoA plus L‐carnitine in
combination with malate one can assess the capacity of ‐oxidation. The measured oxygen
consumption rate is expressed per milligram mitochondrial protein and is used as a
measure for the intrinsic mitochondrial function.
Introduction
17
In vivo 31P MRS
31P MRS offers a non‐invasive approach of recording concentrations of phosphorylated
metabolites and intracellular pH. 31P MRS has widely been used to study skeletal muscle
metabolism in living mammalian tissues, and is also one of the major types of
measurement methods used in this thesis. 31P MR spectra of skeletal muscle typically show
five major resonances: an inorganic phosphate (Pi) peak, a phosphocreatine (PCr) peak and
three ATP peaks, from the , and phosphate groups of ATP (Figure 8). In some cases
also peaks from phosphomonoesters (PME) and phosphodiesters (PDE) are visible.
Additionally, metabolic information can be derived indirectly from the 31P MR spectra, i.e.
the intracellular pH and the free concentration of ADP. Tissue pH can be deduced from the
chemical shift of the Pi peak [84], which actually originates from both H2PO4‐ and HPO4
2‐.
However, as these compounds are in rapid chemical exchange, a single peak is observed,
with a chemical shift dependent on the ratio of the H2PO4‐ and HPO4
2‐ concentrations,
which in this way serves as an indicator of intracellular pH. Free levels of ADP in the cell can
be calculated indirectly by use of the CK equilibrium (Eq. 1) [85]). The concentration of ADP
in healthy, resting or moderately active muscle is typically in the tens of micromolar range,
which is too low to allow direct detection with 31P MRS. The knowledge of ADP levels is
relevant, because ADP is an important regulator of the mitochondrial ATP synthesis flux.
31P MR spectra of resting skeletal muscle are relatively constant, even in diseased states,
and to assess impairments in mitochondrial energy production one needs to perturb either
the chemical or the magnetic equilibrium. Most important 31P MR techniques in studies on
the role of mitochondrial function in IR and T2D, including those in this thesis, are
saturation transfer (ST) MRS, to measure (resting) creatine kinase and ATP synthesis fluxes,
and the dynamic measurement of post‐exercise recovery of PCr.
Figure 8. 31P MR spectrum
measured in resting tibialis anterior muscle of a Wistar rat, showing the peaks of inorganic phosphate (Pi), phosphocreatine
(PCr) and the three (, and ) phosphate groups in ATP.
Chapter 1
18
Saturation transfer
Magnetization transfer experiments allow to determine fluxes between compounds that
are in chemical exchange, as long as the system is in a metabolic steady state. One variant
of the magnetization transfer technique relies on the selective saturation of the
magnetization of one of the reactants with a long, frequency‐selective RF pulse, i.e. a
saturation transfer (ST) experiment. Due to the chemical exchange of the reactants, this
results in a decrease of magnetization of the exchange partner. With equation 3 the
exchange rate constant kAB between reactants A and can be determined.
1,
, Eq. 3
M’A and M0A are the magnitudes of magnetization of compound A, when compound B is
selectively saturated and when compound B is not saturated, respectively. The apparent
longitudinal relaxation time of compound A (T1A’) can be determined from an inversion
recovery experiment with saturation of B prior to and during the inversion time. To
calculate the flux from compound A to B, kAB has to be multiplied by the concentration of
compound A.
One of the chemical exchange processes most‐studied using ST, concerns the flux through
the creatine kinase (CK) reaction [86]. CK catalyzes a phosphate exchange reaction, in which
a phosphate moiety is exchanged between the position of ATP to PCr and vice versa. As was described earlier in this chapter, the CK reaction is of high metabolic relevance, since
PCr is an important energy buffer for keeping ATP levels constant. The CK flux can be
determined by complete saturation of the ‐ATP resonance and the quantification of the reduction in the signal intensity of PCr. Measurements of CK flux in skeletal muscle as a
function of workload have shown that CK kinetics is rather insensitive to alterations in ATP
demand and that CK flux greatly exceeds maximal ATP turnover rates [87‐88].
31P ST MRS has also been used to measure the mitochondrial ATP synthesis flux in rodent
and human muscle, under resting conditions [48‐50, 89‐92]. Instead of the signal intensity
of PCr, the magnetization of Pi has to be quantified to calculate the ATP synthesis flux.
Assuming that the muscle ATP synthesis flux, as measured by 31P ST MRS, is predominantly
reflecting oxidative ATP synthesis by the F1F0‐ATP synthase in the mitochondria, muscle ATP
synthesis flux has been taken as a measure for mitochondrial function. Studies have shown
decreased resting ATP synthesis fluxes in skeletal muscle of insulin‐resistant subjects and
insulin‐resistant animal models.
Introduction
19
However, the interpretation of 31P ST MRS data, particularly at rest, is not straightforward.
The lower ATP synthesis flux in resting muscle of insulin‐resistant subjects could actually
reflect a normal regulatory response to a lower energy demand, e.g. caused by impaired
insulin signaling, rather than an impairment of intrinsic mitochondrial function [55‐56, 93].
In order to interpret the ST data in terms of mitochondrial function it is necessary to take
the error signals, i.e. the concentrations of ADP and Pi, into account. Moreover, the Pi
ATP fluxes obtained from 31P ST measurements are comprised of both mitochondrial F1F0‐
ATP synthesis flux and flux through other Pi ATP pathways, in particular the reactions
catalyzed by the glycolytic enzymes glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)
and phosphoglycerate kinase (PGK) [88, 92, 94‐95]. Although the net glycolytic contribution
to the production of ATP (via GAPDH and PGK) versus that of oxidative phosphorylation is
small [96], these enzymes catalyze a coupled near‐equilibrium reaction, and consequently,
exchange between Pi and ATP may greatly exceed the net glycolytic flux [92]. The
interpretation of resting ATP synthesis flux as a measure for mitochondrial function is
therefore relatively complicated. In order to create better conditions for detecting a defect
in mitochondrial oxidative phosphorylation, the ST experiment would need to be done in
exercising muscle [88].
Post‐exercise PCr recovery 31P MRS provides the opportunity of acquiring data during skeletal muscle exercise and
recovery in a non‐invasive manner and with a time resolution of seconds. Figure 9 shows
the concentrations of PCr, Pi and ATP derived from a time series of 31P MR spectra acquired
during rest, exercise and subsequent recovery in rat tibialis anterior (TA) muscle. When
muscle contraction starts, an immediate decrease in PCr concentration and increase in Pi
concentration can be observed (Figure 9), whereas ATP remains constant, although ATP is
used for the muscle contractions. After the muscle contractions stop, the PCr buffer is
restored and Pi levels normalize. The ATP used for resynthesis of PCr is originating
predominantly from oxidative phosphorylation in the mitochondria [97‐98]. Because the CK
reaction is much faster than the ATP synthase reaction [87‐88], the rate of PCr recovery
mirrors the rate of oxidative ATP synthesis and therefore the oxidative capacity of the
mitochondria. A short PCr recovery time constant, corresponding to a fast recovery, reflects
a high oxidative capacity, while slow recovery kinetics may indicate impaired oxidative
capacity. PCr recovery is determined by mitochondrial content, intrinsic mitochondrial
function and other factors like the supply of oxygen and substrates [99‐100]. This implies
that additional (in vitro) measurements are necessary to identify the origin of changes in
oxidative capacity measured with PCr recovery.
One of the first who proved that PCr recovery kinetics provides information about oxidative
ATP synthesis was Mahler et al., who reported that post‐exercise PCr resynthesis is
inversely proportional to the rate of oxygen consumption [101]. This finding was
Chapter 1
20
strengthened by the observation that the PCr concentration does not restore during
ischemic recovery [97‐98]. Furthermore, the correlations of post‐exercise PCr recovery
kinetics with citrate synthase activity [102], succinate dehydrogenase activity and peak
oxygen consumption (VO2peak) [103] are additional findings demonstrating that post‐
exercise PCr resynthesis changes reflects aerobic metabolism.
An important complication in interpreting post‐exercise PCr recovery data is that strenuous
exercise‐induced acidosis negatively affects PCr resynthesis [104‐114]. Therefore it has
been recommended to avoid low end‐exercise pH values, which is difficult to achieve in
combination with a significant PCr depletion. Other strategies to deal with the acidification
problem is to aim at similar end‐exercise pH values within different groups or applying
general correction factors for the influence of pH [106, 109, 111].
Several MR compatible exercise set‐ups have been built for humans, most often for
voluntary arm or leg exercise. In anaesthetized animals, muscle contractions are usually
induced by electrical stimulation. As an alternative to the invasive, direct nerve‐stimulation
method, a minimally invasive electrical stimulation method was developed that allows
longitudinal studies in rats [115]. By subcutaneously implanting electrodes along the nerve
trajectory of the N. Peroneus Communis, highly specific dorsal flexor muscle contractions
can be induced. This procedure was also used in the present thesis.
Figure 9. Peak concentrations of phosphocreatine (black), inorganic phosphate (light gray) and ATP (dark gray), as measured using
31P MRS during a 3‐min rest, 2‐min electrical stimulation and 10‐min recovery
protocol in rat tibialis anterior muscle. Time resolution was 20 seconds.
Introduction
21
Outline of this thesis
Although a lot of research has been performed to investigate the role of skeletal muscle
mitochondrial function and lipid accumulation in the development of IR, still no consensus
is reached about the exact mechanism(s) leading to IR. The aim of this thesis was to study
the timing and nature of muscle mitochondrial adaptations during the development of IR,
by using both in vivo and in vitro measurement techniques. 31P MRS plays an important role
in these studies. The first part of this thesis deals with methodological aspects of the 31P
MRS techniques, whereas the second part describes research on mitochondrial adaptations
in both patients with T2D and animal models of IR.
Both 31P MRS PCr recovery after muscle contraction and resting ATP synthesis flux have
been used to determine mitochondrial function in vivo. However these methods have
provided ambiguous results. Therefore, PCr recovery and resting ATP synthesis flux were
compared in a rat model of known mitochondrial dysfunction in order to establish which
method is most appropriate to assess in vivo skeletal muscle mitochondrial function in
chapter 2. Mitochondrial dysfunction was induced in rats by daily subcutaneous injections
with diphenyleneiodonium (DPI), which irreversibly inhibits complex I (NADH‐ubiquinone
reductase) of the respiratory chain. 31P MRS PCr recovery after exercise is applied in every
chapter of this thesis as a measure for in vivo muscle oxidative capacity. However, cytosolic
pH has a strong influence on the kinetics of PCr recovery and thereby complicates the
interpretation of PCr recovery data. It has been suggested that PCr recovery should be
normalized for end‐exercise pH. However a general correction can only be applied if there
are no intersubject differences in the pH dependence of PCr recovery. In chapter 3 we
investigated the pH dependence of PCr recovery on a subject‐by‐subject basis in vastus
lateralis muscle of healthy humans. Furthermore, we determined the relation between the
pH dependence of PCr recovery and the kinetics of proton efflux at the start of recovery.
To get more insight in the role of muscle mitochondrial dysfunction and lipotoxicity in the
etiology of IR and T2D, we performed three studies in which we determined muscle lipid
content and mitochondrial function. In a cross‐sectional study, described in chapter 4, we
examined in vivo skeletal muscle mitochondrial function in early and advanced stages of
T2D in human subjects. Long‐standing, insulin‐treated type 2 diabetes patients, subjects
with impaired fasting glucose, impaired glucose tolerance and/or recently diagnosed type 2
diabetes, and healthy, normoglycaemic controls, matched for age and body composition
and with low habitual physical activity levels were studied. In vivo mitochondrial function of
the vastus lateralis muscle was evaluated from post‐exercise PCr recovery kinetics using 31P
MRS. IMCL content was assessed in the same muscle using single‐voxel 1H MRS. Chapter 5
describes a longitudinal study with the aim to gain more insight into the timing and nature
of mitochondrial adaptations during the development of high‐fat diet‐induced IR in a well
known rodent model of IR. Adult Wistar rats were fed a high‐fat diet or normal chow for 2.5
and 25 wk. IMCL was measured with in vivo 1H MRS and acylcarnitine levels were quantified
Chapter 1
22
in vitro using tandem MS/MS. Muscle oxidative capacity was assessed in TA muscle in vivo
using 31P MRS PCr recovery and in vitro by measuring mitochondrial DNA copy number and
oxygen consumption in isolated mitochondria. Currently, it is not clear whether
mitochondrial dysfunction is (1) a cause of lipid accumulation due to a decreased capacity
to oxidize FAs or (2) a consequence of the accumulation of lipid intermediates as a result of
increased incomplete FA oxidation. Carnitine is known to stimulate FA oxidation and
export. The aim of the study described in chapter 6 was to test the hypothesis that
carnitine supplementation reduces high‐fat diet‐induced lipotoxicity, improves in vivo
muscle mitochondrial function and ameliorates insulin resistance. Wistar rats were fed
either normal chow or a high‐fat diet for 15 wk and one group of high‐fat diet fed rats was
supplemented with L‐carnitine during the last 8 wk. Muscle mitochondrial function was
measured in vivo by 31P MRS (PCr recovery and resting ATP synthase and creatine kinase
fluxes) and in vitro by high‐resolution respirometry. Muscle lipid levels were determined by 1H MRS (IMCL) and tandem mass spectrometry (acylcarnitines).
Chapter 7 provides a summary with the main outcomes of the studies described in this
thesis. The results are compared to the current hypotheses and future perspectives for
research in the field of muscle mitochondrial function in relation to the development of IR
are discussed.
Introduction
23
References 1. Koves, T.R., J.R. Ussher, R.C. Noland, D. Slentz, M. Mosedale, O. Ilkayeva, J. Bain, R. Stevens, J.R. Dyck,
C.B. Newgard, G.D. Lopaschuk, and D.M. Muoio. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab, 2008. 7(1): 45‐56.
2. Randle, P.J., P.B. Garland, C.N. Hales, and E.A. Newsholme. The glucose fatty‐acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet, 1963. 1(7285): 785‐9.
3. Shulman, G.I. Cellular mechanisms of insulin resistance. J Clin Invest, 2000. 106(2): 171‐6. 4. Wild, S., G. Roglic, A. Green, R. Sicree, and H. King. Global prevalence of diabetes: estimates for the year
2000 and projections for 2030. Diabetes Care, 2004. 27(5): 1047‐53. 5. Goodpaster, B.H. and D.E. Kelley. Skeletal muscle triglyceride: marker or mediator of obesity‐induced
insulin resistance in type 2 diabetes mellitus? Curr Diab Rep, 2002. 2(3): 216‐22. 6. DeFronzo, R.A., R. Gunnarsson, O. Bjorkman, M. Olsson, and J. Wahren. Effects of insulin on peripheral
and splanchnic glucose metabolism in noninsulin‐dependent (type II) diabetes mellitus. J Clin Invest, 1985. 76(1): 149‐55.
7. Weickert, M.O. and A.F. Pfeiffer. Signalling mechanisms linking hepatic glucose and lipid metabolism. Diabetologia, 2006. 49(8): 1732‐41.
8. Rabol, R., R. Boushel, and F. Dela. Mitochondrial oxidative function and type 2 diabetes. Appl Physiol Nutr Metab, 2006. 31(6): 675‐83.
9. Morino, K., K.F. Petersen, and G.I. Shulman. Molecular mechanisms of insulin resistance in humans and their potential links with mitochondrial dysfunction. Diabetes, 2006. 55 Suppl 2: S9‐S15.
10. Sreekumar, R. and K.S. Nair. Skeletal muscle mitochondrial dysfunction & diabetes. Indian J Med Res, 2007. 125(3): 399‐410.
11. Schrauwen‐Hinderling, V.B., M. Roden, M.E. Kooi, M.K. Hesselink, and P. Schrauwen. Muscular mitochondrial dysfunction and type 2 diabetes mellitus. Curr Opin Clin Nutr Metab Care, 2007. 10(6): 698‐703.
12. Szendroedi, J. and M. Roden. Mitochondrial fitness and insulin sensitivity in humans. Diabetologia, 2008. 51(12): 2155‐67.
13. Holloszy, J.O. Skeletal muscle "mitochondrial deficiency" does not mediate insulin resistance. Am J Clin Nutr, 2009. 89(1): 463S‐6S.
14. Guyton, A.C. and J.E. Hall, Textbook of Medical Physiology 10th ed. 2000, Philadelphia: Saunders company.
15. Kraegen, E.W. and G.J. Cooney. Free fatty acids and skeletal muscle insulin resistance. Curr Opin Lipidol, 2008. 19(3): 235‐41.
16. Randle, P.J. Regulatory interactions between lipids and carbohydrates: the glucose fatty acid cycle after 35 years. Diabetes Metab Rev, 1998. 14(4): 263‐83.
17. Boden, G., X. Chen, J. Ruiz, J.V. White, and L. Rossetti. Mechanisms of fatty acid‐induced inhibition of glucose uptake. J Clin Invest, 1994. 93(6): 2438‐46.
18. Walker, M., G.R. Fulcher, C. Catalano, G. Petranyi, H. Orskov, and K.G. Alberti. Physiological levels of plasma non‐esterified fatty acids impair forearm glucose uptake in normal man. Clin Sci (Lond), 1990. 79(2): 167‐74.
19. Nuutila, P., V.A. Koivisto, J. Knuuti, U. Ruotsalainen, M. Teras, M. Haaparanta, J. Bergman, O. Solin, L.M. Voipio‐Pulkki, U. Wegelius, and et al. Glucose‐free fatty acid cycle operates in human heart and skeletal muscle in vivo. J Clin Invest, 1992. 89(6): 1767‐74.
20. Santomauro, A.T., G. Boden, M.E. Silva, D.M. Rocha, R.F. Santos, M.J. Ursich, P.G. Strassmann, and B.L. Wajchenberg. Overnight lowering of free fatty acids with Acipimox improves insulin resistance and glucose tolerance in obese diabetic and nondiabetic subjects. Diabetes, 1999. 48(9): 1836‐41.
21. Boden, G., B. Lebed, M. Schatz, C. Homko, and S. Lemieux. Effects of acute changes of plasma free fatty acids on intramyocellular fat content and insulin resistance in healthy subjects. Diabetes, 2001. 50(7): 1612‐7.
22. Pan, D.A., S. Lillioja, A.D. Kriketos, M.R. Milner, L.A. Baur, C. Bogardus, A.B. Jenkins, and L.H. Storlien. Skeletal muscle triglyceride levels are inversely related to insulin action. Diabetes, 1997. 46(6): 983‐8.
23. Krssak, M., K. Falk Petersen, A. Dresner, L. DiPietro, S.M. Vogel, D.L. Rothman, M. Roden, and G.I. Shulman. Intramyocellular lipid concentrations are correlated with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabetologia, 1999. 42(1): 113‐6.
24. Garland, P.B., P.J. Randle, and E.A. Newsholme. Citrate as an Intermediary in the Inhibition of Phosphofructokinase in Rat Heart Muscle by Fatty Acids, Ketone Bodies, Pyruvate, Diabetes, and Starvation. Nature, 1963. 200: 169‐70.
25. Petersen, K.F. and G.I. Shulman. New insights into the pathogenesis of insulin resistance in humans using magnetic resonance spectroscopy. Obesity (Silver Spring), 2006. 14 Suppl 1: 34S‐40S.
Chapter 1
24
26. Boden, G., F. Jadali, J. White, Y. Liang, M. Mozzoli, X. Chen, E. Coleman, and C. Smith. Effects of fat on insulin‐stimulated carbohydrate metabolism in normal men. J Clin Invest, 1991. 88(3): 960‐6.
27. Dresner, A., D. Laurent, M. Marcucci, M.E. Griffin, S. Dufour, G.W. Cline, L.A. Slezak, D.K. Andersen, R.S. Hundal, D.L. Rothman, K.F. Petersen, and G.I. Shulman. Effects of free fatty acids on glucose transport and IRS‐1‐associated phosphatidylinositol 3‐kinase activity. J Clin Invest, 1999. 103(2): 253‐9.
28. Yu, C., Y. Chen, G.W. Cline, D. Zhang, H. Zong, Y. Wang, R. Bergeron, J.K. Kim, S.W. Cushman, G.J. Cooney, B. Atcheson, M.F. White, E.W. Kraegen, and G.I. Shulman. Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate‐1 (IRS‐1)‐associated phosphatidylinositol 3‐kinase activity in muscle. J Biol Chem, 2002. 277(52): 50230‐6.
29. Holland, W.L., J.T. Brozinick, L.P. Wang, E.D. Hawkins, K.M. Sargent, Y. Liu, K. Narra, K.L. Hoehn, T.A. Knotts, A. Siesky, D.H. Nelson, S.K. Karathanasis, G.K. Fontenot, M.J. Birnbaum, and S.A. Summers. Inhibition of ceramide synthesis ameliorates glucocorticoid‐, saturated‐fat‐, and obesity‐induced insulin resistance. Cell Metab, 2007. 5(3): 167‐79.
30. Dobbins, R.L., L.S. Szczepaniak, B. Bentley, V. Esser, J. Myhill, and J.D. McGarry. Prolonged inhibition of muscle carnitine palmitoyltransferase‐1 promotes intramyocellular lipid accumulation and insulin resistance in rats. Diabetes, 2001. 50(1): 123‐30.
31. Kuhlmann, J., C. Neumann‐Haefelin, U. Belz, J. Kalisch, H.P. Juretschke, M. Stein, E. Kleinschmidt, W. Kramer, and A.W. Herling. Intramyocellular lipid and insulin resistance: a longitudinal in vivo 1H‐spectroscopic study in Zucker diabetic fatty rats. Diabetes, 2003. 52(1): 138‐44.
32. Perseghin, G., P. Scifo, F. De Cobelli, E. Pagliato, A. Battezzati, C. Arcelloni, A. Vanzulli, G. Testolin, G. Pozza, A. Del Maschio, and L. Luzi. Intramyocellular triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H‐13C nuclear magnetic resonance spectroscopy assessment in offspring of type 2 diabetic parents. Diabetes, 1999. 48(8): 1600‐6.
33. Szczepaniak, L.S., E.E. Babcock, F. Schick, R.L. Dobbins, A. Garg, D.K. Burns, J.D. McGarry, and D.T. Stein. Measurement of intracellular triglyceride stores by H spectroscopy: validation in vivo. Am J Physiol, 1999. 276(5 Pt 1): E977‐89.
34. Hulver, M.W., J.R. Berggren, R.N. Cortright, R.W. Dudek, R.P. Thompson, W.J. Pories, K.G. MacDonald, G.W. Cline, G.I. Shulman, G.L. Dohm, and J.A. Houmard. Skeletal muscle lipid metabolism with obesity. Am J Physiol Endocrinol Metab, 2003. 284(4): E741‐7.
35. van Loon, L.J. and B.H. Goodpaster. Increased intramuscular lipid storage in the insulin‐resistant and endurance‐trained state. Pflugers Arch, 2006. 451(5): 606‐16.
36. Goodpaster, B.H., J. He, S. Watkins, and D.E. Kelley. Skeletal muscle lipid content and insulin resistance: evidence for a paradox in endurance‐trained athletes. J Clin Endocrinol Metab, 2001. 86(12): 5755‐61.
37. van Loon, L.J., R. Koopman, R. Manders, W. van der Weegen, G.P. van Kranenburg, and H.A. Keizer. Intramyocellular lipid content in type 2 diabetes patients compared with overweight sedentary men and highly trained endurance athletes. Am J Physiol Endocrinol Metab, 2004. 287(3): E558‐65.
38. van Loon, L.J., R. Koopman, J.H. Stegen, A.J. Wagenmakers, H.A. Keizer, and W.H. Saris. Intramyocellular lipids form an important substrate source during moderate intensity exercise in endurance‐trained males in a fasted state. J Physiol, 2003. 553(Pt 2): 611‐25.
39. He, J., S. Watkins, and D.E. Kelley. Skeletal muscle lipid content and oxidative enzyme activity in relation to muscle fiber type in type 2 diabetes and obesity. Diabetes, 2001. 50(4): 817‐23.
40. Savage, D.B., K.F. Petersen, and G.I. Shulman. Disordered lipid metabolism and the pathogenesis of insulin resistance. Physiol Rev, 2007. 87(2): 507‐20.
41. Simoneau, J.A., J.H. Veerkamp, L.P. Turcotte, and D.E. Kelley. Markers of capacity to utilize fatty acids in human skeletal muscle: relation to insulin resistance and obesity and effects of weight loss. FASEB J, 1999. 13(14): 2051‐60.
42. Mootha, V.K., C.M. Lindgren, K.F. Eriksson, A. Subramanian, S. Sihag, J. Lehar, P. Puigserver, E. Carlsson, M. Ridderstrale, E. Laurila, N. Houstis, M.J. Daly, N. Patterson, J.P. Mesirov, T.R. Golub, P. Tamayo, B. Spiegelman, E.S. Lander, J.N. Hirschhorn, D. Altshuler, and L.C. Groop. PGC‐1alpha‐responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet, 2003. 34(3): 267‐73.
43. Patti, M.E., A.J. Butte, S. Crunkhorn, K. Cusi, R. Berria, S. Kashyap, Y. Miyazaki, I. Kohane, M. Costello, R. Saccone, E.J. Landaker, A.B. Goldfine, E. Mun, R. DeFronzo, J. Finlayson, C.R. Kahn, and L.J. Mandarino. Coordinated reduction of genes of oxidative metabolism in humans with insulin resistance and diabetes: Potential role of PGC1 and NRF1. Proc Natl Acad Sci U S A, 2003. 100(14): 8466‐71.
44. Ritov, V.B., E.V. Menshikova, J. He, R.E. Ferrell, B.H. Goodpaster, and D.E. Kelley. Deficiency of subsarcolemmal mitochondria in obesity and type 2 diabetes. Diabetes, 2005. 54(1): 8‐14.
45. Kelley, D.E., B.H. Goodpaster, and L. Storlien. Muscle triglyceride and insulin resistance. Annu Rev Nutr, 2002. 22: 325‐46.
Introduction
25
46. Phielix, E., V.B. Schrauwen‐Hinderling, M. Mensink, E. Lenaers, R. Meex, J. Hoeks, M.E. Kooi, E. Moonen‐Kornips, J.P. Sels, M.K. Hesselink, and P. Schrauwen. Lower intrinsic ADP‐stimulated mitochondrial respiration underlies in vivo mitochondrial dysfunction in muscle of male type 2 diabetic patients. Diabetes, 2008. 57(11): 2943‐9.
47. Mogensen, M., K. Sahlin, M. Fernstrom, D. Glintborg, B.F. Vind, H. Beck‐Nielsen, and K. Hojlund. Mitochondrial respiration is decreased in skeletal muscle of patients with type 2 diabetes. Diabetes, 2007. 56(6): 1592‐9.
48. Petersen, K.F., S. Dufour, D. Befroy, R. Garcia, and G.I. Shulman. Impaired mitochondrial activity in the insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(7): 664‐71.
49. Petersen, K.F., D. Befroy, S. Dufour, J. Dziura, C. Ariyan, D.L. Rothman, L. DiPietro, G.W. Cline, and G.I. Shulman. Mitochondrial dysfunction in the elderly: possible role in insulin resistance. Science, 2003. 300(5622): 1140‐2.
50. Petersen, K.F., S. Dufour, and G.I. Shulman. Decreased insulin‐stimulated ATP synthesis and phosphate transport in muscle of insulin‐resistant offspring of type 2 diabetic parents. PLoS Med, 2005. 2(9): e233.
51. Schrauwen‐Hinderling, V.B., M.E. Kooi, M.K. Hesselink, J.A. Jeneson, W.H. Backes, C.J. van Echteld, J.M. van Engelshoven, M. Mensink, and P. Schrauwen. Impaired in vivo mitochondrial function but similar intramyocellular lipid content in patients with type 2 diabetes mellitus and BMI‐matched control subjects. Diabetologia, 2007. 50(1): 113‐20.
52. De Feyter, H.M., N.M. van den Broek, S.F. Praet, K. Nicolay, L.J. van Loon, and J.J. Prompers. Early or advanced stage type 2 diabetes is not accompanied by in vivo skeletal muscle mitochondrial dysfunction. Eur J Endocrinol, 2008. 158(5): 643‐53.
53. Boushel, R., E. Gnaiger, P. Schjerling, M. Skovbro, R. Kraunsoe, and F. Dela. Patients with type 2 diabetes have normal mitochondrial function in skeletal muscle. Diabetologia, 2007. 50(4): 790‐6.
54. Hawley, J.A. and S.J. Lessard. Mitochondrial function: use it or lose it. Diabetologia, 2007. 50(4): 699‐702. 55. Short, K.R., K.S. Nair, and C.S. Stump. Impaired mitochondrial activity and insulin‐resistant offspring of
patients with type 2 diabetes. N Engl J Med, 2004. 350(23): 2419‐21; author reply 2419‐21. 56. Wagenmakers, A.J. Insulin resistance in the offspring of parents with type 2 diabetes. PLoS Med, 2005.
2(9): e289. 57. Brehm, A., M. Krssak, A.I. Schmid, P. Nowotny, W. Waldhausl, and M. Roden. Increased lipid availability
impairs insulin‐stimulated ATP synthesis in human skeletal muscle. Diabetes, 2006. 55(1): 136‐40. 58. Sreekumar, R., P. Halvatsiotis, J.C. Schimke, and K.S. Nair. Gene expression profile in skeletal muscle of
type 2 diabetes and the effect of insulin treatment. Diabetes, 2002. 51(6): 1913‐20. 59. Asmann, Y.W., C.S. Stump, K.R. Short, J.M. Coenen‐Schimke, Z. Guo, M.L. Bigelow, and K.S. Nair. Skeletal
muscle mitochondrial functions, mitochondrial DNA copy numbers, and gene transcript profiles in type 2 diabetic and nondiabetic subjects at equal levels of low or high insulin and euglycemia. Diabetes, 2006. 55(12): 3309‐19.
60. Koves, T.R., P. Li, J. An, T. Akimoto, D. Slentz, O. Ilkayeva, G.L. Dohm, Z. Yan, C.B. Newgard, and D.M. Muoio. Peroxisome proliferator‐activated receptor‐gamma co‐activator 1alpha‐mediated metabolic remodeling of skeletal myocytes mimics exercise training and reverses lipid‐induced mitochondrial inefficiency. J Biol Chem, 2005. 280(39): 33588‐98.
61. Noland, R.C., T.R. Koves, S.E. Seiler, H. Lum, R.M. Lust, O. Ilkayeva, R.D. Stevens, F.G. Hegardt, and D.M. Muoio. Carnitine insufficiency caused by aging and overnutrition compromises mitochondrial performance and metabolic control. J Biol Chem, 2009. 284(34): 22840‐52.
62. Adams, S.H., C.L. Hoppel, K.H. Lok, L. Zhao, S.W. Wong, P.E. Minkler, D.H. Hwang, J.W. Newman, and W.T. Garvey. Plasma acylcarnitine profiles suggest incomplete long‐chain fatty acid beta‐oxidation and altered tricarboxylic acid cycle activity in type 2 diabetic African‐American women. J Nutr, 2009. 139(6): 1073‐81.
63. Huffman, K.M., S.H. Shah, R.D. Stevens, J.R. Bain, M. Muehlbauer, C.A. Slentz, C.J. Tanner, M. Kuchibhatla, J.A. Houmard, C.B. Newgard, and W.E. Kraus. Relationships between circulating metabolic intermediates and insulin action in overweight to obese, inactive men and women. Diabetes Care, 2009. 32(9): 1678‐83.
64. Lien, L.F., A.M. Haqq, M. Arlotto, C.A. Slentz, M.J. Muehlbauer, R.L. McMahon, J. Rochon, D. Gallup, J.R. Bain, O. Ilkayeva, B.R. Wenner, R.D. Stevens, D.S. Millington, D.M. Muoio, M.D. Butler, C.B. Newgard, and L.P. Svetkey. The STEDMAN project: biophysical, biochemical and metabolic effects of a behavioral weight loss intervention during weight loss, maintenance, and regain. OMICS, 2009. 13(1): 21‐35.
65. Newgard, C.B., J. An, J.R. Bain, M.J. Muehlbauer, R.D. Stevens, L.F. Lien, A.M. Haqq, S.H. Shah, M. Arlotto, C.A. Slentz, J. Rochon, D. Gallup, O. Ilkayeva, B.R. Wenner, W.S. Yancy, Jr., H. Eisenson, G. Musante, R.S. Surwit, D.S. Millington, M.D. Butler, and L.P. Svetkey. A branched‐chain amino acid‐related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab, 2009. 9(4): 311‐26.
Chapter 1
26
66. Muoio, D.M. and T.R. Koves. Lipid‐induced metabolic dysfunction in skeletal muscle. Novartis Found Symp, 2007. 286: 24‐38; discussion 38‐46, 162‐3, 196‐203.
67. Guo, Z., P. Mishra, and S. Macura. Sampling the intramyocellular triglycerides from skeletal muscle. J Lipid Res, 2001. 42(7): 1041‐8.
68. Koopman, R., G. Schaart, and M.K. Hesselink. Optimisation of oil red O staining permits combination with immunofluorescence and automated quantification of lipids. Histochem Cell Biol, 2001. 116(1): 63‐8.
69. Levin, K., H. Daa Schroeder, F.P. Alford, and H. Beck‐Nielsen. Morphometric documentation of abnormal intramyocellular fat storage and reduced glycogen in obese patients with Type II diabetes. Diabetologia, 2001. 44(7): 824‐33.
70. Goodpaster, B.H., D.E. Kelley, F.L. Thaete, J. He, and R. Ross. Skeletal muscle attenuation determined by computed tomography is associated with skeletal muscle lipid content. J Appl Physiol, 2000. 89(1): 104‐10.
71. Boesch, C. and R. Kreis. Dipolar coupling and ordering effects observed in magnetic resonance spectra of skeletal muscle. NMR Biomed, 2001. 14(2): 140‐8.
72. Bartlett, K., S.J. Eaton, and M. Pourfarzam. New developments in neonatal screening. Arch Dis Child Fetal Neonatal Ed, 1997. 77(2): F151‐4.
73. Bartlett, K. and M. Pourfarzam. Tandem mass spectrometry‐‐the potential. J Inherit Metab Dis, 1999. 22(4): 568‐71.
74. An, J., D.M. Muoio, M. Shiota, Y. Fujimoto, G.W. Cline, G.I. Shulman, T.R. Koves, R. Stevens, D. Millington, and C.B. Newgard. Hepatic expression of malonyl‐CoA decarboxylase reverses muscle, liver and whole‐animal insulin resistance. Nat Med, 2004. 10(3): 268‐74.
75. van Vlies, N., L. Tian, H. Overmars, A.H. Bootsma, W. Kulik, R.J. Wanders, P.A. Wood, and F.M. Vaz. Characterization of carnitine and fatty acid metabolism in the long‐chain acyl‐CoA dehydrogenase‐deficient mouse. Biochem J, 2005. 387(Pt 1): 185‐93.
76. Simoneau, J.A. and D.E. Kelley. Altered glycolytic and oxidative capacities of skeletal muscle contribute to insulin resistance in NIDDM. J Appl Physiol, 1997. 83(1): 166‐71.
77. Kelley, D.E., J. He, E.V. Menshikova, and V.B. Ritov. Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes, 2002. 51(10): 2944‐50.
78. Heilbronn, L.K., S.K. Gan, N. Turner, L.V. Campbell, and D.J. Chisholm. Markers of mitochondrial biogenesis and metabolism are lower in overweight and obese insulin‐resistant subjects. J Clin Endocrinol Metab, 2007. 92(4): 1467‐73.
79. Hojlund, K., K. Wrzesinski, P.M. Larsen, S.J. Fey, P. Roepstorff, A. Handberg, F. Dela, J. Vinten, J.G. McCormack, C. Reynet, and H. Beck‐Nielsen. Proteome analysis reveals phosphorylation of ATP synthase beta ‐subunit in human skeletal muscle and proteins with potential roles in type 2 diabetes. J Biol Chem, 2003. 278(12): 10436‐42.
80. Morino, K., K.F. Petersen, S. Dufour, D. Befroy, J. Frattini, N. Shatzkes, S. Neschen, M.F. White, S. Bilz, S. Sono, M. Pypaert, and G.I. Shulman. Reduced mitochondrial density and increased IRS‐1 serine phosphorylation in muscle of insulin‐resistant offspring of type 2 diabetic parents. J Clin Invest, 2005. 115(12): 3587‐93.
81. Szendroedi, J., A.I. Schmid, M. Chmelik, C. Toth, A. Brehm, M. Krssak, P. Nowotny, M. Wolzt, W. Waldhausl, and M. Roden. Muscle mitochondrial ATP synthesis and glucose transport/phosphorylation in type 2 diabetes. PLoS Med, 2007. 4(5): e154.
82. Sirikul, B., B.A. Gower, G.R. Hunter, D.E. Larson‐Meyer, and B.R. Newcomer. Relationship between insulin sensitivity and in vivo mitochondrial function in skeletal muscle. Am J Physiol Endocrinol Metab, 2006. 291(4): E724‐8.
83. Lanza, I.R. and K.S. Nair. Functional assessment of isolated mitochondria in vitro. Methods Enzymol, 2009. 457: 349‐72.
84. Moon, R.B. and J.H. Richards. Determination of intracellular pH by 31P magnetic resonance. J Biol Chem, 1973. 248(20): 7276‐8.
85. Lawson, J.W. and R.L. Veech. Effects of pH and free Mg2+ on the Keq of the creatine kinase reaction and other phosphate hydrolyses and phosphate transfer reactions. J Biol Chem, 1979. 254(14): 6528‐37.
86. Wallimann, T., M. Wyss, D. Brdiczka, K. Nicolay, and H.M. Eppenberger. Intracellular compartmentation, structure and function of creatine kinase isoenzymes in tissues with high and fluctuating energy demands: the 'phosphocreatine circuit' for cellular energy homeostasis. Biochem J, 1992. 281 ( Pt 1): 21‐40.
87. Shoubridge, E.A., J.L. Bland, and G.K. Radda. Regulation of creatine kinase during steady‐state isometric twitch contraction in rat skeletal muscle. Biochim Biophys Acta, 1984. 805(1): 72‐8.
Introduction
27
88. Brindle, K.M., M.J. Blackledge, R.A. Challiss, and G.K. Radda. 31P NMR magnetization‐transfer measurements of ATP turnover during steady‐state isometric muscle contraction in the rat hind limb in vivo. Biochemistry, 1989. 28(11): 4887‐93.
89. Laurent, D., B. Yerby, R. Deacon, and J. Gao. Diet‐induced modulation of mitochondrial activity in rat muscle. Am J Physiol Endocrinol Metab, 2007. 293(5): E1169‐77.
90. Trenell, M.I., K.G. Hollingsworth, E.L. Lim, and R. Taylor. Increased Daily Walking Improves Lipid Oxidation Without Changes in Mitochondrial Function in Type 2 Diabetes. Diabetes Care, 2008. 31(8): 1644‐1649.
91. Yerby, B., R. Deacon, V. Beaulieu, J. Liang, J. Gao, and D. Laurent. Insulin‐stimulated mitochondrial adenosine triphosphate synthesis is blunted in skeletal muscles of high‐fat‐fed rats. Metabolism, 2008. 57(11): 1584‐90.
92. Brindle, K.M. and G.K. Radda. 31P‐NMR saturation transfer measurements of exchange between Pi and ATP in the reactions catalysed by glyceraldehyde‐3‐phosphate dehydrogenase and phosphoglycerate kinase in vitro. Biochim Biophys Acta, 1987. 928(1): 45‐55.
93. Kemp, G.J. The interpretation of abnormal 31P magnetic resonance saturation transfer measurements of Pi/ATP exchange in insulin‐resistant skeletal muscle. Am J Physiol Endocrinol Metab, 2008. 294(3): E640‐642.
94. Campbell‐Burk, S.L., K.A. Jones, and R.G. Shulman. 31P NMR saturation‐transfer measurements in Saccharomyces cerevisiae: characterization of phosphate exchange reactions by iodoacetate and antimycin A inhibition. Biochemistry, 1987. 26(23): 7483‐92.
95. Kingsley‐Hickman, P.B., E.Y. Sako, P. Mohanakrishnan, P.M. Robitaille, A.H. From, J.E. Foker, and K. Ugurbil. 31P NMR studies of ATP synthesis and hydrolysis kinetics in the intact myocardium. Biochemistry, 1987. 26(23): 7501‐10.
96. Hood, D.A., J. Gorski, and R.L. Terjung. Oxygen cost of twitch and tetanic isometric contractions of rat skeletal muscle. Am J Physiol, 1986. 250(4 Pt 1): E449‐56.
97. Taylor, D.J., P.J. Bore, P. Styles, D.G. Gadian, and G.K. Radda. Bioenergetics of intact human muscle. A 31P nuclear magnetic resonance study. Mol Biol Med, 1983. 1(1): 77‐94.
98. Quistorff, B., L. Johansen, and K. Sahlin. Absence of phosphocreatine resynthesis in human calf muscle during ischaemic recovery. Biochem J, 1993. 291 ( Pt 3): 681‐6.
99. van den Broek, N.M., H.M. De Feyter, L. de Graaf, K. Nicolay, and J.J. Prompers. Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics in muscle after exercise are due to differences in proton efflux rates. Am J Physiol Cell Physiol, 2007. 293(1): C228‐37.
100. Haseler, L.J., A.P. Lin, and R.S. Richardson. Skeletal muscle oxidative metabolism in sedentary humans: 31P‐MRS assessment of O2 supply and demand limitations. J Appl Physiol, 2004. 97(3): 1077‐81.
101. Mahler, M. The relationship between initial creatine phosphate breakdown and recovery oxygen consumption for a single isometric tetanus of the frog sartorius muscle at 20 degrees C. J Gen Physiol, 1979. 73(2): 159‐74.
102. Larson‐Meyer, D.E., B.R. Newcomer, G.R. Hunter, D.R. Joanisse, R.L. Weinsier, and M.M. Bamman. Relation between in vivo and in vitro measurements of skeletal muscle oxidative metabolism. Muscle Nerve, 2001. 24(12): 1665‐76.
103. Praet, S.F., H.M. De Feyter, R.A. Jonkers, K. Nicolay, C. van Pul, H. Kuipers, L.J. van Loon, and J.J. Prompers. 31P MR spectroscopy and in vitro markers of oxidative capacity in type 2 diabetes patients. Magma, 2006. 19(6): 321‐31.
104. Argov, Z., N. De Stefano, and D.L. Arnold. ADP recovery after a brief ischemic exercise in normal and diseased human muscle‐‐a 31P MRS study. NMR Biomed, 1996. 9(4): 165‐72.
105. Arnold, D.L., P.M. Matthews, and G.K. Radda. Metabolic recovery after exercise and the assessment of mitochondrial function in vivo in human skeletal muscle by means of 31P NMR. Magn Reson Med, 1984. 1(3): 307‐15.
106. Bendahan, D., S. Confort‐Gouny, G. Kozak‐Reiss, and P.J. Cozzone. Heterogeneity of metabolic response to muscular exercise in humans. New criteria of invariance defined by in vivo phosphorus‐31 NMR spectroscopy. FEBS Lett, 1990. 272(1‐2): 155‐8.
107. Boska, M.D., J.A. Nelson, N. Sripathi, Pipinos, II, A.D. Shepard, and K.M. Welch. 31P MRS studies of exercising human muscle at high temporal resolution. Magn Reson Med, 1999. 41(6): 1145‐51.
108. Iotti, S., R. Lodi, C. Frassineti, P. Zaniol, and B. Barbiroli. In vivo assessment of mitochondrial functionality in human gastrocnemius muscle by 31P MRS. The role of pH in the evaluation of phosphocreatine and inorganic phosphate recoveries from exercise. NMR Biomed, 1993. 6(4): 248‐53.
109. Lodi, R., G.J. Kemp, S. Iotti, G.K. Radda, and B. Barbiroli. Influence of cytosolic pH on in vivo assessment of human muscle mitochondrial respiration by phosphorus magnetic resonance spectroscopy. Magma, 1997. 5(2): 165‐71.
Chapter 1
28
110. McCully, K.K., S. Iotti, K. Kendrick, Z. Wang, J.D. Posner, J. Leigh, Jr., and B. Chance. Simultaneous in vivo measurements of HbO2 saturation and PCr kinetics after exercise in normal humans. J Appl Physiol, 1994. 77(1): 5‐10.
111. Roussel, M., D. Bendahan, J.P. Mattei, Y. Le Fur, and P.J. Cozzone. 31P magnetic resonance spectroscopy study of phosphocreatine recovery kinetics in skeletal muscle: the issue of intersubject variability. Biochim Biophys Acta, 2000. 1457(1‐2): 18‐26.
112. Takahashi, H., M. Inaki, K. Fujimoto, S. Katsuta, I. Anno, M. Niitsu, and Y. Itai. Control of the rate of phosphocreatine resynthesis after exercise in trained and untrained human quadriceps muscles. Eur J Appl Physiol Occup Physiol, 1995. 71(5): 396‐404.
113. Taylor, D.J., P. Styles, P.M. Matthews, D.A. Arnold, D.G. Gadian, P. Bore, and G.K. Radda. Energetics of human muscle: exercise‐induced ATP depletion. Magn Reson Med, 1986. 3(1): 44‐54.
114. Walter, G., K. Vandenborne, K.K. McCully, and J.S. Leigh. Noninvasive measurement of phosphocreatine recovery kinetics in single human muscles. Am J Physiol, 1997. 272(2 Pt 1): C525‐34.
115. De Feyter, H.M., E. Lenaers, S.M. Houten, P. Schrauwen, M.K. Hesselink, R.J. Wanders, K. Nicolay, and J.J. Prompers. Increased intramyocellular lipid content but normal skeletal muscle mitochondrial oxidative capacity throughout the pathogenesis of type 2 diabetes. FASEB J, 2008. 22(11): 3947‐55.
29
30
Adapted from
N. M. A. van den Broek, J. Ciapaite, K. Nicolay, J. J. Prompers.
Comparison of in vivo post‐exercise phosphocreatine recovery and resting ATP synthesis flux
for the assessment of skeletal muscle mitochondrial function.
Am J Physiol Cell Physiol, in press.
Chapter Comparison of in vivo post‐exercise
phosphocreatine recovery and resting ATP
synthesis flux for the assessment of
skeletal muscle mitochondrial function
Chapter 2
32
Abstract 31P magnetic resonance spectroscopy (MRS) has been used to assess skeletal muscle
mitochondrial function in vivo by measuring (1) phosphocreatine (PCr) recovery after
exercise or (2) resting ATP synthesis flux with saturation transfer (ST). In this study, we
compared both parameters in a rat model of mitochondrial dysfunction with the aim of
establishing the most appropriate method for the assessment of in vivo muscle
mitochondrial function. Mitochondrial dysfunction was induced in adult Wistar rats by daily
subcutaneous injections with the complex I inhibitor diphenyleneiodonium (DPI) for two
weeks. In vivo 31P MRS measurements were supplemented by in vitro measurements of
oxygen consumption in isolated mitochondria. Two weeks of DPI treatment induced
mitochondrial dysfunction, as evidenced by a 20% lower maximal ADP‐stimulated oxygen
consumption rate in isolated mitochondria from DPI‐treated rats oxidizing pyruvate plus
malate. This was paralleled by a 46% decrease in in vivo oxidative capacity, determined
from post‐exercise PCr recovery. Interestingly, no significant difference in resting, ST‐based
ATP synthesis flux was observed between DPI‐treated rats and controls. These results show
that PCr recovery after exercise has a more direct relationship with skeletal muscle
mitochondrial function than the ATP synthesis flux measured with 31P ST MRS in the resting
state.
Assessment of in vivo muscle mitochondrial function
33
Introduction
Mitochondria play a pivotal role in many cellular processes, the most important function
being the production of energy in the form of ATP through a process termed oxidative
phosphorylation. The last decade, mitochondria gained interest in the field of insulin
resistance (IR) and type 2 diabetes (T2D) [1‐6]. Based on the in vivo observation that ATP
synthesis flux in resting skeletal muscle is lower in insulin‐resistant subjects and offspring of
T2D patients compared to healthy controls [7‐8], it has been hypothesized that skeletal
muscle mitochondrial dysfunction is a predisposing factor for IR and/or T2D. The proposed
mechanism links muscle mitochondrial dysfunction to impaired fatty acid metabolism,
which subsequently leads to the accumulation of intramyocellular lipids and lipid
intermediates (e.g. diacylglycerol and ceramides) that interfere with the insulin signaling
cascade [9].
The role of skeletal muscle mitochondrial dysfunction in the development of IR and/or T2D
has been investigated using a variety of techniques [7‐8, 10‐22]. In vitro methodologies, like
the determination of gene expression levels, enzyme activities, mitochondrial content,
morphology and respiration, provide specific information on different aspects of
mitochondrial energy production, but the results cannot be directly translated to in vivo
mitochondrial function. 31P magnetic resonance spectroscopy (MRS) provides a non‐
invasive tool to monitor the energetic status of the cell in vivo by measuring intracellular
phosphorous containing metabolites, i.e. phosphocreatine (PCr), ATP and inorganic
phosphate (Pi). 31P MR spectra of resting skeletal muscle are relatively constant, even in
diseased states, and to assess impairments in mitochondrial energy production one needs
to perturb either the chemical or the magnetic equilibrium as described below.
The resting Pi ATP flux (VATP) can be determined by saturating the ‐ATP peak and monitoring the effect of this perturbation on the Pi magnetization in a so‐called 31P
saturation transfer (ST) experiment [23‐24]. Assuming that VATP is predominantly reflecting
oxidative ATP synthesis by the F1F0‐ATPase in the mitochondria, VATP has been taken as a
measure for mitochondrial function [8, 19]. However, the interpretation of 31P ST data is
not straightforward. The lower ATP synthesis rates in resting muscle of insulin‐resistant
subjects [7‐8] could actually reflect a normal regulatory response to a lower energy
demand caused by impaired insulin signaling, rather than an impairment of intrinsic
mitochondrial function [25‐28]. Moreover, the Pi ATP fluxes obtained from 31P ST
measurements are comprised of both ATP synthase flux and glycolytic exchange flux, with
the latter contributing by as much as 80% at rest [24, 26, 29‐31]. Therefore, decreased
resting VATP does not necessarily reflect a mitochondrial defect.
As an alternative to the resting state ST experiment described above, the metabolic steady
state of the muscle, i.e. the chemical equilibrium, can be disturbed during exercise. During
recovery from exercise, the PCr pool is replenished purely through oxidative ATP synthesis
Chapter 2
34
[32‐34]. Because the creatine kinase reaction is much faster than oxidative ATP production
[35], the measurement of PCr recovery using dynamic 31P MRS after exercise provides an
alternative method to determine the rate of oxidative ATP synthesis. The PCr recovery rate
constant (kPCr) thus reflects in vivo muscle mitochondrial oxidative capacity, i.e. the maximal
capacity for oxidative ATP production, which is typically one order of magnitude higher
than the ATP synthesis rate at rest. Therefore, the post‐exercise PCr recovery rate constant
might be a more suitable measure for in vivo mitochondrial function as compared to the
resting ATP synthesis flux.
In this study, we compared in vivo 31P MRS post‐exercise PCr recovery and resting ATP
synthesis flux in a rat model of mitochondrial dysfunction with the aim of establishing the
most appropriate method for the assessment of in vivo skeletal muscle mitochondrial
function. Mitochondrial dysfunction was induced by daily subcutaneous injections with
diphenyleneiodonium (DPI), which irreversibly inhibits complex I (NADH‐ubiquinone
reductase) of the respiratory chain [36‐39]. In vivo measurements were supplemented by in
vitro measurements of oxygen consumption in isolated mitochondria to confirm inhibition
of complex I and to compare in vivo and in vitro measurements of mitochondrial function.
Materials and Methods
Animals
Adult male Wistar rats (364 ± 18 g, 14 weeks old, n = 16, Charles River Laboratories, The
Netherlands) were housed in pairs at 20°C and 50% humidity, with a 12‐h light‐dark cycle.
Rats were given ad libitum access to water and normal chow (9.2% calories from fat, 67.2%
calories from carbohydrates, 23.6% calories from proteins (R/M‐H diet, Ssniff Spezialdiäten
GmbH, Soest, Germany)). Diphenyleneiodonium sulphate (DPI, Toronto Research Chemicals
inc., North York, Ontario (ON), Canada) was dissolved in a warm 5% (w/v) glucose solution
(1.3 mg/ml) and was injected daily subcutaneously at 1 mg/kg body weight for 11 ± 2 days
(n = 8) [39]. Control animals received similar volumes of subcutaneously injected 5% (w/v)
glucose solution (n = 8).
The day after in vivo MRS measurements, rats were sacrificed by incising the inferior vena
cava under anesthesia. Immediately thereafter, half of one tibialis anterior (TA) was freeze
clamped using liquid nitrogen‐cooled tongs [40] and stored at ‐80°C for ATP determination.
The other part of the TA was frozen in liquid nitrogen and stored at ‐80°C for the
determination of mitochondrial DNA (mtDNA) content. The second TA muscle (fresh) was
used for isolation of mitochondria. Animal handling conformed to the rules of the Animal
Ethics Committee of Maastricht University.
Assessment of in vivo muscle mitochondrial function
35
31P magnetic resonance spectroscopy
All magnetic resonance spectroscopy (MRS) measurements were performed on a 6.3 Tesla
horizontal Bruker MR system (Bruker, Ettlingen, Germany), 14 days after the start of
treatment. Animals were anaesthetized using isoflurane (Forene®) (1‐2%) with medical air
(0.6 L/min) and body temperature was maintained at 35.5 ± 0.5C using heating pads. Respiration was monitored using a pressure sensor registering thorax movement (Rapid
Biomedical, Rimpar, Germany).
Resting ATP synthesis flux (VATP) and the post‐exercise PCr recovery rate constant (kPCr)
were measured in the TA by in vivo 31P MRS. 31P MRS was applied using a combination of a
circular 1H surface coil (40 mm) for making MR images and localized shimming and an
ellipsoid 31P MRS surface coil (10/18 mm), positioned over the TA as described previously
[41]. 31P MR spectra were acquired applying an adiabatic excitation pulse with a flip angle
of 90. A fully relaxed (TR = 25 s, 32 averages) spectrum was measured at rest, followed by
the ST experiment in resting TA muscle to determine VATP. Two spectra (TR = 10.4 s, 128
averages) were acquired for each saturation transfer experiment: A spectrum with
frequency‐selective saturation of the ‐ATP peak yielding the steady state Pi magnetization
in the presence of saturation (M’) and a reference spectrum with saturation at a downfield
frequency, equidistant from Pi, yielding the equilibrium Pi magnetization (M0), both with a
saturation pulse length of 10 s. The apparent longitudinal relaxation time of Pi (T1’) was
determined by performing an 8‐point inversion recovery experiment with an adiabatic full
passage pulse for inversion and with ‐ATP saturation prior to (10 s) and during the inversion delay (inversion times = 0.01, 1, 2, 4, 6.5, 9.5, 13 and 17 s, 32 averages). The total
duration of the ST and inversion recovery experiments was about 2 hours.
After the ST experiments, time series of 31P MR spectra (TR = 5 s, 4 averages) before, during
and after muscle contractions were acquired. Muscle contractions were induced by
electrical stimulation of the TA, via subcutaneously implanted electrodes positioned along
the distal N. Peroneus Communis [41]. The stimulation protocol consisted of a series of
stimulation pulses, applied every second, for a duration of 2 min. Stimulation pulse length
was 100 ms, frequency was 80 Hz and stimulation voltage varied between 2.5 and 4 Volt.
Recovery was followed for 15 minutes. Three to four time series were measured for each
rat.
31P MR spectra were fit in the time domain by using an advanced magnetic resonance
(AMARES) nonlinear least squares algorithm in the jMRUI software package (jMRUI V2.1).
PCr and Pi peaks were fit to Lorentzian line shapes, whereas ‐, ‐, and ‐ATP signals were fit with Gaussian line shapes. Besides the cytosolic Pi signal a second, smaller Pi peak was
observed in the spectra at a frequency ~0.3 ppm downfield from the cytosolic Pi resonance.
The two Pi signals were separately fitted and the line widths were constrained with respect
to the line width of the PCr signal. For the dynamic MRS spectra, the two Pi signals could
Chapter 2
36
not be distinguished and a single Pi peak was fit to a Gaussian line shape. Concentrations of
PCr and Pi were determined relative to the ATP concentration, which was determined in
vitro from freeze clamped tissue as described below. ADP concentrations were calculated
using the creatine kinase equilibrium [42] and intracellular pH was calculated from the
chemical shift difference between PCr and the cytosolic Pi resonance [34].
The T1’ relaxation time of Pi was determined by fitting the inversion recovery data with a 3‐
parameter mono‐exponential function using Origin (OriginPro 7.5 SR0, OriginLab
Corporation, Northampton, MA, USA). The Pi ATP exchange rate constant ( ) was
calculated from the T1’ of Pi and the fractional reduction of Pi magnetization upon selective
saturation of ‐ATP according to: = 1 / . VATP was then calculated by
multiplying with the Pi concentration at rest, as determined from the fully relaxed
spectrum. The intrinsic T1 of Pi was calculated as 1/ . The data of PCr
recovery were fit to a mono‐exponential function using Matlab (version 7.04, Mathworks,
Natick, MA, USA) yielding a rate constant, kPCr. Results from two time series with end‐
stimulation pH values higher than 6.92 [43] were averaged.
TA muscle cross‐sectional area in the midbelly region of the muscle was determined from
the MR images using Mathematica (version 7.0, Wolfram Research, Champaign, IL, USA).
Determination of ATP concentration
Freeze clamped TA muscle was powdered in liquid nitrogen, mixed with 3.25 ml/g
perchloric acid (6% w/v) and homogenized for 30 s using an Ultra‐Turrax® high‐
performance disperser (IKA® Werke GmbH & Co. KG, Staufen, Germany). Next, the
homogenate was neutralized with 5 M K2CO3 to pH 7.4 and centrifuged at 3000 g for 10 min
at 4C. ATP concentration in the supernatant was determined spectrophotometrically using
the hexokinase/glucose‐6‐phosphate dehydrogenase coupled assay as described in detail in
[44].
Isolation of mitochondria
Mitochondria were isolated from one whole TA muscle through a differential centrifugation
procedure as previously described [45]. Briefly, TA muscle was excised, washed in ice cold
0.9% KCl, freed of connective and adipose tissue, weighed and minced with scissors in ice
cold medium A (5 ml for 1 g tissue) containing 150 mM sucrose, 75 mM KCl, 50 mM MOPS,
1 mM KH2PO4, 5 mM MgCl2, 1 mM EGTA, 0.4 mg/ml bacterial proteinase type XXIV, pH 7.4.
Next, 20 ml of medium B containing 250 mM sucrose, 0.1 mM EGTA and 20 mM MOPS, 2
mg/ml BSA, pH 7.4 was added and the mixture was homogenized using Potter‐Elvehjem
Assessment of in vivo muscle mitochondrial function
37
homogenizer. The homogenate was centrifuged at 800 g for 10 min, 4C. The resulting supernatant was centrifuged at 10000 g for 10 min, 4C. The pellet was resuspended in 15 ml of fresh ice cold medium B and centrifuged at 10000 g for 10 min, 4C. Mitochondrial
pellet was resuspended in 100 l of medium B. Protein content was determined using BCA
protein assay kit (Pierce, Thermo Fisher Scientific Inc., Rockford, IL, USA).
Measurement of oxygen consumption
Oxygen consumption rate was measured at 37 °C using a two‐channel high‐resolution
Oroboros oxygraph‐2 k (Oroboros, Innsbruck, Austria). Maximal ADP‐stimulated (state 3)
oxygen consumption rate was determined in the assay medium containing 110 mM KCl, 20
mM Tris, 2.3 mM MgCl2, 5 mM KH2PO4, 50 mM creatine, 4.4 U/ml creatine phosphokinase,
1 mM ATP, 1 mg/ml BSA, pH 7.3. All measurements were performed in 1 ml of assay
medium containing 0.15 mg/ml of mitochondrial protein. To assess NADH‐ (through
complex I) and FADH2‐driven (through complex II) oxygen consumption, 5 mM pyruvate
plus 5 mM malate and 5 mM succinate plus 1 µM rotenone was used, respectively. Oxygen
consumption in resting state (state 4) was determined after addition of 1.25 M
carboxyatractyloside. The signals from the oxygen electrode were recorded at 0.5 s
intervals. Data acquisition and analysis was performed using Oxygraph‐2k‐DatLab software
version 4.2 (Oroboros, Innsbruck, Austria).
Determination of the relative mitochondrial copy number
Genomic DNA was isolated from a 25 mg transversal slice of mid‐belly TA using GenElute™
Mammalian Genomic DNA Miniprep Kit (Sigma‐Aldrich, Zwijndrecht, The Netherlands).
Mitochondrial DNA (mtDNA) content relative to peroxisome proliferator‐activated
receptor‐ coactivator 1 (PGC‐1) gene was measured using real‐time PCR as described in
[46]. Primers for mtDNA were: forward primer – 5’‐ACACCAAAAGGACGAACCTG‐3’, reverse
primer ‐ 5’‐ATGGGGAAGAAGCCCTAGAA‐3’, and for PGC‐1: forward primer – 5’‐
ATGAATGCAGCGGTCTTAGC‐3’, reverse primer – 5’‐AACAATGGCAGGGTTTGTTC‐3’. The
relative mtDNA copy number was calculated using Ct method as described in [47].
Statistical analysis
Data are presented as means ± SD. The listed n values represent the number of animals
used for a particular experiment. Statistical significance of the differences was assessed
using unpaired Student’s t‐tests in SPSS 16.0 statistical package (SPSS Inc., Chicago, IL, USA).
Level of statistical significance was set at p < 0.05.
Chapter 2
38
Results
Animal model
Daily DPI treatment resulted in a decreased food intake (22.1 ± 1.0 and 10.8 ± 2.9 g/day for
control and DPI‐treated rats, respectively; p < 0.001) and a lower body weight (BW) at the
end of the DPI treatment (375 ± 24 and 330 ± 21 g for control and DPI‐treated rats,
respectively; p = 0.001). In parallel with the 12% lower BW, TA muscle cross‐sectional area
was 10% smaller in DPI‐treated rats compared to controls at the end of the treatment (0.43
± 0.03 and 0.38 ± 0.03 cm2 for control and DPI‐treated rats, respectively; p = 0.01).
Mitochondrial characteristics in vitro
To assess the effect of DPI on the intrinsic in vitro mitochondrial function we determined
oxygen consumption rates in mitochondria isolated from the TA muscle (results are
summarized in Table 1). Maximal ADP‐stimulated (state 3) oxygen consumption rate in the
isolated mitochondria oxidizing pyruvate plus malate was 20% lower (p < 0.001) in the DPI‐
treated group compared to controls. The effect of DPI treatment on succinate oxidation
was less pronounced; oxygen consumption rate in state 3 was 12% lower (p = 0.03)
compared to controls. For both substrates oxygen consumption rates in the absence of ATP
synthesis (state 4) were similar in DPI‐treated and control groups.
The effect of chronic inhibition of intrinsic mitochondrial function by DPI on mitochondrial
biogenesis was assessed by determining mtDNA copy number in TA muscle. The relative
mtDNA copy number was not significantly different in the TA muscle of control (2200 ±
323) and DPI‐treated (2190 ± 369) rats.
Table 1. Oxygen consumption rates in isolated TA mitochondria oxidizing different substrates in different
metabolic states.
Pyruvate plus malate Succinate plus rotenone
Control DPI Control DPI
State 3 (nmol O2min
‐1∙mg protein
‐1) 408 ± 21 326 ± 38* 349 ± 28 308 ± 37*
State 4 (nmol O2min‐1∙mg protein
‐1) 19 ± 3 20 ± 4 55 ± 12 52 ± 8
Data are from n = 7 control and n = 8 DPI‐treated animals and are expressed as means ± SD. DPI,
diphenyleneiodonium‐treated rats; State 3, maximal ADP‐stimulated oxygen consumption; State 4,
oxygen consumption in the absence of ATP synthesis. *p < 0.05 when compared to control group.
Assessment of in vivo muscle mitochondrial function
39
Muscle ATP concentration
ATP concentrations in TA muscle were 12.5 ± 3.2 mM for control rats (n = 5) and 11.0 ± 2.5
mM for DPI‐treated rats (n = 7) and were not significantly different (p = 0.4). Because DPI
treatment had no significant effect on the muscle ATP concentration, the average value for
both groups combined (11.6 ± 2.8 mM) was taken as a reference for the in vivo MRS data
analysis.
In vivo 31P MRS
Concentrations of PCr, Pi and ADP and the intracellular pH obtained from the 31P MR
spectra recorded in resting TA muscle are summarized in Table 2. DPI treatment had no
significant effect on any of these parameters.
Table 2. Metabolite concentrations and pH in TA measured by in vivo 31P MRS.
Rest End stimulation
Control DPI Control DPI
pH 7.19 ± 0.02 7.21 ± 0.02 7.07 ± 0.06 7.01 ± 0.06
[PCr] (mM)
47.5 ± 2.8 48.8 ± 2.9 22.8 ± 2.7 20.9 ± 1.2
[Pi] (mM) 3.0 ± 0.4 3.0 ± 0.8 28.6 ± 3.5 31.9 ± 2.8
[ADP] (μM) 13.4 ± 0.6 13.9 ± 0.4 84.8 ± 10.2 89.5 ± 16.7
Data are from n = 8 animals for each group and are expressed as means ± SD. PCr, phosphocreatine; Pi,
inorganic phosphate.
Figure 1. Example of 31P saturation transfer
spectra (saturation pulse length = 10 s, TR =
10.4 s, 128 averages) with saturation on ‐ATP (black) and with saturation at a downfield
frequency, equidistant from Pi (grey),
measured in TA muscle of a DPI‐treated rat.
Chapter 2
40
Figure 1 shows a representative example of a 31P ST spectrum measured in resting TA
muscle with frequency‐selective saturation of the ‐ATP peak (black) and the corresponding reference spectrum with saturation at a downfield frequency equidistant from Pi (grey).
From the 31P ST spectra, the ratio of Pi magnetization with and without saturation of ‐ATP ( ’/ ) was determined and this ratio was not significantly different between the groups
(Table 3). The apparent T1 (T1’) of Pi in case of saturation of ‐ATP was determined from an
8‐point inversion recovery experiment and a representative example of Pi inversion
recovery data and the corresponding mono‐exponential fit is shown in Figure 2. T1’ values
of Pi were similar for control and DPI‐treated animals (Table 3). The exchange rate constant
determined from T1’ and ’/ , VATP determined from and the Pi
concentration, and the intrinsic T1 of Pi were also not significantly different between control
and DPI‐treated animals (Table 3 and Figure 3A).
Figure 2. Pi amplitudes obtained
from an inversion recovery
experiment with inversion times of
0.01, 1, 2, 4, 6.5, 9.5, 13 and 17 s and
with saturation of the ‐ATP peak. A mono‐exponential 3‐paramenter
function (dark line) was fit to the
actual data (filled circles) to
determine T1’, which was 3.75 s for
this example.
Table 3. Parameters determined from 31P saturation transfer MRS.
Control DPI
’/ 0.73 ± 0.06 0.67 ± 0.08
(s) 3.87 ± 0.37 3.76 ± 0.79
(s) 5.38 ± 0.79 5.69 ± 1.33
(s‐1) 0.07 ± 0.02 0.09 ± 0.04
VATP (μmol/g/min) 8.33 ± 1.53 10.45 ± 2.82
Data are from n = 7 control and n = 8 DPI‐treated animals and are expressed as means ± SD. M’,
magnetization of Pi with saturation of ‐ATP; M0, magnetization of Pi with saturation at a downfield
frequency, equidistant from Pi; , apparent longitudinal relaxation time of Pi with saturation of ‐ATP;
, intrinsic longitudinal relaxation time of Pi; , exchange rate constant; VATP, resting ATP synthesis
flux.
Assessment of in vivo muscle mitochondrial function
41
Dynamic 31P MRS measurements during and after recovery from electrical stimulation of
the TA were performed to determine the rate of PCr recovery after stimulation.
Concentrations of PCr, Pi and ADP and the intracellular pH at the end of the stimulation
were not significantly different between groups (Table 2). PCr concentrations during
recovery were fitted with a mono‐exponential function (Figure 4) yielding kPCr. kPCr was 46%
lower in DPI‐treated animals as compared to control rats (Figure 3B, p < 0.001). There was
no relationship between VATP and kPCr (R2 = 0.15, p = 0.15).
Discussion
In this study, we compared in vivo 31P MRS post‐exercise PCr recovery and resting ATP
synthesis flux in a rat model of mitochondrial dysfunction to establish the most appropriate
method for assessment of in vivo skeletal muscle mitochondrial function. Two weeks of
treatment with the complex I inhibitor DPI induced mitochondrial dysfunction, as
evidenced by a 20% lower oxygen consumption rate in isolated mitochondria from DPI‐
treated rats oxidizing pyruvate plus malate. This was paralleled by a 46% decrease in in vivo
oxidative capacity, determined from post‐exercise PCr recovery. Interestingly, no significant
difference in resting ST‐based ATP synthesis flux was observed between DPI‐treated rats
and controls.
DPI is a non‐specific irreversible inhibitor of flavoenzymes, which causes covalent
phenylation of either the flavin or the adjacent amino acid and heme groups of the proteins
due to reduction of DPI to its diphenyleneiodonyl radical form during electron transport
through the flavin moieties of the enzymes [48]. The effects of DPI on mitochondrial
function have been studied extensively in vitro [37, 49‐50]. The acute inhibitory effect of
DPI in submitochondrial particles and isolated mitochondria prepared from different rat
tissues is primarily related to irreversible inhibition of the respiratory chain complex I,
which contains flavin adenine dinucleotide as the enzyme cofactor [37, 49]. It has been
shown that DPI also decreases succinate state 3 oxidation in isolated rat skeletal muscle
mitochondria, suggesting additional inhibition of respiratory chain complex II [51].
Figure 3. (A) Resting ATP synthesis flux
(VATP; data are from n = 7 control and n =
8 DPI‐treated animals) measured by 31P
saturation transfer MRS and (B) PCr
recovery rate constant (kPCr; n = 8 for both
groups) measured with dynamic 31P MRS
after muscle stimulation. Open bars
represent control animals, filled bars
represent DPI‐treated rats. *p < 0.05
compared to controls.
Chapter 2
42
However, the observation that this inhibition is dependent on the chloride concentration in
the assay medium indicates a different potency or mechanism of action of DPI on complex
II [52], which may be related to the fact that complex I and II contain a different flavin
cofactor (i.e. complex II contains flavin mononucleotide). Furthermore, it has been shown
that the transport of respiratory substrates into the mitochondrial matrix in isolated rat
liver mitochondria is not affected by DPI [50, 53].
One may expect that the irreversible modifications of the respiratory chain complexes I and
II observed in vitro also occur in vivo in animals chronically treated with DPI. Indeed, it has
been shown that state 3 oxygen consumption, driven by oxidation of various NADH‐
delivering substrates, was significantly decreased in mitochondria isolated from
gastrocnemius and soleus muscles of rats after 14 days of daily DPI treatment compared to
healthy controls [38]. This effect could be fully explained by specific inhibition of complex I
activity. In agreement, we found that DPI treatment of similar duration as in [38] results in
a significant and comparable decrease in pyruvate plus malate‐driven oxygen consumption
in state 3 in mitochondria isolated from TA muscle. In addition, we observed a small (12%)
but significant decrease in the succinate‐driven oxygen consumption in state 3. The
available published data on the effects of in vivo DPI treatment on succinate‐driven state 3
oxygen consumption in isolated skeletal muscle mitochondria are scarce and contradictory.
Cooper et al. found no significant impairment of succinate‐driven state 3 oxygen
Figure 4. Representative examples of relative PCr concentrations during rest, muscle stimulation and recovery (time resolution = 20 s) for a control animal (open symbols) and a DPI‐treated rat (filled symbols). PCr concentrations are expressed as a percentage of the resting PCr concentration (left y‐axis is for the control rat, right y‐axis for the DPI‐treated rat). Mono‐exponential functions (dark lines) were fit to the recovery data and the PCr recovery rate constants were 0.61 and 0.31 min
‐1 for
the control and DPI‐treated animal, respectively.
Assessment of in vivo muscle mitochondrial function
43
consumption in isolated skeletal muscle mitochondria [38], while Hayes et al. reported an
80% decrease in state 3 succinate oxidation in skeletal muscle mitochondria isolated from
DPI‐treated animals compared to controls [51]. The different results are likely due to
differences in the concentration of chloride in the assay medium, which affects the potency
of DPI to inhibit succinate‐driven state 3 oxygen consumption (see above) [52]. Because of
the low intracellular chloride concentration, the most important mode of action of DPI in
vivo is probably inhibition of complex I.
In the present study, mtDNA copy number in TA muscle was similar for DPI‐treated and
control rats, indicating that DPI treatment did not affect mitochondrial biogenesis. This is in
agreement with Cooper et al. who showed similar activities of citrate synthase and other
TCA cycle enzymes in the muscle of DPI‐treated animals compared to healthy controls [38].
The observation that DPI treatment does not cause compensatory stimulation of
mitochondrial biogenesis simplifies the interpretation of the in vivo determined PCr
recovery rate constant kPCr. It implies that the decrease in kPCr in TA muscle of DPI‐treated
animals results from impaired intrinsic mitochondrial function rather than decreased
mitochondrial number. However, the discrepancy in the magnitude of the effect of DPI
observed on the in vivo determined kPCr (46% decrease) and the in vitro mitochondrial
function (20% decrease of state 3 oxygen consumption with pyruvate plus malate and 12%
decrease of state 3 oxygen consumption with succinate plus malate) suggests that other
factors than inhibition of the respiratory chain complex I and complex II may additionally
affect PCr recovery in vivo. For example, inhibition of the creatine kinase reaction by DPI
could slow down the recovery of PCr in vivo. However, from our 31P ST data no such
inhibitory effect was observed for the PCr ATP flux (data not shown), which is in
agreement with observations in a previous study [51]. Alternatively, a lowered muscle
tissue perfusion could be responsible for a slower PCr recovery by reducing the availability
of oxygen and substrate. DPI has been shown to inhibit endothelial nitric oxide synthase
(eNOS), an enzyme involved in vascular vasodilation [54]. Inhibition of eNOS lowers muscle
blood flow during recovery from exercise in humans [55‐56] and therefore we cannot
exclude that a lower muscle perfusion in DPI‐treated animals may have affected kPCr.
In contrast to the oxygen consumption in isolated mitochondria and the in vivo PCr
recovery, resting ST‐based ATP synthesis flux, VATP, was not affected by DPI treatment.
Mitochondrial ATP synthesis is a demand driven process regulated by several error signals.
Most investigated is the feedback control loop involving changes in the extra‐mitochondrial
concentrations of the ATP hydrolysis products ADP [57‐59] and, to a lesser extent, Pi [59]. In
order to interpret ATP synthesis flux data in terms of mitochondrial function, it is essential
to take the error signals, i.e. concentrations of ADP and Pi, into account [26]. A lower ATP
synthesis flux in combination with low error signals simply represents a lower ATP demand.
On the other hand, a lower ATP synthesis flux in combination with normal or high error
signals would indeed imply mitochondrial dysfunction or a decrease in the density of
mitochondria. In our study, DPI‐treated rats had both normal resting ATP synthesis flux and
Chapter 2
44
a normal resting Pi concentration, whereas the ADP concentration was slightly, but not
significantly elevated with respect to controls. These results imply that the resting ATP
demand was not affected by treatment with DPI. The slightly increased ADP level in DPI‐
treated rats suggests that a higher error signal was needed to maintain a normal ATP
synthesis flux. Alternatively, another, yet unidentified error signal might have been
increased in DPI‐treated rats in order to match the normal ATP demand.
Another factor which complicates the interpretation of 31P MRS ST data is that the Pi ATP
flux obtained from 31P MRS ST measurements is comprised of both mitochondrial F1F0‐ATP
synthase flux and flux through other Pi ATP pathways, in particular the reactions
catalyzed by the glycolytic enzymes glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)
and phosphoglycerate kinase (PGK). Although the net glycolytic contribution to the
production of ATP (via GAPDH and PGK) versus that of oxidative phosphorylation is small
[60], these enzymes catalyze a coupled near‐equilibrium reaction, and consequently,
exchange between Pi and ATP may greatly exceed the net glycolytic flux [29]. Experimental
proof for glycolytic exchange between Pi and ATP came from 31P MRS ST experiments in
yeast. Pi ATP exchange was inhibited in yeast by iodoacetate, an irreversible inhibitor of
GAPDH [30], whereas in anaerobic yeast cells overexpressing PGK, Pi ATP exchange was
increased [61]. 31P MRS ST experiments performed on perfused rat hearts showed that the
glycolytic exchange reaction contributes significantly to the measured Pi ATP flux and
that only when glycolytic activity was inhibited, the measured Pi ATP rates displayed a
linear dependence on oxygen consumption and thus represent mitochondrial F1F0‐ATP
synthase flux [31]. For rat skeletal muscle, values for resting ATP turnover from 31P MRS ST
experiments are about 5 times higher as compared to figures derived from oxygen
consumption rates, whereas in human muscle the difference is even larger (reviewed in
[26]). This indicates that the contribution of glycolytic exchange flux to the measured Pi
ATP flux from 31P MRS ST experiments in resting skeletal muscle can be as much as 80%. Up
to this point, it has been assumed that the mitochondrial F1F0‐ATP synthase flux, which
constitutes only a small part of the total Pi ATP flux measured, is unidirectional.
However, it has been shown that at low rates of respiration (e.g. resting muscle) the F1F0‐
ATP synthase reaction is near to equilibrium and that mitochondrial exchange between Pi
and ATP can make a significant contribution to the measured Pi ATP flux [62]. Taken
together, if there would have been an effect of DPI on the rate of net resting mitochondrial
ATP production, this effect might not be measurable due to the large contribution of
glycolytic exchange flux and possibly also a significant contribution of mitochondrial
exchange flux to the total measured Pi ATP flux. In order to create better conditions for
detecting a defect in skeletal muscle mitochondrial oxidative phosphorylation, 31P MRS ST
experiments would need to be performed during steady‐state muscle contraction, when
the F1F0‐ATP synthase reaction is largely unidirectional and the flux through the F1F0‐ATP
synthase is much larger than the glycolytic exchange flux [24]. However, it is very
challenging to maintain steady‐state muscle contractions during the lengthy 31P MRS ST
Assessment of in vivo muscle mitochondrial function
45
experiments and this might in particular be a problem in patient groups. Moreover, in such
an experiment it is crucial to achieve similar relative workloads in different subjects or
animals.
Most 31P MRS ST studies investigating the role of skeletal muscle mitochondrial function in
the development of insulin resistance and T2D lack a proper discussion about the error
signals, i.e. the concentrations of Pi and ADP. The lower VATP in insulin‐resistant subjects
and animal models and patients with type 2 diabetes [7‐8, 19, 63‐65] might be explained by
a lower Pi concentration [8], caused by reduced sarcolemmal Pi uptake in insulin‐resistant
muscle [66], rather than a decrease in the Pi ATP exchange rate constant .
Furthermore, it has been shown that hyperinsulinemia decreases the intracellular pH [67],
resulting in a lower ADP concentration and therefore a lower error signal for oxidative ATP
synthesis. In addition, all 31P MRS ST studies investigating the role of skeletal muscle
mitochondrial function in the development of IR and T2D were performed in resting
skeletal muscle [7‐8, 19, 63‐65, 68]. As was explained in the previous paragraph, in resting
skeletal muscle the contribution of F1F0‐ATP synthesis flux to the total measured Pi ATP
flux will be relatively low, which further complicates the interpretation of these
measurements. In a recent study, measurements of both PCr recovery and resting ATP
synthesis flux were performed in type 2 diabetes patients and healthy controls [68]. The
half times of PCr recovery were not different between groups and also resting ATP
synthesis flux was not impaired in the patients with type 2 diabetes as compared to
controls, while concentrations of Pi and ADP were similar for both groups. In addition, half
times of PCr recovery were not significantly correlated with resting ATP synthesis fluxes,
which is in agreement with the results of the present study. It should be noted that the
animal model that we used in this study is a model of mitochondrial dysfunction, but that it
does not represent a model for insulin resistance of T2D. On the contrary, it has been
shown that DPI can induce hypoglycemia [50, 52].
In conclusion, the rate constant of PCr recovery measured with dynamic 31P MRS after
exercise provides a more sensitive measure of skeletal muscle mitochondrial function than
the ATP synthesis flux determined with 31P saturation transfer in the resting state. The ATP
synthesis flux itself represents the ATP demand of the muscle and in order to interpret the
data in terms of mitochondrial function it is necessary to take the errors signals, i.e. the
concentrations of ADP and Pi, into account. Moreover, the Pi ATP flux obtained from a 31P saturation transfer experiment in the resting state is dominated by glycolytic exchange
flux. In order to detect a defect in mitochondrial oxidative phosphorylation the latter
experiment would need to be done in exercising muscle.
Chapter 2
46
Acknowledgements
The authors thank Joep van Lier for his support in fitting the 31P MR spectra and dr. Jeroen
Jeneson for stimulating discussions.
The work of J.C. is financed by the Netherlands Consortium for Systems Biology (NCSB)
which is part of the Netherlands Genomics Initiative / Netherlands Organisation for
Scientific Research. J.J.P. is supported by a VIDI grant from the Netherlands Organisation
for Scientific Research (VIDI grant number 700.58.421).
Assessment of in vivo muscle mitochondrial function
47
References 1. Rabol, R., R. Boushel, and F. Dela. Mitochondrial oxidative function and type 2 diabetes. Appl Physiol
Nutr Metab, 2006. 31(6): 675‐83. 2. Morino, K., K.F. Petersen, and G.I. Shulman. Molecular mechanisms of insulin resistance in humans and
their potential links with mitochondrial dysfunction. Diabetes, 2006. 55 Suppl 2: S9‐S15. 3. Sreekumar, R. and K.S. Nair. Skeletal muscle mitochondrial dysfunction & diabetes. Indian J Med Res,
2007. 125(3): 399‐410. 4. Schrauwen‐Hinderling, V.B., M. Roden, M.E. Kooi, M.K. Hesselink, and P. Schrauwen. Muscular
mitochondrial dysfunction and type 2 diabetes mellitus. Curr Opin Clin Nutr Metab Care, 2007. 10(6): 698‐703.
5. Szendroedi, J. and M. Roden. Mitochondrial fitness and insulin sensitivity in humans. Diabetologia, 2008. 51(12): 2155‐67.
6. Holloszy, J.O. Skeletal muscle "mitochondrial deficiency" does not mediate insulin resistance. Am J Clin Nutr, 2009. 89(1): 463S‐6S.
7. Petersen, K.F., D. Befroy, S. Dufour, J. Dziura, C. Ariyan, D.L. Rothman, L. DiPietro, G.W. Cline, and G.I. Shulman. Mitochondrial dysfunction in the elderly: possible role in insulin resistance. Science, 2003. 300(5622): 1140‐2.
8. Petersen, K.F., S. Dufour, D. Befroy, R. Garcia, and G.I. Shulman. Impaired mitochondrial activity in the insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(7): 664‐71.
9. Yu, C., Y. Chen, G.W. Cline, D. Zhang, H. Zong, Y. Wang, R. Bergeron, J.K. Kim, S.W. Cushman, G.J. Cooney, B. Atcheson, M.F. White, E.W. Kraegen, and G.I. Shulman. Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate‐1 (IRS‐1)‐associated phosphatidylinositol 3‐kinase activity in muscle. J Biol Chem, 2002. 277(52): 50230‐6.
10. Mootha, V.K., C.M. Lindgren, K.F. Eriksson, A. Subramanian, S. Sihag, J. Lehar, P. Puigserver, E. Carlsson, M. Ridderstrale, E. Laurila, N. Houstis, M.J. Daly, N. Patterson, J.P. Mesirov, T.R. Golub, P. Tamayo, B. Spiegelman, E.S. Lander, J.N. Hirschhorn, D. Altshuler, and L.C. Groop. PGC‐1alpha‐responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet, 2003. 34(3): 267‐73.
11. Patti, M.E., A.J. Butte, S. Crunkhorn, K. Cusi, R. Berria, S. Kashyap, Y. Miyazaki, I. Kohane, M. Costello, R. Saccone, E.J. Landaker, A.B. Goldfine, E. Mun, R. DeFronzo, J. Finlayson, C.R. Kahn, and L.J. Mandarino. Coordinated reduction of genes of oxidative metabolism in humans with insulin resistance and diabetes: Potential role of PGC1 and NRF1. Proc Natl Acad Sci U S A, 2003. 100(14): 8466‐71.
12. Ritov, V.B., E.V. Menshikova, J. He, R.E. Ferrell, B.H. Goodpaster, and D.E. Kelley. Deficiency of subsarcolemmal mitochondria in obesity and type 2 diabetes. Diabetes, 2005. 54(1): 8‐14.
13. He, J., S. Watkins, and D.E. Kelley. Skeletal muscle lipid content and oxidative enzyme activity in relation to muscle fiber type in type 2 diabetes and obesity. Diabetes, 2001. 50(4): 817‐23.
14. Heilbronn, L.K., S.K. Gan, N. Turner, L.V. Campbell, and D.J. Chisholm. Markers of mitochondrial biogenesis and metabolism are lower in overweight and obese insulin‐resistant subjects. J Clin Endocrinol Metab, 2007. 92(4): 1467‐73.
15. Hojlund, K., K. Wrzesinski, P.M. Larsen, S.J. Fey, P. Roepstorff, A. Handberg, F. Dela, J. Vinten, J.G. McCormack, C. Reynet, and H. Beck‐Nielsen. Proteome analysis reveals phosphorylation of ATP synthase beta ‐subunit in human skeletal muscle and proteins with potential roles in type 2 diabetes. J Biol Chem, 2003. 278(12): 10436‐42.
16. Kelley, D.E., J. He, E.V. Menshikova, and V.B. Ritov. Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes, 2002. 51(10): 2944‐50.
17. Mogensen, M., K. Sahlin, M. Fernstrom, D. Glintborg, B.F. Vind, H. Beck‐Nielsen, and K. Hojlund. Mitochondrial respiration is decreased in skeletal muscle of patients with type 2 diabetes. Diabetes, 2007. 56(6): 1592‐9.
18. Morino, K., K.F. Petersen, S. Dufour, D. Befroy, J. Frattini, N. Shatzkes, S. Neschen, M.F. White, S. Bilz, S. Sono, M. Pypaert, and G.I. Shulman. Reduced mitochondrial density and increased IRS‐1 serine phosphorylation in muscle of insulin‐resistant offspring of type 2 diabetic parents. J Clin Invest, 2005. 115(12): 3587‐93.
19. Petersen, K.F., S. Dufour, and G.I. Shulman. Decreased insulin‐stimulated ATP synthesis and phosphate transport in muscle of insulin‐resistant offspring of type 2 diabetic parents. PLoS Med, 2005. 2(9): e233.
20. Simoneau, J.A. and D.E. Kelley. Altered glycolytic and oxidative capacities of skeletal muscle contribute to insulin resistance in NIDDM. J Appl Physiol, 1997. 83(1): 166‐71.
21. De Feyter, H.M., N.M. van den Broek, S.F. Praet, K. Nicolay, L.J. van Loon, and J.J. Prompers. Early or advanced stage type 2 diabetes is not accompanied by in vivo skeletal muscle mitochondrial dysfunction. Eur J Endocrinol, 2008. 158(5): 643‐53.
Chapter 2
48
22. Schrauwen‐Hinderling, V.B., M.E. Kooi, M.K. Hesselink, J.A. Jeneson, W.H. Backes, C.J. van Echteld, J.M. van Engelshoven, M. Mensink, and P. Schrauwen. Impaired in vivo mitochondrial function but similar intramyocellular lipid content in patients with type 2 diabetes mellitus and BMI‐matched control subjects. Diabetologia, 2007. 50(1): 113‐20.
23. Brown, T.R., K. Ugurbil, and R.G. Shulman. 31P nuclear magnetic resonance measurements of ATPase kinetics in aerobic Escherichia coli cells. Proc Natl Acad Sci U S A, 1977. 74(12): 5551‐3.
24. Brindle, K.M., M.J. Blackledge, R.A. Challiss, and G.K. Radda. 31P NMR magnetization‐transfer measurements of ATP turnover during steady‐state isometric muscle contraction in the rat hind limb in vivo. Biochemistry, 1989. 28(11): 4887‐93.
25. Brehm, A., M. Krssak, A.I. Schmid, P. Nowotny, W. Waldhausl, and M. Roden. Increased lipid availability impairs insulin‐stimulated ATP synthesis in human skeletal muscle. Diabetes, 2006. 55(1): 136‐40.
26. Kemp, G.J. The interpretation of abnormal 31P magnetic resonance saturation transfer measurements of Pi/ATP exchange in insulin‐resistant skeletal muscle. Am J Physiol Endocrinol Metab, 2008. 294(3): E640‐642.
27. Short, K.R., K.S. Nair, and C.S. Stump. Impaired mitochondrial activity and insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(23): 2419‐21; author reply 2419‐21.
28. Wagenmakers, A.J. Insulin resistance in the offspring of parents with type 2 diabetes. PLoS Med, 2005. 2(9): e289.
29. Brindle, K.M. and G.K. Radda. 31P‐NMR saturation transfer measurements of exchange between Pi and ATP in the reactions catalysed by glyceraldehyde‐3‐phosphate dehydrogenase and phosphoglycerate kinase in vitro. Biochim Biophys Acta, 1987. 928(1): 45‐55.
30. Campbell‐Burk, S.L., K.A. Jones, and R.G. Shulman. 31P NMR saturation‐transfer measurements in Saccharomyces cerevisiae: characterization of phosphate exchange reactions by iodoacetate and antimycin A inhibition. Biochemistry, 1987. 26(23): 7483‐92.
31. Kingsley‐Hickman, P.B., E.Y. Sako, P. Mohanakrishnan, P.M. Robitaille, A.H. From, J.E. Foker, and K. Ugurbil. 31P NMR studies of ATP synthesis and hydrolysis kinetics in the intact myocardium. Biochemistry, 1987. 26(23): 7501‐10.
32. Quistorff, B., L. Johansen, and K. Sahlin. Absence of phosphocreatine resynthesis in human calf muscle during ischaemic recovery. Biochem J, 1993. 291 ( Pt 3): 681‐6.
33. Sahlin, K. Intracellular pH and energy metabolism in skeletal muscle of man. With special reference to exercise. Acta Physiol Scand Suppl, 1978. 455: 1‐56.
34. Taylor, D.J., P.J. Bore, P. Styles, D.G. Gadian, and G.K. Radda. Bioenergetics of intact human muscle. A 31P nuclear magnetic resonance study. Mol Biol Med, 1983. 1(1): 77‐94.
35. Vicini, P. and M.J. Kushmerick. Cellular energetics analysis by a mathematical model of energy balance: estimation of parameters in human skeletal muscle. Am J Physiol Cell Physiol, 2000. 279(1): C213‐24.
36. Majander, A., M. Finel, and M. Wikstrom. Diphenyleneiodonium inhibits reduction of iron‐sulfur clusters in the mitochondrial NADH‐ubiquinone oxidoreductase (Complex I). J Biol Chem, 1994. 269(33): 21037‐42.
37. Ragan, C.I. and D.P. Bloxham. Specific labelling of a constituent polypeptide of bovine heart mitochondrial reduced nicotinamide‐adenine dinucleotide‐ubiquinone reductase by the inhibitor diphenyleneiodonium. Biochem J, 1977. 163(3): 605‐15.
38. Cooper, J., R. Petty, D. Hayes, J. Morgan‐Hughes, and J. Clark. Chronic administration of the oral hypoglycaemic agent diphenyleneiodonium to rats. An animal model of impaired oxidative phosphorylation (mitochondrial myopathy). Biochemical Pharmacology, 1988. 37(4): 687‐694.
39. Cooper, J.M., R.K. Petty, D.J. Hayes, R.A. Challiss, M.J. Brosnan, E.A. Shoubridge, G.K. Radda, J.A. Morgan‐Hughes, and J.B. Clark. An animal model of mitochondrial myopathy: a biochemical and physiological investigation of rats treated in vivo with the NADH‐CoQ reductase inhibitor, diphenyleneiodonium. J Neurol Sci, 1988. 83(2‐3): 335‐47.
40. Palladino, G.W., J.J. Wood, and H.J. Proctor. Modified freeze clamp technique for tissue assay. J Surg Res, 1980. 28(2): 188‐90.
41. De Feyter, H.M., E. Lenaers, S.M. Houten, P. Schrauwen, M.K. Hesselink, R.J. Wanders, K. Nicolay, and J.J. Prompers. Increased intramyocellular lipid content but normal skeletal muscle mitochondrial oxidative capacity throughout the pathogenesis of type 2 diabetes. FASEB J, 2008. 22(11): 3947‐55.
42. Lawson, J.W. and R.L. Veech. Effects of pH and free Mg2+ on the Keq of the creatine kinase reaction and other phosphate hydrolyses and phosphate transfer reactions. J Biol Chem, 1979. 254(14): 6528‐37.
43. van den Broek, N.M., H.M. De Feyter, L. de Graaf, K. Nicolay, and J.J. Prompers. Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics in muscle after exercise are due to differences in proton efflux rates. Am J Physiol Cell Physiol, 2007. 293(1): C228‐37.
44. Trautschold, I., W. Lamprecht, and G. Schweitzer, Nucleotides, Coenzymes and related compounds, in Methods of enzymatic analysis, H. Bergmeyer, Editor. 1985, , Wiley‐VCH.
Assessment of in vivo muscle mitochondrial function
49
45. van den Broek, N.M., J. Ciapaite, H.M. De Feyter, S.M. Houten, R.J. Wanders, J.A. Jeneson, K. Nicolay, and J.J. Prompers. Increased mitochondrial content rescues in vivo muscle oxidative capacity in long‐term high‐fat‐diet‐fed rats. FASEB J, 2010. 24(5): 1354‐64.
46. Ciapaite, J., S.J. Bakker, G. Van Eikenhorst, M.J. Wagner, T. Teerlink, C.G. Schalkwijk, M. Fodor, D.M. Ouwens, M. Diamant, R.J. Heine, H.V. Westerhoff, and K. Krab. Functioning of oxidative phosphorylation in liver mitochondria of high‐fat diet fed rats. Biochim Biophys Acta, 2007. 1772(3): 307‐16.
47. Szuhai, K., J. Ouweland, R. Dirks, M. Lemaitre, J. Truffert, G. Janssen, H. Tanke, E. Holme, J. Maassen, and A. Raap. Simultaneous A8344G heteroplasmy and mitochondrial DNA copy number quantification in myoclonus epilepsy and ragged‐red fibers (MERRF) syndrome by a multiplex molecular beacon based real‐time fluorescence PCR. Nucleic Acids Res, 2001. 29(3): E13.
48. O'Donnell, V.B., G.C. Smith, and O.T. Jones. Involvement of phenyl radicals in iodonium inhibition of flavoenzymes. Mol Pharmacol, 1994. 46(4): 778‐85.
49. Gatley, S.J. and S.A. Sherratt. The effects of diphenyleneiodonium on mitochondrial reactions. Relation of binding of diphenylene[125I]iodonium to mitochondria to the extent of inhibition of oxygen uptake. Biochem J, 1976. 158(2): 307‐15.
50. Holland, P.C. and H.S. Sherratt. Biochemical effects of the hypoglycaemic compound diphenyleneiodonnium. Catalysis of anion‐hydroxyl ion exchange across the inner membrane of rat liver mitochondria and effects on oxygen uptake. Biochem J, 1972. 129(1): 39‐54.
51. Hayes, D.J., E. Byrne, E.A. Shoubridge, J.A. Morgan‐Hughes, and J.B. Clark. Experimentally induced defects of mitochondrial metabolism in rat skeletal muscle. Biological effects of the NADH: coenzyme Q reductase inhibitor diphenyleneiodonium. Biochem J, 1985. 229(1): 109‐17.
52. Holland, P.C., M.G. Clark, D.P. Bloxham, and H.A. Lardy. Mechanism of action of the hypoglycemic agent diphenyleneiodonium. J Biol Chem, 1973. 248(17): 6050‐6.
53. Gatley, S.J. and H.S. Sherratt. The effects of diphenyleneiodonium and of 2,4‐dichlorodiphenyleneiodonium on mitochondrial reactions. Mechanism of the inhibition of oxygen uptake as a consequence of the catalysis of the chloride/hydroxyl‐ion exchange. Biochem J, 1976. 158(2): 317‐26.
54. Stuehr, D.J., O.A. Fasehun, N.S. Kwon, S.S. Gross, J.A. Gonzalez, R. Levi, and C.F. Nathan. Inhibition of macrophage and endothelial cell nitric oxide synthase by diphenyleneiodonium and its analogs. FASEB J, 1991. 5(1): 98‐103.
55. Clifford, P.S. and Y. Hellsten. Vasodilatory mechanisms in contracting skeletal muscle. J Appl Physiol, 2004. 97(1): 393‐403.
56. Saltin, B., G. Radegran, M.D. Koskolou, and R.C. Roach. Skeletal muscle blood flow in humans and its regulation during exercise. Acta Physiol Scand, 1998. 162(3): 421‐36.
57. Jeneson, J.A., J.P. Schmitz, N.M. van den Broek, N.A. van Riel, P.A. Hilbers, K. Nicolay, and J.J. Prompers. Magnitude and control of mitochondrial sensitivity to ADP. Am J Physiol Endocrinol Metab, 2009. 297(3): E774‐84.
58. Kemp, G.J. Interactions of mitochondrial ATP synthesis and the creatine kinase equilibrium in skeletal muscle. J Theor Biol, 1994. 170(3): 239‐46.
59. Wu, F., J.A. Jeneson, and D.A. Beard. Oxidative ATP synthesis in skeletal muscle is controlled by substrate feedback. Am J Physiol Cell Physiol, 2007. 292(1): C115‐24.
60. Hood, D.A., J. Gorski, and R.L. Terjung. Oxygen cost of twitch and tetanic isometric contractions of rat skeletal muscle. Am J Physiol, 1986. 250(4 Pt 1): E449‐56.
61. Brindle, K.M. 31P NMR magnetization‐transfer measurements of flux between inorganic phosphate and adenosine 5'‐triphosphate in yeast cells genetically modified to overproduce phosphoglycerate kinase. Biochemistry, 1988. 27(16): 6187‐96.
62. Sheldon, J.G., S.P. Williams, A.M. Fulton, and K.M. Brindle. 31P NMR magnetization transfer study of the control of ATP turnover in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A, 1996. 93(13): 6399‐404.
63. Laurent, D., B. Yerby, R. Deacon, and J. Gao. Diet‐induced modulation of mitochondrial activity in rat muscle. Am J Physiol Endocrinol Metab, 2007. 293(5): E1169‐77.
64. Szendroedi, J., A.I. Schmid, M. Chmelik, C. Toth, A. Brehm, M. Krssak, P. Nowotny, M. Wolzt, W. Waldhausl, and M. Roden. Muscle mitochondrial ATP synthesis and glucose transport/phosphorylation in type 2 diabetes. PLoS Med, 2007. 4(5): e154.
65. Yerby, B., R. Deacon, V. Beaulieu, J. Liang, J. Gao, and D. Laurent. Insulin‐stimulated mitochondrial adenosine triphosphate synthesis is blunted in skeletal muscles of high‐fat‐fed rats. Metabolism, 2008. 57(11): 1584‐90.
66. Polgreen, K.E., G.J. Kemp, B. Leighton, and G.K. Radda. Modulation of Pi transport in skeletal muscle by insulin and IGF‐1. Biochim Biophys Acta, 1994. 1223(2): 279‐84.
Chapter 2
50
67. Taylor, D.J., S.W. Coppack, T.A. Cadoux‐Hudson, G.J. Kemp, G.K. Radda, K.N. Frayn, and L.L. Ng. Effect of insulin on intracellular pH and phosphate metabolism in human skeletal muscle in vivo. Clin Sci (Lond), 1991. 81(1): 123‐8.
68. Trenell, M.I., K.G. Hollingsworth, E.L. Lim, and R. Taylor. Increased Daily Walking Improves Lipid Oxidation Without Changes in Mitochondrial Function in Type 2 Diabetes. Diabetes Care, 2008. 31(8): 1644‐1649.
51
52
Adapted from
N. M. A. van den Broek, H. M. M. L. De Feyter, L. de Graaf, K. Nicolay and J. J. Prompers.
Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics in
muscle after exercise are due to differences in proton efflux rates.
Am J Physiol Cell Physiol, 2007. 293(1):C228‐37.
Used with permission from The Am Physiol Soc.DOI: 10.1152/ajpcell.00023.2007.
Chapter Intersubject differences in the effect of acidosis
on phosphocreatine recovery kinetics
in muscle after exercise are due to
differences in proton efflux rates
Chapter 3
54
Abstract 31P magnetic resonance spectroscopy provides the possibility to obtain bio‐energetic data
during skeletal muscle exercise and recovery. The time constant of phosphocreatine (PCr)
recovery (PCr) has been used as a measure of mitochondrial function. However, cytosolic
pH has a strong influence on the kinetics of PCr recovery and it has been suggested that PCr should be normalized for end‐exercise pH. A general correction can only be applied if there
are no intersubject differences in the pH dependence of PCr. We investigated the pH
dependence of PCr on a subject‐by‐subject basis. Furthermore, we determined the kinetics
of proton efflux at the start of recovery. Intracellular acidosis slowed PCr recovery and the
pH dependence of PCr differed among subjects, ranging from ‐33.0 to ‐75.3 s per pH unit.
The slope of the relation between PCr and end‐exercise pH was positively correlated with both the proton efflux rate and the apparent proton efflux rate constant, indicating that
subjects with a smaller pH dependence of PCr have a higher proton efflux rate. Our study implies that simply correcting PCr for end‐exercise pH is not adequate, in particular when comparing patients and controls, as certain disorders are characterized by altered proton
efflux from muscle fibers.
Effect of acidosis on PCr recovery related to proton efflux
55
Introduction 31P magnetic resonance spectroscopy (MRS) provides the possibility to obtain bio‐energetic
data during skeletal muscle exercise and recovery in a non‐invasive manner and with a time
resolution of a few seconds. This has made a major contribution to our understanding of
energy metabolism, its control and the way in which it can be affected in disease [1‐7].
During recovery from exercise, phosphocreatine (PCr) is resynthesized purely as a
consequence of oxidative ATP synthesis [8‐10] and therefore measurement of the time
constant of PCr recovery (PCr) provides information about mitochondrial function. This
technology was very recently applied to study in vivo mitochondrial function in patients
with type 2 diabetes [11].
Several studies have shown that cytosolic pH has a strong influence on the kinetics of PCr
recovery [1, 12‐21]. The slower PCr recovery in the presence of intracellular acidosis could
reflect a decreased mitochondrial respiration at low pH. However, there are conflicting
data about the effects of low pH on respiratory rates, ranging from inhibition [22‐24], a
very small or no significant effect [19, 25‐29] to even an increased effectiveness [30‐31].
PCr recovery in the presence of intracellular acidosis could also be slowed down due to
factors downstream of oxidative phosphorylation, i.e. increased ATP consumption by
cellular ion pumps [21, 32‐34] and/or a pH‐dependent shift in the creatine kinase
equilibrium [35‐36].
As an alternative to PCr, the time constant of ADP recovery (ADP) can be used to assess oxidative capacity. The concentration of ADP is the principal error signal in the feedback
loop controlling mitochondrial oxidation and therefore ADP recovery is one of the most
sensitive MRS indices of mitochondrial function [37]. Finally, the maximum aerobic capacity
(Qmax), which can be calculated from the 31P MRS recovery data, provides a parameter for
mitochondrial function. Both ADP [1‐2, 17‐18, 20] and Qmax [18‐19, 21] have been shown to
be independent of end‐exercise pH. A drawback of the use of these parameters compared
to PCr is that they are indirectly derived from the PCr recovery data, using a number of
assumptions.
It has been suggested that PCr can be normalized for end‐exercise pH [13, 18, 21].
However, a general correction for pH can only be applied if there are no intersubject
differences in the pH dependence of PCr recovery kinetics. In previous studies, data of
different subjects have been grouped to investigate the effect of pH on PCr recovery [1, 12‐
21]. We investigated the effect of acidosis on PCr in the M. vastus lateralis on a subject‐by‐
subject basis. To this end, each subject performed 10‐13 exercise protocols of different
intensity to reach different levels of acidification. Furthermore, we studied the pH
dependence of ADP and Qmax and we determined the kinetics of proton efflux at the start of
recovery to obtain a measure of the rate of pH recovery.
Chapter 3
56
Materials and Methods
Subjects
Four male and two female healthy subjects participated in this study. The nature and the
risks of the experimental procedures were explained to the subjects and all gave their
written informed consent to participate in the study, which was approved by the local
Medical Ethical Committee of the Máxima Medical Center, Veldhoven, The Netherlands.
Subjects varied in age (mean age: 31 ± 12 years; five subjects in the range of 20‐33 years
and one subject of 53 years), body mass index (BMI) (mean BMI: 21.1 ± 1.8 kg/cm2; range:
17.9‐22.6) and daily activity level, i.e. level of activities during daily living, work and leisure
time (e.g. sports). However, none of the subjects was highly trained.
31P Magnetic resonance spectroscopy
31P MRS of the M. vastus lateralis was performed by using a 1.5‐Tesla whole‐body scanner
(Gyroscan S15/ACS, Philips Medical Systems, Best, The Netherlands). Subjects were
measured in a supine position. After collecting transversal and sagittal scout images, the
magnetic field homogeneity was optimized by localized shimming on the proton signal
using the body coil. The 31P signals were collected using a 6‐cm diameter surface coil placed
over the M. vastus lateralis. From the dimension of the coil and the size and geometry of a
typical upper leg, it was estimated that the majority of the signal in the unlocalized 31P MRS
measurements originated from the M. vastus lateralis, with minimal contaminations from
the adjacent M. rectus femoris and underlying M. vastus intermedius. Data were acquired
following a 90 adiabatic excitation pulse with a sweep width of 2 kHz and 1024 data points. Spectra were acquired using a repetition time of 3 s during a rest‐exercise‐recovery
protocol (2 scans/spectrum yielding a time resolution of 6 s, total of 150 spectra/15 min).
The first 20 spectra (2 min) were measured at rest, after which the subjects started the
exercise (see below). The duration of the exercise varied per subject, but never exceeded 9
min, so that at least 4 min of recovery were recorded.
Exercise protocol inside magnet
All subjects performed a single leg extension exercise in the supine position inside the
magnet, which has been shown to be limited to the four muscles of the quadriceps [38].
The exercise was conducted by rhythmically lifting a lever (resting on the lower leg,
proximal of the foot) connected to an ergometer. The upper leg was supported with the hip
joint in a 30 degrees ante flexed position and immobilized with two 3 cm wide Velcro©
straps. One contraction was performed every 1.5 s acoustically guided by a digital
metronome. The initial workload varied per subject and ranged between 7.5 and 12.5 W.
Effect of acidosis on PCr recovery related to proton efflux
57
This level was maintained for the first min and the workload was then increased by 5 W
each min. To achieve different levels of metabolic activation, and hence different degrees
of cytosolic acidification, subjects performed exercises of different durations. Each subject
performed 10‐13 different protocols during 4‐9 different sessions in a randomized order,
with at least 15 min rest between different protocols within one session. The position of
the 31P surface coil was marked on the leg during the first session and the coil was placed at
the same location during the next sessions.
The reproducibility of the 31P MRS measurements was determined in one subject. This
subject performed 10 times the same protocol during 5 different sessions. In one other
subject, it was tested whether the position of the 31P surface coil on the M. vastus lateralis,
when varied in the proximal and distal direction, influenced the 31P MRS measurements.
For this purpose, the subject performed 5 times the same protocol, with a maximal
difference of 15 cm of the position of the 31P surface coil in the proximo‐distal direction.
Data analysis
Spectra were fitted in the time domain by using a nonlinear least squares algorithm
(AMARES) in the jMRUI software package [39]. PCr, Pi and ATP signals were fitted to
Lorentzian line shapes. The three ATP peaks were fitted as two doublets and one triplet,
with equal amplitudes and line widths and prior knowledge for the J‐coupling constant (17
Hz). For the time series, the PCr line width during recovery was constrained to the average
PCr line width during recovery (excluding the first 10 data points), obtained from a prior,
unconstrained fit.
Absolute concentrations of the phosphorylated metabolites were calculated after
correction for partial saturation and assuming that [ATP] is 8.2 mM at rest [12]. Intracellular
pH was calculated from the chemical shift difference between the Pi and PCr resonances (; measured in parts per million) using the following formula [9]:
6.753.27
5.63 . 1
The free cytosolic ADP concentration ([ADP]) was calculated from pH and [PCr] using a
creatine kinase equilibrium constant (Keq) of 1.66×109 M‐1 [35] and assuming that 15% of
the total creatine is unphosphorylated at rest [33], using the equation:
Chapter 3
58
. 2
Recoveries of PCr and ADP were fitted to mono‐exponential functions using Matlab
(version 6.1, Mathworks, Natick, Massachusetts, USA). Results are expressed as the
metabolite’s time constant of recovery, i.e. PCr and ADP.
Calculation of the initial rate of PCr recovery (VPCr) was based on the PCr recovery rate
(1/PCr) and the difference between the resting and end‐exercise PCr concentrations (PCr) [40]:
1· ∆ . 3
The calculation of the maximum aerobic capacity (Qmax) was either based on the ADP‐
control model [41], in which VPCr has a hyperbolic dependence on the end‐exercise ADP
concentration ([ADP]end) according to Michaelis‐Menten kinetics with a Km of 30 M [40]:
· 1 . 4
or on a linear approximation of the ADP‐control model
1
. 5
which is equivalent to the non‐equilibrium thermodynamic control model [37, 42].
However, this approximation is only valid when pH changes are small.
Effect of acidosis on PCr recovery related to proton efflux
59
The proton efflux rate at the start of recovery was calculated as described by Kemp et al.
[43‐44] from the changes in the PCr concentration and pH during the first 12 s of recovery
according to
. 6
where is the amount of protons consumed per mole of PCr hydrolysis ( 1/ 1
10 . [45]), m is the amount of protons produced per mole of oxidative ATP
synthesis ( 0.16/ 1 10 . [46]) and is the cytosolic buffering capacity (20 slykes (i.e. mmol l‐1 pH‐1) plus the calculated contribution of Pi, which is given by 2.3 · ·
1 [47]). The apparent efflux rate constant was calculated as / Δ , where
pH is defined as [37]. Both the proton efflux rate E and the apparent efflux
rate constant at the start of recovery depend on the end‐exercise pH (pHend) [44] and
therefore only the data sets with a pHend between 6.6 and 6.8 were used to calculate an
average value for E and for each subject. These data sets with rather low pHend values
were chosen for two reasons: (1) they had a greater PCr depletion and therefore a larger
PCr resynthesis rate and proton production rate at the start of recovery and (2) they had a
greater increase in Pi and therefore a better visible Pi peak at the start of recovery, which is
important for an accurate and precise pH determination.
Figure 1. Typical M. vastus lateralis 31P
MR spectra for one subject at rest
(panel A, number of scans = 60), at the
end of exercise (panel B, number of
scans = 2) and at 15 and 93 s of
recovery (panels C and D, respectively,
number of scans = 2). Spectra were
processed with 5 Hz line broadening. Pi
indicates inorganic phosphate; PDE,
phosphodiesters; PCr, phospho‐
creatine; and , and indicate the three phosphate groups of ATP. For
this subject the PCr depletion at the
end of exercise (panel b) was 65% and
the corresponding pHend was 6.93.
Note that the Pi signal is not discernible
in the spectrum in panel d.
Chapter 3
60
Statistics
All data are expressed as means ± standard deviation (SD). Reproducibility is reported as
the coefficient of variation · 100 . Linear regression analyses were
performed using the SPSS 14.0 software package (SPSS Inc, Chicago, IL, USA). Level of
statistical significance was set at p < 0.05.
Results
Reproducibility
Figure 1 shows typical examples of 31P MR spectra from a subject’s vastus lateralis muscle
at rest, at the end of exercise and at two time points during recovery. For the same data
set, PCr and ADP concentrations are plotted as a function of time in Figure 2. At the end of
recovery, both PCr and ADP concentrations are identical to the resting condition. Figure 2
also illustrates the mono‐exponential fits of the PCr and ADP recoveries. The reproducibility
of the determination of 31P MRS parameters for mitochondrial function, i.e. PCr, ADP, VPCr,
Qmax‐ADP and Qmax‐lin, was determined in one subject and the results are shown in Table 1.
Table 1. Reproducibility of 31P MRS recovery parameters.
PCr (s) ADP (s) VPCr (mM/s) Qmax‐ADP
(mM/s)
Qmax‐lin
(mM/s)
all mean 29.4 11.8 0.86 1.17 1.38
(n = 10) SD 3.3 0.8 0.09 0.10 0.14
CV (%) 11.3 6.7 10.7 8.4 10.4
pHend > 6.9 mean 27.9 12.2 0.88 1.18 1.43
(n = 7) SD 1.9 0.5 0.09 0.09 0.12
CV (%) 6.9 4.0 9.8 7.9 8.4
PCr, phosphocreatine (PCr) recovery time constant; ADP, adenosine diphosphate (ADP) recovery time
constant; VPCr, initial rate of PCr recovery based on the PCr recovery rate (1/PCr) and the difference between the resting and end‐exercise PCr concentrations; Qmax‐ADP, maximum rate of oxidative ATP
synthesis calculated according to the ADP‐control model, i.e. based on VPCr related to end‐exercise ADP
concentration and an assumed Km of 30 μM; Qmax‐lin, maximum rate of oxidative ATP synthesis calculated
by a linear approximation of the ADP‐control model. Coefficient of variation (CV) was calculated as
/ · 100. The upper 3 rows are the results from all 10 data sets. The lower 3 rows are the results
from the 7 data sets with an end‐exercise pH (pHend) higher than 6.9.
Effect of acidosis on PCr recovery related to proton efflux
61
When including all 10 measurements, the CV ranged from 6.7% for ADP to 11.3% for PCr. However, even though the exercise protocol was identical for all 10 measurements, there
was still some variation in pHend. For 7 of the 10 data sets pHend ranged between 6.91 and
6.96 (mean pHend: 6.93 ± 0.02), while for 3 data sets pHend was lower than 6.9, i.e. 6.87, 6.84
and 6.81, respectively. When only the 7 data sets with pHend above 6.9 were considered,
the CV for PCr became 6.9%, which is comparable to other studies [19, 34]. No systematic
differences were observed for the two measurements performed during one session.
In one other subject, it was tested whether the position of the 31P surface coil on the M.
vastus lateralis, when varied in the proximal and distal direction, influenced the 31P MRS
measurements. For the 5 measurements, with a maximal difference of 15 cm of the
position of the 31P surface coil in the proximo‐distal direction, the CVs for PCr and ADP were 4.5 and 3.0%, respectively. Therefore, it can be concluded that within a certain range the
exact positioning of the 31P surface coil does not affect the parameters for mitochondrial
function and that regional variations in fiber type composition in the proximo‐distal
direction of the M. vastus lateralis are probably small.
End‐exercise status
To achieve different levels of metabolic activation, and hence different degrees of cytosolic
acidification, subjects performed 10‐13 exercises of different durations. For all
measurements, homeostasis of ATP was maintained throughout the exercise protocol.
Figure 2. PCr (panel A) and ADP (panel B) concentrations during rest, exercise and recovery obtained
from the data set that was also used in Figure 1 with a time resolution of 6 s. The recoveries of PCr and
ADP (starting at t = 0) were fitted to mono‐exponential functions (solid lines). The time constants for PCr
and ADP recovery were 26.9 and 11.8 s, respectively.
Chapter 3
62
Figure 3. PCr, [ADP]end, PCr, ADP, VPCr, Qmax‐ADP and Qmax‐lin plotted as a function of pHend for three
subjects. Linear regression analysis results (solid lines) are shown for significant correlations (p < 0.05).
Effect of acidosis on PCr recovery related to proton efflux
63
None of the subjects showed a split Pi peak during exercise or recovery and therefore
acidosis was not extremely heterogeneous in the measured muscle tissue. The ranges of
pHend, PCr and [ADP]end reached for each subject are summarized in Table 2. The smallest
range in pHend values was obtained for subject 4 and covered 0.3 pH units, while the largest
range was obtained for subject 2 and covered 0.6 pH units. For subject 5, one protocol
resulted in a rather low [ADP]end of 25 M. For all other measurements, [ADP]end was well
above the accepted Km value of 30 M for oxidative ATP synthesis. In the two top rows of
Figure 3, PCr and [ADP]end are plotted as a function of pHend, from which it can be seen
that PCr was negatively correlated with pHend for all three subjects, whereas [ADP]end was
not significantly correlated with pHend.
Table 2. Ranges of pHend, PCr and [ADP]end reached at the end of exercise.
subject pHend PCr (mM) [ADP]end (M)
1 6.56 ‐ 7.02 13.6 ‐ 28.1 48.2 ‐ 143.4
2 6.41 ‐ 6.99 17.2 ‐ 33.2 55.3 ‐ 123.2
3 6.53 ‐ 7.02 17.3 ‐ 34.5 39.5 ‐ 174.2
4 6.65 ‐ 6.96 17.4 ‐ 33.7 45.9 ‐ 128.3
5 6.50 ‐ 6.95 9.8 ‐ 29.9 25.1 ‐ 149.1
6 6.63 ‐ 7.01 21.1 ‐ 39.1 59.6 ‐ 192.3
pHend, intracellular muscle pH at the end of exercise; PCr, difference between the resting and end‐exercise PCr concentrations; [ADP]end, ADP concentration at the end of exercise.
Table 3. Averages of 31P MRS recovery parameters for the different exercise protocols.
subject PCr (s) ADP (s) VPCr (mM/s) Qmax‐ADP (mM/s) Qmax‐lin (mM/s)
1 32.7 ± 6.3 11.1 ± 2.5 0.71 ± 0.12 0.96 ± 0.08 1.10 ± 0.22
2 38.4 ± 10.8 10.4 ± 1.3 0.70 ± 0.09 0.97 ± 0.14 1.06 ± 0.29
3 44.4 ± 12.0 12.3 ± 1.4 0.60 ± 0.11 0.83 ± 0.16 0.94 ± 0.31
4 37.9 ± 6.0 12.6 ± 1.7 0.66 ± 0.07 0.90 ± 0.07 1.00 ± 0.15
5 35.0 ± 4.9 13.0 ± 2.1 0.68 ± 0.15 0.97 ± 0.13 1.12 ± 0.18
6 51.0 ± 9.6 16.1 ± 2.5 0.56 ± 0.05 0.76 ± 0.07 0.86 ± 0.18
Data are presented as means ± SD.
Chapter 3
64
Recovery
Recoveries of PCr and ADP could be satisfactorily described by mono‐exponential functions,
also at low pHend values (average R2 values for the mono‐exponential fits were 0.971 ±
0.025 and 0.915 ± 0.063 for PCr and ADP recovery data, respectively). Table 3 lists the
average values for PCr, ADP, VPCr, Qmax‐ADP and Qmax‐lin for all subjects. For each subject, there
was a strong negative linear relationship between PCr and pHend. The third row of Figure 3
shows the correlation between PCr and pHend for three of the subjects. Around pHend 7, PCr was very similar for these three subjects, but at lower pHend values PCr differed. Therefore, the pH dependence of PCr differed, with subject 1 showing the weakest pH dependence and subject 3 showing the strongest pH dependence. The results of the linear regression
analyses for all subjects are shown in Table 4. The slope of the relation between PCr and
Table 4. Correlation of 31P MRS recovery parameters with end‐exercise pH.
subject 1
(n = 10)
subject 2
(n = 12)
subject 3
(n = 13)
subject 4
(n = 12)
subject 5
(n = 13)
subject 6
(n = 11)
grouped
data
(n = 71)
PCr R ‐0.99* ‐0.98* ‐0.94* ‐0.87* ‐0.85* ‐0.94* ‐0.74*
slope (s/U) ‐42.9 ‐57.9 ‐75.3 ‐56.2 ‐33.0 ‐62.9 ‐54.1
PCr at pH 7 (s) 25.3 22.8 26.4 26.8 28.7 39.6 28.6
ADP R 0.86* 0.40 0.53 0.78* 0.13 0.89* 0.41*
slope (s/U) 14.8 2.9 4.9 14.2 2.3 15. 8 8.0
VPCr R 0.13 0.74* 0.55 ‐0.26 0.06 0.47 0.24*
slope (mM/s/U) 0.11 0.36 0.41 ‐0.20 0.07 0.16 0.20
Qmax‐ADP R 0.52 0.88* 0.83* 0.40 0.14 0.79* 0.47*
slope (mM/s/U) 0.30 0.67 0.86 0.31 0.14 0.37 0.46
Qmax‐lin R 0.96* 0.97* 0.92* 0.85* 0.85* 0.95* 0.83*
slope (mM/s/U) 1.45 1.56 1.88 1.41 1.24 1.18 1.41
R, correlation coefficient determined from a linear regression analysis, *p < 0.01; slope is representing the
steepness of the correlation; U, pH unit; PCr at pH 7 was obtained from the linear relation between PCr and end‐exercise pH.
Effect of acidosis on PCr recovery related to proton efflux
65
pHend ranged from ‐33.0 to ‐75.3 s per pH unit. For five of the six subjects, PCr at pHend 7
calculated from the linear relation between PCr at pHend was very similar. Only the older
subject, subject 6, had a longer PCr at pHend 7.
The post‐exercise ADP recovery was faster than the PCr recovery (Table 3). For subjects 1‐5,
ADP was again very similar, while subject 6 had a longer ADP (Table 3). In the fourth row of Figure 3, ADP is plotted against pHend for three of the subjects. For subjects 2, 3 and 5, ADP did not depend on pHend (Table 4). However, for subjects 1, 4 and 6, ADP was significantly positively correlated with pHend (Table 4).
VPCr is a measure of the actual mitochondrial ATP synthesis rate and according to the ADP‐
control model [41], VPCr has a hyperbolic dependence on [ADP]end (Eq. [4]). For all but one
measurement, [ADP]end was well above the accepted Km value of 30 M for oxidative ATP
synthesis and therefore differences in [ADP]end will not have a large effect on VPCr. From
Figure 3, it can be seen that PCr and PCr vary in the same direction as a function of pHend.
As a consequence, for five of the subjects, VPCr was independent of pHend (Table 4). Only for
subject 2, a significant positive correlation between VPCr and pHend was found (Figure 3,
Table 4).
The average values of Qmax‐ADP were smaller than the average values of Qmax‐lin (Table 3). In
accordance with the longer PCr (at pH 7) and ADP, subject 6 also showed smaller values for
Qmax‐ADP and Qmax‐lin compared to the other subjects (Table 3). Qmax‐ADP showed significant
positive correlations with pHend for three of the six subjects (Figure 3, Table 4). These are
the subjects with the largest correlation coefficients for VPCr versus pHend. For each subject,
there was a strong positive linear relationship between Qmax‐lin and pHend (Figure 3, Table 4).
Figure 4. Correlations between the slope of the relation between PCr and pHend and (panel A) the proton
efflux rate E (R = 0.91, p = 0.03) and (panel B) the apparent efflux rate constant (R = 0.96, p = 0.01) for five of the six subjects.
Chapter 3
66
For subject 5, the proton efflux rate E and the apparent efflux rate constant could not be determined. For the other five subjects, the mean total cytosolic buffering capacity at the start of recovery amounted 35 ± 1 slykes, the mean proton efflux rate E was 16 ± 3
mM/min and the mean apparent efflux rate constant was 38 ± 6 mM/(min ∙ pH unit). In
Figures 4a and 4b, the slope of the relation between PCr and pHend is plotted against E and
, respectively. The slope of the relation between PCr and pHend was positively correlated
with both E (R = 0.91, p = 0.03) and (R = 0.96, p = 0.01).
Discussion
Several studies have shown that cytosolic pH has a strong influence on the kinetics of PCr
recovery [1, 12‐21]. In order to establish a relationship between e.g. PCr and pHend, one or a
few data points of different subjects have generally been grouped. However, this
procedure will not reveal intersubject differences in the pH dependence of PCr and, moreover, the effect of pH on PCr recovery might be exaggerated by a systematic bias. We
investigated the effect of acidosis on PCr recovery on a subject‐by‐subject basis by
collecting 10‐13 data sets per subject, using protocols of different intensity and duration
resulting in different degrees of cytosolic acidification. We showed that for each subject
there is a strong negative linear relationship between PCr and pHend, but that the slopes are
different for different subjects, ranging from ‐33.0 to ‐75.3 s per pH unit. This implies that
no general formula can be applied to correct PCr for differences in pHend. Qualitatively and
quantitatively, the results obtained when the data of the different subjects are grouped
(Table 4; last column) are in good agreement with data from the literature on the pH
dependence of PCr recovery kinetics measured in different muscle types, revealing linear
relationships between PCr and pHend (or the minimum pH reached during recovery; pHmin)
with slopes ranging from roughly ‐20 to ‐90 s per pH unit [14, 16‐21].
The slower PCr recovery in the presence of intracellular acidosis could reflect a decreased
mitochondrial respiration at low pH. The mechanisms by which protons affect oxidative
phosphorylation include (1) a direct effect on the mitochondria, i.e. a decreased oxidative
capacity at low pH, or (2) an indirect effect, because a low pH decreases the ADP
concentration through the constraints set by the creatine kinase (CK) equilibrium resulting
in a lower signal for mitochondrial ATP supply. There are conflicting data about the effects
of low pH on respiratory rates. Hypercapnic acidosis has been found to reduce the aerobic
capacity of perfused cat soleus muscle by a factor of three [22]. However, it is not clear
whether this change was caused by acidosis per se or by some other effect of hypercapnic
perfusion. Moreover, it has been shown that in skinned fibers from rat soleus muscle the
rate of respiration is impaired by lactic acidosis and elevated concentrations of Pi [23].
Jubrias et al. showed that intracellular acidosis inhibits oxidative phosphorylation in vivo in
Effect of acidosis on PCr recovery related to proton efflux
67
hand and lower limb muscle and their results suggest that pH has a direct effect on
mitochondrial function, because oxidative flux did not increase during exercise that
generated acidosis despite a significant rise in [ADP] [24]. In contrast, in vitro studies on
isolated mitochondria suggest that the effect of acidosis on oxidative phosphorylation is
very small [25‐27, 29] and not significant in the range pH 6.5‐7.5 [25]. Likewise, an in vivo
study of electrically stimulated rabbit muscle showed that CO2‐induced acidosis (to pH 6.7)
did not decrease the maximum aerobic capacity [28]. Moreover, in human medial
gastrocnemius muscle aerobic ATP synthesis rates were not lowered by acidosis [19]. Other
reports have suggested that mitochondrial respiration is even more effective at low pH [30‐
31].
PCr recovery in the presence of intracellular acidosis could also be slowed down due to
factors downstream of oxidative phosphorylation, consistent with the observation that the
recovery of oxyhemoglobin saturation measured by near‐infrared spectroscopy is not
affected by acidosis [15]. Ion pumping reactions also require ATP and therefore not all the
ATP that is synthesized oxidatively during recovery is available for the CK reaction. At low
pH the amount of ATP that is shuttled to cellular ion pumps might be increased [21, 34] in
order to reestablish pH homeostasis. It has been reported that ion pumping reactions can
consume about 43% of the total ATP produced [32‐33]. The slow PCr recovery at low pH
has also been attributed to a pH‐dependent shift in the CK equilibrium. The CK equilibrium
constant (Keq) depends on proton and metal ion concentrations [35]. Iotti et al. showed that
at the end of muscular exercise, Keq can increase even more than threefold compared to
rest, due to a decrease in pH and an increase in the free Mg2+ concentration [36].
Therefore, net PCr resynthesis throughout recovery behaves as a function of both
intracellular pH and net ATP flux [1, 19]. This was confirmed by a model for ATP production
(according to the ADP‐control model; see Eq. [4]) and pH recovery, which reproduced the
main features of recovery from exercise, including the feature that PCr recovery is slowed
when the pH is low [48]. Finally, with the incremental exercise protocol that we used, part
of the pH dependence of PCr recovery could originate from selective fiber type
recruitment, i.e. recruitment of mainly oxidative type 1 fibers (with short PCr) during the low exercise intensities with a high pHend and recruitment of relatively more type 2 fibers
(with long PCr) during the higher exercise intensities with a low pHend.
The observed intersubject differences in the pH dependence of PCr are likely to reflect differences in the rate of pH recovery. Unfortunately, the recovery of pH could not be
investigated, because the Pi peak consistently disappeared within the noise after about 1
min of recovery (Fig. 1d) and for the exercises at higher intensities was not fully recovered
by the end of the time series. This phenomenon has been reported before in the literature
[9, 14, 17, 20, 34, 49] and has been attributed to sequestering of Pi inside the mitochondria
where it becomes “NMR‐invisible” [1, 50], or trapping of Pi into the glycogenolytic pathway
during exercise leading to phosphomonoester (PME) production [51]. The recovery of pH is
Chapter 3
68
much slower than PCr recovery [1, 17, 19, 52] and therefore it was attempted to increase
the signal‐to‐noise ratio of the Pi peak by the summation of spectra during the recovery
phase. However, even when 4 spectra were added, yielding a time resolution of 24 s, the
position of the Pi peak could not be accurately determined.
The recovery of cytosolic pH to the resting value is a function of net proton efflux [9].
Several mechanisms are responsible for proton efflux, such as sodium/proton exchange,
sodium‐dependent chloride/bicarbonate exchange, efflux of undissociated lactic acid, and
outward proton/lactate cotransport. The change in proton concentration in the cell can be
calculated from the change in pH multiplied by the cytosolic buffer capacity and equals the
proton efflux rate minus the rate of proton generation by PCr resynthesis and aerobic ATP
production [43]. We calculated proton efflux rates E and apparent proton efflux rate
constants at the start of recovery. Both E and are pH dependent [44] and therefore only the data sets with a pHend between 6.6 and 6.8 were used to calculate an average value for
E and for each subject. The mean values that we found for the total cytosolic buffering
capacity , E and correspond with values reported by Kemp et al., calculated for a similar
pHend range with exactly the same formulas [44]. The slope of the relation between PCr and pHend was positively correlated with both E (R = 0.91, p = 0.03) and (R = 0.96, p = 0.01), indicating that subjects with a smaller pH dependence of PCr have a higher proton efflux rate, most likely as a result of a better blood flow due to e.g. an increased capillary density
possibly related to the subject’s fiber type composition. Higher proton efflux rates will lead
to faster pH recovery and therefore the observed correlations support our hypothesis that
the intersubject differences in the pH dependence of PCr are related to differences in the rate of pH recovery.
To overcome the problems of pH determination during recovery associated with the
transient loss of Pi signal, Chen et al. modeled the pH recovery based on the CK equilibrium
by considering the transition from exercise to recovery as a step function input [53]. The
entire pH recovery was characterized by calculating the time required for pH recovery
(tpHrec) and a strong linear correlation was observed between tpHrec and the half‐time of PCr
recovery in normal subjects (average pHend ~ 6.7). This strong correlation corroborates the
link between the PCr recovery rate and the overall pH recovery rate. Moreover, the large
variation in tpHrec within normal subjects (tpHrec ranged from about 2‐18 min) implies that
differences in the pH dependence of PCr can be significant, as we demonstrated in the
present study. Certain disorders, e.g. hypertension and mitochondrial myopathy [43] and
dermatomyositis and polymyositis [54], are associated with altered proton efflux from
muscle fibers, which will affect the rate of pH recovery. Therefore, when comparing PCr
recovery measurements between patients and controls, differences in proton efflux rates
or rates of pH recovery should be considered, as these might lead to systematic changes of
PCr, in particular when pHend is low.
Effect of acidosis on PCr recovery related to proton efflux
69
As an alternative to PCr, the kinetics of ADP recovery can be used to assess oxidative capacity. As [ADP] is the principal error signal in the feedback loop controlling
mitochondrial oxidation, ADP recovery is one of the most sensitive MRS indices of
mitochondrial function [37]. It has been shown that in contrast to PCr, ADP is independent of end‐exercise pH [1‐2, 17‐18, 20], which was confirmed by a theoretical model [48]. We
investigated the pH dependence of ADP for each subject. For 3 subjects ADP was not significantly correlated with pHend, while for the 3 other subjects ADP was positively correlated with pHend, i.e. ADP recovery became faster at low pH. This phenomenon was
also observed by Larson‐Meyer et al. [34], but we doubt that it has any physiological
meaning. It has been reported that the recovery of ADP is not always mono‐exponential
and that [ADP] can decrease below the resting concentration during the second min of
recovery [17]. The size and duration of this so‐called ADP undershoot was found to
correlate with pHmin [17]. We also observed an ADP undershoot in some of our data sets.
However, it was difficult to quantify this effect, as the undershoot occurs in the period
during which the Pi peak becomes invisible, resulting in a less reliable pH estimation. When
the ADP recovery data with an undershoot are fitted with a mono‐exponential function, the
time constant will be underestimated [55] and this could explain the positive correlation
between ADP and pHend. Furthermore, it was assumed that the CK equilibrium constant was
not affected by the different metabolic conditions present after exercise and therefore
changes in pH and the free Mg2+ concentration [56], in particular for the exercises at higher
intensities, are sources of error for the [ADP] calculation which might lead to deviations in
ADP.
VPCr is a measure of the actual mitochondrial ATP synthesis rate and therefore does not
represent an absolute measure of oxidative capacity. Still, a number of studies have
reported that VPCr is independent of pH [18‐19, 21]. This is a consequence of [ADP]end being
either similar for different degrees of acidification [21] or well above the accepted Km value
of 30 M for oxidative ATP synthesis [18‐19]. The latter was also the case for most of our
measurements and for five of the subjects VPCr was independent of pHend. VPCr can be
calculated from the product of 1/PCr and PCr (Eq. [3]) [18, 21, 40], like in our study, or VPCr
can be measured directly from the first data points (typically 10‐14 s) [19‐20, 34]. Boska et
al. applied both methods and found good correlations between calculated and measured
VPCr in controls (R = 0.753) and patients with peripheral vascular occlusive disease (R =
0.646) [20]. However, Walter et al. observed that at low pH (pHend 6.45) the calculated VPCr
was about two times smaller than the measured VPCr [19]. Under conditions in which
intracellular pH is decreased, any model that relies on PCr is no longer valid and VPCr should
be measured directly from the initial phase of recovery. However, this method is extremely
sensitive to sampling rate and signal‐to‐noise values [19]. Moreover, it has been shown that
VPCr measured from the first 10 s of recovery is underestimated by up to 56% and that a 1‐2
s time window is needed for the determination of VPCr, requiring very high time‐resolution 31P MRS data [57]. The fact that VPCr calculated from PCr was still independent of pHend in
Chapter 3
70
five of our subjects results from the fact that 1/PCr and PCr vary in opposite directions as a function of pHend (Fig. 3). For the sixth subject changes in these two quantities apparently
did not compensate each other completely.
In the literature, Qmax‐ADP has been found to be independent of pH [18, 21], although in the
study of Walter et al. only Qmax‐ADP based on the measured VPCr was pH independent [19]. In
our study, Qmax‐ADP was based on the calculated VPCr and [ADP]end (Eq. [4]). Although VPCr and
[ADP]end were not significantly correlated with pHend (except for one subject), these
parameters tend to vary in opposite directions as a function of pHend (Fig. 3), resulting in
significant positive correlations between Qmax‐ADP and pHend for three of the six subjects. For
each subject, there was a strong positive linear relationship between Qmax‐lin and pHend. Qmax‐
lin was calculated as the inverse of PCr multiplied by [PCr]rest (Eq. [5]). Within one subject,
[PCr]rest was more or less constant within the time span of the study and therefore Qmax‐lin
was equivalent to the inverse of PCr, showing an equally strong relationship with pHend. The
slopes of the relation between Qmax‐ADP and pHend were much smaller than for Qmax‐lin and
thus at low pH the error is smaller for Qmax‐ADP. However, in accordance with Walter et al.
[19], our data show that none of the models that rely on PCr are reliable to predict Qmax in
the presence of intracellular acidosis [19].
In conclusion, intracellular acidosis slowed PCr recovery and the pH dependence of PCr differed among subjects, ranging from ‐33.0 to ‐75.3 s per pH unit. The effect of acidosis on
PCr recovery kinetics after exercise correlated with the kinetics of proton efflux at the start
of recovery, strongly indicating that the intersubject differences in the pH dependence of
PCr reflect differences in the rate of pH recovery. Our study implies that simply correcting
PCr for end‐exercise pH using a general formula is not adequate, in particular when
comparing patients and controls, as certain disorders are characterized by altered proton
efflux from muscle fibers. Also, matching for end‐exercise pH is not sufficient when subject
groups systematically differ in proton efflux kinetics. Therefore, PCr can only be used as a measure of mitochondrial function when end‐exercise pH is close to resting values.
Avoiding a decrease in intracellular pH along with a sufficient drop in PCr to model PCr
recovery may however be difficult in untrained subjects or patients [19]. An exercise
protocol that progressively increases work, like we used in this study, has been reported to
be successful in decreasing PCr without severe acidification as opposed to sustained‐load
exercise [19, 24]. Indeed, we obtained data sets with pHend close to 7 with a drop in PCr of
roughly 50%. Alternatively, one could use an exercise protocol of short duration (9 s) with
very rapid contractions, which has the advantage that it is believed to simultaneously
recruit all fibers [19], or a gated protocol in which the acquisition is gated to contractions of
short duration without significant muscle acidification that are repeated in a steady state
for as many times as necessary to obtain the desired signal‐to‐noise ratio [58]. The kinetics
of ADP recovery is independent of end‐exercise pH. Disadvantages of using ADP as a measure of mitochondrial function are the complex time‐dependent undershoot of ADP
Effect of acidosis on PCr recovery related to proton efflux
71
during recovery and the assumptions that have to be made to calculate [ADP] [17]. Qmax can
only be used when based on VPCr directly measured from the initial recovery data points.
However, the reproducibility of the latter parameter is much lower than for PCr and ADP [19].
Acknowledgements
We are very grateful to Jan van Ooyen and Peter Coolen for their continuing support in
maintaining the MR scanner. We also like to express our appreciation to the research
volunteers who participated in this study.
Chapter 3
72
References 1. Arnold, D.L., P.M. Matthews, and G.K. Radda. Metabolic recovery after exercise and the assessment of
mitochondrial function in vivo in human skeletal muscle by means of 31P NMR. Magn Reson Med, 1984. 1(3): 307‐15.
2. Arnold, D.L., D.J. Taylor, and G.K. Radda. Investigation of human mitochondrial myopathies by phosphorus magnetic resonance spectroscopy. Ann Neurol, 1985. 18(2): 189‐96.
3. Kemp, G.J., D.J. Taylor, C.H. Thompson, L.J. Hands, B. Rajagopalan, P. Styles, and G.K. Radda. Quantitative analysis by 31P magnetic resonance spectroscopy of abnormal mitochondrial oxidation in skeletal muscle during recovery from exercise. NMR Biomed, 1993. 6(5): 302‐10.
4. Kent‐Braun, J.A., R.G. Miller, and M.W. Weiner. Magnetic resonance spectroscopy studies of human muscle. Radiol Clin North Am, 1994. 32(2): 313‐35.
5. Radda, G.K., J. Odoom, G. Kemp, D.J. Taylor, C. Thompson, and P. Styles. Assessment of mitochondrial function and control in normal and diseased states. Biochim Biophys Acta, 1995. 1271(1): 15‐9.
6. Mattei, J.P., D. Bendahan, and P. Cozzone. P‐31 magnetic resonance spectroscopy. A tool for diagnostic purposes and pathophysiological insights in muscle diseases. Reumatismo, 2004. 56(1): 9‐14.
7. Chance, B., J. Im, S. Nioka, and M. Kushmerick. Skeletal muscle energetics with PNMR: personal views and historic perspectives. NMR Biomed, 2006. 19(7): 904‐26.
8. Sahlin, K. Intracellular pH and energy metabolism in skeletal muscle of man. With special reference to exercise. Acta Physiol Scand Suppl, 1978. 455: 1‐56.
9. Taylor, D.J., P.J. Bore, P. Styles, D.G. Gadian, and G.K. Radda. Bioenergetics of intact human muscle. A 31P nuclear magnetic resonance study. Mol Biol Med, 1983. 1(1): 77‐94.
10. Quistorff, B., L. Johansen, and K. Sahlin. Absence of phosphocreatine resynthesis in human calf muscle during ischaemic recovery. Biochem J, 1993. 291( Pt 3): 681‐6.
11. Schrauwen‐Hinderling, V.B., M.E. Kooi, M.K. Hesselink, J.A. Jeneson, W.H. Backes, C.J. van Echteld, J.M. van Engelshoven, M. Mensink, and P. Schrauwen. Impaired in vivo mitochondrial function but similar intramyocellular lipid content in patients with type 2 diabetes mellitus and BMI‐matched control subjects. Diabetologia, 2007. 50(1): 113‐20.
12. Taylor, D.J., P. Styles, P.M. Matthews, D.A. Arnold, D.G. Gadian, P. Bore, and G.K. Radda. Energetics of human muscle: exercise‐induced ATP depletion. Magn Reson Med, 1986. 3(1): 44‐54.
13. Bendahan, D., S. Confort‐Gouny, G. Kozak‐Reiss, and P.J. Cozzone. Heterogeneity of metabolic response to muscular exercise in humans. New criteria of invariance defined by in vivo phosphorus‐31 NMR spectroscopy. FEBS Lett, 1990. 272(1‐2): 155‐8.
14. Iotti, S., R. Lodi, C. Frassineti, P. Zaniol, and B. Barbiroli. In vivo assessment of mitochondrial functionality in human gastrocnemius muscle by 31P MRS. The role of pH in the evaluation of phosphocreatine and inorganic phosphate recoveries from exercise. NMR Biomed, 1993. 6(4): 248‐53.
15. McCully, K.K., S. Iotti, K. Kendrick, Z. Wang, J.D. Posner, J. Leigh, Jr., and B. Chance. Simultaneous in vivo measurements of HbO2 saturation and PCr kinetics after exercise in normal humans. J Appl Physiol, 1994. 77(1): 5‐10.
16. Takahashi, H., M. Inaki, K. Fujimoto, S. Katsuta, I. Anno, M. Niitsu, and Y. Itai. Control of the rate of phosphocreatine resynthesis after exercise in trained and untrained human quadriceps muscles. Eur J Appl Physiol Occup Physiol, 1995. 71(5): 396‐404.
17. Argov, Z., N. De Stefano, and D.L. Arnold. ADP recovery after a brief ischemic exercise in normal and diseased human muscle‐‐a 31P MRS study. NMR Biomed, 1996. 9(4): 165‐72.
18. Lodi, R., G.J. Kemp, S. Iotti, G.K. Radda, and B. Barbiroli. Influence of cytosolic pH on in vivo assessment of human muscle mitochondrial respiration by phosphorus magnetic resonance spectroscopy. Magma, 1997. 5(2): 165‐71.
19. Walter, G., K. Vandenborne, K.K. McCully, and J.S. Leigh. Noninvasive measurement of phosphocreatine recovery kinetics in single human muscles. Am J Physiol, 1997. 272(2 Pt 1): C525‐34.
20. Boska, M.D., J.A. Nelson, N. Sripathi, Pipinos, II, A.D. Shepard, and K.M. Welch. 31P MRS studies of exercising human muscle at high temporal resolution. Magn Reson Med, 1999. 41(6): 1145‐51.
21. Roussel, M., D. Bendahan, J.P. Mattei, Y. Le Fur, and P.J. Cozzone. 31P magnetic resonance spectroscopy study of phosphocreatine recovery kinetics in skeletal muscle: the issue of intersubject variability. Biochim Biophys Acta, 2000. 1457(1‐2): 18‐26.
22. Harkema, S.J. and R.A. Meyer. Effect of acidosis on control of respiration in skeletal muscle. Am J Physiol, 1997. 272(2 Pt 1): C491‐500.
23. Walsh, B., T. Tiivel, M. Tonkonogi, and K. Sahlin. Increased concentrations of P(i) and lactic acid reduce creatine‐stimulated respiration in muscle fibers. J Appl Physiol, 2002. 92(6): 2273‐6.
24. Jubrias, S.A., G.J. Crowther, E.G. Shankland, R.K. Gronka, and K.E. Conley. Acidosis inhibits oxidative phosphorylation in contracting human skeletal muscle in vivo. J Physiol, 2003. 553(Pt 2): 589‐99.
Effect of acidosis on PCr recovery related to proton efflux
73
25. Chance, B. and H. Conrad. Acid‐linked functions of intermediates in oxidative phosphorylation. II. Experimental studies of the effect of pH upon respiratory, phosphorylative and transfer activities of liver and heart mitochondria. J Biol Chem, 1959. 234(6): 1568‐70.
26. Tobin, R.B., C.R. Mackerer, and M.A. Mehlman. pH effects on oxidative phosphorylation of rat liver mitochondria. Am J Physiol, 1972. 223(1): 83‐8.
27. Mitchelson, K.R. and F.J. Hird. Effect of pH and halothane on muscle and liver mitochondria. Am J Physiol, 1973. 225(6): 1393‐8.
28. Nioka, S., Z. Argov, G.P. Dobson, R.E. Forster, H.V. Subramanian, R.L. Veech, and B. Chance. Substrate regulation of mitochondrial oxidative phosphorylation in hypercapnic rabbit muscle. J Appl Physiol, 1992. 72(2): 521‐8.
29. Willis, W.T. and M.R. Jackman. Mitochondrial function during heavy exercise. Med Sci Sports Exerc, 1994. 26(11): 1347‐53.
30. Connett, R.J. Analysis of metabolic control: new insights using scaled creatine kinase model. Am J Physiol, 1988. 254(6 Pt 2): R949‐59.
31. Funk, C.I., A. Clark, Jr., and R.J. Connett. A simple model of aerobic metabolism: applications to work transitions in muscle. Am J Physiol, 1990. 258(6 Pt 1): C995‐1005.
32. Baker, A.J., R. Brandes, T.M. Schendel, S.D. Trocha, R.G. Miller, and M.W. Weiner. Energy use by contractile and noncontractile processes in skeletal muscle estimated by 31P‐NMR. Am J Physiol, 1994. 266(3 Pt 1): C825‐31.
33. Boska, M. ATP production rates as a function of force level in the human gastrocnemius/soleus using 31P MRS. Magn Reson Med, 1994. 32(1): 1‐10.
34. Larson‐Meyer, D.E., B.R. Newcomer, G.R. Hunter, H.P. Hetherington, and R.L. Weinsier. 31P MRS measurement of mitochondrial function in skeletal muscle: reliability, force‐level sensitivity and relation to whole body maximal oxygen uptake. NMR Biomed, 2000. 13(1): 14‐27.
35. Lawson, J.W. and R.L. Veech. Effects of pH and free Mg2+ on the Keq of the creatine kinase reaction and other phosphate hydrolyses and phosphate transfer reactions. J Biol Chem, 1979. 254(14): 6528‐37.
36. Iotti, S., C. Frassineti, A. Sabatini, A. Vacca, and B. Barbiroli. Quantitative mathematical expressions for accurate in vivo assessment of cytosolic [ADP] and DeltaG of ATP hydrolysis in the human brain and skeletal muscle. Biochim Biophys Acta, 2005. 1708(2): 164‐77.
37. Kemp, G.J. and G.K. Radda. Quantitative interpretation of bioenergetic data from 31P and 1H magnetic resonance spectroscopic studies of skeletal muscle: an analytical review. Magn Reson Q, 1994. 10(1): 43‐63.
38. Richardson, R.S., L.R. Frank, and L.J. Haseler. Dynamic knee‐extensor and cycle exercise: functional MRI of muscular activity. Int J Sports Med, 1998. 19(3): 182‐7.
39. Vanhamme, L., A. van den Boogaart, and S. Van Huffel. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson, 1997. 129(1): 35‐43.
40. Kemp, G.J., C.H. Thompson, P.R. Barnes, and G.K. Radda. Comparisons of ATP turnover in human muscle during ischemic and aerobic exercise using 31P magnetic resonance spectroscopy. Magn Reson Med, 1994. 31(3): 248‐58.
41. Chance, B., J.S. Leigh, Jr., B.J. Clark, J. Maris, J. Kent, S. Nioka, and D. Smith. Control of oxidative metabolism and oxygen delivery in human skeletal muscle: a steady‐state analysis of the work/energy cost transfer function. Proc Natl Acad Sci U S A, 1985. 82(24): 8384‐8.
42. Meyer, R.A. A linear model of muscle respiration explains monoexponential phosphocreatine changes. Am J Physiol, 1988. 254(4 Pt 1): C548‐53.
43. Kemp, G.J., D.J. Taylor, P. Styles, and G.K. Radda. The production, buffering and efflux of protons in human skeletal muscle during exercise and recovery. NMR Biomed, 1993. 6(1): 73‐83.
44. Kemp, G.J., C.H. Thompson, D.J. Taylor, and G.K. Radda. Proton efflux in human skeletal muscle during recovery from exercise. Eur J Appl Physiol Occup Physiol, 1997. 76(5): 462‐71.
45. Wolfe, C.L., H.F. Gilbert, K.M. Brindle, and G.K. Radda. Determination of buffering capacity of rat myocardium during ischemia. Biochim Biophys Acta, 1988. 971(1): 9‐20.
46. Mainwood, G.W. and J.M. Renaud. The effect of acid‐base balance on fatigue of skeletal muscle. Can J Physiol Pharmacol, 1985. 63(5): 403‐16.
47. Roos, A. and W.F. Boron. Intracellular pH. Physiol Rev, 1981. 61(2): 296‐434. 48. Styles, P., G.J. Kemp, and G.K. Radda. A model for metabolic control which reproduces the main features
of recovery from exercise which are observed by 31P MRS. Proceedings of the 11th Annual Meeting of the Society of Magnetic Resonance in Medicine, Berlin, Germany, , 1992: p. 2702.
49. Rossiter, H.B., S.A. Ward, F.A. Howe, J.M. Kowalchuk, J.R. Griffiths, and B.J. Whipp. Dynamics of intramuscular 31P‐MRS P(i) peak splitting and the slow components of PCr and O2 uptake during exercise. J Appl Physiol, 2002. 93(6): 2059‐69.
Chapter 3
74
50. Iotti, S., R. Funicello, P. Zaniol, and B. Barbiroli. Kinetics of post‐exercise phosphate transport in human skeletal muscle: an in vivo 31P‐MR spectroscopy study. Biochem Biophys Res Commun, 1991. 176(3): 1204‐9.
51. Bendahan, D., S. Confort‐Gouny, G. Kozak‐Reiss, and P.J. Cozzone. Pi trapping in glycogenolytic pathway can explain transient Pi disappearance during recovery from muscular exercise. A 31P NMR study in the human. FEBS Lett, 1990. 269(2): 402‐5.
52. Walter, G., K. Vandenborne, M. Elliott, and J.S. Leigh. In vivo ATP synthesis rates in single human muscles during high intensity exercise. J Physiol, 1999. 519 (Pt 3): 901‐10.
53. Chen, J.T., T. Taivassalo, Z. Argov, and D.L. Arnold. Modeling in vivo recovery of intracellular pH in muscle to provide a novel index of proton handling: application to the diagnosis of mitochondrial myopathy. Magn Reson Med, 2001. 46(5): 870‐8.
54. Cea, G., D. Bendahan, D. Manners, D. Hilton‐Jones, R. Lodi, P. Styles, and D.J. Taylor. Reduced oxidative phosphorylation and proton efflux suggest reduced capillary blood supply in skeletal muscle of patients with dermatomyositis and polymyositis: a quantitative 31P‐magnetic resonance spectroscopy and MRI study. Brain, 2002. 125(Pt 7): 1635‐45.
55. Chen, J.T., Z. Argov, R.E. Kearney, and D.L. Arnold. Fitting cytosolic ADP recovery after exercise with a step response function. Magn Reson Med, 1999. 41(5): 926‐32.
56. Iotti, S., C. Frassineti, L. Alderighi, A. Sabatini, A. Vacca, and B. Barbiroli. In vivo (31)P‐MRS assessment of cytosolic [Mg(2+)] in the human skeletal muscle in different metabolic conditions. Magn Reson Imaging, 2000. 18(5): 607‐14.
57. Newcomer, B.R., M.D. Boska, and H.P. Hetherington. Non‐P(i) buffer capacity and initial phosphocreatine breakdown and resynthesis kinetics of human gastrocnemius/soleus muscle groups using 0.5 s time‐resolved (31)P MRS at 4.1 T. NMR Biomed, 1999. 12(8): 545‐51.
58. Slade, J.M., T.F. Towse, M.C. Delano, R.W. Wiseman, and R.A. Meyer. A gated 31P NMR method for the estimation of phosphocreatine recovery time and contractile ATP cost in human muscle. NMR Biomed, 2006. 19(5): 573‐80.
75
76
Adapted from
H. M. M. L. De Feyter, N. M. A. van den Broek, S. F. E. Praet, K. Nicolay, L. J. C. van Loon and
J. J. Prompers. Early or advanced stage type 2 diabetes is not accompanied by in vivo
skeletal muscle mitochondrial dysfunction.
Eur J Endocrinol, 2008. 158(5):643‐53.
Chapter Early or advanced stage type 2 diabetes is
not accompanied by in vivo skeletal muscle
mitochondrial dysfunction
Chapter 4
78
Abstract
Objective: Several lines of evidence support a potential role of skeletal muscle
mitochondrial dysfunction in the pathogenesis of insulin resistance and/or type 2 diabetes.
However, it remains to be established whether mitochondrial dysfunction represents either
cause or consequence of the disease. We examined in vivo skeletal muscle mitochondrial
function in early and advanced stages of type 2 diabetes, with the aim to gain insight in the
proposed role of mitochondrial dysfunction in the etiology of insulin resistance and/or type
2 diabetes.
Methods: 10 long‐standing, insulin‐treated type 2 diabetes patients, 11 subjects with
impaired fasting glucose, impaired glucose tolerance and/or recently diagnosed type 2
diabetes, and 12 healthy, normoglycaemic controls, matched for age and body composition
and with low habitual physical activity levels were studied. In vivo mitochondrial function of
the vastus lateralis muscle was evaluated from post‐exercise phosphocreatine (PCr)
recovery kinetics using 31P magnetic resonance spectroscopy (MRS). Intramyocellular lipid
(IMCL) content was assessed in the same muscle using single‐voxel 1H MRS.
Results: IMCL content tended to be higher in type 2 diabetes patients compared to
normoglycaemic controls (p = 0.06). 31P MRS parameters for mitochondrial function, i.e. PCr
and ADP recovery time constants and maximum aerobic capacity, did not differ between
groups.
Conclusions: The finding that in vivo skeletal muscle oxidative capacity does not differ
between long‐standing, insulin‐treated type 2 diabetes patients, subjects with early stage
type 2 diabetes, and sedentary, normoglycaemic controls suggests that mitochondrial
dysfunction does not necessarily represent either cause or consequence of insulin
resistance and/or type 2 diabetes.
In vivo mitochondrial function in type 2 diabetes
79
Introduction
Insulin resistance is an early event in the pathogenesis of type 2 diabetes. However, the
exact processes leading to insulin resistance remain unresolved. Previous studies have
reported a strong correlation between intramyocellular lipid (IMCL) content and insulin
resistance [1‐2]. However, the proposed relationship between IMCL accumulation and
skeletal muscle insulin resistance is not unambiguous, as it is strongly influenced by training
status and/or habitual physical activity [3‐4]. Nonetheless, several lines of evidence indicate
that mitochondrial dysfunction, presumably associated with a reduced capacity to oxidize
fatty acids, might stimulate IMCL accretion and, as such, contribute to the development of
skeletal muscle insulin resistance [5].
Data to support the proposed role of skeletal muscle mitochondrial dysfunction in the
development of insulin resistance and/or type 2 diabetes have been obtained with various
in vitro methods, including measurements of oxidative enzyme activities [6‐10], mRNA
and/or protein expression of OXPHOS (oxidative phosphorylation) genes [10‐14] as well as
mitochondrial content, morphology and respiration [8‐9, 14‐15]. Furthermore, in vivo
magnetic resonance spectroscopy (MRS) measurements of basal mitochondrial adenosine
triphosphate (ATP) synthesis rates [16‐18] and post‐exercise phosphocreatine (PCr) [19]
and adenosine diphosphate (ADP) [20] recovery kinetics also point towards a potential role
for mitochondrial dysfunction in the etiology of insulin resistance and/or type 2 diabetes.
However, it was recently shown that in vitro mitochondrial respiration of permeabilized
muscle fibers from biopsies of type 2 diabetes patients and healthy controls did not differ
between groups when the data was normalized for mitochondrial DNA content or citrate
synthase activity [21]. In other words, the type 2 diabetes patients showed normal intrinsic
mitochondrial function, but an impaired oxidative capacity that was entirely attributed to a
lower mitochondrial content. Environmental factors play an important role in regulating
skeletal muscle oxidative capacity, and the lower mitochondrial content in type 2 diabetes
patients might simply be the result of a reduced habitual physical activity level [21‐23].
Furthermore, mitochondrial dysfunction in type 2 diabetes might be secondary to impaired
insulin signaling [24‐26] and/or abnormal blood glucose, insulin [27‐28] and free fatty acid
(FFA) [26] levels. Therefore, the debate continues as to whether mitochondrial dysfunction
represents either cause or consequence of insulin resistance and/or type 2 diabetes.
The aim of this study was to assess whether in vivo mitochondrial function is impaired in
the early and/or overt diabetes state. Therefore, we examined in vivo muscle mitochondrial
function in 3 groups of subjects representative for different stages in the development of
type 2 diabetes: long‐standing, insulin‐treated type 2 diabetes patients, subjects with
impaired fasting glucose, impaired glucose tolerance and/or recently diagnosed type 2
diabetes, and healthy, normoglycaemic controls, all matched for age and body composition
and with low habitual physical activity levels. In vivo mitochondrial function was
determined from post‐exercise PCr recovery kinetics measured with 31P MRS [29‐31].
Chapter 4
80
Materials and methods
Subjects
Ten long‐standing, insulin‐treated type 2 diabetes patients (Diabetes group), 11 subjects
with impaired fasting glucose, impaired glucose tolerance and/or recently diagnosed type 2
diabetes (pre‐Diabetes group) and 12 healthy, normoglycaemic controls (Control group)
were selected to participate in this study. All subjects were male. The diabetes patients had
been diagnosed with type 2 diabetes for over 5 years, established by a fasting plasma
glucose concentration larger than or equal to 7.0 mmol/l at the time of diagnosis as defined
by the World Health Organisation (WHO) [32]. Subjects in the Control and pre‐Diabetes
groups had no family history of diabetes and were selected based on an oral glucose
tolerance test (OGTT) according to WHO criteria [32]. Control subjects showed normal
fasting glucose concentrations and normal glucose tolerance (fasting glucose < 6.1 mmol/l
and 2 h glucose < 7.8 mmol/l). Subjects in the pre‐Diabetes group had either elevated
fasting plasma glucose concentrations (fasting glucose ≥ 6.1 and < 7.0 mmol/l, and 2 h
glucose < 7.8 mmol/l; n = 5), or impaired glucose tolerance (fasting glucose < 7.0 mmol/l,
and 2 h glucose ≥ 7.8 and < 11.1 mmol/l; n = 2), or recently diagnosed type 2 diabetes (< 1
month; fasting glucose ≥ 7.0 mmol/l or 2 h glucose ≥ 11.1 mmol/l; n = 4).
All diabetes patients were on exogenous insulin treatment and had been on a stable
regimen of diabetes medication over the last 3 months before being recruited. Out of the
10 participating diabetes patients, 7 patients were treated with short (Novorapid®, n = 6) or
rapid acting insulin (Humulin®, n = 1) before each meal either in combination with NPH
insulin (Insulatard®, n = 5), premixed biphasic isophane insulin (Mixtard 30/70® in
combination with metformin, n = 1), or a very long‐acting insulin analogue (insulin glargine,
n = 1), all administered before bedtime. Two patients were treated with premixed biphasic
isophane insulin twice a day (Mixtard 30/70®) in combination with metformin. One patient
used NPH insulin (Humulin NPH®) once a day before breakfast in combination with
metformin and a sulphonylurea (glimepiride). Patients using thiazolidinediones were
excluded from participation. None of the subjects in the pre‐Diabetes group used blood
glucose lowering medication and all showed glycosylated haemoglobin (HbA1c) contents
below 6%. Subjects using ‐blockers for less than 6 months and subjects with impaired liver
function, renal failure, severe retinopathy or a history of severe cardiovascular problems
were excluded from participation. Subjects with clinical signs of peripheral vascular disease
(Fontaine stage II or higher) were excluded based on a thorough history taking for signs of
intermittent claudication, as well as a physical examination of the peripheral vascular
system (no arterial bruits over abdominal aorta, iliac artery and/or diminished pulses of
femoral artery, dorsalis pedis and posterior tibial artery). The nature and the risks of the
experimental procedures were explained to the subjects and all gave their written
informed consent to participate in the study, which was approved by the local Medical
Ethical Committee of the Máxima Medical Center, Veldhoven, The Netherlands. This study
In vivo mitochondrial function in type 2 diabetes
81
is part of a larger project that studies mitochondrial function in chronic metabolic disease
[33‐34].
Body composition
Body mass index (BMI) and waist circumference were measured using an analogue weight
scale and standard measuring tape. Whole‐body fat free mass (FFM) and truncal fat mass
were determined using whole‐body dual energy X‐ray absorptiometry (DXA) (Hologic QDR‐
4500 Discovery A, software version 12.3:3, Hologic Inc. Bedford, MA, USA).
Habitual physical activity level
Habitual physical activity level was assessed with the Tecumseh and Minnesota
Occupational and Leisure Time Activity Questionnaire [35]. The activity level was expressed
in metabolic equivalents (MET), which is a scale of the energy cost of various physical
activities in multiples of the resting metabolic rate.
Whole‐body oxygen uptake capacity
Maximal whole‐body oxygen uptake capacity (VO2peak) and maximal workload capacity
(Wmax) were measured during an incremental exercise test until exhaustion, performed on a
cycle ergometer (Medifit Ergometer, Medifit systems, Maarn, The Netherlands) using a
ramp protocol [36]. Gas exchange measurements were performed continuously (Ergostar II,
PMS Professional Medical Systems, Basel, Switzerland). Maximal whole‐body oxygen
uptake capacity was defined as the VO2 value remaining unchanged or increasing less than
1 ml∙min‐1∙kg‐1 for 30 s or more despite an increment in workload [37]. Cardiac function was
monitored using a 12‐lead electrocardiogram with heart rate being recorded continuously
(Polar Electro, Kempele, Finland). The age‐predicted maximal heart rate was calculated
according to the prediction model of Gellish et al. [38].
Blood sampling and analyses
Subjects reported at the laboratory at 8.00 a.m. after an overnight fast. After 5‐10 min of
supine rest, fasting samples of venous blood were collected from an antecubital vein.
Subsequently, a standard OGTT was performed for all subjects, except for the diabetes
patients, and blood samples were collected 2 h after ingestion of the glucose load. Blood
plasma samples were collected into EDTA containing tubes and centrifuged for 10 min at
4C. Aliquots of plasma were frozen immediately in liquid nitrogen and stored at ‐80C until
Chapter 4
82
further analyses. Plasma concentrations of glucose (Roche, Basel, Switzerland) and NEFA
(Wako Chemicals, Neuss, Germany) were analyzed with a COBAS semi‐automatic analyzer
(Roche). Plasma insulin was determined in duplicate by radioimmunoassay (Linco, St.
Charles, MO, USA) for Control and pre‐Diabetes groups. Cross sensitivity of exogenously
administered insulin with this radioimmunoassay prohibited the detection of endogenously
produced insulin in the insulin‐treated diabetes patients. Blood HbA1c content was analyzed
by high‐performance liquid chromatography (Bio‐Rad Diamat, Munich, Germany). For the
Control and pre‐Diabetes groups, the homeostasis model assessment (HOMA) index [39]
was calculated.
MRS measurements
MRS measurements were performed with a 1.5‐Tesla whole‐body scanner (Gyroscan
S15/ACS, Philips Medical Systems, Best, The Netherlands) during 2 sessions. On the evening
before the first MRS session, subjects received a standardized meal (mean ± SD: 41.0±15.1
kJ per kg body weight, containing 40.7 energy% (En%) fat, 14.6 En% protein, and 44.7 En%
carbohydrate) after which subjects remained fasted and were allowed to drink water only.
Subjects travelled by car or public transport and reported at the laboratory at 8.30 a.m.,
where they received a standardized breakfast. First, 1H MRS measurements were
performed for the quantification of IMCL. After a short break, the 31P MRS protocol was
carried out to familiarize the subjects to the in‐magnet exercise and to determine the
optimal exercise intensity for the second visit. During the second MRS session, scheduled
within one week, 31P MRS measurements were performed to assess skeletal muscle
mitochondrial function. For the diabetes patients, blood glucose lowering medication was
not withdrawn before the MRS measurements.
1H MRS
IMCL was measured in the M. vastus lateralis with image‐guided single‐voxel 1H MRS using
the body coil for transmission and a 8‐cm diameter surface coil for signal reception. For
each subject, 5 voxels with a size of 10×10×15 mm3 were measured at different positions
within the M. vastus lateralis. The voxels were carefully placed to avoid subcutaneous fat
and visible interstitial fat using standard T1‐weighted images (Fig. 1A). Spectra were
recorded with a point resolved spectroscopy (PRESS) sequence (repetition time, 1500 ms;
echo time, 35 ms; spectral width, 2000 Hz; number of data points, 2048; 128 averages)
using chemical shift selective (CHESS) saturation for water suppression. Unsuppressed
water spectra (32 averages) were recorded from the same voxels and used as an internal
reference.
In vivo mitochondrial function in type 2 diabetes
83
1H MRS data analysis
All spectra were fitted in the time domain by using a nonlinear least squares algorithm
(AMARES) in the jMRUI software package [40] without further post‐processing, except for
manual phasing. The unsuppressed water spectrum was phased and fitted to a Lorentzian
line shape. The zero‐order phase correction from the water spectrum was applied to the
corresponding water suppressed spectrum and the total creatine (tCr) CH3 peak was
referenced to 3.02 ppm. In the water suppressed spectrum, peaks from
trimethylammonium (TMA) and tCr CH3, extramyocellular lipid (EMCL) and IMCL CH2 and
EMCL and IMCL CH3 protons (see Fig. 1B) were fitted to Gaussian line shapes. The positions
and areas of the EMCL and IMCL CH3 peaks were constrained with respect to the positions
and areas of the EMCL and IMCL CH2 peaks, respectively [41]. To increase the reliability of
the fit, the linewidth of the IMCL CH2 peak ( ) was constrained with respect to the
linewidth of the water peak ( ) according to
Figure 1. (A) Transversal T1‐weighted spin‐echo image of the upper leg of a type 2 diabetes patient with
voxel positioning within the M. vastus lateralis for single‐voxel 1H MRS. The size of the voxel was
10×10×15 mm3. The arrow points at a phantom which was located in the centre of the 8‐cm diameter
surface coil. (B) 1H MR spectrum from the voxel depicted in panel A. The spectrum was processed with 1
Hz line broadening. Peak annotations: tCr CH2/CH3, CH2/CH3 protons of total creatine; TMA, CH3 protons
of trimethylammonium groups; EMCL CH2/CH3, CH2/CH3 protons of extramyocellular lipid, IMCL CH2/CH3,
CH2/CH3 protons of intramyocellular lipid.
Chapter 4
84
0.09 1.02 · . 1
This constraint was derived from previously recorded data sets with a well resolved IMCL
peak (number of data sets = 20, R = 0.821, p < 0.0001). A soft constraint, which was
empirically determined from previously recorded data sets, was applied to the linewidth of
the EMCL CH2 peak ( ), i.e.. 3.33 . When the resonance
frequency () of the IMCL CH2 peak could not be accurately fitted (i.e. ≤ 1.25 ppm or ≥ 1.31 ppm), the position of the IMCL CH2 peak was constrained with respect to the position
of the tCr CH3 peak (chemical shift difference = 3.02 – 1.28 ppm). IMCL was expressed as a
percentage of the water signal measured in the same voxel without water suppression.
IMCL levels determined from different voxels of one subject were averaged.
31P MRS 31P MRS was performed as described previously [33]. In short, 31P signals were collected
using a 6‐cm diameter surface coil placed over the M. vastus lateralis (spectral width, 2000
Hz; number of data points, 1024). A fully relaxed spectrum was measured at rest with a
repetition time of 30 s and 24 scans. Then, spectra were acquired during a rest‐exercise‐
recovery protocol with a repetition time of 3 s and 2 scans yielding a time resolution of 6 s.
The first 20 spectra (2 min) were measured at rest, after which the subjects started the
exercise. Subjects performed an dynamic incremental single‐leg extension exercise in the
supine position using a home‐built MR compatible ergometer [33]. The duration of the
exercise varied per subject, but never exceeded 8 min, so that at least 5 min of recovery
were recorded. During the first visit, subjects performed a test run and exercised until
fatigued. During the second session, the duration of the exercise was chosen to deplete PCr
by about 50%, while aiming to avoid the intracellular pH to drop below 6.8.
31P MRS data analysis
The 31P MRS data were analyzed as described previously [33]. In short, spectra were fitted
in the time domain by using a nonlinear least squares algorithm (AMARES) in the jMRUI
software package [40]. PCr, inorganic phosphate (Pi), ATP and phosphodiester (PDE) signals
were fitted to Lorentzian line shapes. Absolute concentrations of the phosphorylated
metabolites were calculated after correction for partial saturation and assuming that the
ATP concentration is 8.2 mM at rest [42]. Intracellular pH was calculated from the chemical
shift difference between the Pi and PCr resonances [43]. The free cytosolic ADP
concentration was calculated from the PCr concentration and pH using a creatine kinase
equilibrium constant (Keq) of 1.66×109 M‐1 [44]. Recoveries of PCr and ADP were fitted to
In vivo mitochondrial function in type 2 diabetes
85
mono‐exponential functions using Matlab (version 7.3., Mathworks, Natick, Massachusetts,
USA). Results are expressed as the metabolite’s time constant of recovery, i.e. PCr and ADP, which are both measures of mitochondrial function [43, 45‐46]. Calculation of the initial
rate of PCr recovery (VPCr) was based on the PCr recovery rate constant (1/PCr) and the difference between the resting and end‐exercise PCr concentrations [47]. Calculation of the
maximum aerobic capacity (Qmax) was based on the ADP‐control model [48], in which VPCr
has a hyperbolic dependence on the end‐exercise ADP concentration according to
Michaelis‐Menten kinetics with a Km of 30 M [47].
Statistics
All data are expressed as means ± SD. Statistical analyses were performed using the SPSS
14.0 software package (SPSS Inc, Chicago, IL, USA). An ANOVA test was performed to test
for overall differences between the Control, pre‐Diabetes and Diabetes groups and a
Bonferroni post‐hoc test was applied for pair‐wise comparisons between the 3 groups. For
comparisons between 2 groups, differences were determined with an unpaired t‐test. To
evaluate the relationships between variables, Pearson correlation coefficients were
calculated. All tests were carried out in a 2‐sided way and the level of significance was set
at p < 0.05.
Results
Subjects’ characteristics
The subjects’ characteristics are shown in Table 1. Subjects were carefully matched for age
and BMI. In accordance, FFM and truncal fat mass did not differ between the Control, pre‐
Diabetes and Diabetes groups (p = 0.29 and p = 0.87, respectively). Wmax and VO2peak were
similar for the Control and pre‐Diabetes groups (p = 1.00), but significantly lower in the
Diabetes group. The latter was paralleled by a lower maximal heart rate and a lower
habitual physical activity level, but the physical activity levels were not significantly
different between groups (p = 0.12). Wmax and VO2peak correlated with both maximal heart
rate (Pearson’s R between 0.59 and 0.68, p < 0.01) and percentage of the age‐predicted
maximal heart rate (Pearson’s R between 0.56 and 0.62, p < 0.01).
Chapter 4
86
Table 1. Subjects’ characteristics. Data are expressed as means ± SD.
Control pre‐Diabetes Diabetes
n 12 11 10
Age (yrs) 56.5 ± 6.0 58.5 ± 5.0 58.8 ± 7.6
BMI (kg/m2) 32.9 ± 4.6 32.1 ± 3.2 31.8 ± 4.0
Body weight (kg) 101.0 ± 14.7 103.9 ± 10.0 95.0 ± 14.4
FFM (kg) 70.7 ± 7.1 72.4 ± 6.9 67.2 ± 8.4
Truncal fat mass (kg) 15.5 ± 5.5 16.3 ± 2.9 15.4 ± 4.5
Wmax (W) 247 ± 42 244 ± 30 156 ± 38*†
Wmax per kg BW (W/kg) 2.48 ± 0.47 2.39 ± 0.47 1.66 ± 0.36*†
VO2peak per kg BW (ml∙min‐1∙kg
‐1) 32.3 ± 5.4 33.8 ± 5.9 25.2 ± 3.5*
†
VO2peak per kg FFM (ml∙min‐1∙kg
‐1) 45.6 ± 6.0 48.5 ± 8.3 35.5 ± 5.2*
†
HRmax (b/min) 166 ± 13 171 ± 12 140 ± 23*†
% predicted HRmax 99 ± 8 103 ± 7 84 ± 14*†
Activity level (MET h∙day‐1) 19.3 ± 7.4 19.5 ± 8.7 13.6 ± 4.3
Fasting glucose (mmol/l) 5.7 ± 0.2 6.6 ± 0.5 11.0 ± 2.5*†
Fasting insulin (µU/ml) 16.2 ± 9.1 22.2 ± 9.6 nd
HOMA index 4.1 ± 2.2 6.5 ± 2.7#
nd
2‐h glucose (mmol/l) 5.7 ± 1.3 8.4 ± 2.6‡
nd
2‐h insulin (µU/ml) 74.8 ± 40.5 146.2 ± 64.5‡
nd
HbA1c (%) 5.3 ± 0.3 5.5 ± 0.2 7.7 ± 1.0*†
FFA (mmol/l) 0.31 ± 0.10 0.39 ± 0.13 0.46 ± 0.26
Years with type 2 diabetes na na 12 ± 7
Years of insulin therapy na na 7 ± 8
BMI, body mass index; FFM, fat free mass; Wmax, maximal workload capacity; Wmax per kg BW, Wmax per kg
body weight; VO2peak, maximal oxygen uptake per kg body weight or per kg FFM; HRmax, maximal heart rate
during the VO2peak test; % predicted HRmax, percentage of the age‐predicted HRmax; MET, metabolic
equivalents; HOMA, homeostasis model assessment; 2‐h glucose/insulin, glucose/insulin concentration 2 h
after ingestion of glucose load during the oral glucose tolerance test; HbA1c, glycosylated haemoglobin;
FFA, free fatty acids; na, not applicable; nd, not determined. * Significantly different from Control (ANOVA,
p < 0.01). † Significantly different from pre‐Diabetes (ANOVA, p < 0.01).
# Significantly different from
Control (t‐test, p < 0.05). ‡ Significantly different from Control (t‐test, p < 0.01).
In vivo mitochondrial function in type 2 diabetes
87
Fasting plasma glucose and insulin levels were not significantly different for the Control and
pre‐Diabetes groups (p = 0.39 and p = 0.14, respectively), but the HOMA index was
significantly higher for the pre‐Diabetes group than for the Control group. Also the glucose
and insulin concentrations measured 2 h after ingestion of the glucose load during the
OGTT were significantly higher for the pre‐Diabetes group than for the Control group. HbA1c
contents were similar for the Control and pre‐Diabetes groups (p = 1.00) and all subjects
had HbA1c contents below 6%. Fasting glucose concentrations and HbA1c content were
significantly higher in the Diabetes group (glucose ≥ 8 mmol/l and HbA1c ≥ 6.3%) compared
to the Control and pre‐Diabetes groups. Diabetes patients had been diagnosed with type 2
diabetes for 12 ± 7 years and had been on exogenous insulin therapy for 7 ± 8 years. Plasma
FFA concentrations did not differ between groups (p = 0.16).
Wmax and VO2peak correlated negatively with both fasting plasma glucose (Pearson’s R
between ‐0.48 and ‐0.57, p < 0.01) and HbA1c (Pearson’s R between ‐0.58 and ‐0.70, p <
0.01).
IMCL content
Due to excessive EMCL contamination in the spectra, IMCL content could not be
determined for 2 subjects in the Control group and 1 subject in the Diabetes group. For all
other subjects, spectra from at least 2 out of the 5 voxels were quantified. IMCL content
tended to be higher for the Diabetes group compared to the Control group (1.3 ± 0.4 (n =
10), 1.6 ± 0.7 (n = 11) and 2.0 ± 0.7 (n = 9) % of the water signal in Control, pre‐Diabetes
and Diabetes groups, respectively, p = 0.06; Fig. 2).
Figure 2. IMCL content in Control, pre‐Diabetes
and Diabetes groups. Bars indicate the mean
values for the 3 groups and filled circles
represent individual data points. IMCL content
tended to be higher for the Diabetes group
compared to the Control group (ANOVA: p =
0.06).
Chapter 4
88
Skeletal muscle mitochondrial function
Figure 3a and 3b show typical examples of 31P MR spectra from a subject’s vastus lateralis
muscle at rest and at the end of exercise, respectively. Table 2 summarizes the baseline,
end‐exercise and recovery 31P MRS results for the Control, pre‐Diabetes and Diabetes
groups. At rest, PCr, Pi, ADP and PDE concentrations and intracellular pH were not
significantly different for the Control, pre‐Diabetes and Diabetes groups (p = 0.93, p = 0.50,
p = 0.09, p = 0.42 and p = 0.10, respectively). The end‐exercise status has to be taken into
account when analysing the recovery data. End‐exercise metabolite concentrations were
similar for the 3 groups, except for the end‐exercise ADP concentration, which was lower in
the Diabetes group. However, for all groups the end‐exercise ADP concentration was well
above the accepted Km value of 30 M for oxidative ATP synthesis. The average PCr
depletion was 54 ± 10, 56 ± 7 and 46 ± 8% for Control, pre‐Diabetes and Diabetes groups,
respectively. For none of the subjects did the end‐exercise pH drop below 6.75. The
average end‐exercise pH was not different for the 3 groups (p = 0.72), which is a
prerequisite for comparing PCr.
PCr and ADP recoveries could be satisfactorily described by mono‐exponential functions
(average R2 values for the mono‐exponential fits were 0.96 ± 0.02 and 0.91 ± 0.06 for PCr
and ADP recovery data, respectively). Figure 4 illustrates both the raw data and mono‐
Figure 3. Typical M. vastus lateralis 31P MR spectra for a normoglycaemic control subject at rest (panel A,
number of scans = 60) and at the end of exercise (panel B, number of scans = 2). Spectra were processed
with 5 Hz line broadening. Pi indicates inorganic phosphate; PDE, phosphodiesters; PCr,
phosphocreatine; and , and indicate the 3 phosphate groups of ATP. For this subject the PCr depletion at the end of exercise (panel B) was 61% and the corresponding end‐exercise pH was 6.96.
In vivo mitochondrial function in type 2 diabetes
89
exponential fits of the PCr and ADP recoveries for one subject. Figure 5 shows the mean
values and the distribution of PCr and Qmax for the Control, pre‐Diabetes and Diabetes
groups. The 31P MRS parameters for mitochondrial function, i.e. PCr, ADP and Qmax, did not
differ between the Control, pre‐Diabetes and Diabetes groups (p = 0.62, p = 0.29 and p =
0.24, respectively; Table 2). VPCr was significantly lower for the Diabetes group compared to
the pre‐Diabetes group as a result of the lower end‐exercise ADP concentration.
All 31P MRS parameters for mitochondrial function correlated significantly with both Wmax
(absolute Pearson’s R between 0.45 and 0.60, p < 0.01) and VO2peak (absolute Pearson’s R
between 0.35 and 0.59, p < 0.05). However, PCr, ADP and Qmax did not correlate with fasting
plasma glucose or HbA1c.
Table 2. 31P MRS parameters during rest, end of exercise and recovery. Data are expressed as means ± SD.
Control pre‐Diabetes Diabetes
Rest [PCr] (mM) 37.3 ± 2.3 37.7 ± 2.6 37.2 ± 4.2
[Pi] (mM) 4.8 ± 0.6 5.0 ± 0.7 4.7 ± 0.8
[ADP] (M) 10.0 ± 0.3 10.4 ± 0.5 10.1 ± 0.4
[PDE] (mM) 6.0 ± 0.8 6.0 ± 0.9 5.5 ± 0.9
pH 7.06 ± 0.01 7.07 ± 0.02 7.06 ± 0.02
End‐ exercise [PCr] (mM) 16.9 ± 3.5 16.7 ± 2.8 20.0 ± 4.5
[Pi] (mM) 23.3 ± 4.9 24.2 ± 3.5 19.8 ± 4.0
[ADP] (M) 68.2 ± 18.2 70.2 ± 11.3 47.0 ± 10.4*†
pH 6.91 ± 0.08 6.92 ± 0.07 6.89 ± 0.12
Recovery PCr (s) 44.5 ± 10.5 41.7 ± 6.2 46.6 ± 16.5
ADP (s) 19.1 ± 3.5 16.9 ± 5.0 20.2 ± 6.0
VPCr (mM/s) 0.47 ± 0.09 0.51 ± 0.08 0.39 ± 0.09†
Qmax (mM/s) 0.69 ± 0.12 0.74 ± 0.11 0.65 ± 0.13
PCr, phosphocreatine; Pi, inorganic phosphate; ADP, adenosine diphosphate; PDE, phosphodiesters; pH,
intracellular muscle pH; PCr, PCr recovery time constant; ADP, ADP recovery time constant; VPCr, initial rate
of PCr recovery; Qmax, maximum rate of oxidative ATP synthesis. * Significantly different from Control
(ANOVA, p < 0.01). † Significantly different from pre‐Diabetes (ANOVA, p < 0.01).
Chapter 4
90
Discussion
In this study, it was shown that in vivo skeletal muscle oxidative capacity, as determined
from post‐exercise PCr recovery kinetics using 31P MRS, does not differ between long‐
standing, insulin‐treated type 2 diabetes patients, subjects with early stage type 2 diabetes,
and healthy, normoglycaemic controls, all matched for age and body composition and with
low habitual physical activity levels. Therefore, our results suggest that skeletal muscle
mitochondrial dysfunction does not necessarily represent either cause or consequence of
type 2 diabetes.
Maximal workload capacity and maximal whole‐body oxygen uptake capacity were
significantly lower in type 2 diabetes patients and correlated negatively with both plasma
glucose and blood HbA1c contents. The 31P MRS recovery parameters, PCr, ADP or Qmax,
correlated significantly with Wmax and VO2peak, but not with plasma glucose or blood HbA1c
levels. Therefore, Wmax and VO2peak seem to represent markers of the disease status,
whereas PCr, ADP and Qmax report only on local muscle mitochondrial function. The Pearson
correlation coefficient for the correlation between Qmax and VO2peak per kg fat free mass
(VO2peak per kg FFM) was 0.47 (R2 = 0.22). Therefore, only 22% of the variance in VO2peak per kg FFM
can be explained by Qmax. Qmax was measured locally in the M. vastus lateralis during a
dynamic single‐leg extension exercise, while whole‐body VO2peak was determined on a cycle
ergometer. The remaining variance in VO2peak per kg FFM could be accounted for by differences
in cardiovascular capacity at peak work rates, which are not expected to play a significant
role during the local exercise in the MR scanner. The apparent discrepancy between the
Figure 4. PCr (A) and ADP (B) recovery curves from the data set that was also used in Figure 3. Mono‐
exponential functions (dark lines) were fit to the actual data (filled circles) obtained every 6 s. The time
constants for PCr and ADP recovery were 41.8 and 19.8 s, respectively.
In vivo mitochondrial function in type 2 diabetes
91
results for Qmax and VO2peak per kg FFM might therefore be explained by differences in cardiac
output. The latter concept seems to be supported by a significant correlation between the
maximal heart rate and VO2peak per kg FFM (Pearson’s R = 0.64, p < 0.01).
In the present study, 31P MRS was applied to assess in vivo skeletal muscle mitochondrial
function from measurements during recovery from exercise. During recovery from exercise,
PCr is resynthesized purely as a consequence of oxidative ATP synthesis [43, 45]. Therefore,
PCr provides information about mitochondrial function. Recently, Schrauwen‐Hinderling et
al. applied the same technique to study mitochondrial function in the M. vastus lateralis in
overweight type 2 diabetes patients and BMI‐matched control subjects [19]. Contrary to
our results, PCr was 45% longer in the type 2 diabetes compared to the control group (i.e.
39.4 ± 17.5 vs 27.0 ± 3.9 s), suggestive of impaired in vivo mitochondrial function in the
type 2 diabetes patients. It should be noted that different exercise protocols were chosen
for the 31P MRS measurements in the 2 studies. We applied an exercise protocol that
progressively increased work to deplete PCr by about 50% without severe acidification. In
the previous study, exercise was performed with a single load, until a steady state was
reached with an average PCr depletion of 28%. The larger PCr depletion in the current
study was accompanied by a slightly larger drop in pH (change in pH was on average 0.16 in
this study vs 0.04 in [19]). Several studies have shown that PCr recovery is slowed in the
presence of intracellular acidosis [30, 49‐50]. However, we carefully matched the end‐
exercise pH for the Control, pre‐Diabetes and Diabetes groups, to allow a direct comparison
Figure 5. PCr (A) and Qmax (B) in Control, pre‐Diabetes and Diabetes groups. Bars indicate the mean values
for the 3 groups and filled circles represent individual data points. PCr and Qmax were not significantly
different between Control, pre‐Diabetes and Diabetes groups (ANOVA: p = 0.62 and p = 0.24,
respectively).
Chapter 4
92
of PCr between groups. As an alternative to PCr, both ADP and Qmax can also be used to
asses in vivo mitochondrial function and these parameters have been shown to be
independent of pH [30, 50‐53]. In accordance with the PCr data matched for end‐exercise
pH, ADP and Qmax also did not differ between groups.
As was suggested in the literature, PCr can be normalized for pH using a correction factor of
‐46 s per pH unit [49]. Recalculating the values from Schrauwen‐Hinderling et al. to a 0.12
lower end‐exercise pH results in PCr’s of about 45 and 32 s for the type 2 diabetes and control group, respectively. This implies that the mitochondrial function in our Control, pre‐
Diabetes and Diabetes groups is similar to that of the type 2 diabetes patients in the study
of Schrauwen‐Hinderling et al., but that the controls in the latter study show a greater
mitochondrial oxidative capacity. For other studies in healthy, elderly subjects previously
reported in the literature [54‐56], average values for PCr (recalculated at an end‐exercise pH of 6.9 if necessary) have been shown to range between 43 and 46 s, which is well in line
with the average PCr in our control group. As such, muscle mitochondrial function in these
overweight normoglycaemic control subjects does not seem to be substantially impaired.
As an alternative to measuring muscle oxidative capacity from dynamic 31P MRS
experiments after exercise, Petersen et al. applied 31P MRS saturation transfer experiments
to measure mitochondrial ATP synthesis rates in resting skeletal muscle of healthy, young,
lean, insulin‐resistant offspring of type 2 diabetes patients and insulin‐sensitive control
subjects matched for age, height, weight and physical activity [17]. ATP synthesis rates
were approximately 30% lower in insulin‐resistant subjects than in controls and it was
concluded that the insulin‐resistant offspring might have an inherited defect in
mitochondrial phosphorylation. However, as commented by Short et al. [24] and
Wagenmakers [25], the lower ATP synthesis rates in insulin‐resistant subjects could actually
be caused by the impaired insulin signaling, jeopardizing insulin‐dependent mitochondrial
processes, rather than the reverse. This view is supported by studies that show that high‐
dose insulin infusions increase mRNA transcript levels of genes involved in mitochondrial
function, mitochondrial protein synthesis and mitochondrial ATP production rates in
healthy people, but not in type 2 diabetes patients [28, 57‐58]. In summary, it seems likely
that the decreased basal ATP synthesis rates in type 2 diabetes patients are a result of the
decreased insulin sensitivity and do not necessarily reflect any intrinsic mitochondrial
defect.
The diabetes patients in the current study were on exogenous insulin treatment for more
than 5 years and continued their medication during the study. As blood glucose and insulin
levels can affect measurements of mitochondrial function [27‐28], the higher plasma insulin
levels in the Diabetes group as a result of exogenous insulin treatment might be a
confounding factor in this study. Despite the insulin treatment, our diabetes patients were
still hyperglycaemic (fasting plasma glucose concentration: 11.0 ± 2.5 mmol/l). Therefore,
In vivo mitochondrial function in type 2 diabetes
93
hyperglycaemia might also have affected the measurement of in vivo muscle mitochondrial
function in the diabetes patients.
Rabøl et al. recently reviewed the experimental data on mitochondrial dysfunction in type 2
diabetes and concluded that evidence of an intrinsic defect in the mitochondria of type 2
diabetes patients is far from convincing [23]. Considering that type 2 diabetes patients are
generally physically inactive, the impairments in oxidative metabolism in type 2 diabetes
patients might simply be attributed to their sedentary lifestyle. In this regard, it is
important to note that in most studies physical activity has not been (strictly) controlled
for. For studies in which physical activity was taken into account, the results suggest that
the abnormalities in oxidative metabolism in type 2 diabetes patients can at least partly be
attributed to physical inactivity [14, 28, 59‐60]. In accordance, recent data from respiration
measurements on permeabilized muscle fibers show that when O2 flux is being normalized
for mitochondrial DNA content or citrate synthase activity, no differences in mitochondrial
respiration rate are observed between type 2 diabetes patients and healthy controls [21].
These results imply that type 2 diabetes patients have normal intrinsic mitochondrial
function, but an impaired oxidative capacity due to a reduced mitochondrial content, most
likely secondary to lower habitual physical activity levels [21‐22]. More recently, Turner et
al. examined markers of muscle mitochondrial fatty acid oxidative capacity in rodent
models of lipid‐induced insulin resistance [61]. Surprisingly, fatty acid oxidative capacity
and protein expression of peroxisome proliferator‐activated receptor‐ coactivator 1 (PGC‐1) and mitochondrial respiratory chain subunits appeared to be upregulated. As
such, the authors concluded that, at least in these rodent models, high lipid availability
does not lead to intramuscular lipid accumulation and insulin resistance by decreasing
muscle mitochondrial fatty acid oxidative capacity. In accordance, in the present study we
observed no differences in in vivo muscle mitochondrial function between long‐standing
type 2 diabetes patients, subjects with early stage type 2 diabetes and sedentary,
normoglycaemic controls.
In conclusion, subjects with early stage type 2 diabetes as well as long‐standing, insulin‐
treated type 2 diabetes patients do not show signs of in vivo skeletal muscle mitochondrial
dysfunction. The latter implies that mitochondrial dysfunction does not necessarily
represent either cause or consequence of insulin resistance and/or type 2 diabetes.
Impairments in oxidative metabolism in type 2 diabetes patients observed in previous
studies are likely to be secondary to a less active lifestyle and/or impaired insulin signaling.
Acknowledgements
We are very grateful to Larry de Graaf for his technical assistance with the MR scanner.
Chapter 4
94
References 1. Pan, D.A., S. Lillioja, A.D. Kriketos, M.R. Milner, L.A. Baur, C. Bogardus, A.B. Jenkins, and L.H. Storlien.
Skeletal muscle triglyceride levels are inversely related to insulin action. Diabetes, 1997. 46(6): 983‐8. 2. Krssak, M., K. Falk Petersen, A. Dresner, L. DiPietro, S.M. Vogel, D.L. Rothman, M. Roden, and G.I.
Shulman. Intramyocellular lipid concentrations are correlated with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabetologia, 1999. 42(1): 113‐6.
3. van Loon, L.J., R. Koopman, R. Manders, W. van der Weegen, G.P. van Kranenburg, and H.A. Keizer. Intramyocellular lipid content in type 2 diabetes patients compared with overweight sedentary men and highly trained endurance athletes. Am J Physiol Endocrinol Metab, 2004. 287(3): E558‐65.
4. van Loon, L.J. and B.H. Goodpaster. Increased intramuscular lipid storage in the insulin‐resistant and endurance‐trained state. Pflugers Arch, 2006. 451(5): 606‐16.
5. Morino, K., K.F. Petersen, and G.I. Shulman. Molecular mechanisms of insulin resistance in humans and their potential links with mitochondrial dysfunction. Diabetes, 2006. 55 Suppl 2: S9‐S15.
6. Simoneau, J.A. and D.E. Kelley. Altered glycolytic and oxidative capacities of skeletal muscle contribute to insulin resistance in NIDDM. J Appl Physiol, 1997. 83(1): 166‐71.
7. He, J., S. Watkins, and D.E. Kelley. Skeletal muscle lipid content and oxidative enzyme activity in relation to muscle fiber type in type 2 diabetes and obesity. Diabetes, 2001. 50(4): 817‐23.
8. Kelley, D.E., J. He, E.V. Menshikova, and V.B. Ritov. Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes, 2002. 51(10): 2944‐50.
9. Ritov, V.B., E.V. Menshikova, J. He, R.E. Ferrell, B.H. Goodpaster, and D.E. Kelley. Deficiency of subsarcolemmal mitochondria in obesity and type 2 diabetes. Diabetes, 2005. 54(1): 8‐14.
10. Heilbronn, L.K., S.K. Gan, N. Turner, L.V. Campbell, and D.J. Chisholm. Markers of mitochondrial biogenesis and metabolism are lower in overweight and obese insulin‐resistant subjects. J Clin Endocrinol Metab, 2007. 92(4): 1467‐73.
11. Mootha, V.K., C.M. Lindgren, K.F. Eriksson, A. Subramanian, S. Sihag, J. Lehar, P. Puigserver, E. Carlsson, M. Ridderstrale, E. Laurila, N. Houstis, M.J. Daly, N. Patterson, J.P. Mesirov, T.R. Golub, P. Tamayo, B. Spiegelman, E.S. Lander, J.N. Hirschhorn, D. Altshuler, and L.C. Groop. PGC‐1alpha‐responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet, 2003. 34(3): 267‐73.
12. Patti, M.E., A.J. Butte, S. Crunkhorn, K. Cusi, R. Berria, S. Kashyap, Y. Miyazaki, I. Kohane, M. Costello, R. Saccone, E.J. Landaker, A.B. Goldfine, E. Mun, R. DeFronzo, J. Finlayson, C.R. Kahn, and L.J. Mandarino. Coordinated reduction of genes of oxidative metabolism in humans with insulin resistance and diabetes: Potential role of PGC1 and NRF1. Proc Natl Acad Sci U S A, 2003. 100(14): 8466‐71.
13. Hojlund, K., K. Wrzesinski, P.M. Larsen, S.J. Fey, P. Roepstorff, A. Handberg, F. Dela, J. Vinten, J.G. McCormack, C. Reynet, and H. Beck‐Nielsen. Proteome analysis reveals phosphorylation of ATP synthase beta ‐subunit in human skeletal muscle and proteins with potential roles in type 2 diabetes. J Biol Chem, 2003. 278(12): 10436‐42.
14. Morino, K., K.F. Petersen, S. Dufour, D. Befroy, J. Frattini, N. Shatzkes, S. Neschen, M.F. White, S. Bilz, S. Sono, M. Pypaert, and G.I. Shulman. Reduced mitochondrial density and increased IRS‐1 serine phosphorylation in muscle of insulin‐resistant offspring of type 2 diabetic parents. J Clin Invest, 2005. 115(12): 3587‐93.
15. Mogensen, M., K. Sahlin, M. Fernstrom, D. Glintborg, B.F. Vind, H. Beck‐Nielsen, and K. Hojlund. Mitochondrial respiration is decreased in skeletal muscle of patients with type 2 diabetes. Diabetes, 2007. 56(6): 1592‐9.
16. Petersen, K.F., D. Befroy, S. Dufour, J. Dziura, C. Ariyan, D.L. Rothman, L. DiPietro, G.W. Cline, and G.I. Shulman. Mitochondrial dysfunction in the elderly: possible role in insulin resistance. Science, 2003. 300(5622): 1140‐2.
17. Petersen, K.F., S. Dufour, D. Befroy, R. Garcia, and G.I. Shulman. Impaired mitochondrial activity in the insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(7): 664‐71.
18. Szendroedi, J., A.I. Schmid, M. Chmelik, C. Toth, A. Brehm, M. Krssak, P. Nowotny, M. Wolzt, W. Waldhausl, and M. Roden. Muscle mitochondrial ATP synthesis and glucose transport/phosphorylation in type 2 diabetes. PLoS Med, 2007. 4(5): e154.
19. Schrauwen‐Hinderling, V.B., M.E. Kooi, M.K. Hesselink, J.A. Jeneson, W.H. Backes, C.J. van Echteld, J.M. van Engelshoven, M. Mensink, and P. Schrauwen. Impaired in vivo mitochondrial function but similar intramyocellular lipid content in patients with type 2 diabetes mellitus and BMI‐matched control subjects. Diabetologia, 2007. 50(1): 113‐20.
20. Sirikul, B., B.A. Gower, G.R. Hunter, D.E. Larson‐Meyer, and B.R. Newcomer. Relationship between insulin sensitivity and in vivo mitochondrial function in skeletal muscle. Am J Physiol Endocrinol Metab, 2006. 291(4): E724‐8.
In vivo mitochondrial function in type 2 diabetes
95
21. Boushel, R., E. Gnaiger, P. Schjerling, M. Skovbro, R. Kraunsoe, and F. Dela. Patients with type 2 diabetes have normal mitochondrial function in skeletal muscle. Diabetologia, 2007. 50(4): 790‐6.
22. Hawley, J.A. and S.J. Lessard. Mitochondrial function: use it or lose it. Diabetologia, 2007. 50(4): 699‐702. 23. Rabol, R., R. Boushel, and F. Dela. Mitochondrial oxidative function and type 2 diabetes. Appl Physiol
Nutr Metab, 2006. 31(6): 675‐83. 24. Short, K.R., K.S. Nair, and C.S. Stump. Impaired mitochondrial activity and insulin‐resistant offspring of
patients with type 2 diabetes. N Engl J Med, 2004. 350(23): 2419‐21; author reply 2419‐21. 25. Wagenmakers, A.J. Insulin resistance in the offspring of parents with type 2 diabetes. PLoS Med, 2005.
2(9): e289. 26. Brehm, A., M. Krssak, A.I. Schmid, P. Nowotny, W. Waldhausl, and M. Roden. Increased lipid availability
impairs insulin‐stimulated ATP synthesis in human skeletal muscle. Diabetes, 2006. 55(1): 136‐40. 27. Sreekumar, R., P. Halvatsiotis, J.C. Schimke, and K.S. Nair. Gene expression profile in skeletal muscle of
type 2 diabetes and the effect of insulin treatment. Diabetes, 2002. 51(6): 1913‐20. 28. Asmann, Y.W., C.S. Stump, K.R. Short, J.M. Coenen‐Schimke, Z. Guo, M.L. Bigelow, and K.S. Nair. Skeletal
muscle mitochondrial functions, mitochondrial DNA copy numbers, and gene transcript profiles in type 2 diabetic and nondiabetic subjects at equal levels of low or high insulin and euglycemia. Diabetes, 2006. 55(12): 3309‐19.
29. Radda, G.K., P.J. Bore, D.G. Gadian, B.D. Ross, P. Styles, D.J. Taylor, and J. Morgan‐Hughes. 31P NMR examination of two patients with NADH‐CoQ reductase deficiency. Nature, 1982. 295(5850): 608‐9.
30. Arnold, D.L., P.M. Matthews, and G.K. Radda. Metabolic recovery after exercise and the assessment of mitochondrial function in vivo in human skeletal muscle by means of 31P NMR. Magn Reson Med, 1984. 1(3): 307‐15.
31. Chance, B., J. Im, S. Nioka, and M. Kushmerick. Skeletal muscle energetics with PNMR: personal views and historic perspectives. NMR Biomed, 2006. 19(7): 904‐26.
32. World Health Organization and International Diabetes Federation, Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia: report of a WHO/IDF consultation. 2006, Geneva, Switzerland: WHO Press.
33. Praet, S.F., H.M. Feyter, R.A. Jonkers, K. Nicolay, C. Pul, H. Kuipers, L.J. Loon, and J.J. Prompers. (31)P MR spectroscopy and in vitro markers of oxidative capacity in type 2 diabetes patients. Magma, 2006. 19(6): 321‐31.
34. De Feyter, H.M., S.F. Praet, N.M. van den Broek, H. Kuipers, C.D. Stehouwer, K. Nicolay, J.J. Prompers, and L.J. van Loon. Exercise training improves glycemic control in long‐standing insulin‐treated type 2 diabetic patients. Diabetes Care, 2007. 30(10): 2511‐3.
35. Montoye, H.J., H.C.G. Kemper, W.H.M. Saris, and R.A. Washburn, Measuring physical activity and energy expenditure. 1996, Champaign, Illinois: Human Kinetics.
36. Zhang, Y.Y., M.C. Johnson, 2nd, N. Chow, and K. Wasserman. Effect of exercise testing protocol on parameters of aerobic function. Med Sci Sports Exerc, 1991. 23(5): 625‐30.
37. Weber, C.T., J.S. Janicki, and P.A. McElroy, Cardiopulmonary exercise testing (CPX) testing. 1986, Philadelphia: W. B. Saunders.
38. Gellish, R.L., B.R. Goslin, R.E. Olson, A. McDonald, G.D. Russi, and V.K. Moudgil. Longitudinal modeling of the relationship between age and maximal heart rate. Med Sci Sports Exerc, 2007. 39(5): 822‐9.
39. Matthews, D.R., J.P. Hosker, A.S. Rudenski, B.A. Naylor, D.F. Treacher, and R.C. Turner. Homeostasis model assessment: insulin resistance and beta‐cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 1985. 28(7): 412‐9.
40. Vanhamme, L., A. van den Boogaart, and S. Van Huffel. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson, 1997. 129(1): 35‐43.
41. Boesch, C., J. Slotboom, H. Hoppeler, and R. Kreis. In vivo determination of intra‐myocellular lipids in human muscle by means of localized 1H‐MR‐spectroscopy. Magn Reson Med, 1997. 37(4): 484‐93.
42. Taylor, D.J., P. Styles, P.M. Matthews, D.A. Arnold, D.G. Gadian, P. Bore, and G.K. Radda. Energetics of human muscle: exercise‐induced ATP depletion. Magn Reson Med, 1986. 3(1): 44‐54.
43. Taylor, D.J., P.J. Bore, P. Styles, D.G. Gadian, and G.K. Radda. Bioenergetics of intact human muscle. A 31P nuclear magnetic resonance study. Mol Biol Med, 1983. 1(1): 77‐94.
44. Lawson, J.W. and R.L. Veech. Effects of pH and free Mg2+ on the Keq of the creatine kinase reaction and other phosphate hydrolyses and phosphate transfer reactions. J Biol Chem, 1979. 254(14): 6528‐37.
45. Quistorff, B., L. Johansen, and K. Sahlin. Absence of phosphocreatine resynthesis in human calf muscle during ischaemic recovery. Biochem J, 1993. 291 ( Pt 3): 681‐6.
46. Kemp, G.J. and G.K. Radda. Quantitative interpretation of bioenergetic data from 31P and 1H magnetic resonance spectroscopic studies of skeletal muscle: an analytical review. Magn Reson Q, 1994. 10(1): 43‐63.
Chapter 4
96
47. Kemp, G.J., C.H. Thompson, P.R. Barnes, and G.K. Radda. Comparisons of ATP turnover in human muscle during ischemic and aerobic exercise using 31P magnetic resonance spectroscopy. Magn Reson Med, 1994. 31(3): 248‐58.
48. Chance, B., J.S. Leigh, Jr., B.J. Clark, J. Maris, J. Kent, S. Nioka, and D. Smith. Control of oxidative metabolism and oxygen delivery in human skeletal muscle: a steady‐state analysis of the work/energy cost transfer function. Proc Natl Acad Sci U S A, 1985. 82(24): 8384‐8.
49. Iotti, S., R. Lodi, C. Frassineti, P. Zaniol, and B. Barbiroli. In vivo assessment of mitochondrial functionality in human gastrocnemius muscle by 31P MRS. The role of pH in the evaluation of phosphocreatine and inorganic phosphate recoveries from exercise. NMR Biomed, 1993. 6(4): 248‐53.
50. van den Broek, N.M., H.M. De Feyter, L. de Graaf, K. Nicolay, and J.J. Prompers. Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics in muscle after exercise are due to differences in proton efflux rates. Am J Physiol Cell Physiol, 2007. 293(1): C228‐37.
51. Argov, Z., N. De Stefano, and D.L. Arnold. ADP recovery after a brief ischemic exercise in normal and diseased human muscle‐‐a 31P MRS study. NMR Biomed, 1996. 9(4): 165‐72.
52. Lodi, R., G.J. Kemp, S. Iotti, G.K. Radda, and B. Barbiroli. Influence of cytosolic pH on in vivo assessment of human muscle mitochondrial respiration by phosphorus magnetic resonance spectroscopy. Magma, 1997. 5(2): 165‐71.
53. Roussel, M., D. Bendahan, J.P. Mattei, Y. Le Fur, and P.J. Cozzone. 31P magnetic resonance spectroscopy study of phosphocreatine recovery kinetics in skeletal muscle: the issue of intersubject variability. Biochim Biophys Acta, 2000. 1457(1‐2): 18‐26.
54. Kent‐Braun, J.A. and A.V. Ng. Skeletal muscle oxidative capacity in young and older women and men. J Appl Physiol, 2000. 89(3): 1072‐8.
55. Conley, K.E., S.A. Jubrias, and P.C. Esselman. Oxidative capacity and ageing in human muscle. J Physiol, 2000. 526(Pt 1): 203‐10.
56. Scheuermann‐Freestone, M., P.L. Madsen, D. Manners, A.M. Blamire, R.E. Buckingham, P. Styles, G.K. Radda, S. Neubauer, and K. Clarke. Abnormal cardiac and skeletal muscle energy metabolism in patients with type 2 diabetes. Circulation, 2003. 107(24): 3040‐6.
57. Stump, C.S., K.R. Short, M.L. Bigelow, J.M. Schimke, and K.S. Nair. Effect of insulin on human skeletal muscle mitochondrial ATP production, protein synthesis, and mRNA transcripts. Proc Natl Acad Sci U S A, 2003. 100(13): 7996‐8001.
58. Petersen, K.F., S. Dufour, and G.I. Shulman. Decreased insulin‐stimulated ATP synthesis and phosphate transport in muscle of insulin‐resistant offspring of type 2 diabetic parents. PLoS Med, 2005. 2(9): e233.
59. Ortenblad, N., M. Mogensen, I. Petersen, K. Hojlund, K. Levin, K. Sahlin, H. Beck‐Nielsen, and M. Gaster. Reduced insulin‐mediated citrate synthase activity in cultured skeletal muscle cells from patients with type 2 diabetes: evidence for an intrinsic oxidative enzyme defect. Biochim Biophys Acta, 2005. 1741(1‐2): 206‐14.
60. Timmons, J.A., J. Norrbom, C. Scheele, H. Thonberg, C. Wahlestedt, and P. Tesch. Expression profiling following local muscle inactivity in humans provides new perspective on diabetes‐related genes. Genomics, 2006. 87(1): 165‐72.
61. Turner, N., C.R. Bruce, S.M. Beale, K.L. Hoehn, T. So, M.S. Rolph, and G.J. Cooney. Excess lipid availability increases mitochondrial fatty acid oxidative capacity in muscle: evidence against a role for reduced fatty acid oxidation in lipid‐induced insulin resistance in rodents. Diabetes, 2007. 56(8): 2085‐92.
97
98
Adapted from
N. M. A. van den Broek, J. Ciapaite, H. M. M. L De Feyter, S. M. Houten, R. J. A. Wanders,
J. A. L. Jeneson, K. Nicolay, J. J. Prompers. Increased mitochondrial content rescues in vivo
muscle oxidative capacity in long‐term high‐fat diet fed rats.
FASEB J, 2010. 24(5):1354‐64.
Chapter Increased mitochondrial content rescues
in vivo muscle oxidative capacity
in long‐term high‐fat diet fed rats
Chapter 5
100
Abstract
Mitochondria are thought to play a crucial role in the etiology of muscle insulin resistance
(IR). The aim of this study was to gain more insight into the timing and nature of
mitochondrial adaptations during the development of high‐fat diet (HFD)‐induced IR. Adult
Wistar rats were fed HFD or normal chow for 2.5 and 25 weeks. Intramyocellular lipids
(IMCL) were quantified in vivo using 1H magnetic resonance spectroscopy (MRS). Muscle
oxidative capacity was assessed in vivo using 31P MRS and in vitro by measuring
mitochondrial DNA copy number and oxygen consumption in isolated mitochondria. MRS in
tibialis anterior muscle revealed 3.3‐fold higher IMCL content and 1.2‐fold increased
oxidative capacity after 2.5 weeks of HFD feeding. The latter result could be fully accounted
for by increased mitochondrial content. After 25 weeks of HFD, maximal ADP‐stimulated
oxygen consumption in isolated mitochondria oxidizing pyruvate plus malate remained
unaffected, while IMCL and mitochondrial content had further increased compared to
controls (5.1‐fold and 1.4‐fold, respectively). Interestingly, in vivo oxidative capacity at this
time point was identical to controls. These results show that skeletal muscle in HFD‐
induced IR accompanied by IMCL accumulation requires a progressively larger
mitochondrial pool size to maintain normal oxidative capacity in vivo.
Mitochondrial function and insulin resistance
101
Introduction
High caloric diets and an inactive lifestyle contribute to the worldwide increase in
prevalence of obesity, one of the main risk factors for the development of insulin resistance
(IR) and type 2 diabetes [1]. In obesity, non‐adipose tissues are exposed to abnormally high
levels of fat‐derived substrates, leading to accumulation of intracellular lipids [2]. A
negative correlation between the content of intramyocellular lipids (IMCL) and whole‐body
insulin sensitivity in obese and diabetic individuals has been demonstrated in several
studies [3‐5], indicating that lipids play a key role in the development of IR.
Based on the in vivo observation that basal ATP synthesis rates in skeletal muscle are lower
in IR subjects and their offspring compared to healthy controls [6‐7], it has been
hypothesized that the accumulation of IMCL is a consequence of impaired mitochondrial
function. Presumably, this leads to accumulation of long‐chain acyl‐CoA esters,
diacylglycerols and ceramides, which interfere with the insulin signaling cascade [8].
The nature of mitochondrial dysfunction and the cause and effect relationship between
mitochondrial dysfunction and the development of IR remain elusive. Several human
studies have provided evidence for dysfunctional muscle mitochondria in insulin resistant
states by showing down‐regulation of genes encoding mitochondrial enzymes [9‐10],
decreased mitochondrial content and lower mitochondrial respiratory chain activity [11],
while other human studies did not reveal mitochondrial dysfunction in IR muscle [12‐13].
However, the observation that improved glycemic control achieved by insulin treatment
results in normalized expression of at least some genes involved in mitochondrial
metabolism [14] suggests that the different outcome of the aforementioned studies may
be due to a varying level of glycemic control in the study‐subjects and implies that the
down‐regulation of mitochondrial genes, which has been associated with mitochondrial
dysfunction, is not the cause but rather the effect of IR.
Similarly, reports on mitochondrial adaptations in skeletal muscle during the development
of IR in rodent models of diet‐induced obesity and IR have been inconclusive and have
shown altered mitochondrial morphology and impaired function [15‐16] as well as
improved mitochondrial capacity to oxidize fat‐derived substrates and increased
mitochondrial number [17‐18].
These contrasting findings, at least in rodent models, may be partially explained by
differences in study design (e.g. dietary lipid content and composition, duration of the
feeding). Indeed, some reports suggest time‐dependence of mitochondrial adaptations
during the development of IR [19], reinforcing the need for longitudinal studies. Moreover,
few attempts have been made to evaluate to what extent changes in mitochondrial
number and function observed in vitro translate into changes in oxidative capacity in vivo
and, consequently, how significant these changes are for whole‐body glucose homeostasis
in vivo. In this study we therefore aimed to obtain a complete picture of changes in skeletal
Chapter 5
102
muscle mitochondrial number and function in early and later stages of high‐fat diet (HFD)‐
induced IR (after 2.5 and 25 weeks of diet, respectively) by combining in vitro and in vivo
methodologies.
Materials and Methods
Animals
Adult male Wistar rats (366 ± 16 g, 14 weeks old, n = 38, Charles River Laboratories, The
Netherlands) were housed in pairs at 20°C and 50% humidity, on a 12‐h light‐dark cycle.
Rats were given ad libitum access to water and to either a high‐fat diet (HFD, 45.7% calories
from fat (predominantly lard), 35.5% calories from carbohydrate, 18.3% calories from
protein (diet no. 58V8, Testdiet, Richmond, IN, USA)) (n = 19) or normal chow (NC, 9.2%
calories from fat, 67.2% calories from carbohydrate, 23.6% calories from protein (R/M‐H
diet, Ssniff Spezialdiäten GmbH, Soest, Germany)) (n = 19) for 25 weeks. After 2.5 wk n = 9
rats from each diet group were killed and tissues were collected for in vitro analyses. The
remaining rats (n = 10 per diet group) were used for in vivo measurements after 2.5 and 25
wk of diet and in vitro measurements after 25 wk of diet. Rats were killed by incising the
inferior vena cava under anesthesia. Animal handling conformed to the rules of the Animal
Ethics Committee of Maastricht University.
Plasma parameters
Oral glucose tolerance tests (OGTT) were performed after 15 days and 25 wk of diet, two
days before the in vivo measurements. After an overnight fast (20h) rats received an oral
glucose bolus of 1g/kg body weight. Blood samples were taken without anesthesia from the
vena saphena at 0, 15, 30, 60, 90 and 120 min after the bolus. Blood glucose concentration
was determined using an automatic glucometer (FreeStyle, Abbott, IL, USA). Plasma insulin
concentration was determined using an ultrasensitive rat insulin ELISA kit (Mercodia,
Uppsala, Sweden).
To quantify whole‐body insulin resistance, the area under the glucose curve (AUCg)
obtained during OGTT was multiplied by the area under the insulin curve (AUCi) measured
during the OGTT [20].
The concentration of plasma free fatty acids (FFA) and triacylglycerols (TAG) was
determined using NEFA C kit (Wako Chemicals GmbH, Neuss, Germany) and serum
triglyceride determination kit (Sigma‐Aldrich, Zwijndrecht, The Netherlands), respectively.
Mitochondrial function and insulin resistance
103
Determination of muscle acylcarnitine content
The content of acylcarnitines was determined in freeze‐dried lateral gastrocnemius muscle
as described previously [21].
Magnetic resonance spectroscopy
All magnetic resonance spectroscopy (MRS) measurements were performed on a 6.3 Tesla
horizontal Bruker MR system (Bruker, Ettlingen, Germany). Animals were anaesthetized
using isoflurane (Forene®) (1‐2%) with medical air (0.6 L/min). Body temperature was
maintained at 36 ± 1C using heating pads. Respiration was monitored using a pressure
sensor registering thorax movement (Rapid Biomedical, Rimpar, Germany).
IMCL content in TA after 2.5 weeks and 25 weeks of diet was determined in vivo using
single‐voxel localized 1H MRS as described previously [22]. Voxels of 222 mm3 were
measured in white (ventral) and red (dorsal) TA using an ellipsoid 1H surface coil (18/22
mm). The 1H MR spectra were acquired using the LASER sequence [23] (repetition time TR =
1 s, echo time TE = 16 ms, SWAMP water suppression [24], 512 averages).
In vivo muscle oxidative capacity was measured in TA after 2.5 weeks and 25 weeks of diet
using 31P MRS. A circular 1H surface coil (40 mm) and an ellipsoid 31P MRS surface coil
(10/18 mm) were positioned over the TA as described previously [25] and 31P spectra were
acquired by applying an adiabatic excitation pulse with a flip angle of 90. Fully relaxed (TR = 20 s, 32 averages) and partially saturated spectra (TR = 5 s, 128 averages) were measured
at rest, followed by the acquisition of a time series of spectra (TR = 5 s, 4 averages) before,
during and after electrical stimulation of TA via subcutaneously implanted electrodes
positioned along the distal nerve trajectory of the N. Peroneus Communis [25]. A time
series consisted of 3 min rest, 2 min of electrical stimulation and 10 min of recovery.
Stimulation pulse length was 100 ms, frequency was 80 Hz, stimulation voltage was about 3
V and pulses were applied 1/s. Three to 4 time series were measured for each rat during a
single session.
31P and 1H MR spectra were fitted in the time domain using a nonlinear least squares
algorithm (AMARES) in the jMRUI software package [26] as described previously [22, 25].
IMCL was expressed as a percentage of the water signal measured in the same voxel. PCr
resynthesis data were fitted to a mono‐exponential function using Matlab (version 7.04,
Mathworks, Natick, MA, USA) yielding a rate constant kPCr. For each animal, the results of
two time series with similar end‐exercise pH values were averaged.
Chapter 5
104
Isolation of mitochondria
Mitochondria were isolated from one whole TA muscle through a differential centrifugation
procedure [27]. Briefly, TA muscle was excised, washed in ice cold 0.9% KCl, freed of
connective and adipose tissue, weighed and minced with scissors in ice cold medium A (5
ml for 1 g tissue) containing 150 mM sucrose, 75 mM KCl, 50 mM MOPS, 1 mM KH2PO4, 5
mM MgCl2, 1 mM EGTA, 0.4 mg/ml bacterial proteinase type XXIV, pH 7.4. Next, 20 ml of
medium B containing 250 mM sucrose, 0.1 mM EGTA and 20 mM MOPS, 2 mg/ml BSA, pH
7.4 was added and the mixture was homogenized using Potter‐Elvehjem homogenizer. The
homogenate was centrifuged at 800 g for 10 min, 4C. The resulting supernatant was centrifuged at 10000 g for 10 min, 4C. The pellet was resuspended in 15 ml of fresh ice
cold medium B and centrifuged at 10000 g for 10 min, 4C. Mitochondrial pellet was
resuspended in 100 l of medium B. Protein content was determined using BCA protein
assay kit (Pierce, Thermo Fisher Scientific Inc., Rockford, IL, USA).
Measurement of oxygen consumption
Oxygen consumption rate was measured at 37C using a two‐channel high‐resolution Oroboros oxygraph‐2 k (Oroboros, Innsbruck, Austria). Mitochondria (0.15 mg/ml) were
incubated in 1 ml of assay medium containing 110 mM KCl, 20 mM Tris, 2.3 mM MgCl2, 5
mM KH2PO4, 1 mg/ml BSA, pH 7.3, and supplemented with either 5 mM pyruvate plus 5
mM malate or 25 M palmitoyl‐L‐carnitine plus 2.5 mM malate as the oxidizable substrate.
An ADP‐regenerating system consisting of excess hexokinase (4.8 U/ml) and glucose (12.5
mM) was used to maintain steady‐state oxygen consumption rates. 1 mM of ATP was
added to initiate state 3 respiration. Oxygen consumption in resting state (state 4) and
uncoupled state (state U) was determined after addition of 1.25 M carboxyatractyloside
and 1 M carbonyl cyanide 3‐chlorophenyl hydrazone, respectively. The quality of
mitochondrial preparation was assessed by determining the effect of 20 M cytochrome c
and 1 M coenzyme Q10 on oxygen consumption rate in state 3 and state 4. The signals
from the oxygen electrode were recorded at 0.5 s intervals. Data acquisition and analysis
was performed using Oxygraph‐2k‐DatLab software version 4.2 (Oroboros, Innsbruck,
Austria).
Determination of the relative mitochondrial copy number
Genomic DNA was isolated from a 25 mg transversal slice of mid‐belly TA using GenElute™
Mammalian Genomic DNA Miniprep Kit (Sigma‐Aldrich, Zwijndrecht, The Netherlands).
Mitochondrial DNA (mtDNA) content relative to peroxisome proliferator‐activated
receptor‐γ coactivator 1 (PGC‐1) gene was measured using real‐time PCR as described in
[28]. Primers for mtDNA were: forward primer ‐5′‐ACACCAAAAGGACGAACCTG‐3′, reverse
Mitochondrial function and insulin resistance
105
primer ‐5′‐ATGGGGAAGAAGCCCTAGAA‐3′, and for PGC‐1: forward primer ‐5′‐
ATGAATGCAGCGGTCTTAGC‐3′, reverse primer ‐5′‐AACAATGGCAGGGTTTGTTC‐3′. The
relative mtDNA copy number was calculated using Ct method as described in [29].
Determination of enzyme activities
Homogenates of transversal sections of mid‐belly TA muscle were prepared as described in
[25]. Succinate dehydrogenase (SDH) activity was determined as described in [30].
Cytochrome c oxidase (CCO) activity was measured using a method described by
Cooperstein and Lazarow [31]. Citrate synthase (CS) activity was measured as described in
[32]. Very long chain acyl‐CoA dehydrogenase (VLCAD) activity was measured as described
in [33].
Immunoprecipitation and Western blotting
Frozen transversal sections of mid‐belly TA (100‐150 mg) were homogenized in 10 volumes
of cold RIPA buffer (1% Nonidet P40 substitute, 0.5% Na deoxycholate, 0.1% Na dodecyl
sulfate, 150 mM NaCl, 50 mM Tris, pH 8) supplemented with protease inhibitor cocktail
(dilution 1:200) (Sigma‐Aldrich, Zwijndrecht, The Netherlands). Homogenates were
solubilized for 2 h at 4C and centrifuged at 12000 g for 15 min at 4C. Supernatants (10 g of total protein) were resolved by SDS‐PAGE. Furthermore, 1 ml of diluted supernatant (1
mg/ml of total protein) was immunoprecipitated with 4 µg of anti‐IRS1 antibody (Sigma‐
Aldrich, Zwijndrecht, The Netherlands) overnight at 4C, followed by 2 hour incubation with protein A‐agarose beads (Sigma‐Aldrich, Zwijndrecht, The Netherlands). Washed beads
were mixed with Laemmli buffer, heated at 98C for 5 min and centrifuged at 14000 g for 1
min at 4C and the supernatant was resolved by SDS‐PAGE.
After electrophoresis proteins were transferred to polyvinylidene difluoride membranes
(Millipore). After blocking membranes were incubated with anti‐UCP3 antibody (Sigma‐
Aldrich, Zwijndrecht, The Netherlands), anti‐PGC1 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), anti‐IRS1 or anti‐phospho‐IRS1 (pIRS1, pSer312, Sigma‐Aldrich, Zwijndrecht,
The Netherlands) overnight at 4C. After washing with PBS containing 0.1% Tween, membranes were treated with horse‐radish peroxidase conjugated goat anti‐rabbit IgG
(Pierce, Thermo Fisher Scientific Inc., Rockford, IL, USA) for 1 h at room temperature. The
immunocomplexes were detected using SuperSignal West Dura Extended Duration
Substrate (Pierce, Thermo Fisher Scientific Inc., Rockford, IL, USA) and quantified using
Quantity One 1‐D analysis software version 4.4.0 (Bio‐Rad Laboratories Inc., Hercules, CA,
USA). The level of IRS1 phosphorylation was expressed as the fraction of pIRS1 in the total
pool of IRS1.
Chapter 5
106
Statistical analysis
Data are presented as means ± SD. The listed n values represent the number of animals
used for a particular experiment. The statistical significance of the diet and age effects was
assessed using either ANOVA for repeated measures with one between‐subjects factor
(diet) and one within‐subjects factor (age) (body weight, energy intake, plasma glucose
concentrations and MRS data) or two‐way ANOVA with two between‐subjects factor (diet
and age) (the remaining data). Only if the interaction term between the factors was found
to be significant, the effect of each factor was analyzed separately using Student’s t‐test
(paired or unpaired, depending on the type of the measurement). All analyses were
performed with SPSS 16.0 (SPSS Inc., Chicago, IL, USA). Level of statistical significance was
set at p < 0.05.
Figure 1. The effect of high‐fat diet on glucose homeostasis. (A) Glucose concentrations during OGTT (n
= 9 and n = 10, for 2.5 and 25 wk data, respectively). Symbols denoting statistically significant
differences were excluded for simplicity. (B) Insulin concentrations during OGTT (n = 5‐10 per time
point). Symbols denoting statistically significant differences were excluded for simplicity. (C)
Development of insulin resistance in HFD‐fed rats. Insulin resistance was quantified by multiplying AUCg
with AUCi (Table 1). *p < 0.05 vs. NC‐fed rats of the same age; #p < 0.05 vs. the same diet group after 2.5
wk of diet. (D) Phosphorylation of IRS1 at ser307 in TA. The level of IRS1 phosphorylation was expressed
as the fraction of phosphorylated IRS1 in the total pool of IRS1 (n = 6 per diet group). There was a
significant effect of HFD (p < 0.001) and age (p < 0.001) on IRS1 phosphorylation level. Data are means ±
SD.
Mitochondrial function and insulin resistance
107
Results
Animal model
Animal characteristics after 2.5 and 25 wk of diet are summarized in Table 1. Body weight
and energy intake of HFD‐fed rats was significantly higher (p < 0.001) compared to NC‐fed
rats at both time points. Fasting blood glucose concentration was not affected by HFD, but
significantly increased with age (p < 0.01). Fasting insulin concentration was significantly
higher (p < 0.001) in HFD‐fed animals compared to controls at both time points. OGTT
revealed progressive glucose intolerance in HFD‐fed animals as indicated by increased AUCg
(p < 0.001) (Figure 1A, Table 1) and AUCi (p < 0.01) (Figure 1B, Table 1) compared to
controls. In addition, plasma FFA, TAG and glycerol concentrations increased significantly (p
< 0.001) in response to HFD feeding.
Table 1. Animal characteristics.
2.5 wk 25 wk
NC HFD NC HFD
Body weight (g) 385 ± 18 426 ± 20
#505 ± 30
$655 ± 55
#,$
Food intake (kJ/week) 2014 ± 118 2716 ± 270#
2018 ± 89 2310 ± 82#,$
Fasting glucose (mM)*
4.3 ± 0.5 4.6 ± 0.4 4.9 ± 0.5 5.0 ± 0.7
AUCg (mM∙h)*,†
10.3 ± 0.8 12.2 ± 1.1 10.6 ± 0.8 13.5 ± 1.1
Fasting insulin (pM)*,†
137 ± 25 317 ± 140 200 ± 165 517 ± 25
AUCi (pM∙h) 313 ± 56 756 ± 220‡
408 ± 20 1517 ± 125#,$
AUCg*AUCi (mM∙h*pM∙h) 3189 ± 631 9628 ± 2789‡
3485 ± 1113 20839 ± 2482#,◊
Plasma FFA (mM)*,†
0.07 ± 0.03 0.30 ± 0.07 0.18 ± 0.05 0.33 ± 0.06
Plasma TAG (mM)*,†
0.7 ± 0.2 1.2 ± 0.3 1.1 ± 0.3 1.5 ± 0.4
Plasma glycerol (mM)*,†
0.07 ± 0.02 0.15 ± 0.04 0.09 ± 0.03 0.19 ± 0.05
Data are from n = 9 and n = 10 animals for 2.5 wk and 25 wk, respectively, and are expressed as means ±
SD. NC, normal chow; HFD, high‐fat diet; AUCg and AUCi, area under the glucose and insulin curve,
respectively; FFA, free fatty acids; TAG, triacylglycerols. *General age effect (p < 0.05), †General effect of
HFD (p < 0.05), #p < 0.001 vs. NC‐fed rats of the same age,
‡p < 0.05 vs. NC‐fed rats of the same age,
$p <
0.001 vs. the same diet group after 2.5 wk of diet, ◊p < 0.05 vs. the same diet group after 2.5 wk of diet.
Chapter 5
108
Figure 2. Muscle IMCL content and acylcarnitine profile. Representative examples of 1H MR spectra,
measured in the dorsal TA of the same rat after 2.5 wk (A) and 25 wk (B) of HFD and of the same rat after
2.5 wk (C) and 25 wk (D) of NC. Cr, creatine; tCr, total creatine; EMCL, extramyocellular lipid; IMCL,
intramyocellular lipid. (E) IMCL content in the dorsal TA (n = 10 for each diet group). IMCL content was
expressed as a percentage of the water signal. *P < 0.001 vs. NC‐fed rats of the same age, #P < 0.001 vs.
the same diet group after 2.5 wk of diet. (F) Acylcarnitine content in gastrocnemius muscle (n = 9 and n =
10, for 2.5 and 25 wk data, respectively). *p < 0.001 vs. NC‐fed rats of the same age, #p < 0.001 vs. the
same diet group after 2.5 wk of diet, †p = 0.06 vs. NC‐fed rats of the same age. Data are means ± SD.
Mitochondrial function and insulin resistance
109
HFD feeding led to development of whole‐body insulin resistance as indicated by 2.9 (p <
0.01) and 4.7 (p < 0.001) fold higher product of AUCg and AUCi (Figure 1C, Table 1) in HFD‐
fed rats compared to controls after 2.5 and 25 wk of diet, respectively. To confirm an
impairment of the insulin signaling cascade in skeletal muscle, we determined the
phosphorylation level of insulin receptor substrate 1 (IRS1) at serine 307 (ser307). IRS1 ser307
phosphorylation was higher (p < 0.001) in tibialis anterior (TA) muscle of HFD‐fed rats
compared to controls both after 2.5 wk and 25 wk (Figure 1D). Irrespective of the diet, the
level of IRS1 ser307 phosphorylation significantly (p < 0.001) increased with age.
IMCL and acylcarnitine content
Next, we assessed IMCL levels in TA using 1H MRS. Representative examples of 1H MR
spectra of the dorsal, oxidative part of TA are shown in Figure 2A‐D. IMCL content in this
part of the TA was 3.3 (p < 0.001) and 5.1 (p < 0.001) fold higher in HFD‐fed compared to
controls after 2.5 wk and 25 wk of diet, respectively (Figure 2E). IMCL content in the
ventral, glycolytic part of TA was lower than that in the oxidative part in both NC‐ and HFD‐
fed rats (data not shown). HFD feeding led to a 3.2 and 3.8 fold higher IMCL content (p <
0.001) in this part of the TA after 2.5 wk and 25 wk of diet, respectively (data not shown).
To verify impairment of lipid metabolism, we determined the content of different
acylcarnitine species in the muscle. The content of free carnitine, acetylcarnitine and
propionylcarnitine was lower in the muscle of HFD‐fed rats compared to controls after 2.5
and 25 wk of diet (data not shown). The content of medium‐ (data not shown) and long‐
chain acylcarnitines in almost all cases was significantly higher in muscle of HFD‐fed rats
both after 2.5 and 25 wk of diet (Figure 2F). The largest increase in response to HFD feeding
was observed in long‐chain acylcarnitine species (C16:0, C18:0 and C18:1), which reflect the
major dietary and stored fatty acids. The relative increase was similar after 2.5 and 25 wk of
diet, but the absolute levels were higher after 25 wk of HFD.
In vivo muscle oxidative capacity
We used 31P MRS to measure post‐exercise phosphocreatine (PCr) recovery rate in TA, an in
vivo measure of the oxidative capacity of this muscle [34]. Representative examples of 31P
MR spectra from TA during rest and at the end of electrical stimulation are shown in Figure
3A and 3B, respectively. Concentrations of metabolites and pH obtained from the 31P MR
spectra are summarized in Table 2. HFD feeding had no significant effect on the metabolite
concentrations in TA at rest and at the end of electrical stimulation, except for [Pi] at rest,
which was significantly higher (p < 0.05) after 25 wk of HFD.
Chapter 5
110
PCr concentrations during the recovery after electrical stimulation were fitted with a mono‐
exponential function (Figure 3C) and recovery rate constants (kPCr), representing in vivo
muscle oxidative capacity, were calculated (Figure 3D). kPCr was significantly higher (20%, p
< 0.01) in HFD‐fed rats compared to NC‐fed rats after 2.5 wk of diet. However, kPCr was
similar for HFD‐ and NC‐fed rats after 25 wk of diet. Irrespective of the diet, kPCr decreased
with age (p < 0.001).
Figure 3. In vivo oxidative capacity of TA assessed by 31P MRS. Representative examples of
31P MR
spectra at rest (A) and at the end of stimulation (B), measured in TA of the same NC‐fed rat after 2.5 wk
of diet. , and γ, phosphate groups of ATP; PCr, phosphocreatine; Pi, inorganic phosphate. (C) Representative example of fitting of PCr concentrations during the recovery after electrical stimulation.
(D) The rate constants of PCr recovery (kPCr) in TA after electrical stimulation (n = 10 for each diet group).
*p < 0.001 vs. NC‐fed rats of the same age, #p < 0.001 vs. the same diet group after 2.5 wk of diet. Data
are means ± SD.
Mitochondrial function and insulin resistance
111
Intrinsic mitochondrial function
Table 3 summarizes the effects of HFD feeding on the intrinsic mitochondrial function in
vitro, represented by oxygen consumption rates in isolated TA mitochondria oxidizing
glucose‐ or fat‐derived substrates under different metabolic states. HFD feeding had no
significant effect on pyruvate plus malate‐driven oxygen consumption in actively
phosphorylating (state 3), resting (state 4) and uncoupled (state U) mitochondria.
Interestingly, a ~30% higher palmitoyl‐L‐carnitine plus malate driven oxygen consumption
rate in state 3 and state U was observed in mitochondria from TA of HFD‐fed rats compared
to controls after 25 wk of diet. Independent of the diet, oxygen consumption rates declined
with age (p < 0.001).
FFA‐ and reactive oxygen species‐induced expression and activation of uncoupling protein 3
(UCP3) [35] may increase proton permeability of the inner mitochondrial membrane and
diminish ATP synthesis efficiency. Figure 4A shows that there was no significant difference
in UCP3 expression in TA of NC‐ and HFD‐fed rats after 2.5 wk of diet, whereas after 25 wk
expression of UCP3 was 76% higher (p < 0.001) in TA of HFD‐fed rats compared to controls
of the same age.
Table 2. Metabolite concentrations and pH in TA measured by in vivo 31P MRS.
Rest 2.5 wk 25 wk
NC HFD NC HFD
pH 7.14 ± 0.01 7.16 ± 0.03 7.13 ± 0.04 7.15 ± 0.03
[PCr] (mM)* 40.1 ± 2.1 38.2 ± 2.5 41.1 ± 1.5 41.1 ± 3.4
[Pi] (mM) 2.6 ± 0.7 2.4 ± 0.6 2.7 ± 0.8 3.6 ± 0.6†,‡
[ADP] (μM) 12.4 ± 0.6 12.4 ± 0.7 11.8 ± 1.0 12.4 ± 0.7
End stimulation 2.5 wk 25 wk
NC HFD NC HFD
pH 7.01 ± 0.05 7.02 ± 0.04 7.04 ± 0.04 7.01 ± 0.04
[PCr] (mM)* 18.0 ± 2.2 16.8 ± 1.2 20.0 ± 3.8 18.5 ± 2.3
[Pi] (mM) 22.4 ± 1.8 20.8 ± 2.4 21.7 ± 2.7 23.3 ± 3.75
[ADP] (μM) 83.9 ± 10.4 88.1 ± 10.2 81.0 ± 24.6 82.2 ± 16.3
Data are from n = 10 animals for each group and are expressed as means ± SD. NC, normal chow;
HFD, high‐ diet; PCr, phosphocreatine; Pi, inorganic phosphate. *General age effect (p < 0.05), †p <
0.05 vs. NC‐fed fat rats of the same age, ‡p < 0.001 vs. the same diet group after 2.5 wk of diet.
Chapter 5
112
Mitochondrial content
Skeletal muscle oxidative capacity in vivo is determined by the intrinsic mitochondrial
function as well as the amount of mitochondria in the tissue. Figure 4B shows that HFD
feeding resulted in a significant increase (p < 0.001) in the relative mitochondrial copy
number in TA compared to controls (23% and 42% after 2.5 wk and 25 wk of HFD diet,
respectively). In addition, the expression of peroxisome proliferator‐activated receptor‐γ
coactivator 1 (PGC1) was significantly higher (p < 0.001) in TA of HFD‐fed rats compared
to controls (36% and 66% after 2.5 and 25 wk of HFD, respectively) (Figure 4C), indicating
that the increase in the relative mitochondrial copy number may be mediated by PGC1.
Enzyme activities
To further study muscle adaptations to HFD feeding, we determined in vitro activities of a
number of mitochondrial enzymes involved in substrate oxidation and the mitochondrial
respiratory chain. The specific activities of citrate synthase (CS), very‐long‐chain acyl‐CoA
dehydrogenase (VLCAD), succinate dehydrogenase (SDH) and cytochrome c oxidase (CCO)
were significantly elevated (p < 0.05) after HFD feeding (Table 4). A substantial increase in
Table 3. Oxygen consumption rates in isolated TA mitochondria oxidizing different substrates in different
metabolic states.
Pyruvate plus malate 2.5 wk 25 wk
NC HFD NC HFD
State 3 (nmol O2min
‐1∙mg protein
‐1)* 641 ± 76 654 ± 113 375 ± 60 375 ± 80
State 4 (nmol O2min‐1∙mg protein
‐1)* 38 ± 4 41 ± 8 31 ± 5 27 ± 6
State U (nmol O2min‐1∙mg protein
‐1)* 784 ± 103 777 ± 145 495 ± 112 475 ± 118
Palmitoyl‐L‐carnitine plus malate 2.5 wk 25 wk
NC HFD NC HFD
State 3 (nmol O2min
‐1∙mg protein
‐1)* 275 ± 96 279 ± 48 103 ± 32 137 ± 34
State 4 (nmol O2min
‐1∙mg protein
‐1)* 37 ± 12 44 ± 13 30 ± 7 25 ± 6
State U (nmol O2min‐1∙mg protein
‐1)* 312 ± 72 319 ± 53 146 ± 42 187 ± 50
Data are from n = 9 and n = 10 animals for 2.5 wk and 25 wk, respectively, and are expressed as means ±
SD. NC, normal chow; HFD, high‐fat diet; State 3, maximal ADP‐stimulated oxygen consumption; State 4,
oxygen consumption in the absence of ATP synthesis; State U, oxygen consumption in uncoupled state.
*General age effect (p < 0.05).
Mitochondrial function and insulin resistance
113
enzyme activities was already observed after 2.5 wk of HFD (CS by 17%, VLCAD by 40%, SDH
by 27% and CCO by 39%), which became more pronounced after 25 wk of HFD (CS by 42%,
VLCAD by 84%, SDH by 49% and CCO by 49%). Increased activities of mitochondrial
enzymes provide further evidence for increased mitochondrial proliferation in response to
HFD feeding. Interestingly, VLCAD was the enzyme with the largest increase, suggesting
increased VLCAD enzyme activity per mitochondrion.
Table 4. Specific enzyme activities in TA homogenates.
2.5 wk 25 wk
NC HFD NC HFD
CS (nmolmin
‐1∙mg protein
‐1)*
,† 193.4 ± 33.7 226.7 ± 34.8 152.2 ± 38.7 216.3 ± 41.9
VLCAD (nmolmin‐1∙mg protein
‐1)*
,† 30.8 ± 9.9 43.0 ± 16.5 11.0 ± 5.0 20.3 ± 5.3
SDH (nmolmin‐1∙mg protein
‐1)*
,† 10.5 ± 2.7 13.4 ± 2.7 6.4 ± 2.3 9.5 ± 2.6
CCO (nmolmin‐1∙mg protein
‐1)† 33.5 ± 6.1 46.5 ± 12.8 36.5 ± 10.0 54.5 ± 13.4
Data are from n = 9 and n = 10 animals for 2.5 wk and 25 wk, respectively, and are expressed as means ±
SD. NC, normal chow; HFD, high‐fat diet; CS, citrate synthase, VLCAD, very‐long‐chain acyl‐CoA
dehydrogenase; SDH, succinate dehydrogenase; CCO, cytochrome c oxidase. *General age effect (p <
0.05), †General effect of HFD (p < 0.05).
Figure 4. The effect of high‐fat diet on mitochondrial content and protein expression in TA. (A)
Expression of UCP3 in TA (n = 4 for each diet group). *p < 0.001 vs. NC‐fed rats of the same age. (B)
Relative mitochondrial copy number in TA as determined by quantitative PCR (n = 9 and n = 10, for 2.5
and 25 wk data, respectively). There was a significant effect of HFD (p < 0.001) and age (p < 0.001) on the
relative mitochondrial copy number in TA. (C) Expression of PGC1 in TA (n = 6 per diet group). There was a significant effect of HFD (p < 0.001) and age (p < 0.01) on PGC1 expression in TA. Data are means ± SD.
Chapter 5
114
Discussion
Several lines of evidence indicate that decreased skeletal muscle oxidative capacity,
presumably associated with a reduced mitochondrial capacity to oxidize fatty acids, might
stimulate IMCL accumulation and contribute to the development of skeletal muscle IR [36].
However, an increasing number of studies in humans and rodents show a dissociation
between mitochondrial dysfunction and insulin resistance [12‐13, 17‐18, 37]. To clarify this
issue, we investigated the adaptations of skeletal muscle oxidative metabolism during the
development of HFD‐induced IR in rats using both in vivo and in vitro approaches. We
showed that HFD feeding does not affect the maximal ADP‐stimulated (i.e. state 3)
pyruvate plus malate‐driven oxygen consumption rate in isolated mitochondria, but leads
to an increased number of muscle mitochondria and an increased capacity to oxidize fat‐
derived substrates in vitro. Surprisingly, this did not result in a higher muscle oxidative
capacity in vivo after long‐term HFD feeding. These findings indicate that the higher
mitochondrial content compensates for a compromised in vivo mitochondrial function in
insulin resistant muscle.
The results of our in vitro measurements are in agreement with previous studies on the
effect of HFD feeding on muscle mitochondrial function in rats. It has repeatedly been
shown that HFD does not affect respiration of rat muscle mitochondria when using
substrates other than fatty acids, i.e. pyruvate/glutamate plus malate and succinate [17,
38‐39], implying that the intrinsic functioning of the mitochondria is not changed in vitro.
Furthermore, a number of studies have shown that HFD feeding in rats induces increased
biogenesis of mitochondria and increases in vitro mitochondrial fatty acid oxidative
capacity in skeletal muscle [15, 17‐18, 40‐42]. The strength of this study is that it combines
in vitro and in vivo methodologies to measure mitochondrial function, resulting in the new
and unexpected finding that the mitochondrial adaptations observed in vitro are not
paralleled by an increased muscle oxidative capacity in vivo after long‐term HFD feeding.
In vivo skeletal muscle oxidative capacity is determined by mitochondrial content, intrinsic
mitochondrial function and other factors like the supply of oxygen and substrates. We
showed that short‐term (2.5 wk) HFD feeding resulted in the accumulation of IMCL and
higher oxidative capacity of TA as indicated by a 20% higher PCr recovery rate constant
(kPCr) compared to controls. kPCr reflects in vivo muscle mitochondrial oxidative capacity
because during recovery from exercise PCr is resynthesized purely as a consequence of
oxidative ATP synthesis [43‐45] and the creatine kinase reaction is much faster than
oxidative ATP production [46]. Since during the recovery from short intense exercise
carbohydrates are preferentially used as the source of oxidizable substrates in the muscle
[47], our observation that pyruvate plus malate‐driven oxygen consumption in isolated TA
mitochondria was similar in NC‐ and HFD‐fed rats implies that the increase in TA oxidative
capacity observed in vivo after 2.5 wk of HFD feeding was not a consequence of improved
intrinsic mitochondrial function. We showed that the relative mitochondrial copy number
Mitochondrial function and insulin resistance
115
as well as CS activity, which has been shown to correlate with mitochondrial volume
density [48‐49], was increased in the same range as kPCr, suggesting that the increase in
mitochondrial content caused the higher in vivo muscle oxidative capacity after short‐term
HFD feeding.
Long‐term HFD feeding caused further accumulation of IMCL in TA. However, in contrast to
short‐term HFD feeding, long‐term (25 wk) HFD feeding resulted in similar TA oxidative
capacity in vivo compared to controls as indicated by similar kPCr. In agreement, pyruvate
plus malate‐driven oxygen consumption rates in isolated TA mitochondria were not
significantly different in the two diet groups at this time point. However, long‐term HFD
feeding caused a further increase in PGC1 expression with a concomitant increase in
mitochondrial copy number and CS activity in TA compared to controls. Therefore, one
would also expect a higher in vivo oxidative capacity in TA of HFD‐fed rats. However, we
showed that 42% more mitochondria were needed in TA of HFD‐fed rats to maintain the
same in vivo muscle oxidative capacity as in controls of the same age. This discrepancy
suggests that the intrinsic functioning of mitochondria is impaired when they are placed in
their natural cellular environment by factors that are not taken into account during in vitro
measurement of intrinsic mitochondrial function. Efficient supply of substrates and oxygen
to the muscle and removal of potentially toxic metabolites is essential for normal
mitochondrial function in the tissue. Insulin regulates muscle blood flow through
stimulation of nitric oxide‐dependent vasodilation [50]. Consequently, the development of
vascular IR in response to HFD‐feeding may result in diminished muscle perfusion and
substrate supply causing sub‐optimal mitochondrial performance in vivo. However, our
observation that during electrical stimulation the rate of acidification and the degree of PCr
depletion was similar in TA of HFD‐fed rats compared to controls at both time points
indicates that the perfusion of TA was not affected by HFD feeding.
Not only the shortage but also the excess of substrates may negatively affect mitochondrial
function in vivo. High concentrations of FFA and their metabolites may impair
mitochondrial ATP production through a number of mechanisms. E.g. long‐chain acyl‐CoA
esters may inhibit mitochondrial adenine nucleotide translocator leading to impaired
exchange of cytosolic ADP for mitochondrial ATP [51]. Moreover, increased availability of
FFA may lead to increased expression [17] as well as activation of UCP3 [35], which in turn
may lead to diminished efficiency of ATP synthesis. We showed that both the content of
long‐chain acylcarnitines, which reflects the long‐chain acyl‐CoA content [52], and the
expression of UCP3 in the muscle were significantly increased after long‐term HFD feeding
suggesting that both proposed mechanisms (i.e. inhibition of adenine nucleotide
translocator and mitochondrial uncoupling) may contribute to mitochondrial functional
impairment in vivo. On the other hand, increased expression of UCP3 in the muscle may be
a trade‐off between lower mitochondrial efficiency (in terms of ATP production) and
lipotoxicity in vivo, which enables muscle cells to burn more fat [53] and to protect
Chapter 5
116
themselves from increased production of reactive oxygen species arising from increased fat
oxidation [35].
Fatty acids are activating ligands of peroxisome proliferator‐activated receptors (PPARs),
which regulate the expression of genes encoding enzymes involved in fatty acid oxidation
[54]. The increase in mitochondrial capacity to oxidize fatty acids observed in this study
may therefore be related to direct activation of PPARs by fatty acids. In addition, PPARs are
co‐activated by PGC1, the master regulator of mitochondrial biogenesis [55]. We showed
that the expression of PGC1 in TA was progressively stimulated by HFD feeding with a
concomitant increase in mitochondrial DNA copy number. However, also high insulin levels
stimulate mitochondrial protein synthesis [56]. Therefore, the HFD‐induced changes in
skeletal muscle oxidative metabolism observed in this study could be caused by increased
levels of FFA or insulin, or both.
As an alternative to measuring muscle oxidative capacity from dynamic 31P MRS
experiments after exercise, Laurent et al. applied 31P MRS saturation transfer experiments
to measure mitochondrial ATP synthesis rates in resting skeletal muscle of HFD‐fed rats
[19]. ATP synthesis rates rapidly declined after the commencement of HFD feeding,
returned to normal values after 2‐3 weeks and decreased again after 1 month of HFD
feeding and the lower rates were taken as evidence for mitochondrial dysfunction [19].
However, the interpretation of 31P saturation transfer data is not straightforward. The
lower basal ATP synthesis rates in insulin‐resistant states [6‐7] could actually reflect a lower
energy demand caused by impaired insulin signaling rather than an impairment of
mitochondrial function [57‐59]. Moreover, the fluxes obtained from 31P saturation transfer
measurements are comprised of both ATP synthesis flux and glycolytic exchange flux, with
the latter contributing by as much as 80% [60‐61]. Therefore, decreased basal ATP
synthesis rates do not necessarily reflect a mitochondrial defect.
It has been suggested that an increased expression of enzymes involved in ‐oxidation without a concomitant upregulation of enzymes involved in the tricarboxylic acid (TCA)
cycle may result in incomplete oxidation of fatty acids and depletion of TCA cycle
intermediates leading to reduced glucose oxidation and IR [62]. Our data on the enzyme
activities and oxygen consumption in isolated mitochondria suggest that such dissociation
may indeed occur in the muscle of HFD‐fed rats. Furthermore, the increase in the content
of medium‐chain acylcarnitines and the decrease in acetylcarnitine content in the muscle of
HFD‐fed rats may imply incomplete oxidation of fatty acids. However, we did not measure
TCA cycle and beta oxidation flux and therefore our interpretation is rather speculative.
Moreover, acetyl‐CoA, the precursor of acetylcarnitine, is also produced during oxidation of
glucose and amino acids. Therefore, the decrease in acetylcarnitine content in the muscle
of HFD‐fed rats may indicate decreased glucose oxidation rather than incomplete oxidation
of fatty acids.
Mitochondrial function and insulin resistance
117
In conclusion, this study revealed the adaptations of rat skeletal muscle mitochondria
characteristic to early and later stages of HFD‐induced IR. We demonstrated that HFD
feeding resulted in progressive increase in mitochondrial content and higher capacity to
oxidize fat‐derived substrates in the long‐term. Most importantly, the combination of in
vitro and in vivo approaches allowed the establishment of the relationship between in vitro
intrinsic mitochondrial function, mitochondrial number, and in vivo muscle oxidative
capacity, showing that an increased number of mitochondria with normal function in vitro
was required to maintain normal oxidative capacity in vivo during later stages of IR. The
observed dissociation between in vivo and in vitro muscle oxidative capacity suggests that
in vivo mitochondrial function in insulin resistant rat muscle is compromised by factors not
taken into account when mitochondrial function is assed under in vitro conditions.
Acknowledgements
We would like to thank J.W. Habets, L.B.P Niesen and R.A.M. Jonkers for technical
assistance.
The work of J.C. and J.A.J. is financed by the Netherlands Consortium for Systems Biology
(NCSB) which is part of the Netherlands Genomics Initiative / Netherlands Organization for
Scientific Research. J.J.P. and S.M.H. are supported by VIDI grants from the Netherlands
Organization for Scientific Research (VIDI grant numbers 700.58.421 and 016.086.336,
respectively).
Chapter 5
118
References 1. Wang, Y., E.B. Rimm, M.J. Stampfer, W.C. Willett, and F.B. Hu. Comparison of abdominal adiposity and
overall obesity in predicting risk of type 2 diabetes among men. Am J Clin Nutr, 2005. 81(3): 555‐63. 2. Boden, G., B. Lebed, M. Schatz, C. Homko, and S. Lemieux. Effects of acute changes of plasma free fatty
acids on intramyocellular fat content and insulin resistance in healthy subjects. Diabetes, 2001. 50(7): 1612‐7.
3. Goodpaster, B.H., J. He, S. Watkins, and D.E. Kelley. Skeletal muscle lipid content and insulin resistance: evidence for a paradox in endurance‐trained athletes. J Clin Endocrinol Metab, 2001. 86(12): 5755‐61.
4. Sinha, R., S. Dufour, K.F. Petersen, V. LeBon, S. Enoksson, Y.Z. Ma, M. Savoye, D.L. Rothman, G.I. Shulman, and S. Caprio. Assessment of skeletal muscle triglyceride content by (1)H nuclear magnetic resonance spectroscopy in lean and obese adolescents: relationships to insulin sensitivity, total body fat, and central adiposity. Diabetes, 2002. 51(4): 1022‐7.
5. van Loon, L.J., R. Koopman, R. Manders, W. van der Weegen, G.P. van Kranenburg, and H.A. Keizer. Intramyocellular lipid content in type 2 diabetes patients compared with overweight sedentary men and highly trained endurance athletes. Am J Physiol Endocrinol Metab, 2004. 287(3): E558‐65.
6. Petersen, K.F., D. Befroy, S. Dufour, J. Dziura, C. Ariyan, D.L. Rothman, L. DiPietro, G.W. Cline, and G.I. Shulman. Mitochondrial dysfunction in the elderly: possible role in insulin resistance. Science, 2003. 300(5622): 1140‐2.
7. Petersen, K.F., S. Dufour, D. Befroy, R. Garcia, and G.I. Shulman. Impaired mitochondrial activity in the insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(7): 664‐71.
8. Yu, C., Y. Chen, G.W. Cline, D. Zhang, H. Zong, Y. Wang, R. Bergeron, J.K. Kim, S.W. Cushman, G.J. Cooney, B. Atcheson, M.F. White, E.W. Kraegen, and G.I. Shulman. Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate‐1 (IRS‐1)‐associated phosphatidylinositol 3‐kinase activity in muscle. J Biol Chem, 2002. 277(52): 50230‐6.
9. Mootha, V.K., C.M. Lindgren, K.F. Eriksson, A. Subramanian, S. Sihag, J. Lehar, P. Puigserver, E. Carlsson, M. Ridderstrale, E. Laurila, N. Houstis, M.J. Daly, N. Patterson, J.P. Mesirov, T.R. Golub, P. Tamayo, B. Spiegelman, E.S. Lander, J.N. Hirschhorn, D. Altshuler, and L.C. Groop. PGC‐1alpha‐responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet, 2003. 34(3): 267‐73.
10. Patti, M.E., A.J. Butte, S. Crunkhorn, K. Cusi, R. Berria, S. Kashyap, Y. Miyazaki, I. Kohane, M. Costello, R. Saccone, E.J. Landaker, A.B. Goldfine, E. Mun, R. DeFronzo, J. Finlayson, C.R. Kahn, and L.J. Mandarino. Coordinated reduction of genes of oxidative metabolism in humans with insulin resistance and diabetes: Potential role of PGC1 and NRF1. Proc Natl Acad Sci U S A, 2003. 100(14): 8466‐71.
11. Ritov, V.B., E.V. Menshikova, J. He, R.E. Ferrell, B.H. Goodpaster, and D.E. Kelley. Deficiency of subsarcolemmal mitochondria in obesity and type 2 diabetes. Diabetes, 2005. 54(1): 8‐14.
12. Boushel, R., E. Gnaiger, P. Schjerling, M. Skovbro, R. Kraunsoe, and F. Dela. Patients with type 2 diabetes have normal mitochondrial function in skeletal muscle. Diabetologia, 2007. 50(4): 790‐6.
13. De Feyter, H.M., N.M. van den Broek, S.F. Praet, K. Nicolay, L.J. van Loon, and J.J. Prompers. Early or advanced stage type 2 diabetes is not accompanied by in vivo skeletal muscle mitochondrial dysfunction. Eur J Endocrinol, 2008. 158(5): 643‐53.
14. Sreekumar, R., P. Halvatsiotis, J.C. Schimke, and K.S. Nair. Gene expression profile in skeletal muscle of type 2 diabetes and the effect of insulin treatment. Diabetes, 2002. 51(6): 1913‐20.
15. Iossa, S., L. Lionetti, M.P. Mollica, R. Crescenzo, M. Botta, A. Barletta, and G. Liverini. Effect of high‐fat feeding on metabolic efficiency and mitochondrial oxidative capacity in adult rats. Br J Nutr, 2003. 90(5): 953‐60.
16. Bonnard, C., A. Durand, S. Peyrol, E. Chanseaume, M.A. Chauvin, B. Morio, H. Vidal, and J. Rieusset. Mitochondrial dysfunction results from oxidative stress in the skeletal muscle of diet‐induced insulin‐resistant mice. J Clin Invest, 2008. 118(2): 789‐800.
17. Turner, N., C.R. Bruce, S.M. Beale, K.L. Hoehn, T. So, M.S. Rolph, and G.J. Cooney. Excess lipid availability increases mitochondrial fatty acid oxidative capacity in muscle: evidence against a role for reduced fatty acid oxidation in lipid‐induced insulin resistance in rodents. Diabetes, 2007. 56(8): 2085‐92.
18. Hancock, C.R., D.H. Han, M. Chen, S. Terada, T. Yasuda, D.C. Wright, and J.O. Holloszy. High‐fat diets cause insulin resistance despite an increase in muscle mitochondria. Proc Natl Acad Sci U S A, 2008. 105(22): 7815‐20.
19. Laurent, D., B. Yerby, R. Deacon, and J. Gao. Diet‐induced modulation of mitochondrial activity in rat muscle. Am J Physiol Endocrinol Metab, 2007. 293(5): E1169‐77.
20. Tran, T.T., N. Gupta, T. Goh, D. Naigamwalla, M.C. Chia, N. Koohestani, S. Mehrotra, G. McKeown‐Eyssen, A. Giacca, and W.R. Bruce. Direct measure of insulin sensitivity with the hyperinsulinemic‐euglycemic clamp and surrogate measures of insulin sensitivity with the oral glucose tolerance test:
Mitochondrial function and insulin resistance
119
correlations with aberrant crypt foci promotion in rats. Cancer Epidemiol Biomarkers Prev, 2003. 12(1): 47‐56.
21. van Vlies, N., L. Tian, H. Overmars, A.H. Bootsma, W. Kulik, R.J. Wanders, P.A. Wood, and F.M. Vaz. Characterization of carnitine and fatty acid metabolism in the long‐chain acyl‐CoA dehydrogenase‐deficient mouse. Biochem J, 2005. 387(Pt 1): 185‐93.
22. De Feyter, H.M., G. Schaart, M.K. Hesselink, P. Schrauwen, K. Nicolay, and J.J. Prompers. Regional variations in intramyocellular lipid concentration correlate with muscle fiber type distribution in rat tibialis anterior muscle. Magn Reson Med, 2006. 56(1): 19‐25.
23. Garwood, M. and L. DelaBarre. The return of the frequency sweep: designing adiabatic pulses for contemporary NMR. J Magn Reson, 2001. 153(2): 155‐77.
24. de Graaf, R.A. and K. Nicolay. Adiabatic water suppression using frequency selective excitation. Magn Reson Med, 1998. 40(5): 690‐6.
25. De Feyter, H.M., E. Lenaers, S.M. Houten, P. Schrauwen, M.K. Hesselink, R.J. Wanders, K. Nicolay, and J.J. Prompers. Increased intramyocellular lipid content but normal skeletal muscle mitochondrial oxidative capacity throughout the pathogenesis of type 2 diabetes. Faseb J, 2008. 22(11): 3947‐55.
26. Vanhamme, L., A. van den Boogaart, and S. Van Huffel. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson, 1997. 129(1): 35‐43.
27. Reichert, M., H. Schaller, W. Kunz, and G. Gerber. The dependence on the extramitochondrial ATP/ADP‐ratio of the oxidative phosphorylation in mitochondria isolated by a new procedure from rat skeletal muscle. Acta Biol Med Ger, 1978. 37(8): 1167‐76.
28. Ciapaite, J., S.J. Bakker, G. Van Eikenhorst, M.J. Wagner, T. Teerlink, C.G. Schalkwijk, M. Fodor, D.M. Ouwens, M. Diamant, R.J. Heine, H.V. Westerhoff, and K. Krab. Functioning of oxidative phosphorylation in liver mitochondria of high‐fat diet fed rats. Biochim Biophys Acta, 2007. 1772(3): 307‐16.
29. Szuhai, K., J. Ouweland, R. Dirks, M. Lemaitre, J. Truffert, G. Janssen, H. Tanke, E. Holme, J. Maassen, and A. Raap. Simultaneous A8344G heteroplasmy and mitochondrial DNA copy number quantification in myoclonus epilepsy and ragged‐red fibers (MERRF) syndrome by a multiplex molecular beacon based real‐time fluorescence PCR. Nucleic Acids Res, 2001. 29(3): E13.
30. Munujos, P., J. Coll‐Canti, F. Gonzalez‐Sastre, and F.J. Gella. Assay of succinate dehydrogenase activity by a colorimetric‐continuous method using iodonitrotetrazolium chloride as electron acceptor. Anal Biochem, 1993. 212(2): 506‐9.
31. Cooperstein, S.J. and A. Lazarow. A microspectrophotometric method for the determination of cytochrome oxidase. J Biol Chem, 1951. 189(2): 665‐70.
32. Srere, P.A., H. Brazil, and L. Gonen. The citrate condensing enzyme of pigeon breast muscle and moth flight muscle. Acta. Chem. Scand. , 1963. 17: S129‐S134.
33. Wanders, R.J., P. Vreken, M.E. den Boer, F.A. Wijburg, A.H. van Gennip, and I.J. L. Disorders of mitochondrial fatty acyl‐CoA beta‐oxidation. J Inherit Metab Dis, 1999. 22(4): 442‐87.
34. Kemp, G.J. and G.K. Radda. Quantitative interpretation of bioenergetic data from 31P and 1H magnetic resonance spectroscopic studies of skeletal muscle: an analytical review. Magn Reson Q, 1994. 10(1): 43‐63.
35. Echtay, K.S., D. Roussel, J. St‐Pierre, M.B. Jekabsons, S. Cadenas, J.A. Stuart, J.A. Harper, S.J. Roebuck, A. Morrison, S. Pickering, J.C. Clapham, and M.D. Brand. Superoxide activates mitochondrial uncoupling proteins. Nature, 2002. 415(6867): 96‐9.
36. Morino, K., K.F. Petersen, and G.I. Shulman. Molecular mechanisms of insulin resistance in humans and their potential links with mitochondrial dysfunction. Diabetes, 2006. 55 Suppl 2: S9‐S15.
37. Turner, N. and L.K. Heilbronn. Is mitochondrial dysfunction a cause of insulin resistance? Trends Endocrinol Metab, 2008. 19(9): 324‐30.
38. Chanseaume, E., A.L. Tardy, J. Salles, C. Giraudet, P. Rousset, A. Tissandier, Y. Boirie, and B. Morio. Chronological approach of diet‐induced alterations in muscle mitochondrial functions in rats. Obesity (Silver Spring), 2007. 15(1): 50‐9.
39. Hoeks, J., J.J. Briede, J. de Vogel, G. Schaart, M. Nabben, E. Moonen‐Kornips, M.K. Hesselink, and P. Schrauwen. Mitochondrial function, content and ROS production in rat skeletal muscle: effect of high‐fat feeding. FEBS Lett, 2008. 582(4): 510‐6.
40. Nemeth, P.M., B.W. Rosser, R.M. Choksi, B.J. Norris, and K.M. Baker. Metabolic response to a high‐fat diet in neonatal and adult rat muscle. Am J Physiol, 1992. 262(2 Pt 1): C282‐6.
41. Garcia‐Roves, P., J.M. Huss, D.H. Han, C.R. Hancock, E. Iglesias‐Gutierrez, M. Chen, and J.O. Holloszy. Raising plasma fatty acid concentration induces increased biogenesis of mitochondria in skeletal muscle. Proc Natl Acad Sci U S A, 2007. 104(25): 10709‐13.
42. McAinch, A.J., J.S. Lee, C.R. Bruce, R.J. Tunstall, J.A. Hawley, and D. Cameron‐Smith. Dietary regulation of fat oxidative gene expression in different skeletal muscle fiber types. Obes Res, 2003. 11(12): 1471‐9.
Chapter 5
120
43. Taylor, D.J., P.J. Bore, P. Styles, D.G. Gadian, and G.K. Radda. Bioenergetics of intact human muscle. A 31P nuclear magnetic resonance study. Mol Biol Med, 1983. 1(1): 77‐94.
44. Sahlin, K. Intracellular pH and energy metabolism in skeletal muscle of man. With special reference to exercise. Acta Physiol Scand Suppl, 1978. 455: 1‐56.
45. Quistorff, B., L. Johansen, and K. Sahlin. Absence of phosphocreatine resynthesis in human calf muscle during ischaemic recovery. Biochem J, 1993. 291( Pt 3): 681‐6.
46. Vicini, P. and M.J. Kushmerick. Cellular energetics analysis by a mathematical model of energy balance: estimation of parameters in human skeletal muscle. Am J Physiol Cell Physiol, 2000. 279(1): C213‐24.
47. van Hall, G., J. Gonzalez‐Alonso, M. Sacchetti, and B. Saltin. Skeletal muscle substrate metabolism during exercise: methodological considerations. Proc Nutr Soc, 1999. 58(4): 899‐912.
48. Andrade, F.H., C.A. McMullen, and R.E. Rumbaut. Mitochondria are fast Ca2+ sinks in rat extraocular muscles: a novel regulatory influence on contractile function and metabolism. Invest Ophthalmol Vis Sci, 2005. 46(12): 4541‐7.
49. Toledo, F.G., E.V. Menshikova, V.B. Ritov, K. Azuma, Z. Radikova, J. DeLany, and D.E. Kelley. Effects of physical activity and weight loss on skeletal muscle mitochondria and relationship with glucose control in type 2 diabetes. Diabetes, 2007. 56(8): 2142‐7.
50. Laakso, M., S.V. Edelman, G. Brechtel, and A.D. Baron. Impaired insulin‐mediated skeletal muscle blood flow in patients with NIDDM. Diabetes, 1992. 41(9): 1076‐83.
51. Lerner, E., A.L. Shug, C. Elson, and E. Shrago. Reversible inhibition of adenine nucleotide translocation by long chain fatty acyl coenzyme A esters in liver mitochondria of diabetic and hibernating animals. J Biol Chem, 1972. 247(5): 1513‐9.
52. Brass, E.P. and C.L. Hoppel. Relationship between acid‐soluble carnitine and coenzyme A pools in vivo. Biochem J, 1980. 190(3): 495‐504.
53. Costford, S.R., S.N. Chaudhry, S.A. Crawford, M. Salkhordeh, and M.E. Harper. Long‐term high‐fat feeding induces greater fat storage in mice lacking UCP3. Am J Physiol Endocrinol Metab, 2008. 295(5): E1018‐24.
54. Gulick, T., S. Cresci, T. Caira, D.D. Moore, and D.P. Kelly. The peroxisome proliferator‐activated receptor regulates mitochondrial fatty acid oxidative enzyme gene expression. Proc Natl Acad Sci U S A, 1994. 91(23): 11012‐6.
55. Lin, J., P. Puigserver, J. Donovan, P. Tarr, and B.M. Spiegelman. Peroxisome proliferator‐activated receptor gamma coactivator 1beta (PGC‐1beta ), a novel PGC‐1‐related transcription coactivator associated with host cell factor. J Biol Chem, 2002. 277(3): 1645‐8.
56. Stump, C.S., K.R. Short, M.L. Bigelow, J.M. Schimke, and K.S. Nair. Effect of insulin on human skeletal muscle mitochondrial ATP production, protein synthesis, and mRNA transcripts. Proc Natl Acad Sci U S A, 2003. 100(13): 7996‐8001.
57. Short, K.R., K.S. Nair, and C.S. Stump. Impaired mitochondrial activity and insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(23): 2419‐21; author reply 2419‐21.
58. Wagenmakers, A.J. Insulin resistance in the offspring of parents with type 2 diabetes. PLoS Med, 2005. 2(9): e289.
59. Kemp, G.J. The interpretation of abnormal 31P magnetic resonance saturation transfer measurements of Pi/ATP exchange in insulin‐resistant skeletal muscle. Am J Physiol Endocrinol Metab, 2008. 294(3): E640‐2; author reply E643‐4.
60. Brindle, K.M., M.J. Blackledge, R.A. Challiss, and G.K. Radda. 31P NMR magnetization‐transfer measurements of ATP turnover during steady‐state isometric muscle contraction in the rat hind limb in vivo. Biochemistry, 1989. 28(11): 4887‐93.
61. Brindle, K.M. and G.K. Radda. 31P‐NMR saturation transfer measurements of exchange between Pi and ATP in the reactions catalysed by glyceraldehyde‐3‐phosphate dehydrogenase and phosphoglycerate kinase in vitro. Biochim Biophys Acta, 1987. 928(1): 45‐55.
62. Koves, T.R., J.R. Ussher, R.C. Noland, D. Slentz, M. Mosedale, O. Ilkayeva, J. Bain, R. Stevens, J.R. Dyck, C.B. Newgard, G.D. Lopaschuk, and D.M. Muoio. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab, 2008. 7(1): 45‐56.
121
122
Adapted from
N. M. A. van den Broek, J. Ciapaite, S. M. Houten, R. J. A. Wanders, K. Nicolay,
J. J. Prompers. Carnitine insufficiency in high‐fat diet fed rats does not contribute to lipid‐
induced impairment of skeletal muscle mitochondrial function in vivo.
In preparation
Chapter Carnitine insufficiency in high‐fat diet fed rats
does not contribute to lipid‐induced impairment
of skeletal muscle mitochondrial function in vivo
Chapter 6
124
Abstract
Muscle lipid overload plays an important role in the development of insulin resistance.
Although it has been hypothesized that mitochondrial dysfunction and an associated
decreased capacity to oxidize fatty acids (FAs) are the primary cause for this disturbance,
recently it has been shown that high‐fat diets lead to increased rather than decreased FA
oxidation, resulting in the accumulation of lipid intermediates causing mitochondrial
dysfunction. Carnitine supplementation might be effective in reducing these toxic lipid
intermediates by increasing FA oxidation and export of lipid metabolites out of the
mitochondria and into the plasma. The aim of the present study was to test the hypothesis
that carnitine supplementation reduces high‐fat diet‐induced lipotoxicity, improves muscle
mitochondrial function and ameliorates insulin resistance. Wistar rats were fed either
normal chow or a high‐fat diet for 15 weeks and one group of high‐fat diet fed rats was
supplemented with L‐carnitine during the last 8 weeks. Muscle mitochondrial function was
measured in vivo by 31P MRS (PCr recovery after contraction and resting ATP synthesis and
creatine kinase fluxes) and in vitro by high‐resolution respirometry. Muscle lipid status was
determined by 1H MRS (intramyocellular lipids) and tandem mass spectrometry
(acylcarnitines). High‐fat diet feeding induced carnitine insufficiency and muscle lipid
overload and led to an increased number of muscle mitochondria with an improved
capacity to oxidize FAs in vitro. At the same time, in vivo mitochondrial function was
compromised, probably by elevated levels of lipid intermediates leading to increased FA‐
induced mitochondrial uncoupling. Despite the restoration of muscle free carnitine
content, carnitine supplementation did not induce improvements in muscle lipid status, in
vivo mitochondrial function or insulin resistance. These results suggest that the decrease in
muscle free carnitine as a result of high‐fat diet feeding did not contribute to the observed
impairment in in vivo muscle mitochondrial function.
Carnitine insufficiency and mitochondrial function
125
Introduction
Diabetes has reached epidemic proportions worldwide [1]. Type 2 diabetes accounts for 85‐
95% of all diabetes cases and is characterized by insulin resistance in major metabolic
tissues [2]. One of the leading hypotheses in the research field of type 2 diabetes is that
inherited or acquired muscle mitochondrial dysfunction, associated with a reduced
mitochondrial capacity to oxidize fatty acids (FAs), leads to a lipid overload in muscle cells,
causing insulin resistance [3‐4]. Recent studies have challenged this theory and link insulin
resistance to an increased capacity to oxidize FAs [5‐8]. However, it has been shown that
the high rates of ‐oxidation in insulin‐resistant states are associated with low rates of complete FA oxidation and therefore increased incomplete ‐oxidation [9‐11]. It has been hypothesized that the accumulation of incompletely metabolized FAs in the mitochondria
causes ‘mitochondrial stress’ and that this contributes to the development of insulin
resistance [10, 12].
Carnitine is an essential nutrient with multiple functions. Its major role is in the formation
of acylcarnitines from long‐chain FAs, which is required for the transport of acyl moieties
into the mitochondrial matrix for ‐oxidation [13‐14]. A second role for carnitine is to increase acyl and acetyl group efflux out of the mitochondria and into the plasma [15].
Moreover, carnitine stimulates the oxidation of pyruvate by lowering the mitochondrial
acetyl‐CoA/CoA ratio in a reaction catalyzed by carnitine acetyltransferase (CAT), which
converts acetyl‐CoA into acetylcarnitine [16].
Carnitine insufficiency is a common feature of insulin‐resistant states and it has been
shown that muscle free carnitine negatively correlates with insulin resistance [11].
Therefore, it has been suggested that carnitine supplementation might be an effective
treatment for type 2 diabetes [17]. Indeed, intravenous infusion of carnitine during a
hyperinsulinaemic‐euglycaemic clamp has been shown to increase whole‐body glucose
disposal in both healthy subjects and type 2 diabetes patients [18‐22]. Moreover, carnitine
supplementation has been shown to attenuate the development of insulin resistance in
mice and rats fed with high‐fat diets [11, 23]. The positive effect of carnitine on insulin
sensitivity can be explained by different mechanisms, depending on the concentration of
carnitine. In the perfused isolated working rat heart, the addition of carnitine to the
perfusion medium increased glucose oxidation by lowering the concentration of acetyl‐CoA
in the mitochondrial matrix through CAT [24]. However, at the lower concentrations
achievable in vivo, carnitine has been shown to stimulate FA oxidation [25‐26] through a
mass‐action effect on the transport of long‐chain FAs into the mitochondrial matrix [27]. At
the same time, carnitine increases the efflux of acylcarnitines from muscle tissue [11].
Therefore, it has been proposed that carnitine supplementation ameliorates insulin
resistance by reducing lipotoxicity both through the increased oxidation and increased
export of muscle lipid metabolites [17].
Chapter 6
126
In a previous study, we showed that rats fed a long‐term high‐fat diet have an increased
number of muscle mitochondria, with an increased capacity to oxidize fat‐derived
substrates and a normal capacity to oxidize glucose‐derived substrates in vitro [8].
Surprisingly, this did not result in a higher muscle oxidative capacity in vivo, as measured by 31P magnetic resonance spectroscopy (MRS). Muscle tissue free carnitine was significantly
decreased after long‐term high‐fat diet feeding, whereas muscle long‐chain acylcarnitines
were significantly increased. This finding indicates that in vivo muscle mitochondrial
function might have been compromised by high concentrations of lipid metabolites. The
aim of the present study was to test the hypothesis that carnitine supplementation reduces
high‐fat diet‐induced lipotoxicity, improves in vivo muscle mitochondrial function and
ameliorates insulin resistance. Wistar rats were fed either normal chow or a high‐fat diet
for 15 weeks and one group of high‐fat diet fed rats was supplemented with L‐carnitine
(300 mg/kg body weight/d) during the last 8 weeks. In vivo muscle mitochondrial function
was measured by several 31P MRS methods and in vitro mitochondrial function was
determined by measuring oxygen consumption in isolated mitochondria. Furthermore,
intramyocellular lipid levels were determined by 1H MRS and free carnitine and
acylcarnitine levels were determined in muscle tissue, blood and urine using tandem mass
spectrometry.
Materials and Methods
Animals
Adult male Wistar rats (348 ± 18 g, 14 weeks old, n = 30, Charles River Laboratories, The
Netherlands) were housed in pairs at 20C and 50% humidity, with a 12‐h light‐dark cycle.
Ad libitum food and water was provided during a period of 15 weeks. The rats were divided
in three groups: A control group receiving normal chow (NC, 10% calories from fat, 70%
calories from carbohydrate, 20% calories from protein (D12450B, Research Diet Services,
Wijk bij Duurstede, the Netherlands), n = 10), a group receiving a high‐fat diet (HFD, 45%
calories from fat (predominantly lard), 35% calories from carbohydrate, 20% calories from
protein (D12451, Research Diet Services, Wijk bij Duurstede, the Netherlands), n = 10), and
a group receiving the same high‐fat diet, supplemented with 300 mg/kg body weight/d L‐
carnitine in their drinking water for the last 8 weeks (HFDC, n = 10). Body weight and food
and water intake were determined weekly.
About two days after the in vivo MRS measurements, rats were sacrificed by incising the
inferior vena cava under anesthesia. One tibialis anterior (TA) muscle (fresh) was used for
isolation of mitochondria. The other TA was frozen in liquid nitrogen and stored at ‐80C for acylcarnitine content and mitochondrial DNA content determinations. Urine samples were
Carnitine insufficiency and mitochondrial function
127
taken and stored at ‐80C for acylcarnitine determination. Animal handling conformed to
the rules of the Animal Ethics Committee of Maastricht University.
Oral glucose tolerance test
Oral glucose tolerance tests (OGTT) were performed after 15 weeks of diet, three to five
days before the in vivo measurements. After a four‐hour fast, rats received an oral glucose
bolus of 1 g/kg body weight. Blood samples were taken without anesthesia from the vena
saphena just before and at 15, 30, 60, 90 and 120 min after the bolus. Blood glucose
concentration was determined using an automatic glucometer (FreeStyle, Abbott, IL, USA).
Plasma insulin concentration was determined using an ultrasensitive rat insulin ELISA kit
(Mercodia, Uppsala, Sweden). Areas under the OGTT curves for both glucose (AUCg) and
insulin (AUCi) were calculated.
Determination of acylcarnitine content
The content of acylcarnitines was determined in freeze‐dried TA muscle, blood and urine by
tandem mass spectrometry as described previously [28‐30].
Magnetic resonance spectroscopy
All magnetic resonance spectroscopy (MRS) measurements were performed on a 6.3 Tesla
horizontal Bruker MR system (Bruker, Ettlingen, Germany). Animals were anaesthetized
using isoflurane (Forene®) (1‐2%) with medical air (0.6 L/min) and body temperature was
maintained at 36 ± 0.5C using heating pads. Respiration was monitored using a pressure
sensor registering thorax movement (Rapid Biomedical, Rimpar, Germany).
Intramyocellular lipid (IMCL) content in TA was measured in vivo using single‐voxel localized 1H MRS. Voxels of 333 mm3 were measured in the dorsal part of the TA, close to the tibia
bone, with a circular 1H surface coil (Ø 40mm) and using the PRESS sequence (repetition
time TR = 1.5 s, echo time TE = 9.4 ms). One spectrum was acquired without water
suppression (16 averages) and one with water suppression (VAPOR water suppression, 512
averages). 1H MR spectra were fit in the time domain using an advanced magnetic
resonance (AMARES) nonlinear least squares algorithm in the jMRUI software package
(jMRUI V2.1) [31] as described previously [32]. IMCL was expressed as a percentage of the
water signal measured in the same voxel.
PCr recovery rate (kPCr), ATP synthesis flux (VATP) and creatine kinase flux (VCK) were
measured in the TA by 31P MRS to determine in vivo mitochondrial function. 31P MRS was
Chapter 6
128
applied using a combination of a circular 1H surface coil (40 mm) for shimming and an
ellipsoid 31P MRS surface coil (10/18 mm), positioned over the TA, as described previously
[32‐33]. 31P MR spectra were acquired applying an adiabatic excitation pulse with a flip
angle of 90 [34]. A fully relaxed (TR = 25 s, 48 averages) spectrum was measured at rest,
followed by the saturation transfer (ST) experiment in resting TA muscle to determine VATP
and VCK. Three spectra (TR = 10.4 s) were acquired for the ST experiments: A spectrum with
frequency selective saturation of the ‐ATP peak yielding the steady state Pi and PCr magnetizations in the presence of saturation ( and , respectively, 2*64 averages), a
reference spectrum with saturation at a downfield frequency, equidistant from Pi, yielding
the equilibrium Pi magnetization ( , , 128 averages), and a reference spectrum with
saturation at a downfield frequency, equidistant from PCr, yielding the equilibrium PCr
magnetization ( , , 64 averages). For all three experiments the saturation pulse length
was set to 10 s. The apparent longitudinal relaxation times of Pi ( , ) and PCr ( , ) were
determined by performing a 7‐point inversion recovery experiment with an adiabatic full
passage pulse for inversion and with ‐ATP saturation prior to (10 s) and during the inversion delay (inversion times = 0.01, 1, 2, 4, 6.5, 10.5 and 17 s, 32 averages). The total
duration of the ST and inversion recovery experiments was about 2 hours.
After the ST experiments, time series of 31P MR spectra (TR = 5 s, 4 averages) before, during
and after muscle contractions were acquired. Muscle contractions were induced by
electrical stimulation of the TA, via subcutaneously implanted electrodes positioned along
the distal N. Peroneus Communis [33]. The stimulation protocol consisted of a series of
stimulation pulses, applied every second, for a duration of 2 min. Stimulation pulse length
was 100 ms, frequency was 80 Hz and stimulation voltage varied between 2.5 and 4 Volt, to
reach similar levels of PCr depletion. Recovery was followed for 10 minutes. Three time
series were measured for each rat.
31P MR spectra were fit in the time domain by using an advanced magnetic resonance
(AMARES) nonlinear least squares algorithm in the jMRUI software package (jMRUI V2.1)
[31]. PCr and Pi peaks were fit to Lorentzian line shapes, whereas ‐, ‐, and ‐ATP signals were fit with Gaussian line shapes. Besides the cytosolic Pi signal, a second, smaller Pi peak
was observed in the spectra at a frequency ~0.3 ppm downfield from the cytosolic Pi
resonance. The two Pi signals were separately fitted and the line widths were constrained
with respect to the line width of the PCr signal. In the dynamic MRS spectra, the two Pi
signals could not be distinguished and a single Pi peak was fit to a Gaussian line shape.
Concentrations of PCr and Pi were determined relative to the ATP concentration, which was
assumed to be 11.6 mM (~7.8 μmol/g wet wt) in resting TA muscle, as was determined
previously (Chapter 2). ADP concentrations were calculated using the creatine kinase
equilibrium [35] and intracellular pH was calculated from the chemical shift difference
between PCr and the cytosolic Pi resonance [36].
Carnitine insufficiency and mitochondrial function
129
The apparent T1 ( ) relaxation times of Pi and PCr were determined by fitting the inversion
recovery data with a 3‐parameter mono‐exponential function using Origin (OriginPro 7.5
SR0, OriginLab Corporation, Northampton, MA, USA). The Pi ATP exchange rate constant
( ) was calculated from , and the fractional reduction of Pi magnetization upon
selective saturation of ‐ATP according to:
1 / ,
,
The resting ATP synthesis flux (VATP) was then calculated by multiplying with the Pi
concentration at rest, as determined from the fully relaxed spectrum. The PCr ATP
exchange rate constant was calculated according to:
1 / ,
,
The creatine kinase flux (VCK) was calculated by multiplying with the resting PCr
concentration.
The data of PCr recovery were fit to a mono‐exponential function using Matlab (version
7.04, Mathworks, Natick, MA, USA) yielding a rate constant, kPCr. Results from two time
series with end‐stimulation pH values higher than 6.92 [37] were averaged.
Isolation of mitochondria
Mitochondria were isolated from one whole TA muscle through a differential centrifugation
procedure as described previously [38]. Briefly, TA muscle was excised, washed in ice cold
0.9% KCl, freed of connective and adipose tissue, weighed and minced with scissors in ice
cold medium A (5 ml for 1 g tissue) containing 150 mM sucrose, 75 mM KCl, 50 mM MOPS,
1 mM KH2PO4, 5 mM MgCl2, 1 mM EGTA, 0.4 mg/ml bacterial proteinase type XXIV, pH 7.4.
Next, 20 ml of medium B containing 250 mM sucrose, 0.1 mM EGTA and 20 mM MOPS, 2
mg/ml BSA, pH 7.4 was added and the mixture was homogenized using a Potter‐Elvehjem
homogenizer. The homogenate was centrifuged at 800 g for 10 min, 4C. The resulting supernatant was centrifuged at 10000 g for 10 min, 4C. The pellet was resuspended in 15 ml of fresh ice cold medium B and centrifuged at 10000 g for 10 min, 4C. Mitochondrial
pellet was resuspended in 100 l of medium B. Protein content was determined using the
BCA protein assay kit (Pierce, Thermo Fisher Scientific Inc., Rockford, IL, USA).
Chapter 6
130
Measurement of oxygen consumption
Oxygen consumption rate was measured at 37°C using a two‐channel high‐resolution
Oroboros oxygraph‐2 k (Oroboros, Innsbruck, Austria). Mitochondria were incubated in the
assay medium containing 110 mM KCl, 20 mM Tris, 2.3 mM MgCl2, 5 mM KH2PO4 and 1
mg/ml BSA, pH 7.3. All measurements were performed in 1 ml of assay medium containing
0.15 mg/ml of mitochondrial protein. Three different combinations of substrates were used
to asses mitochondrial capacity to oxidize tricarboxylic acid cycle (TCA) and ‐oxidation substrates: (i) 5 mM pyruvate plus 5 mM malate (TCA cycle), (ii) 25 M palmitoyl‐L‐
carnitine plus 2.5 mM malate (‐oxidation and TCA cycle), and (iii) 25 M palmitoyl‐CoA
plus 2.5 mM L‐carnitine plus 2.5 mM malate (carnitine palmitoyltransferase 1 (CPT1), ‐oxidation and TCA cycle). An ADP‐regenerating system consisting of excess hexokinase (4.8
U/ml) and glucose (12.5 mM) was used to maintain steady‐state oxygen consumption rates.
Maximal ADP‐stimulated oxygen consumption rate (state 3) was initiated by addition of 1
mM of ATP. Maximal oxygen consumption rate in the uncoupled state (state U) was
determined after addition of 1 M carbonyl cyanide 3‐chlorophenyl hydrazone (CCCP).
Basal oxygen consumption rate in the absence of ATP synthesis (state 4), which reflects
basal proton leak rate across the inner mitochondrial membrane, was measured after fully
blocking ATP synthesis with 1.25 M carboxyatractyloside. The sensitivity of the basal
proton leak rate to fatty acids, which reflects activation of the uncoupling proteins (UCPs)
[39], was determined by measuring stimulation of oxygen uptake rate in state 4 after
addition of 90 µM of palmitic acid (C16:0). The signals from the oxygen electrode were
recorded at 0.5 s intervals. Data acquisition and analysis was performed using Oxygraph‐2k‐
DatLab software version 4.2 (Oroboros, Innsbruck, Austria).
Determination of the relative mitochondrial copy number
Relative mitochondrial copy number was measured as described previously [8]. Briefly,
Genomic DNA was isolated from a 25 mg transversal slice of mid‐belly TA using GenElute™
Mammalian Genomic DNA Miniprep Kit (Sigma‐Aldrich, Zwijndrecht, The Netherlands).
Mitochondrial DNA (mtDNA) content relative to peroxisome proliferator‐activated
receptor‐ coactivator 1 (PGC‐1) gene was measured using real‐time PCR as described in
[40]. Primers for mtDNA were: forward primer – 5’‐ACACCAAAAGGACGAACCTG‐3’, reverse
primer ‐ 5’‐ATGGGGAAGAAGCCCTAGAA‐3’, and for PGC‐1: forward primer – 5’‐
ATGAATGCAGCGGTCTTAGC‐3’, reverse primer – 5’‐AACAATGGCAGGGTTTGTTC‐3’. The
relative mtDNA copy number was calculated using the Ct method as described in [41].
Carnitine insufficiency and mitochondrial function
131
Statistical analysis
Data are presented as means ± SD. The listed n values represent the number of animals
used for a particular experiment. Statistical significance of the differences was assessed by
applying one way Analysis of Variance (ANOVA) using Tukey HSD post hoc analysis in the
SPSS 16.0 statistical package (SPSS Inc., Chicago, IL, USA). The level of statistical significance
was set at p < 0.05.
Results
Animal model
Animal characteristics are summarized in Table 1. Body weight was significantly lower for
the HFD group compared to the HFDC group at the start of the experiment. After 15 wk of
diet, body weight was similar in all groups, whereas body weight gain was significantly
higher in HFD animals compared to NC (p = 0.047) and HFDC (p = 0.02) animals. Average
energy intake did not differ between groups (Table 1).
Fasting plasma insulin and blood glucose levels during the OGTT are presented in Table 1.
After 15 wk of diet, animals had similar fasting blood glucose and plasma insulin levels. The
AUCg from the OGTT was significantly higher for both HFD (p = 0.02) and HFDC (p = 0.003)
groups when compared to controls, whereas the AUCi was not different between the three
groups.
Table 1. Animal characteristics
NC HFD HFDC
Body weight (g) t = 0 349 ± 18 338 ± 18 359 ± 12
†
Body weight (g) t = 15 538 ± 23 559 ± 32 543 ± 25
Delta body weight (g) 189 ± 22 221 ± 38* 185 ± 23†
Food intake (kJ/week) 2239 ± 74 2342 ± 156 2286 ± 187
Fasting glucose (mM) 5.4 ± 0.7 5.6 ± 0.7 5.8 ± 0.5
AUCg (mM∙h)
11.3 ± 0.6 12.4 ± 1.2* 12.7 ± 0.7*
Fasting insulin (pM)
546 ± 222 561 ± 207 521 ± 160
AUCi (pM∙h) 1526 ± 359 1451 ± 370 1534 ± 465
Data are from n = 10 NC, n = 9 HFD and n = 10 HFDC animals after 15 wk of diet and are expressed as
means ± SD. NC, normal chow; HFD, high‐fat diet; HFDC, high‐fat diet supplemented with carnitine;
AUCg and AUCi, area under the glucose and insulin curve, respectively; *p < 0.05 when compared to NC, †p < 0.05 when compared to HFD.
Chapter 6
132
IMCL content
In vivo 1H MRS was applied to evaluate the effect of 15 wk of high‐fat diet and 15 wk of
high‐fat diet in combination with 8 wk of carnitine supplementation on IMCL levels in TA
muscle. As is shown in Figure 1, IMCL levels were two‐fold higher in HFD (p < 0.001) and
HFDC (p < 0.005) rats as compared to controls. Eight wk of carnitine supplementation had
no effect on IMCL content.
Acylcarnitine content
To verify impairment of lipid metabolism, we determined the content of different
acylcarnitine species in TA muscle, blood and urine. Free carnitine and acetylcarnitine levels
and levels of short‐, medium‐ and long‐chain acylcarnitines are represented in Figure 2.
After 15 wk of diet, free carnitine (C0) levels were decreased in muscle (p = 0.003) and
blood (p = 0.001) of HFD animals. Carnitine treatment in rats fed a high‐fat diet restored
muscle free carnitine levels and led to a 1.5 fold higher blood free carnitine level as
compared to NC animals. Muscle acetylcarnitine (C2) levels were not significantly different
between the three groups, whereas acetylcarnitine levels were two‐fold higher in blood of
HFDC animals compared to HFD animals (p < 0.001). Carnitine treatment resulted in
massive increases in urine free carnitine (40‐fold), acetylcarnitine (25‐fold) and
short/medium‐chain acylcarnitine levels. However, no changes in long‐chain acylcarnitines
in urine were observed. Muscle medium‐ and long‐chain acylcarnitines were in almost all
cases significantly elevated in HFD and HFDC groups compared to NC, but similar for HFD
and HFDC groups. For most medium‐ and long‐chain acylcarnitines in blood, a significant
decrease was observed as a result of 15 wk of high‐fat diet, whereas levels of these
medium‐ and long‐chain acylcarnitines were restored or increased in HFDC rats compared
to controls.
Figure 1. IMCL content in the dorsal
part of the TA muscle of NC (n = 10),
HFD (n = 9) and HFDC (n = 10) rats. IMCL
content was expressed as a percentage
of the water signal. *p < 0.05 when
compared to NC.
Carnitine insufficiency and mitochondrial function
133
Mitochondrial function in vivo
31P MRS was used to assess in vivo muscle mitochondrial function. Concentrations of
metabolites and pH obtained from the 31P MR spectra in resting TA muscle are summarized
in Table 2. These parameters were not affected by the high‐fat diet feeding alone or by the
high‐fat diet in combination with carnitine supplementation.
From the 31P ST spectra, the ratios of Pi and PCr magnetization with and without selective
saturation of ‐ATP ( / , and / , , respectively) were determined and these
ratios were not significantly different between groups for both Pi and PCr (Table 3).
relaxation times of Pi and PCr in case of saturation of ‐ATP were determined from a 7‐
point inversion recovery experiment. Although the relaxation times of Pi and PCr seem
to be decreased in HFD and HFDC animals, there were no significant differences between
the three groups (Table 3). The exchange rate constant and ATP synthesis flux, VATP,
Figure 2. Free carnitine and acetylcarnitine content in TA muscle (A), blood (D) and urine (G), short‐ and
medium‐chain acylcarnitine content in TA muscle (B), blood (E) and urine (H) and long‐chain acylcarnitine
content in TA muscle (C), blood (F) and urine (I) of NC (n = 10), HFD (n = 9) and HFDC (n = 10) rats. *p <
0.05 when compared to NC, †p < 0.05 when compared to HFD
Chapter 6
134
were also not significantly different between NC, HFD and HFDC groups (Table 3).
Interestingly, and the CK flux, VCK, were significantly higher in both HFD (p = 0.01
and 0.02 for and VCK, respectively) and HFDC (p = 0.02 and 0.03 for and
VCK, respectively) groups compared to controls.
Dynamic 31P MRS measurements during and after recovery from electrical stimulation of
the TA were performed to determine the rate constant of PCr recovery after contraction.
Concentrations of PCr, Pi and ADP and the intracellular pH at the end of stimulation were
not significantly different between groups (Table 2). PCr concentrations during recovery
Table 2. Metabolite concentrations and pH in TA measured by in vivo 31P MRS.
Rest NC HFD HFDC
pH 7.18 ± 0.05 7.18 ± 0.03 7.18 ± 0.02
[PCr] (mM)
45.8 ± 3.7 46.2 ± 2.5 46.3 ± 3.3
[Pi] (mM) 4.0 ± 1.0 4.0 ± 1.1 3.4 ± 0.5
[ADP] (μM) 13.1 ± 1.3 13.2 ± 0.8 13.1 ± 0.7
End stimulation NC HFD HFDC
pH 7.04 ± 0.05 7.05 ± 0.04 7.07 ± 0.04
[PCr] (mM)
20.6 ± 4.4 23.9 ± 3.0 23.6 ± 1.6
[Pi] (mM) 28.0 ± 5.0 25.0 ± 2.5 24.8 ± 1.4
[ADP] (μM) 92.5 ± 32.7 73.4 ± 14.3 76.8 ± 12.0
Data are from n = 10 NC, n = 9 HFD and n = 10 HFDC animals and are expressed as means ± SD. NC,
normal chow; HFD, high‐fat diet; HFDC, high‐fat diet supplemented with carnitine; PCr,
phosphocreatine; Pi, inorganic phosphate; ADP, adenosine diphosphate. No significant differences
were observed.
Figure 3. The rate constant of PCr recovery (kPCr) in TA after electrical stimulation (A) and relative
mitochondrial DNA copy number (B) measured in TA muscle of NC (n = 10), HFD (n = 9) and HFDC (n =
10) rats. *p < 0.05 when compared to NC, †p < 0.05 when compared to HFD.
Carnitine insufficiency and mitochondrial function
135
were fitted with a mono‐exponential function yielding kPCr. kPCr was similar for the NC, HFD
and HFDC animals (Figure 3a).
Mitochondrial function in vitro
Table 4 summarizes the effects of 15 wk of high‐fat diet feeding and 15 wk of high‐fat diet
feeding in combination with 8 wk of carnitine supplementation on intrinsic mitochondrial
function in vitro, represented by oxygen consumption rates in isolated TA mitochondria
oxidizing glucose‐ or fat‐derived substrates in different metabolic states. Both high‐fat diet
feeding and high‐fat diet feeding in combination with carnitine supplementation had no
significant effect on pyruvate plus malate‐driven oxygen consumption rate in any metabolic
state. Interestingly, a ~15% higher palmitoyl‐L‐carnitine plus malate driven oxygen
consumption rate in state 3 and state U was observed in mitochondria from HFD and HFDC
rats compared to NC. A similar effect was observed when palmitoyl‐CoA plus L‐carnitine
plus malate was used as the oxidizible substrate. Oxygen consumption rate in state 4 was
similar in all three groups with both palmitoyl‐L‐carnitine plus malate and palmitoyl‐CoA
plus L‐carnitine plus malate as the oxidizible substrate.
Table 3. Parameters determined from 31P saturation transfer MRS
NC HFD HFDC
/ , 0.76 ± 0.09 0.79 ± 0.05 0.77 ± 0.05
, (s) 4.25 ± 0.38 3.82 ± 0.74 3.62 ± 0.25
(s‐1) 3.24 ± 1.31 3.35 ± 0.81 3.84 ± 0.84
VATP (μmol/g/min) 9.22 ± 5.53 9.20 ± 4.64 9.16 ± 2.55
NC HFD HFDC
/ , 0.58 ± 0.03 0.54 ± 0.03 0.55 ± 0.02
, (s) 2.09 ± 0.10 2.00 ± 0.05 2.00 ± 0.10
(s‐1) 12.2 ± 1.1 13.5 ± 1.0* 13.5 ± 0.9*
VCK (μmol/g/min) 372 ± 36 426 ± 42* 423 ± 43*
Data in the upper part of the table are from n = 7 NC, n = 8 HFD and n = 9 HFDC animals. Data in the
lower part of the table are from n = 10 NC, n = 9 HFD and n = 9 HFDC animals. All data are expressed as
means ± SD. NC, normal chow; HFD, high‐fat diet; HFDC, high‐fat diet supplemented with carnitine; ,
magnetization of Pi and PCr with saturation of ‐ATP; , , magnetization of Pi and PCr with saturation at
a downfield frequency, equidistant from Pi and PCr, respectively; , , apparent longitudinal relaxation
time of Pi and PCr with saturation of ‐ATP; , Pi ATP exchange rate constant; VATP, resting ATP
synthesis flux; , PCr ATP exchange rate constant; VCK, resting creatine kinase flux. *p < 0.05
when compared to NC.
Chapter 6
136
Table 4 also shows the effect of 90 µM of palmitic acid on the oxygen consumption rate in
state 4 in isolated mitochondria respiring on pyruvate plus malate. Addition of palmitic acid
caused a significantly higher increase in the oxygen consumption rate in the HFD group
compared to NC group (p = 0.02), indicating increased FA‐induced mitochondrial
uncoupling. The effect of palmitic acid in the HFDC group was similar to that in the HFD
group, but not significantly different from the effect in the NC group.
The effect of 15 wk of high‐fat diet feeding and 15 wk of high‐fat diet feeding in
combination with 8 wk of carnitine supplementation on mitochondrial biogenesis was
assessed by determining mtDNA copy number in TA muscle. The relative mtDNA copy
number was significantly increased in HFD (p = 0.03) and HFDC (p = 0.04) animals as
compared to NC animals, whereas no difference was observed between HFD and HFDC
animals (Figure 3b).
Table 4. Oxygen consumption rates in isolated TA mitochondria oxidizing different substrates in different
metabolic states.
Pyruvate plus malate NC HFD HFDC
State 3 (nmol O2min‐1mg protein
‐1) 543 ± 69 564 ± 42 567 ± 74
State 4 (nmol O2min‐1mg protein
‐1) 36 ± 8 34 ± 3 38 ± 5
State U (nmol O2min‐1mg protein
‐1) 649 ± 76 669 ± 49 684 ± 85
Pyruvate plus malate plus palmitic acid NC HFD HFDC
State 4 (nmol O2min‐1mg protein
‐1) 204 ± 29 257 ± 22* 237 ± 56
Palmitoylcarnitine plus malate NC HFD HFDC
State 3 (nmol O2min‐1mg protein
‐1) 146 ± 18 171 ± 20* 166 ± 20
State 4 (nmol O2min‐1mg protein
‐1) 31 ± 5 31 ± 2 32 ± 3
State U (nmol O2min‐1mg protein
‐1) 223 ± 27 251 ± 22 254 ± 29*
PalmitoylCoA plus L‐carnitine plus malate NC HFD HFDC
State 3 (nmol O2min‐1mg protein
‐1) 140 ± 20 167 ± 14* 167 ± 26*
State 4 (nmol O2min‐1mg protein
‐1) 31 ± 5 31 ± 2 31 ± 5
State U (nmol O2min‐1mg protein
‐1) 230 ± 29 265 ± 15* 263 ± 31*
Data are from n = 10 NC, n = 9 HFD and n = 10 HFDC animals and are expressed as means ± SD. NC,
normal chow; HFD, high‐fat diet; HFDC, high‐fat diet supplemented with carnitine; State 3, maximal ADP‐
stimulated oxygen consumption; State 4, oxygen consumption in the absence of ATP synthesis; State U,
oxygen consumption after uncoupling. *p < 0.05 when compared to NC.
Carnitine insufficiency and mitochondrial function
137
Discussion
It is generally accepted that muscle lipid overload leads to insulin resistance. Although it
has been hypothesized that mitochondrial dysfunction and an associated decreased
capacity to oxidize FAs are at the basis of this derangement [3‐4], recently it has been
shown that high‐fat diets lead to increased rather than decreased FA oxidation, resulting in
the accumulation of lipid intermediates which can cause metabolic stress and induce
mitochondrial dysfunction [8‐11]. Carnitine supplementation might be effective in reducing
these toxic lipid intermediates by increasing complete FA oxidation and export of lipid
metabolites out of the mitochondria and into the plasma. Therefore, to investigate
whether carnitine supplementation can reduce high‐fat diet‐induced lipotoxicity, we
determined muscle lipid status, in vivo and in vitro muscle mitochondrial function, and
insulin resistance in rats which received a 15 wk high‐fat diet and in rats which received a
15 wk high‐fat diet in combination with 8 wk carnitine supplementation. In agreement with
previous results, high‐fat diet feeding resulted in a decrease in muscle free carnitine [8, 11],
increased IMCL [8] and muscle, blood and urine acylcarnitine levels [8, 10‐11], an increased
capacity to oxidize fat‐derived substrates in vitro [5‐6, 8, 42], an increased number of
muscle mitochondria [5‐6, 8, 42], but a normal muscle oxidative capacity in vivo [8],
indicating that in vivo mitochondrial function was compromised, possibly by toxic lipid
intermediates. Despite the restoration of muscle free carnitine content, carnitine
supplementation did not induce improvements in muscle lipid status, in vivo mitochondrial
function or insulin resistance. These results indicate that the decrease in muscle free
carnitine as a result of high‐fat diet feeding did not contribute to the observed impairment
in in vivo muscle mitochondrial function.
One of the leading hypotheses in the research field of type 2 diabetes is that inherited or
acquired skeletal muscle mitochondrial dysfunction, associated with a reduced
mitochondrial capacity to oxidize FAs, leads to a lipid overload in muscle cells, causing
insulin resistance [3‐4]. Recent studies have challenged this theory and link insulin
resistance to an increased capacity to oxidize FAs [5‐8]. However, it has been shown in
myotubes and rodent models that high rates of ‐oxidation in insulin‐resistant states are associated with low rates of complete FA oxidation and elevated incomplete ‐oxidation [9‐11]. High rates of incomplete ‐oxidation occur when carbon flux through the ‐oxidation machinery outpaces entry of acetyl‐CoA into the tricarboxylic acid cycle. It has been
hypothesized that the accumulation of incompletely metabolized FAs in the mitochondria
causes ‘mitochondrial stress’ leading to mitochondrial dysfunction and insulin resistance
[10, 12]. In the two contradictory hypotheses mitochondrial dysfunction is put forward
either as a cause or a consequence of excessive accumulation of lipids and lipid
intermediates in muscle cells [9‐11]. The results of our in vitro measurements, showing up‐
regulation of FA oxidation in vitro, are in line with the latter hypothesis. In previous studies
equivalent increases in in vitro mitochondrial FA oxidative capacity have been observed in
skeletal muscle of high‐fat diet fed rats [6, 42‐47]. Our results are also in agreement with
Chapter 6
138
other studies which have shown that high‐fat feeding does not affect respiration of rat
muscle mitochondria when using substrates other than fatty acids, i.e. pyruvate/glutamate
plus malate and succinate [6, 46‐47], implying that the intrinsic functioning of the
mitochondria is not impaired in vitro. The observed 25% increase in mitochondrial copy
number in HFD animals is in agreement with our previous study, in which it was shown that
long‐term high‐fat diet feeding causes an increase in peroxisome proliferator‐activated
receptor‐ coactivator 1 (PGC1) expression with a concomitant increase in mitochondrial
copy number and citrate synthase activity compared to controls [8]. Furthermore, a
number of other studies has likewise shown that high‐fat diet feeding in rats induces
increased biogenesis of mitochondria [5‐6, 42‐45] as an adaptive response to the higher
mitochondrial FA load.
From the observed upregulation of in vitro muscle mitochondrial function and content, one
would expect a concomitant improvement of in vivo muscle mitochondrial function. In vivo
resting ATP synthesis flux (VATP) was not changed in HFD rats compared to NC rats, which is
in contrast with data from a previous study in high‐fat diet fed rats. In the latter study, VATP
decreased by about 40% after 5 wk of high‐fat diet, which was explained as a failure of the
mitochondria to compensate for the relatively long duration of the high‐fat diet [48]. As far
as we know, no VATP data have been presented after longer term high‐fat diet feeding. The
interpretation of resting VATP data as a measure for mitochondrial function is, however, not
straightforward. Mitochondrial ATP synthesis is a demand driven process and a decreased
VATP under resting conditions could simply reflect a normal regulatory response to a lower
energy demand [49‐52]. Moreover, the Pi ATP fluxes obtained from 31P ST
measurements are comprised of both ATP synthesis flux and glycolytic exchange flux, with
the latter contributing by as much as 80% at rest [50, 53‐56]. Besides VATP, we also
determined VCK from the 31P ST experiments and we found that VCK was increased by 15% in
HFD rats compared to controls. However, the physiological relevance of the observed
increase in VCK in HFD animals is not entirely clear. In a previous study, an increase in VCK in
skeletal muscle of high‐fat diet fed rats compared to control rats was explained as a
compensatory mechanism in response to impaired mitochondrial function [57]. However,
this seems unlikely, because in normal conditions the creatine kinase reaction is already
much faster than oxidative ATP production [58]. Alternatively, the 15% higher VCK in HFD
rats can point towards a shift in muscle fiber type towards more oxidative fibers [59].
We have shown in a previous study that the rate of PCr recovery after exercise has a more
direct relationship with skeletal muscle mitochondrial function than resting VATP (Chapter
2). The PCr recovery rate constant (kPCr) was equal for HFD and NC rats, despite the 25%
increase in the number of mitochondria, which is in agreement with our previous results in
long‐term high‐fat diet fed rats [8]. These data indicate that an increased number of
mitochondria with normal or even improved function in vitro is required to maintain
normal oxidative capacity in vivo in HFD rats. The observed dissociation between in vivo
and in vitro muscle oxidative capacity suggests that in vivo mitochondrial function in muscle
Carnitine insufficiency and mitochondrial function
139
of HFD rats is compromised by factors not taken into account when mitochondrial function
is assessed under in vitro conditions.
Despite a similar body weight, HFD rats had 90% more IMCL compared to NC animals. The
acylcarnitine profile likewise showed that muscle of HFD rats was overloaded with lipids.
The vast majority of acylcarnitines is produced in the mitochondria, and acylcarnitine levels
can therefore, in combination with measurements of substrate oxidation and mitochondrial
function, be interpreted as a measure for (incomplete) β‐oxidation. In agreement with
previous studies, even, medium‐chain (C6‐C12) acylcarnitine intermediates, which
represent incompletely oxidized FAs, were elevated in muscle from HFD rats compared to
NC animals [9‐11]. Although the measurement of acylcarnitines provides a comprehensive
snapshot of intermediary metabolism, it is important to emphasize that steady‐state
metabolite concentrations represent the net balance between production, consumption,
import and export. However, from the combination of increased in vitro fat oxidation rates
and increased levels of medium‐chain acylcarnitines in muscle of HFD animals, it can be
speculated that incomplete ‐oxidation is elevated in HFD animals.
It has been shown that elevated levels of FAs may impair mitochondrial ATP production
through a number of mechanisms. E.g. long‐chain acyl‐CoA esters may inhibit the
mitochondrial adenine nucleotide translocator leading to impaired exchange of cytosolic
ADP for mitochondrial ATP [60]. Moreover, increased availability of FAs may lead to
increased expression [6, 8] as well as activation of UCP3 [39], which in turn may lead to
diminished efficiency of ATP synthesis due to increased mitochondrial uncoupling. In order
to test whether the latter mechanism could have been the cause for the observed in vivo
mitochondrial dysfunction, we measured state 4 oxygen consumption rates in isolated
mitochondria respiring on pyruvate plus malate in the presence of palmitic acid. In this
experiment, we observed significantly higher oxygen consumption rates in mitochondria
isolated from HFD rats compared to NC rats, indicating increased mitochondrial uncoupling
due to the presence of FAs. Together with the finding that muscle medium‐ and long‐chain
acylcarnitines were increased in HFD rats compared to controls, these data strongly suggest
that FA‐induced mitochondrial uncoupling may contribute to the observed mitochondrial
functional impairment in HFD rats in vivo.
Concomitantly with increased muscle acylcarnitine levels, a decrease was observed in free
carnitine levels in HFD animals. It has been shown that whole‐body carnitine insufficiency is
a common feature in insulin resistant states such as advanced age, genetic diabetes and
diet‐induced obesity [11]. This can be explained by a decreased biosynthesis in the liver
[10‐11, 61‐62], but also by increased sequestration of carnitine in the muscle acylcarnitine
pool [10‐11]. The insulin resistance related decline in free carnitine has been associated
with impaired mitochondrial function and an imbalance between complete and incomplete
fat oxidation [11, 17]. Most direct evidence linking carnitine insufficiency to mitochondrial
dysfunction comes from juvenile visceral steatosis (jvs) mice, a genetic model of carnitine
Chapter 6
140
deficiency, showing marked derangements in mitochondrial biogenesis and morphology, as
well as whole‐body lipid and energy metabolism [63‐64]. It has been hypothesized that
carnitine supplementation ameliorates mitochondrial function and insulin resistance by
reducing lipotoxicity both through the increased oxidation and increased export of muscle
lipid metabolites [17]. In a recent study by Noland et al., it was shown that 8 wk of carnitine
supplementation in long‐term high‐fat diet fed rats restored the ratio of complete to
incomplete fat oxidation and increased efflux of muscle acylcarnitine intermediates, while
improving glucose tolerance [11].
To investigate the effects of carnitine supplementation on muscle lipid status, muscle
mitochondrial function and insulin sensitivity in high‐fat diet fed rats, we supplemented
them with 300 mg/kg body weight/d carnitine during the last 8 wk of the 15 wk diet. While
carnitine supplementation restored muscle free carnitine to the value found in control
animals fed with normal chow, no differences in in vivo IMCL levels were observed between
HFD and HFDC rats. Similar levels of muscle acylcarnitines in HFD and HFDC rats provided
further evidence of unchanged lipid levels inside the muscle, despite the increased levels of
acylcarnitines in blood and urine in the carnitine supplemented group. The effects of
carnitine supplementation on acylcarnitine profiles in muscle, blood and urine are in
agreement with the results of Noland et al. [11]. In this study it was concluded that
carnitine supplementation facilitates the efflux and excretion of acylcarnitines, possibly
leading to a decrease in tissue long‐chain acyl‐CoAs, but currently there is no proof for this
theory.
In line with the lack of effects on muscle lipid status, carnitine supplementation also did not
improve muscle mitochondrial function or insulin sensitivity in high‐fat diet fed rats.
Oxygen consumption rates of muscle mitochondria respiring on glucose‐ or fat‐derived
substrates were similar for HFD and HFDC rats. Addition of palmitic acid had the same
effect on mitochondria from HFD and HFDC rats, both showing increased FA‐induced
mitochondrial uncoupling compared to mitochondria from NC animals. Both in HFD and
HFDC rats muscle mitochondrial content was increased by about 25% relatively to controls,
but in both groups this did not result in an increased muscle oxidative capacity in vivo as
determined from the rate of PCr recovery, indicating that in both groups in vivo
mitochondrial function was similarly impaired. Our results on muscle mitochondrial
function and insulin sensitivity are in contrast with the findings of Noland et al., who
showed that exactly the same regimen of carnitine supplementation (i.e. 300 mg/kg body
weight/d during 8 wk) improved whole‐body glucose tolerance and reversed mitochondrial
abnormalities in rats fed with a high‐fat diet [11]. A difference between the two studies is
the duration of the high‐fat diet, i.e. 15 wk in this study versus 12 months in the study by
Noland et al. However, the 15 wk and 12 month high‐fat diet feeding led to similar
decreases in muscle free carnitine and comparable increases in muscle acylcarnitines.
Another difference between the two studies resides in the methods to assess muscle
mitochondrial function. Noland et al. measured the ratio of complete to incomplete oleate
Carnitine insufficiency and mitochondrial function
141
oxidation in isolated mitochondria, which was decreased in high‐fat diet fed rats, but
completely restored after carnitine supplementation [11]. In the current study we have no
data on the ratio of complete to incomplete fat oxidation, but we have shown that
carnitine supplementation does not improve in vivo muscle mitochondrial function.
Reduction of the free carnitine pool in response to HFD feeding might have a negative
effect on mitochondrial fatty oxidation only if free carnitine is depleted to an extent that it
becomes limiting for CPT1 activity, resulting in a decreased entry of long‐chain acyl‐CoAs
into the mitochondrial matrix. The absence of an effect of 8 wk carnitine treatment in high‐
fat diet fed animals regarding muscle mitochondrial function suggests that this was not the
case in the present study. Moreover, our observation that state 3 oxygen consumption
rates with palmitoyl‐CoA plus L‐carnitine plus malate and palmitoyl‐L‐carnitine plus malate
as the oxidizible substrates were similar in HFD and HFDC groups suggests that neither
palmitoyl‐CoA transport nor oxidation capacity was affected by carnitine treatment.
Currently it is not known whether it is sufficient to correct an incipient carnitine deficiency
status or whether supra‐physiological carnitine levels in both blood and muscle are
necessary to obtain clinically beneficial effects of carnitine supplementation. Due to the
pharmacokinetic properties of carnitine, including a poor gut absorption and bioavailability,
a high renal clearance rate and uptake into tissues via high affinity transporters, it is
difficult to achieve supra‐physiological levels of free carnitine in the muscle.
In conclusion, we showed that high‐fat diet feeding induces carnitine insufficiency and
muscle lipid overload, leading to an increased number of muscle mitochondria with an
improved capacity to oxidize FAs in vitro, but without a concomitant increase in muscle
oxidative capacity in vivo. These results indicate that in vivo mitochondrial function was
compromised in muscle of high‐fat diet fed rats. We provided evidence that the in vivo
mitochondrial dysfunction in high‐fat diet fed rats was caused by elevated levels of lipid
intermediates, leading to increased FA‐induced mitochondrial uncoupling and therefore
less efficient ATP synthesis. Despite the restoration of muscle free carnitine content,
carnitine supplementation did not induce improvements in muscle lipid status, in vivo
mitochondrial function or insulin resistance. These results indicate that the decrease in
muscle free carnitine as a result of high‐fat diet feeding did not contribute to the observed
impairment in in vivo muscle mitochondrial function.
Acknowledgements
We are grateful to sigma‐tau (Utrecht, the Netherlands) for providing L‐carnitine. The work
of J.C. is financed by the Netherlands Consortium for Systems Biology (NCSB) which is part
of the Netherlands Genomics Initiative / Netherlands Organisation for Scientific Research.
J.J.P. and S.M.H. are supported by VIDI grants from the Netherlands Organisation for
Scientific Research (VIDI grant numbers 700.58.421 and 016.086.336, respectively).
Chapter 6
142
References 1. Wild, S., G. Roglic, A. Green, R. Sicree, and H. King. Global prevalence of diabetes: estimates for the year
2000 and projections for 2030. Diabetes Care, 2004. 27(5): 1047‐53. 2. DeFronzo, R.A. Pathogenesis of type 2 (non‐insulin dependent) diabetes mellitus: a balanced overview.
Diabetologia, 1992. 35(4): 389‐97. 3. Lowell, B.B. and G.I. Shulman. Mitochondrial dysfunction and type 2 diabetes. Science, 2005. 307(5708):
384‐7. 4. Morino, K., K.F. Petersen, and G.I. Shulman. Molecular mechanisms of insulin resistance in humans and
their potential links with mitochondrial dysfunction. Diabetes, 2006. 55 Suppl 2: S9‐S15. 5. Hancock, C.R., D.H. Han, M. Chen, S. Terada, T. Yasuda, D.C. Wright, and J.O. Holloszy. High‐fat diets
cause insulin resistance despite an increase in muscle mitochondria. Proc Natl Acad Sci U S A, 2008. 105(22): 7815‐20.
6. Turner, N., C.R. Bruce, S.M. Beale, K.L. Hoehn, T. So, M.S. Rolph, and G.J. Cooney. Excess lipid availability increases mitochondrial fatty acid oxidative capacity in muscle: evidence against a role for reduced fatty acid oxidation in lipid‐induced insulin resistance in rodents. Diabetes, 2007. 56(8): 2085‐92.
7. Turner, N. and L.K. Heilbronn. Is mitochondrial dysfunction a cause of insulin resistance? Trends Endocrinol Metab, 2008. 19(9): 324‐30.
8. van den Broek, N.M., J. Ciapaite, H.M. De Feyter, S.M. Houten, R.J. Wanders, J.A. Jeneson, K. Nicolay, and J.J. Prompers. Increased mitochondrial content rescues in vivo muscle oxidative capacity in long‐term high‐fat‐diet‐fed rats. FASEB J, 2010. 24(5): 1354‐64.
9. Koves, T.R., P. Li, J. An, T. Akimoto, D. Slentz, O. Ilkayeva, G.L. Dohm, Z. Yan, C.B. Newgard, and D.M. Muoio. Peroxisome proliferator‐activated receptor‐gamma co‐activator 1alpha‐mediated metabolic remodeling of skeletal myocytes mimics exercise training and reverses lipid‐induced mitochondrial inefficiency. J Biol Chem, 2005. 280(39): 33588‐98.
10. Koves, T.R., J.R. Ussher, R.C. Noland, D. Slentz, M. Mosedale, O. Ilkayeva, J. Bain, R. Stevens, J.R. Dyck, C.B. Newgard, G.D. Lopaschuk, and D.M. Muoio. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab, 2008. 7(1): 45‐56.
11. Noland, R.C., T.R. Koves, S.E. Seiler, H. Lum, R.M. Lust, O. Ilkayeva, R.D. Stevens, F.G. Hegardt, and D.M. Muoio. Carnitine insufficiency caused by aging and overnutrition compromises mitochondrial performance and metabolic control. J Biol Chem, 2009. 284(34): 22840‐52.
12. Muoio, D.M. and T.R. Koves. Skeletal muscle adaptation to fatty acid depends on coordinated actions of the PPARs and PGC1 alpha: implications for metabolic disease. Appl Physiol Nutr Metab, 2007. 32(5): 874‐83.
13. Fritz, I.B. and K.T. Yue. Long‐Chain Carnitine Acyltransferase and the Role of Acylcarnitine Derivatives in the Catalytic Increase of Fatty Acid Oxidation Induced by Carnitine. J Lipid Res, 1963. 4: 279‐88.
14. Stephens, F.B., D. Constantin‐Teodosiu, and P.L. Greenhaff. New insights concerning the role of carnitine in the regulation of fuel metabolism in skeletal muscle. J Physiol, 2007. 581(Pt 2): 431‐44.
15. Ventura, F.V., L. Ijlst, J. Ruiter, R. Ofman, C.G. Costa, C. Jakobs, M. Duran, I. Tavares de Almeida, L.L. Bieber, and R.J. Wanders. Carnitine palmitoyltransferase II specificity towards beta‐oxidation intermediates‐‐evidence for a reverse carnitine cycle in mitochondria. Eur J Biochem, 1998. 253(3): 614‐8.
16. Uziel, G., B. Garavaglia, and S. Di Donato. Carnitine stimulation of pyruvate dehydrogenase complex (PDHC) in isolated human skeletal muscle mitochondria. Muscle Nerve, 1988. 11(7): 720‐4.
17. Mynatt, R.L. Carnitine and type 2 diabetes. Diabetes Metab Res Rev, 2009. 25 Suppl 1: S45‐9. 18. Ferrannini, E., G. Buzzigoli, S. Bevilacqua, C. Boni, D. Del Chiaro, M. Oleggini, L. Brandi, and F. Maccari.
Interaction of carnitine with insulin‐stimulated glucose metabolism in humans. Am J Physiol, 1988. 255(6 Pt 1): E946‐52.
19. Capaldo, B., R. Napoli, P. Di Bonito, G. Albano, and L. Sacca. Carnitine improves peripheral glucose disposal in non‐insulin‐dependent diabetic patients. Diabetes Res Clin Pract, 1991. 14(3): 191‐5.
20. De Gaetano, A., G. Mingrone, M. Castagneto, and M. Calvani. Carnitine increases glucose disposal in humans. J Am Coll Nutr, 1999. 18(4): 289‐95.
21. Giancaterini, A., A. De Gaetano, G. Mingrone, D. Gniuli, E. Liverani, E. Capristo, and A.V. Greco. Acetyl‐L‐carnitine infusion increases glucose disposal in type 2 diabetic patients. Metabolism, 2000. 49(6): 704‐8.
22. Mingrone, G., A.V. Greco, E. Capristo, G. Benedetti, A. Giancaterini, A. De Gaetano, and G. Gasbarrini. L‐carnitine improves glucose disposal in type 2 diabetic patients. J Am Coll Nutr, 1999. 18(1): 77‐82.
23. Power, R.A., M.W. Hulver, J.Y. Zhang, J. Dubois, R.M. Marchand, O. Ilkayeva, D.M. Muoio, and R.L. Mynatt. Carnitine revisited: potential use as adjunctive treatment in diabetes. Diabetologia, 2007. 50(4): 824‐32.
Carnitine insufficiency and mitochondrial function
143
24. Broderick, T.L., H.A. Quinney, and G.D. Lopaschuk. Carnitine stimulation of glucose oxidation in the fatty acid perfused isolated working rat heart. J Biol Chem, 1992. 267(6): 3758‐63.
25. Wutzke, K.D. and H. Lorenz. The effect of l‐carnitine on fat oxidation, protein turnover, and body composition in slightly overweight subjects. Metabolism, 2004. 53(8): 1002‐6.
26. Muller, D.M., H. Seim, W. Kiess, H. Loster, and T. Richter. Effects of oral L‐carnitine supplementation on in vivo long‐chain fatty acid oxidation in healthy adults. Metabolism, 2002. 51(11): 1389‐91.
27. Arduini, A., M. Bonomini, V. Savica, A. Amato, and V. Zammit. Carnitine in metabolic disease: potential for pharmacological intervention. Pharmacol Ther, 2008. 120(2): 149‐56.
28. van Vlies, N., L. Tian, H. Overmars, A.H. Bootsma, W. Kulik, R.J. Wanders, P.A. Wood, and F.M. Vaz. Characterization of carnitine and fatty acid metabolism in the long‐chain acyl‐CoA dehydrogenase‐deficient mouse. Biochem J, 2005. 387(Pt 1): 185‐93.
29. Chegary, M., H. Brinke, J.P. Ruiter, F.A. Wijburg, M.S. Stoll, P.E. Minkler, M. van Weeghel, H. Schulz, C.L. Hoppel, R.J. Wanders, and S.M. Houten. Mitochondrial long chain fatty acid beta‐oxidation in man and mouse. Biochim Biophys Acta, 2009. 1791(8): 806‐15.
30. Ventura, F.V., C.G. Costa, E.A. Struys, J. Ruiter, P. Allers, L. Ijlst, I. Tavares de Almeida, M. Duran, C. Jakobs, and R.J. Wanders. Quantitative acylcarnitine profiling in fibroblasts using [U‐13C] palmitic acid: an improved tool for the diagnosis of fatty acid oxidation defects. Clin Chim Acta, 1999. 281(1‐2): 1‐17.
31. Vanhamme, L., A. van den Boogaart, and S. Van Huffel. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson, 1997. 129(1): 35‐43.
32. De Feyter, H.M., G. Schaart, M.K. Hesselink, P. Schrauwen, K. Nicolay, and J.J. Prompers. Regional variations in intramyocellular lipid concentration correlate with muscle fiber type distribution in rat tibialis anterior muscle. Magn Reson Med, 2006. 56(1): 19‐25.
33. De Feyter, H.M., E. Lenaers, S.M. Houten, P. Schrauwen, M.K. Hesselink, R.J. Wanders, K. Nicolay, and J.J. Prompers. Increased intramyocellular lipid content but normal skeletal muscle mitochondrial oxidative capacity throughout the pathogenesis of type 2 diabetes. Faseb J, 2008. 22(11): 3947‐55.
34. Garwood, M. and L. DelaBarre. The return of the frequency sweep: designing adiabatic pulses for contemporary NMR. J Magn Reson, 2001. 153(2): 155‐77.
35. Lawson, J.W. and R.L. Veech. Effects of pH and free Mg2+ on the Keq of the creatine kinase reaction and other phosphate hydrolyses and phosphate transfer reactions. J Biol Chem, 1979. 254(14): 6528‐37.
36. Taylor, D.J., P.J. Bore, P. Styles, D.G. Gadian, and G.K. Radda. Bioenergetics of intact human muscle. A 31P nuclear magnetic resonance study. Mol Biol Med, 1983. 1(1): 77‐94.
37. van den Broek, N.M., H.M. De Feyter, L. de Graaf, K. Nicolay, and J.J. Prompers. Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics in muscle after exercise are due to differences in proton efflux rates. Am J Physiol Cell Physiol, 2007. 293(1): C228‐37.
38. Reichert, M., H. Schaller, W. Kunz, and G. Gerber. The dependence on the extramitochondrial ATP/ADP‐ratio of the oxidative phosphorylation in mitochondria isolated by a new procedure from rat skeletal muscle. Acta Biol Med Ger, 1978. 37(8): 1167‐76.
39. Echtay, K.S., D. Roussel, J. St‐Pierre, M.B. Jekabsons, S. Cadenas, J.A. Stuart, J.A. Harper, S.J. Roebuck, A. Morrison, S. Pickering, J.C. Clapham, and M.D. Brand. Superoxide activates mitochondrial uncoupling proteins. Nature, 2002. 415(6867): 96‐9.
40. Ciapaite, J., S.J. Bakker, G. Van Eikenhorst, M.J. Wagner, T. Teerlink, C.G. Schalkwijk, M. Fodor, D.M. Ouwens, M. Diamant, R.J. Heine, H.V. Westerhoff, and K. Krab. Functioning of oxidative phosphorylation in liver mitochondria of high‐fat diet fed rats. Biochim Biophys Acta, 2007. 1772(3): 307‐16.
41. Szuhai, K., J. Ouweland, R. Dirks, M. Lemaitre, J. Truffert, G. Janssen, H. Tanke, E. Holme, J. Maassen, and A. Raap. Simultaneous A8344G heteroplasmy and mitochondrial DNA copy number quantification in myoclonus epilepsy and ragged‐red fibers (MERRF) syndrome by a multiplex molecular beacon based real‐time fluorescence PCR. Nucleic Acids Res, 2001. 29(3): E13.
42. Garcia‐Roves, P., J.M. Huss, D.H. Han, C.R. Hancock, E. Iglesias‐Gutierrez, M. Chen, and J.O. Holloszy. Raising plasma fatty acid concentration induces increased biogenesis of mitochondria in skeletal muscle. Proc Natl Acad Sci U S A, 2007. 104(25): 10709‐13.
43. Iossa, S., L. Lionetti, M.P. Mollica, R. Crescenzo, M. Botta, A. Barletta, and G. Liverini. Effect of high‐fat feeding on metabolic efficiency and mitochondrial oxidative capacity in adult rats. Br J Nutr, 2003. 90(5): 953‐60.
44. McAinch, A.J., J.S. Lee, C.R. Bruce, R.J. Tunstall, J.A. Hawley, and D. Cameron‐Smith. Dietary regulation of fat oxidative gene expression in different skeletal muscle fiber types. Obes Res, 2003. 11(12): 1471‐9.
45. Nemeth, P.M., B.W. Rosser, R.M. Choksi, B.J. Norris, and K.M. Baker. Metabolic response to a high‐fat diet in neonatal and adult rat muscle. Am J Physiol, 1992. 262(2 Pt 1): C282‐6.
46. Chanseaume, E., A.L. Tardy, J. Salles, C. Giraudet, P. Rousset, A. Tissandier, Y. Boirie, and B. Morio. Chronological approach of diet‐induced alterations in muscle mitochondrial functions in rats. Obesity (Silver Spring), 2007. 15(1): 50‐9.
Chapter 6
144
47. Hoeks, J., J.J. Briede, J. de Vogel, G. Schaart, M. Nabben, E. Moonen‐Kornips, M.K. Hesselink, and P. Schrauwen. Mitochondrial function, content and ROS production in rat skeletal muscle: effect of high‐fat feeding. FEBS Lett, 2008. 582(4): 510‐6.
48. Laurent, D., B. Yerby, R. Deacon, and J. Gao. Diet‐induced modulation of mitochondrial activity in rat muscle. Am J Physiol Endocrinol Metab, 2007. 293(5): E1169‐77.
49. Brehm, A., M. Krssak, A.I. Schmid, P. Nowotny, W. Waldhausl, and M. Roden. Increased lipid availability impairs insulin‐stimulated ATP synthesis in human skeletal muscle. Diabetes, 2006. 55(1): 136‐40.
50. Kemp, G.J. The interpretation of abnormal 31P magnetic resonance saturation transfer measurements of Pi/ATP exchange in insulin‐resistant skeletal muscle. Am J Physiol Endocrinol Metab, 2008. 294(3): E640‐642.
51. Short, K.R., K.S. Nair, and C.S. Stump. Impaired mitochondrial activity and insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004. 350(23): 2419‐21; author reply 2419‐21.
52. Wagenmakers, A.J. Insulin resistance in the offspring of parents with type 2 diabetes. PLoS Med, 2005. 2(9): e289.
53. Brindle, K.M., M.J. Blackledge, R.A. Challiss, and G.K. Radda. 31P NMR magnetization‐transfer measurements of ATP turnover during steady‐state isometric muscle contraction in the rat hind limb in vivo. Biochemistry, 1989. 28(11): 4887‐93.
54. Brindle, K.M. and G.K. Radda. 31P‐NMR saturation transfer measurements of exchange between Pi and ATP in the reactions catalysed by glyceraldehyde‐3‐phosphate dehydrogenase and phosphoglycerate kinase in vitro. Biochim Biophys Acta, 1987. 928(1): 45‐55.
55. Campbell‐Burk, S.L., K.A. Jones, and R.G. Shulman. 31P NMR saturation‐transfer measurements in Saccharomyces cerevisiae: characterization of phosphate exchange reactions by iodoacetate and antimycin A inhibition. Biochemistry, 1987. 26(23): 7483‐92.
56. Kingsley‐Hickman, P.B., E.Y. Sako, P. Mohanakrishnan, P.M. Robitaille, A.H. From, J.E. Foker, and K. Ugurbil. 31P NMR studies of ATP synthesis and hydrolysis kinetics in the intact myocardium. Biochemistry, 1987. 26(23): 7501‐10.
57. Chanseaume, E., G. Bielicki, A.L. Tardy, J.P. Renou, D. Freyssenet, Y. Boirie, and B. Morio. Impaired resting muscle energetics studied by (31)P‐NMR in diet‐induced obese rats. Obesity (Silver Spring), 2008. 16(3): 572‐7.
58. Vicini, P. and M.J. Kushmerick. Cellular energetics analysis by a mathematical model of energy balance: estimation of parameters in human skeletal muscle. Am J Physiol Cell Physiol, 2000. 279(1): C213‐24.
59. Yamashita, K. and T. Yoshioka. Profiles of creatine kinase isoenzyme compositions in single muscle fibres of different types. J Muscle Res Cell Motil, 1991. 12(1): 37‐44.
60. Lerner, E., A.L. Shug, C. Elson, and E. Shrago. Reversible inhibition of adenine nucleotide translocation by long chain fatty acyl coenzyme A esters in liver mitochondria of diabetic and hibernating animals. J Biol Chem, 1972. 247(5): 1513‐9.
61. Li, X., B. Monks, Q. Ge, and M.J. Birnbaum. Akt/PKB regulates hepatic metabolism by directly inhibiting PGC‐1alpha transcription coactivator. Nature, 2007. 447(7147): 1012‐6.
62. Mikhail, M.M. and M.M. Mansour. The relationship between serum carnitine levels and the nutritional status of patients with schistosomiasis. Clin Chim Acta, 1976. 71(2): 207‐14.
63. Kaido, M., H. Fujimura, A. Ono, K. Toyooka, H. Yoshikawa, T. Nishimura, K. Ozaki, I. Narama, and M. Kuwajima. Mitochondrial abnormalities in a murine model of primary carnitine deficiency. Systemic pathology and trial of replacement therapy. Eur Neurol, 1997. 38(4): 302‐9.
64. Tein, I. Carnitine transport: pathophysiology and metabolism of known molecular defects. J Inherit Metab Dis, 2003. 26(2‐3): 147‐69.
145
146
Chapter Summary and future perspectives
Summary and future perspectives
148
Summary
Diabetes has reached epidemic proportions worldwide. Type 2 diabetes (T2D) accounts for
about 90% of all diabetes cases and is characterized by insulin resistance (IR) in major
metabolic tissues. The dramatic rise in T2D is associated with the increased occurrence of
obesity and excessive ectopic lipid accumulation, in particular in skeletal muscle, due to
excessive caloric intake and decreased physical activity. However, the exact processes
leading to IR remain unresolved. One of the leading hypotheses in the research field of type
2 diabetes is that inherited or acquired skeletal muscle mitochondrial dysfunction,
associated with a reduced mitochondrial capacity to oxidize fatty acids (FAs), leads to a lipid
overload in muscle cells. High levels of lipid intermediates then induce IR by activating a
protein kinase, which impairs insulin signaling by phosphorylation of serine/threonine sites
on the insulin receptor. Recent studies have challenged this theory and link insulin
resistance to an increased capacity to oxidize FAs rather than the reverse. The high ‐oxidation rates observed in insulin‐resistant muscle are associated with low rates of
complete FA oxidation and elevated incomplete ‐oxidation. It has been hypothesized that the accumulation of incompletely metabolized FAs in the mitochondria causes
‘mitochondrial stress’ leading to insulin resistance. To clarify this issue, the timing and
nature of muscle mitochondrial adaptations during the development of IR and T2D have
been investigated in this thesis using both in vivo and in vitro approaches.
The ambiguous results on the role of muscle mitochondrial dysfunction in T2D may be due
to the different measurement methods used to determine muscle mitochondrial function.
In vitro methodologies, like the determination of gene expression levels, enzyme activities,
mitochondrial content, morphology and respiration, provide specific information on
different aspects of mitochondrial energy production, but the results cannot be directly
translated to in vivo mitochondrial function. 31P magnetic resonance spectroscopy (MRS)
provides a non‐invasive tool to monitor the energetic status of the cell in vivo by measuring
intracellular phosphorous containing metabolites, i.e. phosphocreatine (PCr), ATP and
inorganic phosphate (Pi). 31P MRS has been used to assess skeletal muscle mitochondrial
function in vivo by measuring (1) resting ATP synthesis flux with saturation transfer (ST) or
(2) the rate of PCr recovery after exercise. However, both methods measure completely
different parameters, i.e. basal metabolic rate and maximal oxidative capacity, respectively,
which might explain the contradictory results that have been obtained. In chapter 2, we
compared both parameters in rats treated with the complex I inhibitor
diphenyleneiodonium (DPI) with the aim of establishing the most appropriate method for
the assessment of in vivo muscle mitochondrial function. In vivo 31P MRS measurements
were supplemented by in vitro measurements of oxygen consumption in isolated
mitochondria. Two weeks of DPI treatment induced mitochondrial dysfunction, as
evidenced by a 20% lower maximal ADP‐stimulated oxygen consumption rate in isolated
mitochondria from DPI‐treated rats oxidizing pyruvate plus malate. This was paralleled by a
46% decrease in in vivo oxidative capacity, determined from post‐exercise PCr recovery.
Summary and future perspectives
149
Interestingly, no significant difference in resting, ST‐based ATP synthesis flux was observed
between DPI‐treated rats and controls. These results show that the rate constant of PCr
recovery measured with dynamic 31P MRS after exercise provides a more sensitive measure
of skeletal muscle mitochondrial function than the ATP synthesis flux determined with 31P
ST in the resting state. The ATP synthesis flux itself represents the ATP demand of the
muscle and in order to interpret the data in terms of mitochondrial function it is necessary
to take the error signals, i.e. the concentrations of ADP and Pi, into account. Moreover, the
Pi ATP flux obtained from a 31P ST experiment in the resting state is dominated by
glycolytic exchange flux.
The second methodological chapter (chapter 3) describes the pH dependence of the PCr
recovery time constant (PCr, i.e. the inverse of the PCr recovery rate constant). It has been shown that muscle tissue acidification at the end of exercise severely prolongs PCr
recovery, which is an important complication in the interpretation of post‐exercise PCr
recovery data. It has been proposed that PCr can be normalized for the end‐exercise pH, in
order to use it as a measure for mitochondrial function. However, a general correction for
pH can only be applied if there are no intersubject differences in the pH dependence of PCr
recovery kinetics. In this chapter the pH dependence of PCr was investigated in healthy human volunteers on a subject‐by‐subject basis and it turned out that the effect of acidosis
on PCr recovery kinetics after exercise was different between subjects. The pH dependence
of PCr correlated with the proton efflux rate at the start of recovery, indicating that subjects with a smaller pH dependence of PCr have a higher rate of pH recovery. Therefore, PCr can only be used as a measure of mitochondrial function when end‐exercise pH is close
to resting values. Simply correcting PCr for end‐exercise pH is not adequate, in particular when comparing patients and controls, as certain disorders are characterized by altered
proton efflux from muscle fibers.
In chapters 4‐6, 31P MRS was applied to investigate the role of muscle mitochondrial
dysfunction in the development of IR and T2D. Chapter 4 describes a cross‐sectional human
study, in which we examined in vivo skeletal muscle mitochondrial function in healthy,
normoglycaemic controls, subjects with early stage T2D and long‐standing, insulin‐treated
T2D patients. It was shown that PCr recovery after exercise was not different between
groups, indicating that in vivo oxidative capacity was not impaired in both early and late
stages of T2D. These results imply that mitochondrial dysfunction does not necessarily
represent either cause or consequence of IR and/or T2D. It was suggested that impairments
in oxidative metabolism in type 2 diabetes patients observed in previous studies are likely
to be secondary to a less active lifestyle and/or impaired insulin signaling.
A cross‐sectional study, as presented in chapter 4, only provides correlative data and does
not provide detailed information about the time course of events during the development
of T2D. Therefore, the aim of the next study, described in chapter 5, was to gain more
insight into the timing and nature of mitochondrial adaptations during the development of
Summary and future perspectives
150
high‐fat diet‐induced insulin resistance in rats. Adult Wistar rats were fed a high‐fat diet or
normal chow for 2.5 and 25 weeks. Muscle oxidative capacity was assessed in vivo from 31P
MRS measurements of PCr recovery and in vitro by measuring mitochondrial DNA copy
number and oxygen consumption in isolated mitochondria. Muscle lipid status was
determined by 1H MRS (intramyocellular lipids, IMCL) and tandem mass spectrometry
(acylcarnitines). The short‐term high‐fat diet induced increases in IMCL content and muscle
medium‐ and long‐chain acylcarnitines, together with an increased in vivo oxidative
capacity. The latter result could be fully accounted for by increased mitochondrial content.
The long‐term high‐fat diet resulted in even higher IMCL and acylcarnitine levels, a further
increase in the number of muscle mitochondria and an increased capacity to oxidize fat‐
derived substrates in vitro. Surprisingly, this did not result in a higher muscle oxidative
capacity in vivo. These findings show that skeletal muscle in high‐fat diet‐induced IR
requires a progressively larger mitochondrial pool size to maintain normal oxidative
capacity in vivo.
Comparable to the results of the 25 wk high‐fat diet are the results observed in Wistar rats
on a similar high‐fat diet for 15 wk, as presented in chapter 6. The observed dissociation
between in vivo and in vitro determinations of muscle mitochondrial function after long‐
term high‐fat diet feeding in both chapter 5 and 6 suggests that in vivo mitochondrial
function in insulin‐resistant rat muscle is compromised by factors not taken into account in
vitro. In both studies, muscle tissue free carnitine was significantly decreased, whereas
muscle medium‐ and long‐chain acylcarnitines were significantly increased. This finding
indicates that in vivo muscle mitochondrial function might have been compromised by high
concentrations of lipid metabolites in vivo. This was corroborated by the finding that
mitochondria from high‐fat diet fed rats were more susceptible to FA‐induced
mitochondrial uncoupling, which can explain the observed in vivo mitochondrial
dysfunction.
Next, it was investigated whether the observed carnitine insufficiency, a common feature
of insulin‐resistant states, contributes to the accumulation of muscle lipid intermediates
and the lipid‐induced impairment of skeletal muscle mitochondrial function in vivo.
Carnitine supplementation might reduce toxic lipid intermediates by increasing FA
oxidation and export of lipid metabolites out of the mitochondria and into the plasma,
thereby relieving mitochondrial stress. Despite the normalization of muscle free carnitine
content, carnitine supplementation did not induce improvements in muscle lipid status, in
vivo mitochondrial function or insulin resistance. These results indicate that the decrease in
muscle free carnitine as a result of high‐fat diet feeding did not contribute to the observed
impairment in in vivo muscle mitochondrial function.
Summary and future perspectives
151
Conclusions
Although it has been hypothesized that skeletal muscle mitochondrial dysfunction and an
associated decreased capacity to oxidize FAs is the primary cause for muscle lipid
accumulation and IR, the findings in this thesis are in agreement with recent studies which
have shown that high‐fat diets lead to increased rather than decreased FA oxidation,
resulting in the accumulation of lipid intermediates causing mitochondrial dysfunction.
Furthermore, from the results presented in this thesis it is concluded that it is essential to
use a combination of several techniques, preferably both in vivo and in vitro, for a thorough
investigation of muscle mitochondrial function. Compelling evidence for this is provided in
chapters 5 and 6, where it was shown that a normal in vivo oxidative capacity, as measured
by PCr recovery, can mask an impairment in intrinsic mitochondrial function when it is
accompanied by increased mitochondrial content.
Future perspectives
In vivo PCr recovery kinetics is determined by mitochondrial content, intrinsic
mitochondrial function and other factors like the supply of oxygen and substrates. In this
thesis, mitochondrial content and intrinsic mitochondrial function were additionally
determined in vitro. Intrinsic mitochondrial function was assessed from measurements on
isolated mitochondria. An alternative method would be to measure oxygen consumption in
permeabilized muscle fibers, which more closely reflect the in vivo environment. However,
inhomogeneity of muscles in terms of fiber type composition relates to inhomogeneities in
mitochondrial content. Therefore, inhomogeneous fiber preparation may in turn bias the
outcome of the measurements. This is in particular a problem when studying rodent
muscle, due to the marked regionalization of different fiber types within a single muscle.
For a comprehensive interpretation of changes in PCr recovery kinetics the assessment of
muscle perfusion, regulating the supply of oxygen and substrates, is of great value. Three
possible options for measuring muscle perfusion and oxygenation are briefly discussed:
arterial spin labeling (ASL), 1H MR myoglobin (Mb) spectroscopy and near‐infrared
spectroscopy (NIRS). ASL is an MR‐based technique used to assess tissue perfusion [1‐2]
relying on differences in T1 relaxation rates of tissue, caused by inflowing blood, when using
slice‐selective or non‐selective inversion pulses. This technique has been successfully
applied in brain and heart tissues (1‐2). A low perfusion hampers the application of ASL in
skeletal muscle at rest. However, during exercise muscle perfusion increases significantly
and is therefore more easy to measure with ASL [3]. A recent study demonstrated ASL
during muscle stimulation in combination with 31P MRS in mice [4]. An MR method to check
if oxygen flux is limiting PCr recovery is 1H MR spectroscopy of Mb, which monitors
intramyocellular oxygenation by measuring the signal from deoxy‐Mb at 71‐72 ppm
Summary and future perspectives
152
downfield of water [5]. Interleaved 31P/1H spectra obtained in human quadriceps have
enabled detailed studies on the relation between muscle intracellular oxygenation and PCr
recovery [6]. In vivo Mb spectroscopy data from skeletal muscles of small animals have
never been reported. The main explanation for this, though possibly not the only one, is
probably the lower Mb levels in small rodents [7]. NIRS is a non‐MR alternative to measure
blood flow and local tissue oxygenation levels. NIRS is primarily based on the absorption of
light at wavelengths in the near‐infrared range (400‐1000 nm) where oxygenated
hemoglobin (oxy‐Hb) and deoxygenated hemoglobin (deoxy‐Hb) display different light
absorption characteristics. With MR‐compatible adaptations, complementary NIRS and
MRS measurements can be done simultaneously, as was shown in chronic heart failure
patients [8].
Supplementary in vivo measurements can provide new insights in the complex in vivo
mitochondrial functioning. In chapter 2, it was concluded that the interpretation of the ATP
synthesis flux, particularly at rest, is not straightforward, partially due to the relatively large
contribution of glycolytic exchange flux. In order to create better conditions for detecting a
defect in mitochondrial function, the MR saturation transfer experiment would need to be
done in exercising muscle. In this case, the ATP synthesis flux will be significantly higher,
resulting in less interference of glycolytic exchange flux. However, in such an experiment it
is crucial to sustain steady‐state contractions during the whole duration of the saturation
transfer experiment and it is important to achieve similar relative workloads in different
subjects or animals.
In addition to measuring ATP synthesis flux, it would be valuable to specifically measure
processes more upstream in oxidative metabolism. Koves et al. suggested that lipid
overload causes excessive ‐oxidation without concomitant increases in tricarboxylic acid
(TCA) cycle activity, resulting in accumulation of incompletely oxidized lipid intermediates
[9]. To further elucidate this promising hypothesis, it would be valuable to measure TCA
cycle and ‐oxidation activity in vivo. TCA cycle flux has repeatedly been assessed in brain with 13C MRS, during the infusion of 13C labeled glucose or acetate. By measuring 13C
incorporation into glutamate and glutamine in time, TCA cycle activity can be determined.
Previous studies showed that performing such measurements in human [10] and rat [11]
muscle is possible, in spite of the practical complications regarding the sensitivity and the
low TCA cycle flux in resting conditions. Similarly to what was proposed for 31P saturation
transfer measurements of ATP synthesis flux, steady‐state muscle contractions would
provide a means to circumvent these difficulties by increasing the TCA cycle flux. Moreover,
it would be interesting to combine the TCA cycle and ATP synthesis flux measurements
obtained during steady state muscle contractions. These data would reflect the efficiency
of ATP synthesis [11], i.e. mitochondrial coupling, which might be more representative of in
vivo intrinsic muscle mitochondrial function.
Summary and future perspectives
153
The presence of a relationship between mitochondrial dysfunction and lipotoxicity is
generally accepted. However, information about the underlying processes is scarce. Lipid
accumulation results from an imbalance between lipid storage and oxidation. In vivo 1H‐
[13C] localized MRS in combination with the administration of 13C enriched lipids is a
promising method to assess the incorporation of the 13C label into the IMCL pool, thereby
providing interesting additional information about the rate of uptake and storage of
ingested lipids in vivo. Furthermore, the rate of disappearance of 13C label from the IMCL
pool can be assumed to reflect the rate of ‐oxidation in vivo. Although IMCL and
acylcarnitine levels have been assumed to reflect the content of muscle lipid intermediates
which are supposed to induce IR, directly measuring DAG, ceramides and long‐chain acyl‐
CoA’s in muscle tissue would improve our understanding of the molecular mechanisms
underlying lipid‐induced mitochondrial dysfunction and IR.
Summary and future perspectives
154
References 1. Kober, F., I. Iltis, P.J. Cozzone, and M. Bernard. Myocardial blood flow mapping in mice using high‐
resolution spin labeling magnetic resonance imaging: influence of ketamine/xylazine and isoflurane anesthesia. Magn Reson Med, 2005. 53(3): 601‐6.
2. Streif, J.U., M. Nahrendorf, K.H. Hiller, C. Waller, F. Wiesmann, E. Rommel, A. Haase, and W.R. Bauer. In vivo assessment of absolute perfusion and intracapillary blood volume in the murine myocardium by spin labeling magnetic resonance imaging. Magn Reson Med, 2005. 53(3): 584‐92.
3. Carlier, P.G., C. Brillault‐Salvat, E. Giacomini, C. Wary, and G. Bloch. How to investigate oxygen supply, uptake, and utilization simultaneously by interleaved NMR imaging and spectroscopy of the skeletal muscle. Magn Reson Med, 2005. 54(4): 1010‐3.
4. Baligand, C., H. Gilson, J.C. Menard, O. Schakman, C. Wary, J.P. Thissen, and P.G. Carlier. Functional assessment of skeletal muscle in intact mice lacking myostatin by concurrent NMR imaging and spectroscopy. Gene Ther, 2010. 17(3): 328‐37.
5. Wang, Z.Y., E.A. Noyszewski, and J.S. Leigh, Jr. In vivo MRS measurement of deoxymyoglobin in human forearms. Magn Reson Med, 1990. 14(3): 562‐7.
6. Duteil, S., C. Bourrilhon, J.S. Raynaud, C. Wary, R.S. Richardson, A. Leroy‐Willig, J.C. Jouanin, C.Y. Guezennec, and P.G. Carlier. Metabolic and vascular support for the role of myoglobin in humans: a multiparametric NMR study. Am J Physiol Regul Integr Comp Physiol, 2004. 287(6): R1441‐9.
7. Ohno, T. and A. Kuroshima. Muscle myoglobin as determined by electrophoresis in thermally acclimated rat. Jpn J Physiol, 1986. 36(4): 733‐44.
8. Kemps, H.M., J.J. Prompers, B. Wessels, W.R. De Vries, M.L. Zonderland, E.J. Thijssen, K. Nicolay, G. Schep, and P.A. Doevendans. Skeletal muscle metabolic recovery following submaximal exercise in chronic heart failure is limited more by O(2) delivery than O(2) utilization. Clin Sci (Lond), 2010. 118(3): 203‐10.
9. Koves, T.R., J.R. Ussher, R.C. Noland, D. Slentz, M. Mosedale, O. Ilkayeva, J. Bain, R. Stevens, J.R. Dyck, C.B. Newgard, G.D. Lopaschuk, and D.M. Muoio. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab, 2008. 7(1): 45‐56.
10. Befroy, D.E., K.F. Petersen, S. Dufour, G.F. Mason, R.A. de Graaf, D.L. Rothman, and G.I. Shulman. Impaired mitochondrial substrate oxidation in muscle of insulin‐resistant offspring of type 2 diabetic patients. Diabetes, 2007. 56(5): 1376‐81.
11. Jucker, B.M., J. Ren, S. Dufour, X. Cao, S.F. Previs, K.S. Cadman, and G.I. Shulman. 13C/31P NMR assessment of mitochondrial energy coupling in skeletal muscle of awake fed and fasted rats. Relationship with uncoupling protein 3 expression. J Biol Chem, 2000. 275(50): 39279‐86.
155
156
Nederlandse samenvatting
Dankwoord
List of publications
Curriculum Vitae
Nederlandse samenvatting
158
Samenvatting
“Mitochondriële aanpassingen in insulineresistente spieren”
Diabetes is uitgegroeid tot een wereldwijde epidemie. Type 2 diabetes (T2D) neemt
ongeveer 90% van alle diabetes gevallen in beslag en wordt gekarakteriseerd door
insulineresistentie (IR) in belangrijke metabole weefsels. De enorme stijging in het aantal
T2D patiënten is gerelateerd aan de toename van het aantal mensen met overgewicht en
vervetting van de organen, in het bijzonder de spieren, veroorzaakt door een overmaat aan
voedselconsumptie en een verminderde fysieke activiteit. Echter, de exacte processen die
leiden tot IR zijn nog niet bekend. Een belangrijke hypothese binnen het onderzoeksgebied
naar de oorzaken van T2D is dat een verstoorde functie van de mitochondriën in de
skeletspier leidt tot een verlaagde capaciteit om vetzuren te oxideren, wat resulteert in een
ophoping van vetten in de spiercellen (intramyocellulaire lipiden, IMCL). Verhoogde
concentraties IMCL en intermediairen daarvan induceren daarop IR door een proteïne
kinase te activeren, welke de insuline signalering en daarmee glucose opname verstoort.
Recente studies hebben deze theorie in twijfel getrokken and linken IR aan een
toegenomen in plaats van afgenomen capaciteit om vetzuren te oxideren. De hoge
vetzuuroxidatie snelheden zoals waargenomen in insulineresistente spier zijn geassocieerd
met een lage complete vetzuuroxidatie en een verhoogde incomplete ‐oxidatie. Daarom is
gesuggereerd dat de ophoping van incompleet gemetaboliseerde vetzuren in de
mitochondriën ‘mitochondriële stress’ veroorzaakt, wat zou leiden tot IR. Ter verheldering
van deze kwestie worden in dit proefschrift de mitochondriële aanpassingen in skeletspier
tijdens de ontwikkeling van IR and T2D nader onderzocht middels het gebruik van in vivo
(31P en 1H magnetische resonantie spectroscopie (MRS)) en in vitro (o.a. mitochondriële
respiratie metingen en bepaling van mitochondriaal DNA en acylcarnitine niveaus)
methodes.
De tegenstrijdige resultaten aangaande de rol van mitochondriële disfunctie in T2D zouden
verklaard kunnen worden door de verschillende technieken die gebruikt zijn om de
mitochondriële functie in de spier te bepalen. In vitro methodes, zoals de bepaling van
genexpressie niveaus, enzym activiteiten, het aantal mitochondriën, de morfologie van
mitochondriën en mitochondriële respiratie, geven specifieke informatie over verschillende
aspecten van mitochondriële functie, maar de resultaten kunnen niet direct vertaald
worden naar in vivo mitochondriële functie. Met behulp van 31P MRS kan de energetische
status van de spiercel in vivo worden bepaald door het meten van fosfocreatine (PCr), ATP
en anorganisch fosfaat (Pi). In de literatuur zijn twee verschillende 31P MRS methodes
toegepast om de mitochondriële functie van skeletspier in vivo te bepalen, namelijk het
meten van (1) de ATP synthese flux in rust met behulp van een saturatie transfer (ST)
experiment en (2) de snelheid van PCr herstel na inspanning met dynamische 31P MRS.
Echter, deze methodes leveren totaal verschillende parameters, te weten de snelheid van
het basale metabolisme en de maximale oxidatieve capaciteit. In hoofdstuk 2 van dit
Nederlandse samenvatting
159
proefschrift zijn beide methodes vergeleken in ratten met mitochondriële disfunctie door
behandeling met de complex I remmer diphenyleneiodonium (DPI), met het doel om te
bepalen welke methode het meest geschikt is om de mitochondriële functie van de spier in
vivo te meten. Terwijl de dynamische 31P MRS metingen na inspanning lieten zien dat het
herstel van PCr 46% langzamer was in ratten behandeld met DPI, was de ATP synthese flux
in rust gemeten met ST niet verschillend ten opzichte van controle ratten. Deze resultaten
tonen aan dat in vivo mitochondriële functie in de spier directer en gevoeliger bepaald
wordt door de PCr herstelsnelheid na inspanning dan door ST metingen in rust.
In hoofdstuk 3 is de pH‐afhankelijkheid van de PCr hersteltijdsconstante, PCr, onderzocht. Verzuring van spierweefsel vertraagt de PCr hersteltijd en bemoeilijkt de interpretatie van
PCr als maat voor oxidatieve capaciteit. Door meerdere inspanningsprotocollen met een
variërende intensiteit uit te laten voeren door elke proefpersoon konden we concluderen
dat het effect van verzuring op de PCr herstelsnelheid verschilt tussen personen en dat een
algemene correctiefactor voor verzuring dus niet toepasbaar is. Daarnaast is aangetoond
dat de pH‐afhankelijkheid van PCr correleert met de snelheid van de protonenflux uit de
spier.
In de hoofdstukken 4‐6 is 31P MRS toegepast om de rol van mitochondriële disfunctie in de
spier in de ontwikkeling van IR en T2D nader te onderzoeken. Hoofdstuk 4 beschrijft een
cross‐sectionele studie waarin patiënten, die reeds lange tijd T2D hebben, vergeleken
worden met personen in een vroeg stadium van T2D en met personen met een normale
bloedglucose homeostase. Elke groep bestond uit mannen met gemiddeld dezelfde leeftijd,
lichaamssamenstelling en een laag niveau van dagelijkse lichaamsbeweging. De resultaten
laten zien dat mitochondriële disfunctie niet noodzakelijk een oorzaak of gevolg van IR
en/of T2D is. De verstoring in mitochondriële functie in T2D patiënten zoals in eerdere
studies is aangetoond lijkt eerder een gevolg te zijn van een verlaagde fysieke activiteit
en/of een verstoorde insuline signalering.
Met een diermodel van IR, zoals Wistar ratten op een hoog‐vet dieet, is het mogelijk om
longitudinale studies uit te voeren en daarmee een inzicht te krijgen in het tijdsverloop van
veranderingen in mitochondriële functie tijdens de ontwikkeling van IR. In hoofdstuk 5 is de
mitochondriële functie en vetophoping in de spier in vivo en in vitro bepaald in Wistar
ratten na 2,5 en 25 weken hoog‐vet dieet. Op de korte termijn (2,5 week) leidde het hoog‐
vet dieet tot een stijging van IMCL en acylcarnitines in het spierweefsel en een verhoogde
oxidatieve capaciteit van de spier in vivo, welke volledig verklaard kon worden door een
toename in het aantal mitochondriën. Na 25 weken hoog‐vet dieet was de vetophoping in
de spier verder toegenomen, net zoals het aantal mitochondriën en de capaciteit van de
mitochondriën om in vitro vet te oxideren. Echter, de in vivo oxidatieve capaciteit van de
spier was niet verhoogd, wat wijst op een verlaagde mitochondriële functie in vivo die
gecompenseerd wordt door een verhoogd aantal mitochondriën. De ogenschijnlijke
Nederlandse samenvatting
160
tegenstelling tussen de in vitro en in vivo bepalingen van mitochondriële functie wordt
mogelijk veroorzaakt door factoren die geen rol spelen bij de in vitro metingen.
In hoofdstuk 6 wordt, net zoals in hoofdstuk 5, aangetoond dat in ratten op hoog‐vet dieet
vrij carnitine in de spier significant is verlaagd, terwijl de middellange‐ en lange‐keten
acylcarnitines zijn toegenomen. De verhoogde acylcarnitine niveaus in de spier duiden erop
dat de mitochondriële disfunctie geobserveerd in vivo veroorzaakt kan zijn door een hoge
concentratie vetmetabolieten, wat lijkt te worden bevestigd door de bevinding dat
mitochondriën van ratten op een hoog‐vet dieet vatbaarder zijn voor vetzuur‐geïnduceerde
ontkoppeling in vitro. Bovendien wordt in dit hoofdstuk aangetoond dat carnitine
supplementatie wel het carnitine tekort in de spier normaliseert, maar niet bijdraagt aan
een verbetering van de lipide huishouding en de in vivo mitochondriële functie van de
spier.
De bevindingen die beschreven zijn in dit proefschrift zijn in overeenstemming met eerdere
studies die aan hebben getoond dat IR correleert met een verhoogde in plaats van
verlaagde vetzuuroxidatie, met een ophoping van incompleet gemetaboliseerde vetzuren
en mitochondriële disfunctie als gevolg. Daarnaast is onomstreden vastgesteld dat het
gebruik van een combinatie van verschillende technieken, bij voorkeur zowel in vivo als in
vitro, essentieel is voor een grondig onderzoek naar de mitochondriële functie in
skeletspier.
Dankwoord
161
Dankwoord
Het is gelukt! Vier jaar werk is nu gebundeld in dit proefschrift. Maar dat was me zeker niet
gelukt zonder de samenwerking met en begeleiding van verschillende personen. Het is dan
ook niet overbodig om hen op deze plaats daarvoor te bedanken.
Allereerst natuurlijk Klaas, als professor van de Biomedical NMR groep en promotor van
mijn promotieonderzoek heeft u in een paar jaar tijd van een klein hecht clubje een groep
gemaakt die goed op de kaart staat. Dit heeft er ook voor gezorgd dat u het alleen maar
drukker kreeg, maar toch wist u eigenlijk altijd tijd te maken wanneer het nodig was: van
het bespreken en interpreteren van nieuwe resultaten tot het corrigeren van papers en
hoofdstukken. Uw positieve manier van becommentariëren heb ik altijd erg gewaardeerd.
Jeanine, wat ben jij belangrijk geweest voor de totstandkoming van dit boekje! Je wilde
overal betrokken bij zijn en ik kon altijd bij je binnenlopen. Je persoonlijke benadering om
duidelijk te maken wat wel en wat niet in overeenstemming was met jouw ideeën vond ik
erg prettig, net zoals alle gesprekken over privesituaties. Bedankt voor alles wat je voor me
hebt gedaan, ik heb vooral ook door jou de afgelopen 4 jaar met veel plezier gewerkt!
Jolita, without you the in vitro experiments would not be of such a high quality as they are
now. The dedication with which you work in the lab is really unbelievable and this is
directly resulting in the nicest results. Besides a very good researcher, I learned to know
you as a very pleasant person. I enjoyed our trip to Rome really a lot!
Sander, vooral de laatste tijd hebben onze wegen elkaar gekruist. Met jullie lab technieken
heb je een enorme toegevoegde waarde gehad in onze studies. Zeker ook omdat de
samenwerking altijd zo prettig verliep, vind ik het erg jammer dat je uiteindelijk toch niet in
mijn commissie kunt zitten. De samenwerking tussen jullie lab van het AMC en de
Biomedical NMR groep gaat zeker nog veel mooie studies opleveren!
Voor de humane type 2 diabetes studie heb ik met veel plezier samen gewerkt met Luc van
Loon en Stephan Praet. Luc, eigenlijk ben ik door jou hier op de TU/e terecht gekomen toen
ik bij je aanklopte voor een afstudeerstage. Naast je wetenschappelijke adviezen zal ik
vooral je wijze raad die je gaf na een afsluitingsetentje in Veldhoven over samenwonen
nooit meer vergeten! Stephan, jouw enthousiasme heeft me in het begin weleens
overdonderd, maar ik weet zeker dat dit een belangrijke bijdrage heeft gehad aan de berg
aan data die we hebben geproduceerd en wat geleid heeft tot een mooi resultaat.
Wie ik zeker niet wil vergeten zijn alle vrijwilligers voor de humane studies: zowel voor de
pilot studies; Ad, Ben, Sjaak en Jan, als voor de studie beschreven in hoofdstuk 3; Ewelina,
Jeanine, Gustav, Koen, Ben en Toon (ook bedankt natuurlijk voor alle andere klusjes
waarmee je me geholpen hebt, zoals het plakken van mijn fietsband: ideaal zo’n
Dankwoord
162
conciërge!), en de vrijwilligers die mee hebben gewerkt aan de studie beschreven in
hoofdstuk 4.
Joep van Lier, bedankt voor het delen van jouw ‘fit’ ervaring. Ook al was wat je deed voor
mij soms een grote puzzel, door jouw hulp zijn onze saturatie transfer data zeker
betrouwbaar gefit. Marije, de uurtjes bij de scanner voor onze ‘pH studie’ samen waren een
stuk gezelliger dan wanneer ik er alleen zat. Helaas zijn de data nog niet in dit boekje
verschenen, maar het wordt zeker nog een mooie publicatie. Iedereen in het Scheikunde
lab bedankt voor de hulp bij het uitvoeren van de ELISA’s.
Naast goede wetenschappelijke contacten, zijn niet of minder wetenschappelijke contacten
belangrijk om zo goed mogelijk te presteren. De sfeer in de vakgroep heb ik als erg prettig
ervaren, zeker toen bleek dat er 20 collega’s enthousiast te krijgen waren voor een
korfbaltoernooi. In de afgelopen 4 jaar heb ik een aantal kamergenootjes gehad, waarmee
ik tegenslagen maar zeker ook mooie momenten heb meegemaakt. Ewelina, I am very
happy that you will be very close to me during my defense as my usher, just as you were
during my whole PhD project. Hopefully next year we can arrange our plan to visit Poland
together! Bram, Tessa en Leonie, B2.03 was echt een top kamer. Ik ga de leuke gesprekken
en gezelligheid zeker missen. Geralda, ondertussen zit je alweer een jaar in Utrecht, maar
gelukkig hebben we nog steeds leuk contact. De tenniscarrière, ooit zo mooi begonnen in
Geldrop, kunnen we voorlopig maar beter even in de koelkast zetten. Houden we het toch
gewoon bij een etentje op zijn tijd? Ward, jij blijft toch ook meedoen met de etentjes? Erg
fijn dat de mooie resultaten waar je zo hard voor hebt gewerkt er nu zijn en je boekje er
dan toch echt lijkt te komen!
Dan zijn er nog een aantal collega’s van de Biomedical NMR groep die ik hier wil
vernoemen. Bart, ik zal je spannende weekend verhalen tijdens de pauze gaan missen!
Succes met het voortzetten van ons werk en als opperorganisator van de groepsuitjes.
Richard, soms zit het tegen, maar als het dan eenmaal lukt gaat het echt meezitten en
resulteert alle moeite in erg mooie studies! Hou vol, positief blijven en dan weet ik zeker
dat je een prachtig boekje krijgt. Het blijft natuurlijk wel erg jammer dat onze
spiervezelkleuringen van cross‐secties van de hele TA mijn boekje niet op fleuren. Larry, je
was vaak een rots in de branding als er weer eens brokken waren aan de 1.5T scanner of
wanneer ik als niet‐technicus technische hulp nodig had, bedankt daarvoor. Voor mij als
nuchtere Brabantse meid was je af en toe een vreemde vogel, maar wel een aardige. Jo,
David en in het bijzonder Leonie bedankt voor al jullie biotechnische hulp bij de OGTT’s,
DEC protocollen enz. Hedwig en Ria, ik ben jullie dankbaar voor de hulp met alle
administratieve rompslomp die bij (de afronding van) een promotie project komt kijken.
Jef, ook al luisterden we niet altijd naar je koffie oproep, toch waren de pauzes en BBQ’s na
het Nlaag voetbal toernooi erg gezellig! Henk, door al het werk wat jij hier hebt gedaan,
heb ik een vliegende start kunnen maken en jouw werk door kunnen zetten. Verder wil ik
Dankwoord
163
natuurlijk Gustav, Jeroen, Holger, alle andere collega’s, (ex‐) aio’s en studenten van de
Biomedical NMR groep heel erg bedanken voor de leuke pauzes en groepsuitjes.
Zoals iedereen die mij een beetje kent wel weet, is korfbal een belangrijke bezigheid voor
mij. Ook al lijkt het soms misschien niet zo, toch is het een heerlijke ontspanning en
uitlaatklep. Dat komt voor een groot deel natuurlijk ook door ons gezellige, no‐nonsense,
maar vooral ‘hechte’ team! Naast het zelf spelen mag ik natuurlijk de meiden van de A1
niet vergeten, jullie zijn toppers. Als ik chagrijnig was kwam het waarschijnlijk door de
stress om dit boekje, mijn verontschuldigingen daarvoor. Dimphy, ik vind het fijn dat ik
altijd bij je terecht kan, voor advies, opbeurende woorden, je realistische blik, maar zeker
ook voor gezelligheid! Doreen, there is something, something I want to tell you… je bent
een geweldig vriendinnetje! Ik hoefde dan ook niet lang na te denken wie ik als paranimf
achter me wil hebben staan.
Ons Maastricht/sinterklaas clubje: Anne, Janneke, Marije, Marijn en Nicole, ik hoop dat we
nog lang de vriendinnen blijven die we nu zijn, op naar ons 10‐jarig jubileum!
Frans en Nicolien, vanaf de eerste dag heb ik me welkom gevoeld bij jullie. Jullie interesse
en betrokkenheid in de stappen die Koen en ik namen zijn erg belangrijk voor ons. Luc, ik
ben er nu ook van overtuigd dat de uren die Koen en jij achter de computer doorbrachten
meestal niet door jou kwamen. Succes, maar vooral ook veel plezier nog met je Utrechtse
studentenleventje! We komen binnenkort echt een keer naar je kamer kijken, dus zet die
wereldgerecht pakken maar klaar.
Pap en mam, hierbij wil ik jullie ook ontzettend bedanken voor alle vrijheid en
mogelijkheden die jullie me altijd gegeven hebben. Ook al hebben jullie weinig kaas
gegeten van mitochondriën, jullie zijn geïnteresseerd in alles wat ik doe en meemaak,
waardoor ik altijd graag thuis kom. Zo zijn de voetbalavonden samen met Kees nog steeds
gezellig. Suus, ik ben blij dat je mijn zusje bent en ik heb altijd al geweten dat je zo’n goede
juf zou worden als je nu bent, ik ben trots op je!
Ook al vond je het niet nodig, toch wil en kan ik jou hier niet onvermeld laten Koen. Je bent
belangrijk voor me, om teveel redenen om hier op te noemen. Ik ben benieuwd waar onze
‘(trein) reis’ ons nog allemaal gaat brengen!
List of publications
164
List of publications
International refereed journal publications
N.M.A. van den Broek, J. Ciapaite, K. Nicolay, J.J. Prompers. Comparison of in vivo
post‐exercise phosphocreatine recovery and resting ATP synthesis flux for the
assessment of skeletal muscle mitochondrial function. Am J Physiol Cell Physiol, in
press.
N.M.A. van den Broek, J. Ciapaite, H.M.M.L. De Feyter, S.M. Houten, R.J.A. Wanders,
J.A.L. Jeneson, K. Nicolay and J.J. Prompers. Increased mitochondrial content rescues
in vivo muscle oxidative capacity in long‐term high‐fat‐diet‐fed rats. FASEB, 2010.
24(5):1354‐64.
J.A.L. Jeneson, J.P. Schmitz, N.M.A. van den Broek, N.A. van Riel, P.A. Hilbers, K.
Nicolay, J.J. Prompers. Magnitude and control of mitochondrial sensitivity to ADP.
Am J Physiol Endocrinol Metab, 2009. 297(3): E774‐84.
H.M.M.L. De Feyter, N.M.A. van den Broek, S.F.E. Praet, K. Nicolay, L.J.C. van Loon
and J.J. Prompers. Early or advanced stage type 2 diabetes is not accompanied by in
vivo skeletal muscle mitochondrial dysfunction. Eur J Endocrinol, 2008. 158: 643–
653.
N.M.A. van den Broek, H.M.M.L. de Feyter, L. de Graaf, K. Nicolay, J.J. Prompers.
Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics
in muscle after exercise are due to differences in proton efflux rates. Am J Physiol Cell
Physiol, 2007. 293: C228–C237.
H.M.M.L. De Feyter, S.F.E. Praet, N.M.A. van den Broek, H. Kuipers, C.D. Stehouwer,
K. Nicolay, J.J. Prompers, L.J.C. van Loon. Exercise training improves glycemic control
in longstanding, insulin treated type 2 diabetes patients. Diabetes care, 2007. 30:
2511‐13.
List of publications
165
Selected abstracts at scientific meetings (First author only)
N.M.A. van den Broek, J. Ciapaite, K. Nicolay, J.J. Prompers. Comparison of in vivo post‐
exercise PCr recovery and basal ATP synthesis flux for the assessment of skeletal muscle
mitochondrial function. ISMRM 18th scientific meeting and exhibition, Stockholm, Sweden,
2010.
N.M.A. van den Broek, J. Ciapaite, H.M.M.L. De Feyter, S.M. Houten, R.J.A. Wanders, J.A.L.
Jeneson, K. Nicolay and J.J. Prompers. Metabolic adaptations in skeletal muscle in the
development of insulin resistance measured in vivo by 31P and 1H MRS. NWO‐CW study
sections Nucleic acids, Protein Research and Lipids & Biomembranes, Veldhoven, The
Netherlands, 2008.
N.M.A. van den Broek, J. Ciapaite, H.M.M.L. De Feyter, K. Nicolay, J.J. Prompers. Metabolic
Adaptations in Skeletal Muscle During the Development of Insulin Resistance. EASD 44th
Meeting, Rome, Italy, 2008.
N.M.A. van den Broek, H.M.M.L. de Feyter, K. Nicolay, J.J. Prompers. Metabolic
adaptations in skeletal muscle in the early stage of insulin resistance measured in vivo by 31P
and 1H MRS. ISMRM 16th scientific meeting and exhibition, Toronto, Canada, 2008.
N.M.A. van den Broek, H.M.M.L. de Feyter, K. Nicolay, J.J. Prompers. Metabolic
adaptations in skeletal muscle in the early stage of insulin resistance measured in vivo by 31P
and 1H MRS. NVDO gecombineerde wetenschappelijke jaarvergadering, Doorwerth, The
Netherlands, 2007.
N.M.A. van den Broek, H.M.M.L. de Feyter, K. Nicolay, J.J. Prompers. Metabolic
adaptations in skeletal muscle in the early stage of insulin resistance measured in vivo by 31P
and 1H MRS. JFD fysiologen symposium, Papendal, The Netherlands, 2007.
N.M.A. van den Broek, H.M.M.L. De Feyter, S.F.E. Praet, K. Nicolay, L.J.C. van Loon and J.J.
Prompers. Normal in vivo skeletal muscle mitochondrial function in subjects with impaired
glycemic control and in long‐standing, insulin treated type 2 diabetes patients. ENP 6th
Dutch Endo‐Neuro‐Pshycho meeting, Doorwerth, The Netherlands, 2007.
N.M.A. van den Broek, H.M.M.L. De Feyter, S.F.E. Praet, K. Nicolay, L.J.C. van Loon and J.J.
Prompers. Normal in vivo skeletal muscle mitochondrial function in subjects with impaired
glycemic control and in long‐standing, insulin treated type 2 diabetes patients. ISMRM 15th
scientific meeting and exhibition, Berlin, Germany, 2007.
N.M.A. van den Broek, H.M.M.L. de Feyter, L. de Graaf, K. Nicolay, J.J. Prompers.
Intersubject differences in the effect of acidosis on phosphocreatine recovery kinetics after
exercise. ESMRMB 23rd annual scientific meeting, Warsaw, Poland, 2006.
166
Curriculum Vitae
167
Curriculum Vitae
Nicole van den Broek was born October 23, 1983, in Tilburg, The
Netherlands. She grew up in Diessen and graduated from secondary
school (gymnasium) at the Koning Willem II college in Tilburg in 2001.
In the same year she started her study Health Sciences at the
University of Maastricht. Her master project within the Department
of Movement Sciences of the afore mentioned university under
supervision of dr. Luc van Loon involved the investigation of the pH
dependence of different parameters for muscle oxidative capacity
measured by 31P magnetic resonance spectroscopy (MRS). This
project was performed in cooperation with the Biomedical NMR group of the Department
of Biomedical Engineering of the Eindhoven University of Technology. She obtained her
Master of Science degree in 2005. She continued her research in the Biomedical NMR
group as a junior research scientist and was involved in a study with diabetes patients. Half
a year later she started her PhD project in the same group under supervision of prof. dr.
Klaas Nicolay and dr. Jeanine Prompers. She studied the timing and nature of muscle
mitochondrial adaptations during the development of insulin resistance and type 2
diabetes using both in vivo MRS and in vitro methodology. Her research dealt with
methodological aspects of the MRS techniques as well as the application of these
techniques in human volunteers and animal models of insulin resistance.
168