Download The role of altered fatty acid metabolism in cardiomyopathy and

The role of altered fatty acid metabolism in
cardiomyopathy and heart failure : an in vivo magnetic
resonance imaging and spectroscopy approach
Abdurrachim, D.
DOI:
10.6100/IR782204
Published: 16/12/2014
Document Version
Publisher’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 differences
between the submitted version and the official published version of record. People interested in the research are advised to contact the
author 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
Citation for published version (APA):
Abdurrachim, D. (2014). The role of altered fatty acid metabolism in cardiomyopathy and heart failure : an in vivo
magnetic resonance imaging and spectroscopy approach Eindhoven: Technische Universiteit Eindhoven DOI:
10.6100/IR782204
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and 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 policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Download date: 07. May. 2017
The role of altered fatty acid metabolism in
cardiomyopathy and heart failure
An in vivo magnetic resonance imaging and
spectroscopy approach
Desiree Abdurrachim
A catalogue record is available from the Eindhoven University of Technology.
ISBN: 978-90-386-3732-7
The project described in this thesis was funded by a Vidi grant from the Netherlands
Organisation for Scientific Research (NWO), project number 700.58.421.
Financial support by the Dutch Heart Foundation for the publication of this thesis is
gratefully acknowledged. Financial contribution from Pie Medical Imaging BV is also
gratefully acknowledged.
Cover: MR images of non-diabetic mouse hearts at 12 weeks after transverse aortic
constriction
Cover design: Christiaan Sugiono and Desiree Abdurrachim
Printer: Ipskamp Drukker, Enschede, The Netherlands
Copyright © 2014 by Desiree Abdurrachim
All rights reserved. No part of this book may be reproduced, stored in a database or
retrieval system, or published, in any form or in any way, electronically, mechanically,
by print, photo print, microfilm, or any other means without prior permission by the
author.
The role of altered fatty acid metabolism in
cardiomyopathy and heart failure
An in vivo magnetic resonance imaging and
spectroscopy approach
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 dinsdag 16 december 2014 om 16:00 uur
door
Desiree Abdurrachim
geboren te Bandung, Indonesiё
Dit proefschrift is goedgekeurd door de promotor en de samenstelling van de
promotiecommissie is als volgt:
voorzitter:
prof.dr. P.A.J. Hilbers
promotor:
prof.dr. K. Nicolay
co-promotor:
dr. J.J. Prompers
leden:
prof.dr. J.W.A. Smit (Radboud Universiteit)
prof.dr. J.F.C. Glatz (Universiteit Maastricht)
prof.dr.ir. F.N. van de Vosse
prof.dr. C. Faber (Westfälische Wilhelms-Universität Münster)
dr. S.M. Houten (Icahn School of Medicine at Mount Sinai)
To my beloved parents
Ibu & Papay
Table of contents
Chapter 1
1
Introduction
Chapter 2
9
1
In vivo mouse cardiac MRI , H MRS, and
31
P MRS: Strategies and
current status in the study of cardiac function and metabolism – Review
Chapter 3
25
Good and bad consequences of altered fatty acid metabolism in heart failure:
Evidence from mouse models – Review
Chapter 4
47
High frame-rate retrospectively triggered cine MRI for assessment of
murine diastolic function
Chapter 5
63
In vivo quantification of mouse myocardial metabolite content and
T1 relaxation using 1H MRS
Chapter 6
In vivo
31
77
P MR spectroscopy of the mouse heart using respiratory-gated and
cardiac-triggered 3D ISIS
Chapter 7
93
Cardiac diastolic dysfunction in high-fat diet fed mice is associated with
lipotoxicity without impairment of cardiac energetics in vivo
Chapter 8
119
Reduced Cpt1b sensitivity to malonyl-CoA is associated with disturbed cardiac
energy metabolism and reduced cardiac function in a Cpt1bE3A mouse model
Chapter 9
145
Cardiac metabolic adaptations in diabetic mice prevent the heart from failing upon
pressure overload, as evidenced by a combined in vivo PET, MRI, and MRS approach
Chapter 10
169
Summarizing discussion and future perspective
Acknowledgements
177
List of publications
183
About the author
187
Chapter 1
Introduction
Chapter 1
Cardiovascular disease (CVD) is one of the leading causes of death worldwide (35). The
contribution of CVD to mortality has increased from 10% at the beginning of 20 th century to
30% in 2001 (13). Partly responsible for this increase may be the rising prevalence of diabetes,
as diabetes increases the risk of CVD by two to three times (19). The prevalence of diabetes has
now reached epidemic proportions, with 382 million people (~8.3% of the world’s adult
population) suffering from diabetes in 2013. This figure is expected to grow dramatically in the
next 22 years to an estimated 592 million patients in 2035 (12), making diabetes a growing
problem in today’s society.
The work in this thesis is within the context of diabetes and heart failure. Heart failure is the
end stage of heart diseases, and one of the most common causes of CVD morbidity and
mortality (13). In the following sections, cardiac metabolism, the characteristics of the heart in
diabetes and obesity, and the pathophysiology of heart failure are introduced. Finally, we
present the aim and framework of this thesis.
1.1.
The heart and cardiac metabolism
The heart pumps blood to supply oxygen and nutrients to the tissues in our body. The
importance of the heart was first recognized in the 4 th century BC, by the Greek philosopher
Aristotle from his observation in chick embryos that the heart was formed first before the rest of
the internal organs. However, the anatomy and physiology of the heart was not understood until
1628, when an English physician, William Harvey discovered that “blood by the beat of the
ventricles flows through the lungs and heart and is pumped to the whole body”. He concluded
that the blood in the animal body moves around in a circle continuously and the action or
function of the heart is to accomplish this by pumping (28). Furthermore, he concluded that this
is (the) only reason for the motion and beat of the heart (28).
The human heart beats on average 100,000 times per day, pumping through about 4.7 liters of
blood each minute. To maintain its work, the heart needs a continuous supply of energy, about
more than 6 kg of ATP per day. About 60-70% of the ATP is used for contractile function, and
30-40% is primarily used for maintaining calcium homeostasis and other cellular ion transport
(25, 30). To fuel the metabolism for energy production, the heart is an “omnivore” (22); it can
use various carbon substrates such as fatty acids (FA), glucose, ketone bodies, or lactate,
depending on their availability. However, in the normal heart, the major proportion (about 6070%) of the ATP requirement is provided by FA oxidation (FAO), while the remaining 30-40% is
supplied by glucose oxidation (22).
To adapt to substrate availability and energy demand, the heart is ‘metabolically flexible’ to
switch between FA and glucose (20, 26), an important feature that is usually lost in the
diseased heart, for example in the diabetic heart and in the failing heart. Compared with
glucose oxidation, FAO is more carbon-efficient; one mol of palmitate produces about 3-4 times
more ATP than one mol of glucose does. On the other hand, glucose oxidation is more oxygenefficient compared with FAO, producing 10-12% more ATP for the same amount of oxygen
consumed (22). Therefore, FA substrates are the preferred substrate for cardiac energy
production under conditions of high oxygen and low substrate availability, while utilization of
glucose is generally more favorable at low oxygen and high substrate concentrations.
2
Introduction
1.2.
The heart in diabetes and obesity
The heart in diabetes and obesity differs from the healthy heart. The diabetic heart has an
increased and almost exclusive (90-100%) dependence on FAO to provide energy for its
function (23). Furthermore, the diabetic heart is also thought to be metabolically inflexible to
switch from FA to glucose substrates (22, 32), making the heart more vulnerable to, for
example, conditions of low oxygen supply such as coronary artery disease. More importantly,
even without the presence of any underlying coronary artery diseases or hypertension, diabetic
patients are prone to reduced cardiac function. This condition is known as ‘diabetic
cardiomyopathy’, which is characterized by cardiac diastolic dysfunction in the early stage,
followed by systolic dysfunction and heart failure in later stages (11). Diabetic cardiomyopathy
is prevalent in two-third of diabetic patients (27). The reduced performance of diabetic hearts
may be induced by the alterations in cardiac metabolism in diabetes and obesity (5, 11, 23, 29).
Increased dependence on FAO in the diabetic heart has been associated with detrimental
effects, such as cardiac energy deficiency and accumulation of potentially toxic lipid metabolites,
which may contribute to the development of cardiomyopathy (4, 14). However, the exact
mechanisms and the relative contributions of impaired cardiac energetics and lipotoxicity to the
development of metabolic cardiomyopathy still need to be elucidated.
1.3.
Heart failure
When exposed to stress, the heart compensates for the increased work by thickening its muscle
to maintain a normal cardiac output. If the stress is severe or persistent, the heart will start to
dilate and ultimately fail (3). At failure, the heart cannot pump sufficient blood through the
body. The transition from a normal to a failing heart is usually accompanied by a shift in cardiac
metabolism towards increased glucose oxidation at the expense of FAO (22). However, the
availability of substrates in the blood may still modulate substrate metabolism in the failing
heart. For example, in the presence of high levels of FA, the failing heart exhibits more FAO,
while when the concentration of ketone bodies is increased, FAO is usually suppressed (22).
Moreover, comorbidities such as insulin resistance and diabetes also complicate cardiac
metabolism in heart failure. In heart failure patients without insulin resistance, FAO rate was
shown to be reduced compared with that of healthy volunteers, while in heart failure patients
with insulin resistance, FAO rate was shown to be unchanged (33).
As it is generally accepted that the shift towards glucose oxidation is an adaptive mechanism to
cope with heart failure, heart failure treatments usually aim to inhibit FAO and to stimulate
glucose oxidation. In the diabetic heart, however, the suggested metabolic inflexibility may
prevent a switch to glucose, making the diabetic heart even more vulnerable to failure. On the
other hand, recent evidence suggests that maintained FAO could be protective during heart
failure (21). Therefore, it is not clear whether the shift to glucose oxidation is a cause or a
consequence of heart failure progression. Moreover, metabolic interventions may need to be
tailored to different cardiac metabolic conditions (18), for example due to different metabolic
milieu (e.g. diabetes) or different stages of heart failure. However, a comprehensive view of
cardiac metabolic changes during different stages of heart failure progression and under
different metabolic conditions is currently lacking. Progress in this field critically depends on the
development of non-invasive tools to study cardiac metabolism in vivo.
3
Chapter 1
Figure 1.1. Probing cardiac metabolism and function in vivo using MRS and MRI.
In the heart, fatty acids (FA) and glucose are two main fuel sources to produce ATP. FAO fulfills 60-70%
of the energy requirement of the healthy heart, while the remaining 30-40% is mainly provided by
glucose. Upon uptake of FA into cardiomyocytes, FA can either be stored as lipid droplets for later use,
or oxidized through β-oxidation in the mitochondria to produce energy in the form of ATP. The
concentration of lipids stored in the myocardium can be quantified from cardiac 1H MR spectra, acquired
in the heart septum (voxel shown in white). Using 31P MRS, the ratio of PCr to ATP in the left ventricle
can be quantified as a measure of cardiac energy status. Modulations in the FA metabolic pathways may
result in increased FAO in mouse models, for example through the increase in FA availability in the
blood, such as through high fat diet feeding (1) or mutation in leptin receptor (3), or more downstream
without affecting FA availability and uptake, through the reduction of Cpt1b sensitivity to malonyl-CoA
(2). Malonyl-CoA normally inhibits CPT1 activity, an enzyme which transports FA into the mitochondria.
The effects of altered FAO on cardiac function can be studied using MRI, by acquiring cine movies of the
contracting heart. From such movies, volumetric changes of left ventricular cavity over time within the
cardiac cycle can be calculated. This allows the assessment of systolic function (e.g. EDV, ESV, PER, and
ejection fraction) and diastolic function (e.g. E and A filling rates, shown by the slopes of the red lines).
1
H MRS, 31P MRS, and MRI are therefore powerful tools to study the effects of modulations in cardiac
metabolism, for example in diabetic or obesity-related cardiomyopathy, in which FA oxidation is altered.
ADP: adenosine diphosphate, ATP: adenosine triphosphate, CKc: cytosolic creatine kinase, CKm: mitochondrial creatine
kinase, Cr: creatine, DPG: diphosphoglycerol, EDV: end diastolic volume, ESV: end systolic volume, FAT: fatty acid
translocase, GLUT4: glucose transporter 4, OXPHOS: oxidative phosphorylation, PCr: phosphocreatine, PER: peak
ejection rate, TCA: tricarbocylic acid cycle.
4
Introduction
1.4.
Framework of this thesis
Non-invasive in vivo techniques are of great importance in the research on cardiac function and
metabolism. In vivo measurements allow longitudinal studies to map cardiac metabolic
adaptations at different stages of cardiomyopathy or heart failure progression in the same
animal. Currently available data on cardiac energy metabolism in animal studies mainly
originate from ex vivo isolated perfused heart setups (16, 24). However, these ex vivo
experimental setups do not entirely mimic the complexity of the in vivo situation as the ex vivo
heart is usually perfused with glucose and FA at a constant concentration, In obesity, diabetes,
and heart failure, the concentration of specific substrates and changes in their proportions in the
blood may influence myocardial substrate metabolism.
This thesis aimed to investigate the role of altered cardiac FA metabolism in the development
of metabolic cardiomyopathy and heart failure in mouse models of altered FAO. To this end, we
developed and implemented in vivo magnetic resonance imaging (MRI), 1H magnetic resonance
spectroscopy (1H MRS), and
31
P MRS to study cardiac systolic and diastolic function, myocardial
lipid content, and cardiac energetics, respectively (Fig. 1.1). To gain more insight into cardiac
metabolism, in vivo MRI and MRS data in this thesis were complemented with measurements of
myocardial glucose uptake using in vivo positron emission tomography (PET), substrate
metabolism in ex vivo isolated perfused hearts, ex vivo function of isolated cardiac
mitochondria, myocardial lipid intermediates, and the expression of proteins involved in cardiac
metabolism.
We applied the in vivo MRI and MRS techniques to study the effects of altered FAO in three
different mouse models, with different origins of FA modulation: high fat diet-induced obese
mice, malonyl-CoA insensitive Cpt1b knockin mice, and diabetic db/db mice.
High fat diet (HFD)-induced obese mice
HFD feeding is an intervention to induce a pre-diabetic stage. HFD feeding results in increased
FA availability in the blood, which leads to an upregulation of myocardial FA uptake and FAO.
HFD-fed mice develop cardiomyopathy, the severity of which is dependent on diet compositions
and feeding duration (10).
Malonyl-CoA insensitive Cpt1b knockin mice
Cpt1b is an enzyme which imports FA into the mitochondria. Its activity is inhibited by malonylCoA (Fig. 1.1). The Cpt1bE3A knockin mouse is a novel mouse model, in which Cpt1b is rendered
less sensitive to malonyl-CoA. This results in an increased FAO flux in Cpt1bE3A knockin mouse
hearts, which is accompanied by cardiomyopathy. However, in contrast to HFD-induced obese
mice, whole-body energy balance is not affected in Cpt1bE3A knockin mice (34).
Diabetic db/db mice
Increased FAO in diabetic db/db mice is due to a point mutation in the leptin receptor (8),
leading to disturbed energy regulation. Diabetic db/db mice are morbidly obese, and develop a
similar diabetic phenotype as seen in human diabetic patients (17). Adipose tissue lipolysis is
increased in db/db mice, causing FA availability in the blood, FA uptake into the heart, and
5
Chapter 1
cardiac FAO to be increased (6, 7). Db/db mice have been extensively used to investigate the
etiology of diabetic cardiomyopathy (1, 2, 9, 15, 31, 36).
1.5.
Thesis outline
The first part of this thesis presents reviews of current literature on in vivo MRI and MRS
techniques and the role of altered FAO in cardiomyopathy and heart failure, which are given in
Chapter 2 and Chapter 3, respectively. MRI is the gold standard method to assess cardiac
systolic function, however the time resolution of routine cardiac MRI is usually not sufficient to
visualize the two diastolic filling phases necessary for the assessment of diastolic function. In
chapter 4, the development of high temporal resolution cardiac MRI to measure diastolic
function is described. Reduced cardiac function in the context of altered FAO have been
associated with myocardial lipid accumulation and impaired cardiac energetics. In Chapter 5,
we describe the implementation of a localized
1
H MRS technique for the quantification of
myocardial lipid content in a single voxel in the septum of the in vivo mouse heart. Chapter 6
reports the implementation of a localized
31
P MRS technique to determine the cardiac PCr/ATP
ratio, as a measure of cardiac energy status, in mice.
Using the developed high temporal resolution cardiac MRI, 1H MRS, and
31
P MRS methods, we
investigated the effects of altered cardiac FA uptake and FAO promoted by high-fat diet (HFD)
feeding in mice, which is described in Chapter 7. In Chapter 8, we report the characterization
of cardiac function and metabolism in the Cpt1bE3A knockin mouse. Finally, Chapter 9 presents
a longitudinal study to map metabolic changes during the progression of heart failure in diabetic
and non-diabetic mice, using a combination of MR and PET imaging. The thesis is concluded
with a summarizing discussion and future perspective in Chapter 10.
1.6. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
6
Aasum E, Belke DD, Severson DL, Riemersma RA, Cooper M, Andreassen M, and Larsen TS.
Cardiac function and metabolism in Type 2 diabetic mice after treatment with BM 17.0744, a novel
PPAR-alpha activator. Am J Physiol Heart Circ Physiol 283: H949-957, 2002.
Aasum E, Hafstad AD, Severson DL, and Larsen TS. Age-dependent changes in metabolism,
contractile function, and ischemic sensitivity in hearts from db/db mice. Diabetes 52: 434-441, 2003.
Berry JM, Naseem RH, Rothermel BA, and Hill JA. Models of cardiac hypertrophy and transition to
heart failure. Drug Discovery Today: Disease Models 4: 197-206, 2008.
Boudina S, and Abel ED. Mitochondrial uncoupling: a key contributor to reduced cardiac efficiency in
diabetes. Physiology (Bethesda) 21: 250-258, 2006.
Boyer JK, Thanigaraj S, Schechtman KB, and Perez JE. Prevalence of ventricular diastolic
dysfunction in asymptomatic, normotensive patients with diabetes mellitus. Am J Cardiol 93: 870-875,
2004.
Carley AN, Atkinson LL, Bonen A, Harper ME, Kunnathu S, Lopaschuk GD, and Severson DL.
Mechanisms responsible for enhanced fatty acid utilization by perfused hearts from type 2 diabetic
db/db mice. Arch Physiol Biochem 113: 65-75, 2007.
Carley AN, and Severson DL. Fatty acid metabolism is enhanced in type 2 diabetic hearts. Biochim
Biophys Acta 1734: 112-126, 2005.
Chua SC, Jr., Chung WK, Wu-Peng XS, Zhang Y, Liu SM, Tartaglia L, and Leibel RL. Phenotypes
of mouse diabetes and rat fatty due to mutations in the OB (leptin) receptor. Science 271: 994-996,
1996.
Daniels A, van Bilsen M, Janssen BJA, Brouns AE, Cleutjens JPM, Roemen THM, Schaart G,
van der Velden J, van der Vusse GJ, and van Nieuwenhoven FA. Impaired cardiac functional
reserve in type 2 diabetic db/db mice is associated with metabolic, but not structural, remodelling.
Acta Physiol 200: 11-22, 2010.
Dirkx E, Schwenk RW, Glatz JF, Luiken JJ, and van Eys GJ. High fat diet induced diabetic
cardiomyopathy. Prostaglandins Leukot Essent Fatty Acids 85: 219-225, 2011.
Introduction
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Fang ZY, Prins JB, and Marwick TH. Diabetic cardiomyopathy: Evidence, mechanisms, and
therapeutic implications. Endocr Rev 25: 543-567, 2004.
International Diabetes Federation. IDF Diabetes Atlas, 6th edn. Brussels, Belgium: International
Diabetes Federation, 2013.
Gaziano T, Reddy KS, Paccaud F, Horton S, and Chaturvedi V. Cardiovascular Disease. In:
Disease control priorities in developing countries. Washington, DC: World Bank Publications, 2006.
Goldberg IJ, Trent CM, and Schulze PC. Lipid metabolism and toxicity in the heart. Cell Metab 15:
805-812, 2012.
Greer JJ, Ware DP, and Lefer DJ. Myocardial infarction and heart failure in the db/db diabetic
mouse. Am J Physiol Heart Circ Physiol 290: H146-153, 2006.
How OJ, Aasum E, Severson DL, Chan WY, Essop MF, and Larsen TS. Increased myocardial
oxygen consumption reduces cardiac efficiency in diabetic mice. Diabetes 55: 466-473, 2006.
Hummel KP, Dickie MM, and Coleman DL. Diabetes, a new mutation in the mouse. Science 153:
1127-1128, 1966.
Ingwall JS. Energy metabolism in heart failure and remodelling. Cardiovasc Res 81: 412-419, 2009.
Kannel WB, and McGee DL. Diabetes and cardiovascular risk factors: the Framingham study.
Circulation 59: 8-13, 1979.
Kessler G, and Friedman J. Metabolism of fatty acids and glucose. Circulation 98: 1351, 1998.
Kolwicz SC, Jr., Olson DP, Marney LC, Garcia-Menendez L, Synovec RE, and Tian R. Cardiacspecific deletion of acetyl CoA carboxylase 2 prevents metabolic remodeling during pressure-overload
hypertrophy. Circ Res 111: 728-738, 2012.
Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, and Stanley WC. Myocardial fatty acid
metabolism in health and disease. Physiol Rev 90: 207-258, 2010.
Mandavia CH, Pulakat L, DeMarco V, and Sowers JR. Over-nutrition and metabolic
cardiomyopathy. Metabolism 61: 1205-1210, 2012.
Mazumder PK, O'Neill BT, Roberts MW, Buchanan J, Yun UJ, Cooksey RC, Boudina S, and
Abel ED. Impaired cardiac efficiency and increased fatty acid oxidation in insulin-resistant ob/ob
mouse hearts. Diabetes 53: 2366-2374, 2004.
Nagoshi T, Yoshimura M, Rosano GM, Lopaschuk GD, and Mochizuki S. Optimization of cardiac
metabolism in heart failure. Curr Pharm Des 17: 3846-3853, 2011.
Randle PJ, Garland PB, Hales CN, and Newsholme EA. The glucose fatty-acid cycle. Its role in
insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1: 785-789, 1963.
Redfield MM, Jacobsen SJ, Burnett JC, Jr., Mahoney DW, Bailey KR, and Rodeheffer RJ.
Burden of systolic and diastolic ventricular dysfunction in the community: appreciating the scope of the
heart failure epidemic. JAMA 289: 194-202, 2003.
Ribatti D. William Harvey and the discovery of the circulation of the blood. Journal of Angiogenesis
Research 1: 1-2, 2009.
Rider OJ, Cox P, Tyler D, Clarke K, and Neubauer S. Myocardial substrate metabolism in obesity.
Int J Obes 2012.
Stanley WC, Recchia FA, and Lopaschuk GD. Myocardial substrate metabolism in the normal and
failing heart. Physiol Rev 85: 1093-1129, 2005.
Stuckey DJ, Carr CA, Tyler DJ, Aasum E, and Clarke K. Novel MRI method to detect altered left
ventricular ejection and filling patterns in rodent models of disease. Magn Reson Med 60: 582-587,
2008.
Taegtmeyer H, Golfman L, Sharma S, Razeghi P, and van Arsdall M. Linking gene expression to
function: metabolic flexibility in the normal and diseased heart. Ann N Y Acad Sci 1015: 202-213,
2004.
Tuunanen H, Engblom E, Naum A, Scheinin M, Nagren K, Airaksinen J, Nuutila P, Iozzo P,
Ukkonen H, and Knuuti J. Decreased myocardial free fatty acid uptake in patients with idiopathic
dilated cardiomyopathy: evidence of relationship with insulin resistance and left ventricular
dysfunction. J Card Fail 12: 644-652, 2006.
van Weeghel M. Fatty acid oxidation in health and disease. PhD thesis. Universiteit van Amsterdam,
2013.
WHO (World Health Organization). The World health report: 2002: Reducing the risks, promoting
healthy life. WHO (World Health Organization), 2002.
Yue P, Arai T, Terashima M, Sheikh AY, Cao F, Charo D, Hoyt G, Robbins RC, Ashley EA, Wu J,
Yang PC, and Tsao PS. Magnetic resonance imaging of progressive cardiomyopathic changes in the
db/db mouse. Am J Physiol Heart Circ Physiol 292: H2106-2118, 2007.
7
Chapter 2
In vivo mouse cardiac MRI, 1H MRS, and
31
P MRS:
Strategies and current status in the study of cardiac
function and metabolism
Review
Partly based on:
Small animal MR imaging and spectroscopy of cardiovascular disorders
Desiree Abdurrachim, Adrianus J. Bakermans, Rik P.M. Moonen, Abdallah G. Motaal,
Jeanine J. Prompers, Gustav J. Strijkers, Katrien Vandoorne, Klaas Nicolay
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven,
The Netherlands
Invited review in Progress in NMR Spectroscopy
Chapter 2
2.1. Introduction
Magnetic resonance imaging (MRI) is a non-invasive imaging technology which exploits
magnetic properties of nuclear spins, particularly water protons. MRI is characterized by its
excellent soft-tissue contrast and high spatial resolution, as well as its versatility, as many
different morphological and functional parameters can be measured. In cardiac applications,
MRI has been widely used to assess global cardiac function (77), local cardiac function (18), and
tissue perfusion (88).
While the aforementioned applications derive the information from MR images, magnetic
resonance spectroscopy (MRS) draws information from MR spectra. MRS exploits the different
magnetic environments, which cause nuclei (such as 1H,
31
P,
13
C,
23
Na) in different chemical
environments to appear at different frequencies in the corresponding MR spectra. For example,
in a 1H MR spectrum, creatine protons resonate at different frequencies than triglyceride protons
(Fig. 2.3), while in a
31
P MR spectrum, the phosphorus nucleus in phosphocreatine resonates at
a different frequency than the phosphorus nuclei in ATP (Fig. 2.5). MRS thereby provides a
powerful tool to quantify metabolites and metabolic processes involved in metabolism (reviewed
in (71)).
Initially, cardiac MRI and MRS were developed for human applications. The development of
mouse models to study myocardial metabolism required development of those MRI and MRS
techniques also for the mouse heart. However, major challenges in the in vivo implementation
of mouse cardiac MRI and MRS are the small size and the high heart rate of the mouse heart.
Compared with the human heart, the mouse heart weighs about 2000 times less and beats 10
times
faster.
Moreover,
respiration
also
generates
movement
that
can
disturb
the
measurements. These factors put high demands on acquisition speed, spatial resolution,
sensitivity, as well as the synchronization of the acquisition with the heart and breathing
movement.
Here we review the current status of cardiac cinematic MRI, 1H MRS, and
31
P MRS applications in
vivo in small animals, particularly in mice.
2.2. Cardiac cinematic MRI
Cardiac cinematic (cine) MRI is a technique to acquire a set of cardiac images throughout the
cardiac cycle, thereby generating a movie of the beating heart. Currently, cine MRI is the gold
standard for cardiac function assessment in vivo (19). It has been applied in various murine
models of cardiomyopathy, such as diabetic cardiomyopathy (2, 49, 91), myocardial infarction
(11, 15, 61, 72), and pressure overload-induced heart failure (76, 89). The parameters of
cardiac function are quantified from either 3D cine movies (24, 54), or more commonly, from
cine movies of consecutive short axis slices of the heart, complemented with cine movies of 4chamber and/or 2-chamber long axis slices of the heart (77, 86). The acquisition is typically
performed using gradient echo-based sequences with a low flip angle excitation pulse to allow a
short repetition time (TR). Some examples of these sequences are FLASH (fast low-angle shot)
(28, 38), SSFP (steady-state free precision) (81), and UTE (ultra-short echo time) (42, 58). For
cine MRI, synchronization of the acquisition with the cardiac cycle is essential. To this end, two
10
Cardiac MRI, 1H MRS,
31
P MRS – review
cardiac triggering strategies can be employed, either by using an external ECG signal to
prospectively trigger the acquisition (prospective triggering, Fig. 2.1A-D), or by using an
internal navigator signal acquired during the acquisition, to assign acquired data retrospectively
to the correct cardiac phases (retrospective triggering, Fig. 2.1E-G). The acquisition is
performed during multiple heart beats, which assumes periodicity of the cardiac movement.
2.2.1. Prospective triggering
The prospective triggering method requires the use of an ECG signal (Fig. 2.1A), recorded from
ECG electrodes attached to the paws or the chest of the animal. The acquisition of the first
cardiac frame starts once the ECG R-wave is detected, which generates a trigger (Fig. 2.1C-D).
Within one cardiac cycle, the acquisition is repeated after each TR until nearly the end of the
cardiac cycle. In this case, the number of frames that can be acquired within one cardiac cycle
is defined by the length of the cardiac cycle period divided by the TR.
Within one cardiac cycle, the acquisition fills one k-line of each cardiac frame. In case of parallel
imaging, a number of k-lines can be filled simultaneously. The acquisition is then repeated over
a number of heart beats, until all k-lines of all cardiac frames are filled. One limitation of the
prospective triggering method is that data from the very end of the diastolic phase is usually not
acquired (Fig. 2.1C). The number of frames can be increased so that the acquisition continues
into the next cardiac cycle; however, this will double the acquisition time.
To prevent artefacts due to respiratory motion, the acquisition is often respiratory gated (Fig.
2.1B). The respiratory signal is usually recorded using a pressure balloon positioned on the
abdomen or the back of the animal. Respiratory gating causes pauses in the acquisition, leading
to variations in TR and therefore, disturbances in steady-state magnetization. To maintain
steady-state magnetization, dummy scans can be applied, in which excitation pulses are
continued during respiratory gating, however, without data acquisition (12). However,
maintaining steady-state magnetization during prospective triggering can still be challenging,
because missing or undetected ECG triggers can also cause occasional pauses, which disrupt
maintenance of steady-state magnetization.
2.2.2. Retrospective triggering
In contrast to prospective triggering, with the retrospective triggering strategy, data is acquired
continuously, asynchronous with the cardiac cycle (Fig. 2.1E). After acquisition is completed,
the cine movie is reconstructed by binning the data into a specified number of cardiac frames.
The retrospective triggering scheme utilizes an integral ‘navigator signal’ to assign the data
retrospectively to the corresponding cardiac phase. For this purpose, the sequence is modified
to include the acquisition of a navigator signal (Fig. 2.1F). The navigator signal can be acquired
in the same slice during the refocusing of the slice signal (as shown in Fig. 2.1F) (38), in
another slice (42), or directly from the imaging data, by recording the fluctuations in the center
of k-space when using a radial trajectory (20, 39, 42). The intensity of the navigator signal
reflects both cardiac and respiratory motion (Fig. 2.1G).
11
Chapter 2
Figure 2.1. Strategy to synchronize cine MRI acquisition with cardiac cycle. (A) ECG and (B) respiratory
signal, recorded externally using ECG electrodes and pressure balloon, respectively. (C) In the
prospective triggering scheme, the acquisition is synchronized with the external ECG signal. The number
of frames N per cardiac cycle is defined by the cardiac cycle period L divided by the repetition time TR.
During respiratory movement, one can choose to pause the acquisition, or to acquire dummy scans (in
lighter color) to maintain steady-state magnetization. Cine MRI generated by prospective triggering
often misses the frame at the end of cardiac phase. (D) FLASH sequence typically used with prospective
triggering. (E) In the retrospective triggering scheme, data is acquired continuously, asynchronously
with the cardiac activity. (F) The FLASH sequence is modified to include the acquisition of a navigator
signal (representative navigator signal amplitude is shown in G). The navigator signal is used to
retrospectively bin the data into a number of frames N for cine reconstruction. Continuous data
acquisition, with small variations in the cardiac cycle, creates high probability that data is acquired at
different positions within each cardiac cycle (as illustrated by the two rows in E). In contrast to
prospective triggering, this allows the reconstruction of the data into a higher number of frames, which
is no longer constrained by TR.
Cardiac functional parameters determined from retrospectively triggered cine MRI have been
shown to be comparable with those acquired using prospectively triggered cine MRI (11, 38).
However, retrospective triggering has a few advantages over prospective triggering. First,
because the data is acquired continuously, the number of reconstructed cardiac frames can be
decided after the measurement. More importantly, the number of frames is not limited by
repetition time, which allows reconstruction with higher time resolution (>60 frames/cardiac
cycle), as described in Chapter 4 in this thesis (14). However, due to the stochastic nature of
the k-space filling in retrospective triggering, we have shown that there is a minimum required
number of repetitions to ensure that a sufficient number of k-lines in the k-space of each image
12
Cardiac MRI, 1H MRS,
31
P MRS – review
frame are filled (14). This minimum number of repetitions is dependent on the number of
frames one wishes to reconstruct, which means that a longer acquisition time is required for a
higher time resolution. For high time resolution reconstructions to be practical, acceleration
techniques such as compressed sensing (55, 59) or parallel imaging (74) can significantly
reduce the measurement time.
The continuous nature of the retrospectively triggered acquisition ensures the maintenance of
steady-state magnetization, which is reflected by less signal variations between cardiac frames
(38). This is particularly important for quantitative imaging, for example, when mapping T1
relaxation time. Finally, retrospective triggering eliminates the needs for external ECG signals.
This can be beneficial when ECG signal is of low quality, such as in mouse models of heart
failure or myocardial infarction (16). However, the use of retrospective triggering needs a
reliable navigator signal, which, as we have experienced in our laboratory, can also be
problematic, e.g. in animals with obesity or myocardial infarction (16).
2.2.3. Cine MRI application
The most common parameters derived from cine MRI are the parameters for systolic function.
From the segmentation of the left ventricular (LV) myocardium and cavity (Fig. 2-2A), LV
myocardial mass, end-diastolic volume (EDV), and end-systolic volume (ESV) can be derived.
From this data, systolic function parameters can be calculated, such as stroke volume (SV) =
EDV-ESV, ejection fraction (EF) = (EDV-ESV)/EDV, and cardiac output = SV x heart rate.
Next to systolic function, diastolic function is also an important parameter in cardiac function.
Diabetic or obesity-related cardiomyopathy is often preceded by diastolic dysfunction (23).
Furthermore, heart failure can also occur with preserved systolic function, while diastolic
function is reduced (29). The diastolic phase of the cardiac cycle is characterized by two
separate filling phases: the early filling (E) phase, which is due to active LV relaxation, and the
late filling (A) phase, which is due to atrial contraction. The E rate, A rate, or E/A ratio are
commonly determined by echocardiography (73); unfortunately, the time resolution of cine MRI
in most cases is too low to separate the two diastolic filling phases. Routine cine MRI with
prospective triggering results in a time resolution of typically ~15-25 frames/cardiac cycle,
while the two diastolic filling phases can only be differentiated with a time resolution of ~50-60
frames/cardiac cycle (14) (Fig. 2.2B and C).
The time resolution in the prospective triggering method is constrained by the TR. For the
gradient echo sequence, TR is limited by the lengths of the excitation pulse, the spatial encoding
gradients, spoiler gradients, and echo readout. TR therefore becomes constrained by the desired
spatial resolution, gradient performance, and SNR. At a given spatial resolution, TR can only be
made shorter at the expense of a shorter excitation pulse with a higher slice selection gradient,
shorter and higher spatial encoding and spoiler gradients, and a shorter echo readout with
higher readout gradient and higher receiver bandwidth. However, gradient rise-time and
strength are hardware limited, and increasing the gradient strength may cause gradient duty
cycle problems. Also, a higher receiver bandwidth will lead to lower SNR. Therefore, to achieve
the high time resolution required for diastolic function assessment, a high-performance gradient
system is necessary, such as used in the study of Stuckey et al. of diabetic murine models (82).
13
Chapter 2
As an alternative strategy to obtain high time-resolution cine MRI data, we implemented
retrospective triggering with a long acquisition time (14). As reported in Chapter 4 in this thesis,
we showed that the retrospective triggering method allowed the reconstruction of up to 90
frames/cardiac cycle, revealing a subtle impairment in diastolic function in diabetic mice (14).
We further extended the technique by the use of compressed sensing to accelerate the
acquisitions (59).
Figure 2.2. Analysis of cine MRI for the quantification of cardiac systolic and diastolic function.
(A) Series of cardiac images at different cardiac phases. Blue and red represent the segmentation of
the myocardium and LV cavity, respectively, from which end diastolic volume (EDV) and end systolic
volume (ESV) can be quantified. (B) LV cavity volume-time curve, in which the early (E) and late peak
diastolic filling (A) phases are visible. Red lines indicate the slopes (i.e. rates) of the diastolic filling.
(C) The first derivative of the LV cavity volume-time curve in (B), depicting the LV peak ejection rate
(PER), E rate, and A rate.
2.3. 1H MRS
Proton magnetic resonance spectroscopy (1H MRS) is a powerful tool to gain insight into cardiac
lipid and creatine metabolism (21). Cardiac 1H MR spectra contain a number of peaks associated
with triglycerides (TG) and peaks associated with taurine, choline/carnitine, and creatine (30)
(Fig. 2.3A). The metabolite concentration can be quantified by calculating the peak area,
expressed for example relatively to the water concentration. The accumulation of myocardial TG
has been implicated in the etiology of diabetes- and obesity-related cardiomyopathy (36, 83),
while creatine is an interesting metabolite in heart failure due to its role in buffering the ATP
levels via the creatine kinase system (10, 62, 63).
In the context of lipotoxic cardiomyopathy, cardiac 1H MRS has been applied in vivo in diabetic
db/db mice, which showed increased myocardial TG levels (Fig. 2.3B, unpublished data).
Myocardial TG content has also been measured using 1H MRS in high fat diet mice (1, 36) and
14
Cardiac MRI, 1H MRS,
31
P MRS – review
Figure 2.3. Typical cardiac 1H MR spectra acquired from myocardial septum in (A) a
non-diabetic mouse, (B) a diabetic mouse. The spectra were identically scaled and
contain peaks originating from taurine (tau), choline/carnitine (cho/car), creatine-CH3
(Cr-CH3), and 7 peaks associated with triglycerides (TG): (1) -CH=CH-; (2) -CH=CHCH2-CH=CH-; (3) -CαH2COO; (4) -CH2-CH=CH-CH2-; (5) -CβH2CH2COO; (6) -CH2-;
(7) -CH3. (C) The correlation between TG-CH2 quantified using 1H MRS and TG content
determined biochemically (data are from the studies in (6, 7)).
rats (60), and long-chain acylcarnitine dehydrogenase (LCAD) knockout mice (6, 7), in which
increased myocardial TG content was associated with cardiac hypertrophy and/or dysfunction.
We have previously shown that myocardial TG content measured using in vivo cardiac 1H MRS is
in good agreement with the myocardial TG content determined biochemically (6) (Fig. 2.3C).
For the study of creatine metabolism, cardiac 1H MRS has been applied in mice with abnormal
creatine levels (70, 80), and mice with myocardial infarction (22, 51).
Cardiac 1H MRS data are typically acquired from a 2-4 μL voxel localized in the myocardial
septum during diastole (6, 80) (Fig. 2.4A and B). To synchronize the measurements with
cardiac activity, prospective triggering is used with a certain delay set after the R peak (Fig.
15
Chapter 2
2.4C). Two single-voxel single-shot localization techniques commonly used for
1
H MRS are
PRESS (point resolved spectroscopy; Fig. 2.4D) (9) and STEAM (stimulated echo acquisition
mode; Fig. 2.4E) (26, 27). As the concentration of cardiac proton metabolites is less than 2% of
the water concentration, it is necessary to apply water suppression prior to the PRESS or STEAM
acquisition (Fig. 2.4C). One way to realize this is to use the CHESS (chemical shift selective)
sequence to saturate the water signal by 90° pulse excitation followed by spoiling gradients,
which is usually repeated a number of times. For the purpose of metabolite quantification, the
maintenance of steady-state magnetization is important. This can be achieved by applying
dummy scans during respiratory gating (6, 80) (Fig. 2.4C). In this thesis, the implementation of
PRESS for cardiac metabolite quantification (3) is described in detail in Chapter 5.
Figure 2.4. Diagram for cardiac 1H MRS. (A) Voxel localization in the septum, seen in a 4chamber long axis view and (B) in a short axis view. (C) Timing diagram with respect to ECG
triggering and respiratory gating. Sequence diagram of (D) PRESS and (E) STEAM. Δ: delay after
the trigger, WS: water suppression, P: PRESS or STEAM acquisition, D: dummy scan, TE: echo
time. Adapted from (3).
In humans, both STEAM (10, 75) and, more frequently, PRESS (63, 79, 83, 87) have been
employed to assess myocardial metabolites with localized 1H MRS. In mice, however, almost all
studies were performed using PRESS (3, 6, 21, 51, 70, 80). In both PRESS and STEAM
methods, three frequency-selective radiofrequency (RF) pulses together with three mutually
orthogonal field gradients provide 3D spatial localization in a single scan (Fig. 2.4D and E).
Unwanted transversal magnetization is dephased by crusher gradients, which are balanced in
pairs to ensure proper refocusing of the (stimulated) echo. The need for proper refocusing
makes echo-based sequences such as PRESS and STEAM sensitive to signal loss and signal
contamination due to motion. In this regard, PRESS is less sensitive to motion than STEAM,
because the crusher gradients in PRESS are applied directly around the refocusing pulses, while
in STEAM, the crusher gradients are separated by the mixing time (TM) (57). We have
previously shown that PRESS has a higher signal stability compared to STEAM in the mouse
16
Cardiac MRI, 1H MRS,
31
P MRS – review
heart (4). Furthermore, PRESS resulted in higher SNR compared to STEAM (4). In STEAM, half
of the potential signal is lost after the second 90º pulse, which rotates only half of the
transversal magnetization to the longitudinal axis. This is partially compensated by less T2
relaxation-associated signal loss in STEAM compared to PRESS, because the echo time (TE) in
STEAM is shorter than for PRESS. Although for the aforementioned reasons PRESS might be
more preferable than STEAM, the total duration of single-shot localization with STEAM is shorter
than with PRESS, which may be advantageous for localized 1H-MRS of rapidly moving tissue
such as the heart. Moreover, at higher magnetic field strengths, STEAM may be a favorable
approach considering the higher bandwidth that can be achieved with 90º pulses compared with
180º pulses.
2.4.
31
P MRS
Failing hearts have been associated with low cardiac energy levels (37, 65), analogous to ‘an
engine out of fuel’ (64). Phosphorus magnetic resonance spectroscopy (31P MRS) is uniquely
capable to provide insight into in vivo cardiac energy metabolism, in which phosphocreatine
(PCr) and ATP are the energy currencies. ATP transfer from mitochondria to the utilization site
occurs via the creatine kinase (CK) shuttle. PCr acts as an energy buffer, ensuring sufficient ATP
levels to meet the energy needs. Therefore, the PCr/ATP ratio quantified from cardiac
31
P MR
spectra provides a measure of cardiac energy status (64).
Cardiac
31
P MRS has been applied to study cardiomyopathy or heart failure in mice (1, 5, 34,
67, 85, 90) and rats (50, 68) in vivo, which contributed to the understanding of cardiac
energetics in heart failure (reviewed in (52)). Various readouts of cardiac energetics have been
made available using the application of cardiac
31
P MRS. Apart from the measurement of the
PCr/ATP ratio (13, 67), techniques to determine absolute concentrations of PCr and ATP (34,
41) and ATP turnover kinetics (8, 33) have been implemented. It has been shown that in mice
with heart failure, both PCr and ATP concentrations are reduced, with the degree of reduction
being larger for PCr, resulting in decreased PCr/ATP ratio (34). The lower cardiac energy status
has been linked to a lower CK flux in these mice (33). More importantly, the overexpression of
CK was shown to increase the CK flux and lead to a better survival after heart failure,
suggesting CK as a potential therapeutic target for heart failure (32, 35).
General challenges of cardiac MR in small animals, particularly mice, and low intrinsic sensitivity
of
31
P MRS (6.65% of 1H MRS), limited the implementation of in vivo mouse cardiac
only a few laboratories worldwide (5, 13, 25, 67). Cardiac
31
31
P MRS in
P MRS in small animals was first
implemented in the 1980s in rats, in an open chest application using a solenoid coil surrounding
the heart (47, 66) or in vivo using surface coil localization (31, 84). To date, along with the
advancement in MR hardware and technology and with the availability of higher magnetic fields,
in vivo cardiac
31
P MRS has been implemented non-invasively in mice, using single or multiple
voxel localization techniques. Typically used techniques are 3D image selected in vivo
spectroscopy (ISIS) (5, 67)), 1D chemical shift imaging (CSI) (13), and 2D CSI (25).
Single-voxel localization for cardiac
31
P MRS is usually performed using 3D ISIS, which consists
of three slice-selective 180° pulses for localization followed by a non-selective 90° pulse for
signal detection (69). The localized signal is then obtained by addition and subtraction of signals
17
Chapter 2
obtained from 8 experiments, with different combinations of the three 180° pulses being turned
on or off. Therefore, 3D ISIS is a multi-shot localization technique, which is prone to imperfect
localization in case of motion. For this reason, accurate triggering and gating are very crucial for
the implementation of 3D ISIS in the beating heart. Using 3D ISIS, cardiac
31
P MR spectra are
acquired from a voxel which usually contains the left ventricular myocardium as well as cavity
(Fig. 2.5A). Typical spectra contain a PCr peak originating from myocardium, γ-, α-, β-ATP peaks
from myocardium and blood, and peaks from 2,3-diphosphoglycerate (2,3-DPG) from blood,
which usually obscures the inorganic phosphate (Pi) signal from myocardium (Fig. 2.5B). Due to
potential contamination of left ventricular
31
P MR spectra with ATP signal from blood, one may
need to consider the correction for blood contamination when quantifying the cardiac PCr/ATP
ratio (43, 67). The first implementation of 3D ISIS for in vivo cardiac
31
P MRS in mice was
performed at 2.35 T, with a measurement time of about 3 hours (67). The acquisitions were
cardiac triggered, leading to variations in TR, albeit small, because no respiratory gating was
applied. Although 3D ISIS results in well-defined voxel shapes (17), variation in TR may result
in contamination from outside the voxel (45, 48). In a recent study described in Chapter 6, our
group implemented dummy acquisitions to maintain a constant TR during the 3D ISIS
acquisition with cardiac triggering and respiratory gating (5). The implementation was
performed at 9.4 T, giving the benefit of higher SNR and shorter measurement time (~1 hour).
Figure 2.5. 3D ISIS localized cardiac 31P MRS. (A) 4-chamber and short-axis view of
the heart, with white boxes indicating the ISIS voxel, and (B) typical cardiac 31P
spectrum. DPG: diphosphoglycerate, Pi: inorganic phosphate, PCr: phosphocreatine,
ATP: adenosine triphosphate.
Another localization technique commonly applied for mouse cardiac
31
P MRS is chemical shift
imaging (CSI), which resolves spectra from multiple locations. In the 1D application of cardiac
31
P CSI, the acquisition is performed with a number of phase-encode steps (e.g. 32), resulting
in different spectra for a stack of slices (Fig. 2.6A and B). In the implementation of this method
in the Weiss group (13, 34, 53, 90), the 1D CSI slices are planned in the direction perpendicular
to the plane of the surface coil, in such a way that one slice contains the anterior cardiac wall,
which is used for the quantification of the cardiac PCr and ATP. The localization within the plane
of the slices is realized by surface coil localization. The acquisition in Chacko et al. was
performed at 4.7 T, at constant TR, without cardiac triggering or respiratory gating, resulting in
a total measurement time of 35 minutes (13). Application of CSI has also been performed in a
18
Cardiac MRI, 1H MRS,
31
P MRS – review
2D mode, which resolves spectra from different locations within the myocardium (Fig. 2.7A and
B) (25). Flögel et al. implemented the technique at 9.4 T, with a total measurement of 75
minutes, also without cardiac triggering, assuming minor displacements within spectroscopic
grids between systole and diastole (25). CSI provides the opportunity to study regional
differences in cardiac energetics, for example in myocardial infarction. However, CSI is prone to
contamination due to Fourier bleeding (44).
Figure 2.6. 1D CSI localized cardiac 31P MRS. Left: axial image of a mouse thorax
through the heart at end diastole. Selected 31P spectral slices (indicated in white),
containing (A) the anterior myocardium and (B) the edge of a bulb containing
phenylphosphonic acid, adjacent to the chest. Adapted from Chacko et al. (13).
Figure 2.7. 2D CSI localized cardiac 31P MRS. (A) Axial MR image at end diastole of a
WT mouse with voxels indicated in white, and (B) the corresponding 31P MR spectra of
the septum (#1) and anterior wall (#2). Adapted from Flögel et al. (25), with
permission.
19
Chapter 2
The most recent application of cardiac
31
P MRS involves the determination of ATP turnover
kinetics, by using the saturation transfer technique in combination with 1D CSI (32, 33, 35). As
PCr and ATP are in chemical exchange through CK, applying frequency-selective irradiation
pulses on γ-ATP results in a decrease in the equilibrium longitudinal magnetization of PCr, from
M0,PCr to M0,PCr’. The pseudo first-order CK reaction rate constant kf can be determined from the
relationship between M0,PCr and M0,PCr’, together with the apparent T1 of PCr (T1,PCr’). Using the
TRiST (Triple Repetition Time Saturation Transfer) technique (33, 78), T1,PCr’ and M0,PCr’ can be
estimated from two spectra acquired after saturating the γ-ATP peak (at -2.5 ppm) with a short
TR (Fig. 2.8B) and a long TR (Fig. 2.8C). The magnetization M0,PCr is then determined from a
fully relaxed spectrum, which is acquired with a control saturation at the downfield frequency
mirror-imaged with respect to PCr (at +2.5 ppm; Fig. 2.8D).
Figure 2.8. Saturation transfer 31P MRS for the measurement of cardiac CK flux. (A)
Typical axial MRI of a mouse at the mid LV with the selected 31P spectral slice denoted
between the white lines, (B) 31P MRS with γ-ATP saturation with short TR (TR = 1.5 s,
NEX = 96) and (C) long TR (TR = 6 s, NEX = 32). (D) Control spectrum with TR=10 s
(fully relaxed). Taken from Gupta et al. (33), with permission.
Other important readouts from cardiac
31
P MRS, which have potential for future application in
the heart, are the concentration of ADP and the amount of energy released during ATP
hydrolysis (-ΔGATP). The concentration of ADP can be calculated from the relationship between
the concentration of PCr, ATP, free creatine, and pH, set by the CK equilibrium (90).
Additionally, the concentration of inorganic phosphate (Pi) is required to determine the ΔGATP.
Free creatine can be determined by subtracting phosphocreatine from total creatine. Total
creatine can be quantified in vivo using 1H MRS, as demonstrated in (70). The combination of
31
P MRS and 1H MRS has recently been applied in human studies for this purpose (40). Finally,
while intracellular pH can typically be determined from the shift of the P i peak in skeletal muscle
31
P MR spectra (56), this readout is still very challenging for in vivo cardiac applications. In
20
Cardiac MRI, 1H MRS,
cardiac
31
31
P MRS – review
P MR spectra, the Pi peak is generally very small and in in vivo measurements it is
often obscured by the 2,3-DPG signal from the blood, which in practice limits the use of
31
P MRS
to determine cardiac intracellular pH to perfused heart setups (46).
2.5. Conclusions
To summarize, in the past few years there has been a rapid development in MRI and MRS
techniques for in vivo cardiac applications in small animal studies. The techniques have proven
to be powerful in the characterization of the hearts in mouse models of cardiomyopathy,
providing insight into the alterations in cardiac function and metabolism. These non-invasive
techniques allow longitudinal measurements in the same animal, which is extremely valuable to
study disease progression and to evaluate the effect of treatment.
2.6. Acknowledgement
This work was supported by VIDI grant (700.58.421) from the Netherlands Organisation for
Scientific Research (NWO), and VIDI grant (number: 07952) from the Dutch Technology
Foundation STW, Applied Science Division of NWO and the Technology Program of the Ministry
of Economic Affairs, and the Institute for Imaging Science and Technology Eindhoven (IST/e).
2.7. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Abdurrachim D, Ciapaite J, Wessels B, Nabben M, Luiken JJ, Nicolay K, and Prompers JJ.
Cardiac diastolic dysfunction in high-fat diet fed mice is associated with lipotoxicity without impairment
of cardiac energetics in vivo. Biochim Biophys Acta 1841: 1525-1537, 2014.
Al-Shafei AI, Wise RG, Gresham GA, Carpenter TA, Hall LD, and Huang CL. Magnetic resonance
imaging analysis of cardiac cycle events in diabetic rats: the effect of angiotensin-converting enzyme
inhibition. J Physiol 538: 555-572, 2002.
Bakermans AJ, Abdurrachim D, Geraedts TR, Houten SM, Nicolay K, and Prompers JJ. In vivo
proton T1 relaxation times of mouse myocardial metabolites at 9.4 T. Magn Reson Med doi:
10.1002/mrm.25340, 2014.
Bakermans AJ, Abdurrachim D, Nicolay K, and Prompers JJ. 1H-MRS in the in vivo mouse heart:
A comparison between PRESS and STEAM. In: Proc 21st Ann Meeting ISMRM p. 2011, 2013.
Bakermans AJ, Dodd MS, Nicolay K, Prompers JJ, Tyler DJ, and Houten SM. Myocardial energy
shortage and unmet anaplerotic needs in the fasted long-chain acyl-CoA dehydrogenase knockout
mouse. Cardiovasc Res 100: 441-449, 2013.
Bakermans AJ, Geraedts TR, van Weeghel M, Denis S, Joao Ferraz M, Aerts JM, Aten J,
Nicolay K, Houten SM, and Prompers JJ. Fasting-induced myocardial lipid accumulation in longchain acyl-CoA dehydrogenase knockout mice is accompanied by impaired left ventricular function. Circ
Cardiovasc Imaging 4: 558-565, 2011.
Bakermans AJ, van Weeghel M, Denis S, Nicolay K, Prompers JJ, and Houten SM. Carnitine
supplementation attenuates myocardial lipid accumulation in long-chain acyl-CoA dehydrogenase
knockout mice. J Inherit Metab Dis 36: 973-981, 2013.
Bittl JA, DeLayre J, and Ingwall JS. Rate equation for creatine kinase predicts the in vivo reaction
velocity: 31P NMR surface coil studies in brain, heart, and skeletal muscle of the living rat. Biochemistry
(Mosc) 26: 6083-6090, 1987.
Bottomley PA. Spatial localization in NMR spectroscopy in vivo. Ann N Y Acad Sci 508: 333-348,
1987.
Bottomley PA, and Weiss RG. Non-invasive magnetic-resonance detection of creatine depletion in
non-viable infarcted myocardium. Lancet 351: 714-718, 1998.
Bovens SM, te Boekhorst BC, den Ouden K, van de Kolk KW, Nauerth A, Nederhoff MG,
Pasterkamp G, ten Hove M, and van Echteld CJ. Evaluation of infarcted murine heart function:
comparison of prospectively triggered with self-gated MRI. NMR Biomed 24: 307-315, 2011.
Cassidy PJ, Schneider JE, Grieve SM, Lygate C, Neubauer S, and Clarke K. Assessment of
motion gating strategies for mouse magnetic resonance at high magnetic fields. J Magn Reson Imaging
19: 229-237, 2004.
Chacko VP, Aresta F, Chacko SM, and Weiss RG. MRI/MRS assessment of in vivo murine cardiac
metabolism, morphology, and function at physiological heart rates. Am J Physiol Heart Circ Physiol
279: H2218-2224, 2000.
21
Chapter 2
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
22
Coolen BF, Abdurrachim D, Motaal AG, Nicolay K, Prompers JJ, and Strijkers GJ. High frame
rate retrospectively triggered Cine MRI for assessment of murine diastolic function. Magn Reson Med
69: 648-656, 2013.
Coolen BF, Geelen T, Paulis LE, Nicolay K, and Strijkers GJ. Regional contrast agent
quantification in a mouse model of myocardial infarction using 3D cardiac T1 mapping. J Cardiovasc
Magn Reson 13: 56, 2011.
Coolen BF, Paulis LE, Geelen T, Nicolay K, and Strijkers GJ. Contrast-enhanced MRI of murine
myocardial infarction - part II. NMR Biomed 25: 969-984, 2012.
De Graaf RA. In vivo NMR spectroscopy: principles and techniques. John Wiley & Sons, 2008.
Del-Canto I, Lopez-Lereu MP, Monmeneu JV, Croisille P, Clarysse P, Chorro FJ, Bodi V, and
Moratal D. Characterization of normal regional myocardial function by MRI cardiac tagging. J Magn
Reson Imaging 2013.
Epstein FH. MR in mouse models of cardiac disease. NMR Biomed 20: 238-255, 2007.
Esparza-Coss E, Ramirez MS, and Bankson JA. Wireless self-gated multiple-mouse cardiac cine
MRI. Magn Reson Med 59: 1203-1206, 2008.
Faller KM, Lygate CA, Neubauer S, and Schneider JE. 1H-MR spectroscopy for analysis of cardiac
lipid and creatine metabolism. Heart Fail Rev 18: 657-668, 2013.
Faller KM, Medway DJ, Aksentijevic D, Sebag-Montefiore L, Schneider JE, Lygate CA, and
Neubauer S. Ribose supplementation alone or with elevated creatine does not preserve high energy
nucleotides or cardiac function in the failing mouse heart. PLoS ONE 8: e66461, 2013.
Fang ZY, Prins JB, and Marwick TH. Diabetic cardiomyopathy: Evidence, mechanisms, and
therapeutic implications. Endocr Rev 25: 543-567 2004.
Feintuch A, Zhu Y, Bishop J, Davidson L, Dazai J, Bruneau BG, and Henkelman RM. 4D cardiac
MRI in the mouse. NMR Biomed 20: 360-365, 2007.
Flogel U, Jacoby C, Godecke A, and Schrader J. In vivo 2D mapping of impaired murine cardiac
energetics in NO-induced heart failure. Magn Reson Med 57: 50-58, 2007.
Frahm J, Bruhn H, Gyngell ML, Merboldt KD, Hanicke W, and Sauter R. Localized high-resolution
proton NMR spectroscopy using stimulated echoes: initial applications to human brain in vivo. Magn
Reson Med 9: 79-93, 1989.
Frahm J, Merboldt KD, and Hanicke W. Localized proton spectroscopy using stimulated echoes. J
Magn Reson 72: 502-508, 1987.
Fries P, Massmann A, Seidel R, Muller A, Stroeder J, Custodis F, Reil J, Schneider G, and
Buecker A. Comparison of retrospectively self-gated and prospectively triggered FLASH sequences for
cine imaging of the aorta in mice at 9.4 Tesla. Invest Radiol 47: 259-266, 2012.
Gladden JD, Linke WA, and Redfield MM. Heart failure with preserved ejection fraction. Pflugers
Arch 2014.
Griffin JL, Williams HJ, Sang E, and Nicholson JK. Abnormal lipid profile of dystrophic cardiac
tissue as demonstrated by one- and two-dimensional magic-angle spinning 1H NMR spectroscopy.
Magn Reson Med 46: 249-255, 2001.
Grove TH, Ackerman JJ, Radda GK, and Bore PJ. Analysis of rat heart in vivo by phosphorus
nuclear magnetic resonance. Proc Natl Acad Sci U S A 77: 299-302, 1980.
Gupta A, Akki A, Wang Y, Leppo MK, Chacko VP, Foster DB, Caceres V, Shi S, Kirk JA, Su J,
Lai S, Paolocci N, Steenbergen C, Gerstenblith G, and Weiss RG. Creatine kinase-mediated
improvement of function in failing mouse hearts provides causal evidence the failing heart is energy
starved. J Clin Invest 122: 291-302, 2012.
Gupta A, Chacko VP, Schar M, Akki A, and Weiss RG. Impaired ATP kinetics in failing in vivo
mouse heart. Circ Cardiovasc Imaging 4: 42-50, 2011.
Gupta A, Chacko VP, and Weiss RG. Abnormal energetics and ATP depletion in pressure-overload
mouse hearts: in vivo high-energy phosphate concentration measures by noninvasive magnetic
resonance. Am J Physiol Heart Circ Physiol 297: H59-64, 2009.
Gupta A, Rohlfsen C, Leppo MK, Chacko VP, Wang Y, Steenbergen C, and Weiss RG. Creatine
kinase-overexpression improves myocardial energetics, contractile dysfunction and survival in murine
doxorubicin cardiotoxicity. PLoS ONE 8: e74675, 2013.
Hankiewicz JH, Banke NH, Farjah M, and Lewandowski ED. Early impairment of transmural
principal strains in the left ventricular wall after short-term, high-fat feeding of mice predisposed to
cardiac steatosis. Circ Cardiovasc Imaging 3: 710-717, 2010.
Hardy CJ, Weiss RG, Bottomley PA, and Gerstenblith G. Altered myocardial high-energy
phosphate metabolites in patients with dilated cardiomyopathy. Am Heart J 122: 795-801, 1991.
Heijman E, Graaf Wd, Niessen P, Nauerth A, Eys Gv, Graaf Ld, Nicolay K, and Strijkers GJ.
Comparison between prospective and retrospective triggering for mouse cardiac MRI. NMR Biomed 20:
439-447, 2007.
Hiba B, Richard N, Janier M, and Croisille P. Cardiac and respiratory double self-gated cine MRI in
the mouse at 7 T. Magn Reson Med 55: 506-513, 2006.
Hirsch GA, Bottomley PA, Gerstenblith G, and Weiss RG. Allopurinol acutely increases adenosine
triphospate energy delivery in failing human hearts. J Am Coll Cardiol 59: 802-808, 2012.
Hitchins S, Cieslar JM, and Dobson GP. 31P NMR quantitation of phosphorus metabolites in rat heart
and skeletal muscle in vivo. Am J Physiol Heart Circ Physiol 281: H882-887, 2001.
Hoerr V, Nagelmann N, Nauerth A, Kuhlmann MT, Stypmann J, and Faber C. Cardiacrespiratory self-gated cine ultra-short echo time (UTE) cardiovascular magnetic resonance for
assessment of functional cardiac parameters at high magnetic fields. J Cardiovasc Magn Reson 15: 59,
2013.
Cardiac MRI, 1H MRS,
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
31
P MRS – review
Horn M, Neubauer S, Bomhard M, Kadgien M, Schnackerz K, and Ertl G. 31P-NMR spectroscopy
of human blood and serum: first results from volunteers and patients with congestive heart failure,
diabetes mellitus and hyperlipidaemia. MAGMA 1: 55-60, 1993.
Keevil SF. Spatial localization in nuclear magnetic resonance spectroscopy. Phys Med Biol 51: R579636, 2006.
Keevil SF, Porter DA, and Smith MA. Experimental characterization of the ISIS technique for
volume selected NMR spectroscopy. NMR Biomed 5: 200-208, 1992.
Kolwicz SC, Jr., and Tian R. Assessment of cardiac function and energetics in isolated mouse hearts
using 31P NMR spectroscopy. J Vis Exp 42: 2069, 2010.
Koretsky AP, Wang S, Murphy-Boesch J, Klein MP, James TL, and Weiner MW. 31P NMR
spectroscopy of rat organs, in situ, using chronically implanted radiofrequency coils. Proc Natl Acad Sci
U S A 80: 7491-7495, 1983.
Lawry TJ, Karczmar GS, Weiner MW, and Matson GB. Computer simulation of MRS localization
techniques: an analysis of ISIS. Magn Reson Med 9: 299-314, 1989.
Loganathan R, Bilgen M, Al-Hafez B, Alenezy M, and Smirnova I. Cardiac dysfunction in the
diabetic rat: quantitative evaluation using high resolution magnetic resonance imaging. Cardiovasc
Diabetol 5: 7, 2006.
Luedde M, Flogel U, Knorr M, Grundt C, Hippe HJ, Brors B, Frank D, Haselmann U, Antony C,
Voelkers M, Schrader J, Most P, Lemmer B, Katus HA, and Frey N. Decreased contractility due to
energy deprivation in a transgenic rat model of hypertrophic cardiomyopathy. J Mol Med (Berl) 87:
411-422, 2009.
Lygate CA, Bohl S, ten Hove M, Faller KM, Ostrowski PJ, Zervou S, Medway DJ, Aksentijevic
D, Sebag-Montefiore L, Wallis J, Clarke K, Watkins H, Schneider JE, and Neubauer S.
Moderate elevation of intracellular creatine by targeting the creatine transporter protects mice from
acute myocardial infarction. Cardiovasc Res 96: 466-475, 2012.
Lygate CA, Schneider JE, and Neubauer S. Investigating cardiac energetics in heart failure. Exp
Physiol 98: 601-605, 2013.
Maslov MY, Chacko VP, Stuber M, Moens AL, Kass DA, Champion HC, and Weiss RG. Altered
high-energy phosphate metabolism predicts contractile dysfunction and subsequent ventricular
remodeling in pressure-overload hypertrophy mice. Am J Physiol Heart Circ Physiol 292: H387-391,
2007.
Miraux S, Calmettes G, Massot P, Lefrancois W, Parzy E, Muller B, Arsac LM, Deschodt-Arsac
V, Franconi JM, Diolez P, and Thiaudiere E. 4D retrospective black blood trueFISP imaging of
mouse heart. Magn Reson Med 62: 1099-1105, 2009.
Montesinos P, Abascal JF, Cusso L, Vaquero JJ, and Desco M. Application of the compressed
sensing technique to self-gated cardiac cine sequences in small animals. Magn Reson Med 72: 369380, 2014.
Moon RB, and Richards JH. Determination of intracellular pH by 31P magnetic resonance. J Biol Chem
248: 7276-7278, 1973.
Moonen CT, von Kienlin M, van Zijl PC, Cohen J, Gillen J, Daly P, and Wolf G. Comparison of
single-shot localization methods (STEAM and PRESS) for in vivo proton NMR spectroscopy. NMR
Biomed 2: 201-208, 1989.
Motaal A, Noorman N, de Graaf W, Hoerr V, Florack LJ, Nicolay K, and Strijkers G. Functional
imaging of murine hearts using accelerated self-gated UTE cine MRI. The International Journal of
Cardiovascular Imaging.doi 10.1007/s10554-10014-10531-10558, 2014.
Motaal AG, Coolen BF, Abdurrachim D, Castro RM, Prompers JJ, Florack LM, Nicolay K, and
Strijkers GJ. Accelerated high-frame-rate mouse heart cine-MRI using compressed sensing
reconstruction. NMR Biomed doi: 10.1002/nbm.2883, 2012.
Nagarajan V, Gopalan V, Kaneko M, Angeli V, Gluckman P, Richards AM, Kuchel PW, and
Velan SS. Cardiac function and lipid distribution in rats fed a high-fat diet: in vivo magnetic resonance
imaging and spectroscopy. Am J Physiol Heart Circ Physiol 304: H1495-1504, 2013.
Nahrendorf M, Wiesmann F, Hiller KH, Hu K, Waller C, Ruff J, Lanz TE, Neubauer S, Haase A,
Ertl G, and Bauer WR. Serial cine-magnetic resonance imaging of left ventricular remodeling after
myocardial infarction in rats. J Magn Reson Imaging 14: 547-555, 2001.
Nakae I, Mitsunami K, Matsuo S, Matsumoto T, Morikawa S, Inubushi T, Koh T, and Horie M.
Assessment of myocardial creatine concentration in dysfunctional human heart by proton magnetic
resonance spectroscopy. Magn Reson Med Sci 3: 19-25, 2004.
Nakae I, Mitsunami K, Omura T, Yabe T, Tsutamoto T, Matsuo S, Takahashi M, Morikawa S,
Inubushi T, Nakamura Y, Kinoshita M, and Horie M. Proton magnetic resonance spectroscopy can
detect creatine depletion associated with the progression of heart failure in cardiomyopathy. J Am Coll
Cardiol 42: 1587-1593, 2003.
Neubauer S. The failing heart--an engine out of fuel. N Engl J Med 356: 1140-1151, 2007.
Neubauer S, Krahe T, Schindler R, Horn M, Hillenbrand H, Entzeroth C, Mader H, Kromer EP,
Riegger GA, Lackner K, and et al. 31P magnetic resonance spectroscopy in dilated cardiomyopathy
and coronary artery disease. Altered cardiac high-energy phosphate metabolism in heart failure.
Circulation 86: 1810-1818, 1992.
Nicolay K, Aue WP, Seelig J, van Echteld CJ, Ruigrok TJ, and de Kruijff B. Effects of the anticancer drug adriamycin on the energy metabolism of rat heart as measured by in vivo 31P-NMR and
implications for adriamycin-induced cardiotoxicity. Biochim Biophys Acta 929: 5-13, 1987.
Omerovic E, Basetti M, Bollano E, Bohlooly M, Törnell J, Isgaard J, Hjalmarson Å, Soussi B,
and Waagstein F. In vivo metabolic imaging of cardiac bioenergetics in transgenic mice. Biochem
Biophys Res Commun 271: 222-228, 2000.
23
Chapter 2
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
24
Omerovic E, Bollano E, Mobini R, Kujacic V, Madhu B, Soussi B, Fu M, Hjalmarson A,
Waagstein F, and Isgaard J. Growth hormone improves bioenergetics and decreases catecholamines
in postinfarct rat hearts. Endocrinology 141: 4592-4599, 2000.
Ordidge R, Connelly A, and Lohman J. Image-selected in vivo spectroscopy (ISIS). A new
technique for spatially selective NMR spectroscopy. J Magn Reson 66: 283-294, 1986.
Phillips D, Ten Hove M, Schneider JE, Wu CO, Sebag-Montefiore L, Aponte AM, Lygate CA,
Wallis J, Clarke K, Watkins H, Balaban RS, and Neubauer S. Mice over-expressing the myocardial
creatine transporter develop progressive heart failure and show decreased glycolytic capacity. J Mol
Cell Cardiol 48: 582-590, 2010.
Prompers JJ, Strijkers GJ, and Nicolay K. Magnetic resonance spectroscopy of in vivo tissue
metabolism in small animals. Drug Discovery Today: Technologies 8: e95-e102, 2011.
Protti A, Dong X, Sirker A, Botnar R, and Shah AM. MRI-based prediction of adverse cardiac
remodeling after murine myocardial infarction. Am J Physiol Heart Circ Physiol 303: H309-314, 2012.
Ram R, Mickelsen DM, Theodoropoulos C, and Blaxall BC. New approaches in small animal
echocardiography: imaging the sounds of silence. Am J Physiol Heart Circ Physiol 301: H1765-1780,
2011.
Ratering D, Baltes C, Dorries C, and Rudin M. Accelerated cardiovascular magnetic resonance of
the mouse heart using self-gated parallel imaging strategies does not compromise accuracy of
structural and functional measures. J Cardiovasc Magn Reson 12: 43, 2010.
Rial B, Robson MD, Neubauer S, and Schneider JE. Rapid quantification of myocardial lipid content
in humans using single breath-hold 1H MRS at 3 Tesla. Magn Reson Med 66: 619-624, 2011.
Rudin M, Pedersen B, Umemura K, and Zierhut W. Determination of rat heart morphology and
function in vivo in two models of cardiac hypertrophy by means of magnetic resonance imaging. Basic
Res Cardiol 86: 165-174, 1991.
Ruff J, Wiesmann F, Hiller KH, Voll S, von Kienlin M, Bauer WR, Rommel E, Neubauer S, and
Haase A. Magnetic resonance microimaging for noninvasive quantification of myocardial function and
mass in the mouse. Magn Reson Med 40: 43-48, 1998.
Schar M, El-Sharkawy AM, Weiss RG, and Bottomley PA. Triple repetition time saturation transfer
(TRiST) 31P spectroscopy for measuring human creatine kinase reaction kinetics. Magn Reson Med 63:
1493-1501, 2010.
Schar M, Kozerke S, and Boesiger P. Navigator gating and volume tracking for double-triggered
cardiac proton spectroscopy at 3 Tesla. Magn Reson Med 51: 1091-1095, 2004.
Schneider JE, Tyler DJ, ten Hove M, Sang AE, Cassidy PJ, Fischer A, Wallis J, SebagMontefiore LM, Watkins H, Isbrandt D, Clarke K, and Neubauer S. In vivo cardiac 1H-MRS in the
mouse. Magn Reson Med 52: 1029-1035, 2004.
Srinivasan S, and Ennis DB. Optimal flip angle for high contrast balanced SSFP cardiac cine imaging.
Magn Reson Med 2014.
Stuckey DJ, Carr CA, Tyler DJ, Aasum E, and Clarke K. Novel MRI method to detect altered left
ventricular ejection and filling patterns in rodent models of disease. Magn Reson Med 60: 582-587,
2008.
Szczepaniak LS, Dobbins RL, Metzger GJ, Sartoni-D'Ambrosia G, Arbique D, Vongpatanasin
W, Unger R, and Victor RG. Myocardial triglycerides and systolic function in humans: in vivo
evaluation by localized proton spectroscopy and cardiac imaging. Magn Reson Med 49: 417-423, 2003.
Toyo-oka T, Nagayama K, Umeda M, Eguchi K, and Hosoda S. Rhythmic change of myocardial
phosphate metabolite content in cardiac cycle observed by depth-selected and EKG-gated in vivo 31PNMR spectroscopy in a whole animal. Biochem Biophys Res Commun 135: 808-815, 1986.
Tucci S, Flogel U, Hermann S, Sturm M, Schafers M, and Spiekerkoetter U. Development and
pathomechanisms of cardiomyopathy in very long-chain acyl-CoA dehydrogenase deficient (VLCAD(-/)) mice. Biochim Biophys Acta 1842: 677-685, 2014.
van de Weijer T, van Ewijk PA, Zandbergen HR, Slenter JM, Kessels AG, Wildberger JE,
Hesselink MK, Schrauwen P, Schrauwen-Hinderling VB, and Kooi ME. Geometrical models for
cardiac MRI in rodents: comparison of quantification of left ventricular volumes and function by various
geometrical models with a full-volume MRI data set in rodents. Am J Physiol Heart Circ Physiol 302:
H709-715, 2012.
van der Meer RW, Doornbos J, Kozerke S, Schar M, Bax JJ, Hammer S, Smit JW, Romijn JA,
Diamant M, Rijzewijk LJ, de Roos A, and Lamb HJ. Metabolic imaging of myocardial triglyceride
content: reproducibility of 1H MR spectroscopy with respiratory navigator gating in volunteers.
Radiology 245: 251-257, 2007.
van Nierop BJ, Coolen BF, Dijk WJ, Hendriks AD, de Graaf L, Nicolay K, and Strijkers GJ.
Quantitative first-pass perfusion MRI of the mouse myocardium. Magn Reson Med 69: 1735-1744,
2013.
van Nierop BJ, van Assen HC, van Deel ED, Niesen LB, Duncker DJ, Strijkers GJ, and Nicolay
K. Phenotyping of left and right ventricular function in mouse models of compensated hypertrophy and
heart failure with cardiac MRI. PLoS ONE 8: e55424, 2013.
Weiss RG, Chatham JC, Georgakopolous D, Charron MJ, Wallimann T, Kay L, Walzel B, Wang
Y, Kass DA, Gerstenblith G, and Chacko VP. An increase in the myocardial PCr/ATP ratio in GLUT4
null mice. FASEB J 16: 613-615, 2002.
Yue P, Arai T, Terashima M, Sheikh AY, Cao F, Charo D, Hoyt G, Robbins RC, Ashley EA, Wu J,
Yang PC, and Tsao PS. Magnetic resonance imaging of progressive cardiomyopathic changes in the
db/db mouse. Am J Physiol Heart Circ Physiol 292: H2106-2118, 2007.
Chapter 3
Good and bad consequences of altered fatty acid
metabolism in heart failure:
Evidence from mouse models
Review
Desiree Abdurrachim1, Joost J.F.P. Luiken2, Klaas Nicolay1, Jan F.C. Glatz2,
Jeanine J. Prompers1, and Miranda Nabben1
1
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology,
Eindhoven, The Netherlands
2
Department of Genetics and Cell Biology, Cardiovascular Research Institute Maastricht
(CARIM), Maastricht University, Maastricht, The Netherlands
Submitted
Chapter 3
Abstract
The shift in substrate preference away from fatty acid oxidation (FAO) towards increased
glucose in heart failure has long been interpreted as an oxygen-sparing mechanism. Inhibition
of FAO has therefore evolved as an accepted approach to treat heart failure. However, recent
data indicate that increased reliance on glucose might be detrimental rather than beneficial for
the failing heart. This review discusses new insights into metabolic adaptations in heart failure.
A particular focus lies on data obtained from mouse models with modulations of cardiac FA
metabolism at different levels of the FA metabolic pathway and how these differently affect
cardiac function. Based on studies in which these mouse models were exposed to ischemic and
non-ischemic heart failure, we discuss whether and when modulations in FA metabolism are
protective against heart failure.
Data support that in ischemic heart failure, upregulating glucose utilization may be beneficial to
improve the coupling between glycolysis and glucose oxidation and consequently to improve
cardiac energy efficiency. Upon non-ischemic heart failure, however, it seems that preserved
energetics, rather than increased glucose oxidation or increased FAO per se, is key for
maintaining or restoring cardiac function. A shift towards glucose oxidation may be a rescue
mechanism; however, the increase in glucose utilization often appears insufficient to maintain
cardiac energy status. On the other hand, energetic compensation through high or maintained
FAO can also be a successful strategy to protect the heart from failure. Therefore, the efficacy of
FAO inhibition in the treatment of non-ischemic heart failure is as of yet not evident.
Keywords
heart failure, fatty acid metabolism, genetically altered mouse models, metabolic shift
Abbreviations
AAC: abdominal aortic constriction, ACC: acetyl-CoA carboxylase, ACSL:long-chain acetyl-CoA synthase,
ADP: adenosine diphosphate, AMPK: adenosine monophosphate-activated kinase, ANT: adenine nucleotide
translocase, AsAC: ascending aortic constriction, ATGL: adipose triglycerides lipase, ATP: adenosine
triphosphase, CACT: carnitine acylcarnitine translocase, CD: cluster of differentiation, CPT: carnitine
palmitoyl transferase, Cr: creatine, DG: diglyceride, DGAT: diglyceride acyltransferase, ETC: electron
transport chain, FA: fatty acids, FABPpm:plasma membrane fatty acid binding protein, FABPc:
cardiomyocyte fatty acid binding protein, FADH: flavin adenine dinucleotide, FAT: fatty acid translocase,
FATP: fatty acid transport protein, GLUT: glucose transporter, GPAT: glycerol phosphate acyltransferase,
HIF: hypoxia inducible factor, HSL: hormone sensitive lipase, HFD: high-fat diet, I/R: ischemia
reperfusion, KO: knockout, KI: knockin, LPL: cardiac lipoprotein lipase, MCD: malonyl-CoA carboxylase,
MG: monoglyceride, MGAT: monoglyceride acyltransferase, MGL: monoglycerides lipase, NADH:
nicotinamide adenine dinucleotide, OE: overexpression, PCr: phosphocreatine, PDH: pyruvate
dehydrogenase, PDHK: pyruvate dehydrogenase kinase, PGC1α: Peroxisome proliferator-activated
receptor-γ cofactor-1α, PPARα: Peroxisome proliferator-activated receptor-α, TAC: transverse aortic
constriction, TCA: tricarboxylic acid, TG: triglyceride, UCP: uncoupling protein.
26
Altered fatty acid metabolism – review
3.1. Introduction- Metabolic switch upon heart failure
The heart has sustained high energy demands in order to pump sufficient blood through the
body. In order to obtain this energy, the healthy adult heart primarily relies on the oxidation of
fatty acids (FA) (~60-90%) and to a lesser extent glucose (~10-40%), lactate, and ketone
bodies (67). It is flexible to shift its substrate preference according to physiological and
pathological challenges. It is generally accepted that the transition process towards heart failure
is accompanied by a shift in cardiac substrate preference, with a greater reliance on glucose as
substrate and a concomitant suppression of fatty acid oxidation (FAO) (33, 67). The metabolic
shift towards glucose oxidation has long been interpreted as an oxygen-sparing mechanism of
the heart, and hence inhibition of FAO has evolved as an established approach to treat heart
failure patients (33, 67).
Obese and/or diabetic hearts rely almost exclusively on FAO and are suggested to have lost the
flexibility to switch between substrates. Compared with glucose, FA are oxygen-inefficient
energy substrates. Moreover, FA can induce mitochondrial uncoupling, leading to less efficient
ATP production (15). Given the increased risk of heart failure in type 2 diabetic and obese
patients, it is indeed tempting to hypothesize that high FAO has deleterious effects on the heart.
However, long chain FA (e.g. palmitate) oxidation is 3–4 times more efficient in producing ATP
per carbon atom compared with glucose. Therefore, a substantial increase in glucose oxidation
is necessary to compensate for a loss in energy production, even in case of a small decrease in
FAO. This suggests that the shift away from FAO might not be beneficial for the heart, and that
inhibition of FAO might not always be the best strategy for heart failure treatment.
This review discusses new insights on the role of altered cardiac FA metabolism in ischemic and
non-ischemic heart failure. We review how modulation of different aspects of cardiac FA
metabolism in (genetic) mouse models affects cardiac function. We start with alterations
downstream of the FA metabolic pathway, at the level of FAO, and then move upstream towards
modulations of mitochondrial FA uptake, cellular FA uptake, and eventually to the level of FA
supply from the circulation. Based on studies in which (genetic) mouse models of altered FA
metabolism were exposed to heart failure conditions, we discuss whether and when modulations
in FA metabolism are protective during ischemic and non-ischemic heart failure.
3.2. Cardiac metabolic adaptations
3.2.1. Ischemic heart failure
One of the most common causes of heart failure and cardiac remodeling in patients is ischemia,
a restriction in oxygen and nutrient supply to the heart. Ischemia can occur during myocardial
infarction, angina, or surgery, when the blood flow to a part of the heart is obstructed. During
myocardial ischemia, the limited oxygen supply suppresses aerobic glucose oxidation and FAO.
In this condition, the oxygen-sensing pathway centred on the hypoxia inducible factor (HIF) is
switched on, which induces transcriptional upregulation of glycolytic enzymes (87). As a result,
cardiac substrate metabolism shifts from FAO to glycolysis, which is marked by translocation of
the FA transporter FAT/CD36 away from the sarcolemma, translocation of glucose transporter-4
(GLUT4) toward the sarcolemma, and a decrease in glycogen content (48). As ischemia is
27
Chapter 3
associated with increased FA availability and suppressed FAO, the translocation of FAT/CD36
away from the sarcolemma prevents the accumulation of myocardial lipids (48). It is important
to note, however, that during mild to moderate ischemia FAO remains the main contributor of
the residual oxidative metabolism (35, 75). The increase in glycolysis without a concomitant
increase in glucose oxidation results in an accumulation of potentially harmful catabolites like
lactate and protons, causing intracellular acidosis and intracellular Na+ and Ca2+ overload (26).
This reduces the efficiency of the heart to convert energy into contractile work, because more
energy is needed to restore ion homeostasis.
Upon post-ischemic reperfusion, when oxygen levels are restored, sarcolemmal GLUT4 levels
remain increased. However, as the post-ischemic glycolytic rate is lower, the high sarcolemmal
GLUT4 content during reperfusion is associated with replenishment of glycogen stores (48). FAO
is recovered to (or even exceeds) pre-ischemic levels, while sarcolemmal FAT/CD36 remains
low, leading to a reduced myocardial triglyceride (TG) content (48). The stimulation of FAO has
been thought to be contributed by 5’ AMP-activated protein kinase (AMPK) activation, which
phophorylates and inhibits acetyl-CoA carboxylase-2 (ACC2) (30). The inhibition of ACC2 lowers
the production of malonyl-CoA, which releases its inhibition on carnitine palmitoyltransferase-1
(CPT1) and its consequent inhibition on mitochondrial FA import. However, it has recently been
shown that this AMPK–ACC signaling might not play an obligate role in regulation of myocardial
FAO, and that other mechanisms, perhaps via upregulated uncoupling protein-3 (UCP3)
expression, may contribute to maintained FAO rates in the post-ischemic heart. Either way, an
increase in FAO during post-ischemic reperfusion might negatively impact myocardial efficiency
and function (66), as it further suppresses glucose oxidation rates (via the Randle cycle), and
thus uncouples glucose oxidation from glycolysis, aggravating the intracellular acidosis and
disturbed ion homeostasis.
To summarize, during myocardial ischemia, a shift away from FAO towards anaerobic glycolysis
assists in the ATP production, as limited oxygen supply suppresses aerobic respiration. Upon
reperfusion, FAO returns to pre-ischemic levels. Uncoupling between glycolysis and glucose
oxidation during ischemia and reperfusion may lead to accumulation of harmful catabolites,
causing intracellular acidosis and disturbed ion homeostasis (26). As an increase in glucose
oxidation would reduce the accumulation of harmful catabolites, a shift away from FA towards
glucose oxidation would be beneficial, also during the reperfusion process.
3.2.2. Non-ischemic heart failure
Together with ischemia, (non-ischemic) pressure-overload belongs to the most common causes
of heart failure in patients. In the majority of animal studies, data point to the existence of a
switch to glucose metabolism during the development of non-ischemic heart failure. In mice
with transverse aortic constriction (TAC)-induced pressure overload, it was demonstrated that
myocardial glucose uptake and utilization rates were increased as early as 1 day after TAC, and
that they were further increased at 7 days after TAC (109). Supporting this observation, other
studies have shown increased glycolysis and suppressed FAO at 4–8 weeks after TAC (58, 83,
84). The increase in glucose utilization was associated with increased insulin-independent
GLUT1 and decreased insulin-dependent GLUT4 transporters (64), as observed in the fetal
heart. Increased glucose utilization and reduced FAO were also observed in other animal models
28
Altered fatty acid metabolism – review
of heart failure, such as in rats with abdominal aortic constriction (AAC)-induced pressure
overload (5), volume overload (24), and hypertension-induced heart failure (23), rabbits with
hypertension-induced heart failure (93), and dogs with ventricular pacing-induced heart failure
(60, 63). A few studies, however, do not support the existence of a switch towards glucose
metabolism in non-ischemic heart failure (47, 107, 108), and it remains to be seen whether
such shift would be beneficial.
3.3. The FA metabolic pathway in the normal heart
The long-chain FA metabolic pathway is a multi-step process resulting in the breakdown of FA
for energy production (Fig. 3.1). In the blood stream, FA are either bound to albumin or
esterified into TG and transported in lipoprotein particles. The heart is characterized by a nonfenestrated (continuous) endothelium. Hence, the FA have to be unloaded from albumin and
lipoprotein particles in the circulation, in order to penetrate the endothelium layer and be
available for the heart. FA in the lipoprotein TG are made available via hydrolysis by lipoprotein
lipase (LPL) (40, 101). LPL is mainly present at the vascular side of the endothelium. Although it
is synthesized in cardiomyocytes, it is not present at the cardiomyocyte sarcolemma. LPL
activity and plasma FA availability can be modulated by dietary intake, such as high-fat diet
(HFD) feeding or prolonged fasting (6, 101). FA then cross the endothelial cells through an illunderstood mechanism, whereafter all FA are bound to albumin in the interstitial space.
Subsequently albumin carries the FA through the interstitial space to present these substrates
to the cardiomyocytes, which might be assisted by high-affinity binding sites for albumin.
Thereafter, the FA are taken up mainly by fatty acid translocase (FAT/CD36), with a minor
contribution of passive diffusion. Also other FA transporters (fatty acid transport protein (FATP)
1 and 6, and plasma membrane fatty acid-binding protein (FABPpm)) are present in
cardiomyocytes, but their roles in bulk FA uptake have not been completely clarified (reviewed
in (38)).
Upon CD36-mediated uptake into the soluble cytoplasm, FA are directly transferred from CD36
to small cytoplasmic fatty acid-binding protein (FABPc), which transports the FA from the
sarcolemma to the outer side of the mitochondria for esterification into fatty acyl-CoA (acylCoA) by long-chain fatty acyl-CoA synthetase-1 (ACSL1). FA-CoA can be converted into TG and
stored as lipid droplets in the myocardium or transported into the mitochondria for oxidation. It
is suggested that the majority of acyl-CoA is first shuttled through the TG pool prior to being
oxidized (reviewed in (55)). The stored TG can be made available for oxidation by hydrolysis of
TG into diglycerides (DG) by adipose triglyceride lipase (ATGL) (43), and subsequently into
monoglycerides (MG) by hormone sensitive lipase (HSL) (27, 55). Each of these steps releases
FA, which are converted into acyl-CoA. The turnover of myocardial TG is in fact high, and FA
derived from myocardial TG stores provides a significant proportion of substrates for
mitochondrial FAO (13). FA derived from ATGL-mediated lipolysis have also been shown to be
involved in the regulation of FAO through the activation of peroxisome proliferator-activated
receptor-α (PPARα) and PPARγ coactivator-1α (PGC1α) (44, 55). PPARα functions as a
transcription factor regulating the expression of FAO genes (14), while PGC1α is a key regulator
of energy metabolism, stimulating cardiac mitochondrial biogenesis, regulating substrate
oxidation, and preventing oxidative stress (88).
29
Chapter 3
In order to be oxidized, acyl-CoA needs to be imported into the mitochondria. Whereas shortand medium-chain acyl-CoA can diffuse over the mitochondrial membrane, the transport of
long-chain acyl-CoA requires the carnitine palmitoyl transferase (CPT) shuttle system. CPT1
converts long-chain acyl-CoA into long-chain acylcarnitine, which then crosses the outer
mitochondrial membrane. Acylcarnitine is then transported across the inner mitochondrial
membrane
by
carnitine-acylcarnitine
transporter
(CACT).
Once
in
the
mitochondria,
acylcarnitine is converted back into acyl-CoA by CPT2, and CAT transports free carnitine out of
the mitochondria. The activity of CPT1 is regulated by malonyl-CoA, which is formed by
carboxylation of oxidative metabolism-derived acetyl-CoA by acetyl-CoA carboxylase (ACC).
Malonyl-CoA can be decarboxylated to acetyl-CoA by malonyl-CoA decarboxylase (MCD).
FA β-oxidation is the final step in the FA metabolic pathway, during which acyl-CoA is broken
down to acetyl-CoA. This multi-cycle process takes place inside the mitochondria and involves
several enzymes. The first step in the β-oxidation of long-chain FA is catalyzed by the enzyme
very long-chain acyl-CoA dehydrogenase (VLCAD). FA with shorter chain length are oxidized by
long-chain acyl-CoA dehydrogenase (LCAD) and medium chain acyl-CoA dehydrogenase
(MCAD). During each cycle of β-oxidation an acyl-CoA molecule is shortened by two carbon
atoms, thereby producing acetyl-CoA, which is then fed into the TCA cycle. Both the FA βoxidation pathway and the TCA cycle produce NADH and FADH2, which are used for ATP
production. Additionally, the produced NADH and acetyl-CoA can inhibit glucose oxidation
(shown in grey in Fig. 3.1), by stimulating pyruvate dehydrogenase kinase (PDHK), which in
turn inhibits pyruvate dehydrogenase (PDH), a major regulator of glucose oxidation.
Reciprocally, FA β-oxidation can be inhibited by pyruvate-derived acetyl-CoA, which is
transported to the cytoplasm via reversible conversion into citrate, and subsequent conversion
into malonyl-CoA to inhibit CPT1 (49).
NADH and FADH2 enter the electron transport chain (ETC) where they are reduced into NAD +
and FAD. The resulting H+ and electrons create the electrochemical proton gradient, which is
used to drive ATP synthase to produce energy in the form of ATP. The ATP production is coupled
with oxygen consumption; therefore, the efficiency of the ETC is expressed as the amount of
ATP produced per reduced oxygen atom (P/O). As mentioned earlier, this P/O ratio is lower for
FA compared with glucose oxidation. Moreover, FA have been shown to activate mitochondrial
uncoupling (51), thereby partly dissipating the electrochemical proton gradient as heat instead
of using it to drive ATP production. Adenine nucleotide translocase (ANT) transfers ATP and ADP
into and out of the mitochondria. The transfer of ATP to the utilization site (e.g. myofibril)
involves the creatine kinase (CK) shuttle, which catalyzes the conversion between ATP and
phosphocreatine (PCr).
30
Altered fatty acid metabolism – review
Figure 3.1. Cardiac glucose and long chain FA metabolic pathways. The proteins that are highlighted
with a bold border correspond with the mouse models discussed in this review. FA, besides glucose, are
the main substrates for cardiac energy metabolism. In plasma, FA are available as ‘free’ (unesterified)
FA bound to albumin, or esterified in lipoprotein TG. FA in the lipoprotein TG are made available via
hydrolysis by LPL. Mediated by FAT/CD36, FA are transported into the myocardium. In the myocardium
FA are converted into acyl-CoA, which can either be stored as TG or imported into mitochondria through
the CPT1-CACT-CPT2 carnitine shuttle. CPT1 activity is inhibited by malonyl-CoA, which is formed via
carboxylation of acetyl-CoA by ACC. In the mitochondria, acyl-CoA undergoes β-oxidation, which
produces NADH, FADH2, and acetyl-CoA. NADH and acetyl-CoA can stimulate PDHK, which in turn
inhibits PDH and subsequently glucose oxidation. Together with acetyl-CoA produced through the
glucose metabolism pathway (shown in grey), acetyl-CoA is converted to citrate, which could either
diffuse out of the mitochondria and indirectly inhibit CPT1, or enter the TCA cycle. The TCA cycle
produces NADH and FADH2. The NADH and FADH2 then enter the ETC complexes (I-IV), after which ATP
is produced by F0/F1 ATPase. The coupling between ETC and ATP production can be reduced by UCP and
ANT. ANT transports ATP out of the mitochondria, after which it is further transported to the utilization
site (e.g. the contractile proteins) via the creatine kinase shuttle.
31
Table 3.1. Mouse models of altered cardiac substrate metabolism
Mouse models
References
Cellular
FA
uptake
Cardiac
lipid
content
Lipotoxicity
FAO
Glucose
oxidation
Cardiac
energetics
Cardiac
function
(3, 17, 18, 25,
36, 52, 68,
79, 80, 98)
(9, 10, 54, 77)
↑
↑/=
↑
↑
↓ /=
↓/=
↑/=/↓
↓
↓
n.a.
↓
↑
↓
↓/=
Exacerbated cardiac function after AAC (10). Cardiac
function can be improved by further stimulating
glucose oxidation (54).
(21)
↑
↑/=
↑
↑
↓
n.a.
↓
(50, 59, 91)
↓
↓
↓
↓
↑
=
=/↓
ACSL1-OE
(22, 65)
↑
↑
↑
↑
↑
n.a.
↓
ACSL1-KO
(31, 81)
↓
=
=
↓
↑
=
↓
Lipotoxic cardiomyopathy, female breeders died
during pregnancy or early postpartum.
Protected against HFD-induced lipotoxicity, reduced
(50) or unchanged (59) recovery upon I/R,
exacerbated function upon TAC (91).
Lipotoxic cardiomyopathy (22, 65), higher mortality
rate (65).
Similar response to TAC as controls (31).
TG synthesis
ATGL-OE
(57, 85, 86)
↓
↓
=
↓
↑
=
↑
ATGL-KO
DGAT1-OE
(43, 44, 56)
(39, 65)
↓
↑
↑
↑
↑
↓/↑
↓
↑
↓
↓
↓
n.a.
↓
=/↓
Mitochondrial FA import
CPT1b-KO
(33, 47, 67,
90)
=
=
=
=
=
n.a.
=
n.a.
↑
n.a.
↑
=
↓
↓
FA supply
High fat dietinduced obesity
LPL-KO
Cellular FA uptake
FATP1-OE
FAT/CD36-KO
CPT1bE3A-KI
(2)
Effect on function depends on feeding period/
composition (reviewed in (89)).
Protected against high sucrose- (85) or diabetesinduced (86) cardiomyopathy and TAC (57).
Lipotoxic cardiomyopathy.
Protected against lipotoxic cardiomyopathy at 3-4
months of age (65), lipotoxic cardiomyopathy by one
year of age (39).
After TAC: reduced FAO genes, increased glucose
oxidation genes, exacerbated dysfunction,
lipotoxicity. The homozygous KO is lethal (53, 78) or
has increased mortality (46).
Chapter 3
ACC2-KO
MCD-KO
(58)
(29, 73, 98,
99)
=
↑
=
=
=
=
↑
=/↓
↓
=/↑
=
=/↓
=
=/↓
(32, 37, 95,
96, 105)
(11, 12)
n.a.
↑
↑
=/↓
=/↓
↓
=/↓
n.a.
↑
↑
n.a.
=
=
=
Increased glucose oxidation, reduced cardiac
function and energetics, and lipotoxicity upon
fasting.
Transcriptional control of FAO enzymes
PGC1α-KO
(7, 8, 62)
=
PGC1α-OE
(82)
n.a.
=
n.a.
n.a.
n.a.
↓
↑
↑
↓
↓
=
=
=
PPARα-KO
PPARα-OE
Exacerbated dysfunction after TAC (8).
Preserved vascularity, reduced fibrosis and
apoptosis, but not contractile or mitochondrial
function upon TAC.
Reduced function and energetics at high workload.
Irreversible damage after ischemia.
FAO
VLCAD-KO
LCAD-KO
(72)
(28, 34)
↑
↑
n.a.
=
n.a.
↑
↓
↑
↑
↓
=
n.a.
=
↓
(99)
n.a.
n.a.
n.a.
=
=
n.a.
=
GLUT1-OE
(64, 83)
n.a.
n.a.
=
=
↑
=
=
GLUT1-KO
(82)
n.a.
n.a.
=
↑
↓
=
=
GLUT4-KO
(94)
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
=
Glucose pathway
PDHK-KO
Maintained function and energetics after TAC.
Protected against HFD induced-lipotoxicity (98),
increased glucose oxidation (29), improved recovery
(29) and decreased infarct size (99) upon I/R,
improved performance upon (73). Peri-weaning
mortality and cardiac dysfunction, which resolves
with age (4).
Increased glucose oxidation, improved recovery, and
reduced infarct size after I/R.
Maintained function and energetics after AAC (64),
but not in mice with inducible overexpression after
TAC (83).
Did not accelerate cardiac dysfunction after TAC,
similar lipotoxic effect after TAC.
Lower cardiac function upon fasting, delayed
recovery upon I/R.
Data on glucose oxidation, cellular FA uptake, or FAO are from perfused heart setups or gene/protein expression measurements. Lipotoxicity is indicated by the presence
of myocardial lipid intermediates, or lipid-associated fibrosis, apoptosis or oxidative stress. (↑: increased, ↓: increased, =: unchanged compared with controls, n.a.: data
not available). KO: knockout, KI: knockin, OE: overexpression.
26
Chapter 3
3.4. Animal models of altered FA metabolism
Numerous genetic and non-genetic mouse models exist with interferences at different levels of the
FA metabolic pathway (Table 3.1). Proteins in the FA metabolic pathway which correspond with
mouse models discussed in this review are highlighted with a bold border in Fig. 3.1. They are
extremely useful to study the effect of altered FA metabolism on cardiac function. In this section,
we discuss the findings in these mouse models, starting with modulations downstream of the FA
metabolic pathway, at the level of FAO, after which we move upstream, to conclude with
modulations at the level of FA supply from the circulation.
3.4.1. Modulation of FAO
Identification of patients with inherited acyl-CoA dehydrogenase deficiencies has led to the
development of several mouse models with FAO defects. VLCAD knockout (KO) mice developed
cardiomyopathy and presented with abnormal cardiac electrophysiological changes, which was
associated with myocardial TG accumulation (in case of whole-body KO) (32, 37, 103) and chronic
energy deficiency (95, 105).The cardiac phenotype of VLCAD-KO mice was relatively mild, which
can likely be explained by the overlapping substrate specificity of LCAD (20, 96). In LCAD-KO mice,
a compensatory increase in myocardial glucose utilization was observed (11). Upon fasting, which
normally results in an increased reliance on FAO, myocardial glucose uptake was not decreased
and PDH activity was much less decreased in hearts of LCAD-KO mice compared with controls (11),
most likely in an attempt to maintain cardiac energy levels. However, due to hypoglycaemia, the
sustained myocardial glucose uptake and PDH activity appeared ineffective to maintain metabolic
homeostasis, resulting in reduced in vivo cardiac energy status and function in fasted LCAD-KO
mice (11). This cardiac dysfunction in fasted LCAD-KO mice was further accompanied by increased
accumulation of myocardial TG (12), which indicates the onset of lipotoxicity.
3.4.2. Modulation of FA import into mitochondria
With FA as a major source of energy in the healthy heart, one may speculate that reduced
mitochondrial FA import may deprive the heart of energy and ultimately lead to cardiac
dysfunction. Indeed, reducing mitochondrial FA import through homozygous deletions of CPT1a and
CPT1b, the liver and muscle isoforms of CPT1, respectively, was shown to be embryonically lethal
(53, 78) or lead to increased mortality (46) (depending on gene targeting approach). Mice with
heterozygous deletion of CPT1b, however, do survive, and interestingly do not display altered
cardiac metabolic or functional phenotypes at rest (47). In line with this, inhibition of cardiac CPT1
activity by ~50% in rodents via supplementation of the CPT1 inhibitor etomoxir, did not affect long
chain FA uptake and FAO (70). Together, these findings indicate that CPT1 is necessary but not
rate limiting for regulation of cardiac FAO (69).
Mitochondrial FA import could also be reduced by increasing malonyl-CoA levels, which
consequently inhibits CPT1 activity. Deletion of MCD in mice resulted in increased malonyl-CoA
levels (29). Surprisingly, gene expression of CPT1b in MCD-KO mice was increased, together with
an increase in other PPARα responsive genes, which led to maintained FAO, glucose oxidation, and
cardiac function (29). However, in contrast to wild-type mice, high fat diet (HFD) feeding did not
increase FAO in MCD-KO mice, which prevented accumulation of metabolic intermediates of
34
Altered fatty acid metabolism – review
incomplete FA oxidation, resulting in protection against HFD-induced cardiac dysfunction (98).
Another study recently reported that deletion of MCD did however result in cardiac dysfunction
during the peri-weaning period, when the heart is highly dependent on FA as a substrate (4).
Interestingly, the early cardiac dysfunction was shown to resolve with age (4).
Opposite to the attempts to reduce mitochondrial FA import as discussed above, several mouse
models have been developed with increased mitochondrial FA import. Cardiac specific deletion of
ACC2 in mice resulted in decreased malonyl-CoA levels and increased FAO (58). In diet-induced
obesity (98) and diabetes (1), increased FAO has been associated with the accumulation of
myocardial lipid intermediates and cardiac dysfunction. However, in ACC2-KO mice myocardial lipid
intermediates were not increased, indicating that FA supply and oxidation are in balance. This
resulted in preserved cardiac function and energetics in these mice, even until one year of age
(58). As an alternative to increasing malonyl-CoA levels, mitochondrial FA import can also be
increased by reducing the sensitivity of CPT1b to malonyl-CoA, such as demonstrated in CPT1bE3A
knockin mice. These mice have increased FAO flux, while myocardial TG content was not different
compared with wild-type mice (Chapter 8). In contrast with the ACC2 KO mice (58), however,
CPT1bE3A knockin mice displayed impaired cardiac function and energetics (2), which underscores
the importance of malonyl-CoA control on Cpt1b for cardiac function and metabolic homeostasis.
3.4.3. Modulation of myocardial TG lipolysis / synthesis
Suppression of myocardial TG lipolysis in ATGL-KO mice was shown to reduce PPARα expression
and lead to lethal cardiomyopathy. This was associated with severely reduced FAO, excessive
myocardial TG accumulation, and impaired mitochondrial function (44). Accordingly, treatment of
ATGL-KO mice with a PPARα agonist was able to fully restore cardiac performance (44, 104).
Interestingly, unlike mice with whole-body deletion of ATGL (43, 44), mice with inducible deletion
of cardiac-specific ATGL did not show reduced PPARα levels, but still FAO and cardiac function were
reduced (56). In this case, the lower FAO was not caused by lower expression of PPARα, but most
likely was due to the decrease in FA liberated by TG hydrolysis and the suppression of cellular FA
uptake by CD36 (56). Indeed, CD36-mediated FA uptake is known to be tightly coupled to both TG
synthesis and lipolysis (19, 71).
While deletion of ATGL seems detrimental (43, 44, 56), mice with ATGL overexpression had
improved cardiac function, leading to better adaptation to high workload conditions compared with
control mice (57). Surprisingly, like in ATGL deficiency, ATGL overexpression resulted in reduced
myocardial FA uptake and oxidation and reduced PPARα levels (57). However, in contrast to ATGLKO mice, ATGL-overexpressing mice showed reduced TG content, and the reduced FAO was
compensated for by an increase in glucose oxidation. This shift in substrate reliance resulted in
maintained absolute acetyl-CoA production rates, suggesting no impairment in ATP supply, and
sufficient
ATP
production
to
maintain
normal
cardiac
function
(57).
Furthermore,
ATGL
overexpression was shown to protect the heart against diabetic and high-fat, high-sucrose (HFHS)
diet-induced cardiomyopathy through the attenuation of cardiac steatosis and lower reliance on FA
utilization compared with wild-type controls (85, 86). This lower reliance on FA utilization was
associated with a blunted increase in PPARα and CD36 expression in ATGL-overexpressing mice
compared with controls (85, 86).
35
Chapter 3
Studies in ATGL-KO and overexpression models seem to support the hypothesis that reduced
myocardial TG storage is associated with improved cardiac function. In agreement, HSL
overexpression reduced TG and diglyceride (DG) levels and, consequently, lipotoxicity (92, 97),
and was able to decrease mortality rate in diabetic mice (97). Interestingly, in diacylglycerol
transferase 1 (DGAT1)-overexpressing mice, the enzyme catalysing the synthesis of TG from DG,
increased myocardial TG content was not associated with reduced cardiac function. DGAT1
overexpression attenuated cardiac lipotoxicity by reducing toxic intermediates like DG and
ceramides (65). Findings from another independent study, however, suggest that the beneficial
effects of DGAT1 overexpression are time-course dependent, and that beyond one year of age
DGAT1-overexpressing mice still develop severe cardiomyopathy and myocardial fibrosis (39).
Nevertheless, backcrossing DGAT1-overexpressing mice with a mouse model of lipotoxicity resulted
in improved cardiac function and survival rate, which further supports the finding that DGAT1
overexpression during acute or subacute fatty acid overload may be cardio-protective (65). This
study suggests that rather than the absolute size of myocardial TG storage it is the dynamics of TG
metabolism that plays a role in the cardio-protection. Other animal models with alterations in TG
metabolism have been extensively reviewed in (55).
3.4.4. Modulation of cellular FA import
An increased cellular FA uptake without a concomitant increase in mitochondrial (FA) oxidative flux
can lead to lipotoxicity (41, 55, 102). In mice with cardiac overexpression of FATP1 (21) or ACSL1
(22), an imbalance between FA uptake and oxidation has indeed been shown to cause cardiac
dysfunction. FATP1 overexpression resulted in increased FA uptake and FAO, but also in FA
accumulation, indicating that mitochondrial FAO capacity is insufficient to oxidize excess FA during
lipid overload.
Deletion of CD36 in mice resulted in reduced FAO and lower intramyocardial TG and DG, without
compromised effects on cardiac function and energetics (59, 91), which is ascribed to a substantial
compensatory increase in glucose oxidation (59). Additionally, deletion of CD36 was able to
attenuate HFD-induced increases in TG, DG, and ceramides (91). Similar to CD36-deficient mice,
whole-body and cardiac-specific ACSL1-deficient mice showed a remarkable decrease in FAO and
myocellular lipids, with a concomitant increase in glucose and pyruvate oxidation. This resulted in
hypertrophy and cardiac diastolic dysfunction, whereas systolic function was maintained (31, 81).
3.4.5. Modulation of FA supply from blood
Increased FA availability usually leads to increased FA uptake and FAO in the heart (25, 100). We
recently demonstrated that HFD feeding in mice increased mitochondrial FAO capacity. However,
the increase in FAO could not match the FA supply and uptake, leading to myocardial TG
accumulation and lipotoxicity-associated oxidative stress, fibrosis, impaired calcium homeostasis
and cardiac dysfunction (3). Also other rodent studies have shown that HFD feeding led to
accumulation of myocardial TG and lipid intermediates, and lipotoxicity-associated cardiac
hypertrophy and/or dysfunction (18, 36, 79, 80, 98). Especially this TG accumulation occurred not
only in the presence of increased FAO capacity, but also in the presence of increased FAO flux (80).
Chronically increased FA uptake due to increased sarcolemmal CD36 localization has been found to
be the primary cause of lipid accumulation in hearts of HFD-fed rodents (38, 80).
36
Altered fatty acid metabolism – review
Other than lipotoxicity, impaired cardiac energetics has been suggested as a cause for obesityrelated cardiac dysfunction (16, 25). HFD feeding has been shown to lead to mitochondrial
uncoupling and reduced cardiac efficiency (25). In contrast, we showed that HFD feeding in mice
did not result in impaired cardiac energetics in vivo (3). This is in agreement with ex vivo isolated
perfused heart data (106) that showed decreased contractile function, but unaffected myocardial
PCr and ATP concentrations after HFD-feeding. Although detrimental effects of HFD feeding on the
heart have clearly been shown, cardiac dysfunction was not always observed (17, 52, 68). These
differences in the effect of HFD feeding on cardiac function are likely to be explained by differences
in study design, such as diet composition, duration of the diet, species and gender, as well as the
measurement techniques used (reviewed in (89)).
Given that lipoproteins are a major source of FA supply to the heart, ablation of cardiac LPL will
lead to a marked impairment of this supply. Cardiac LPL ablation was associated with lower CD36
and FATP1 expression, lower myocardial uptake of FA liberated from lipoprotein-TG, and lower
myocardial TG content (9, 10). Additionally, FAO was reduced in cardiac LPL-KO mice, which was
accompanied by increased glucose oxidation, and cardiac hypertrophy and dysfunction (10, 77).
Further stimulation of glucose utilization, by backcrossing cardiac LPL-KO mice with GLUT1overexpressing mice, resulted in improved cardiac function and energetics. This suggests that
cardiomyopathy in cardiac LPL-KO mice was rather due to defective energy production and not the
reduced FA uptake per se (54).
Taken together, data from mouse models with alterations in FA metabolism show that detrimental
effects seem to be caused by insufficient energy production in the heart, for example due to
reduced FAO or uncoupling of the electron transport chain from ATP production. Additionally,
upregulation of upstream components of the FA metabolic pathway that is not paralleled by
concomitant increases downstream of the FA metabolic pathway can result in lipotoxicity and
consequently lead to cardiac dysfunction. To examine whether and when modulations of FA
metabolism are detrimental during the development of heart failure, we review literature on mouse
models of altered FA metabolism upon induction of ischemic and non-ischemic heart failure in the
following section.
3.5. Modulation of FA metabolism in heart failure
3.5.1. Ischemic heart failure
Both during and after a cardiac ischemic insult, a shift away from FA towards glucose utilization
seems beneficial. The importance of glycolysis during ischemia is particularly obvious in mice with
GLUT4 deficiency, which have accelerated cardiac ATP depletion during ischemia and delayed ATP
recovery after ischemia, together with profound cardiac dysfunction (94). Mice with high FAO as a
result of PPARα overexpression (28) or PPARα activation (45) showed exacerbated cardiac
dysfunction after ischemia-reperfusion, which was related to glycogen deposition, apoptosis, and
oxidative stress, in addition to cardiac inefficiency due to glycolysis-glucose oxidation uncoupling.
Furthermore, the detrimental effects of high reliance on FAO have been demonstrated in diabetic
db/db mice, which have lower survival rates after ischemia reperfusion (1, 17, 42). Indirect
stimulation of glucose oxidation as a result of FAO suppression in MCD-KO mice (29, 73, 99) and
37
Chapter 3
direct stimulation of glucose oxidation in PDHK4-KO mice (99) were shown to decrease myocardial
infarct size and to protect the heart against ischemia-reperfusion injury (29). Although upon
coronary artery ligation ATP production rates in MCD-KO mice were shown to be lower than in wildtype mice, cardiac function was better in MCD-KO mice than in wild-type mice, demonstrating
improved coupling between glycolysis and glucose oxidation and improved cardiac efficiency to use
ATP for cardiac work, rather than for maintaining ion homeostasis (73). Indirect and direct
stimulations of glucose oxidation therefore appear effective strategies in the treatment of ischemic
heart failure (61, 99).
3.5.2. Non-ischemic heart failure
Like in ischemic heart failure conditions, the stimulation of glucose utilization has also been proven
beneficial under heart failure conditions with sufficient oxygen supply. Data from mice
overexpressing GLUT1 from birth showed that a substantial increase in glucose uptake and
glycolytic rates can preserve cardiac function and energetics in pressure overload conditions (64).
However, it is worth noting that while the induction of GLUT1 overexpression just 2 days before
induction of pressure overload was sufficient to increase glucose utilization, maintain mitochondrial
function and ATP synthesis, and attenuate the adverse remodeling process, it failed to rescue
cardiac function (83). Data from GLUT1-overexpressing mice suggest that the shift towards glucose
utilization upon pressure overload may be an adaptive mechanism to cope with the extra workload.
However, during stressed conditions such as heart failure, the ability of the heart to sufficiently
increase glucose uptake seems to be impaired (76, 108). This is further demonstrated in PGC1α-KO
mice (62) and PPARα-KO mice (72), which have an increased reliance on carbohydrates as a
substrate for oxidation as a compensation for their decreased FAO. Under normal conditions, hearts
of these mice are able to sustain normal energy metabolism, however, during a high workload
challenge the hearts fail to sustain high contractile performance (62, 72). Interestingly, a further
increase in glucose transport and utilization, by backcrossing the PPARα-KO mice with GLUT1overexpressing mice, corrected the decreased contractile and metabolic reserve in PPARα-KO mice
(72). This suggests that further stimulation of glucose oxidation can help to provide sufficient ATP
levels required during high workload. In line with this, increased glucose oxidation, as
compensation for decreased FAO in cardiac-specific ATGL-overexpressing mice, appeared sufficient
to maintain myocardial energetics and prevent cardiac dysfunction upon TAC-induced pressure
overload (57).
The data above suggest that increased glucose utilization can be a sufficient mechanism to protect
against pressure overload induced cardiac dysfunction. Currently, inhibition of FAO has evolved as
an established approach to treat heart failure patients. However, it can be questioned if inhibition
of FAO without an accompanied increase in glucose oxidation is indeed beneficial for the pressure
overload-challenged heart. Several mouse models with reduced FAO due to direct modulation of FA
metabolism, such as CPT1b-KO (47), CD36-KO (91), PGC1α-KO (8), and LPL-KO (10) mice, show
exacerbated cardiac dysfunction upon pressure overload. In all of these models, the cardiac
functional impairments seemed to be related to insufficient energy production in the heart and/or
cardiac lipotoxicity. In CPT1b-KO mice, the exacerbated cardiac dysfunction upon TAC-induced
pressure overload was associated with reduced FAO, increased TG content and ceramide formation,
as well as increased myocardial fibrosis and apoptosis (47). In cardiac LPL-deficient mice (10),
pressure overload resulted in more severe cardiac dysfunction as compared with wild-type mice,
38
Altered fatty acid metabolism – review
which was most likely due to the impaired FA supply and the associated reductions in FAO and
myocardial ATP content (74). Also PCG1α-KO mice, in which the suppression of cardiac FAO was
associated with significant defects in their ATP balance at rest (7), showed accelerated pressure
overload-induced heart failure (8). Interestingly, when CD36-KO mice were fed a HFD, they were
protected against TAC-induced cardiac dysfunction, which suggests a need for FA substrates during
mechanical stress (91).
Inhibition of FAO might thus not be beneficial when this results in impaired cardiac energetics.
Increasing FA utilization might in fact be an alternative mechanism to treat pressure overloadinduced heart failure. This was demonstrated in a mouse model of cardiac-specific ACC2 deficiency,
which was shown to preserve high FAO, prevent a metabolic shift towards increased reliance on
glycolysis, and better sustain myocardial energetics and function upon pressure overload (58).
However, maintaining the balance between FA supply/uptake and FAO is crucial when stimulating
FA utilization, to prevent the accumulation of TG and lipid intermediates and the associated
lipotoxicity (58). Hence, strategies designed to upregulate (CD36-mediated) FA uptake would be
inadvisable in view of the danger of exceeding the FAO capacity. In addition, it is important that
the increase in FAO is sufficient, not only to provide basal energy levels, but also to meet energy
requirements during high workloads. This was shown in PGC1α-overexpressing and GLUT1-KO
mice, which have high FAO, but which failed to maintain their cardiac energetics and function upon
pressure overload (82). Finally, maintaining coupling between glycolysis and glucose oxidation also
needs to be taken into consideration, because reduced glucose oxidation may occur secondary to
increased FAO (33).
3.6. Summary and future outlook
To summarize (Fig. 3.2), in the ischemic heart, where oxygen supply is limited, a shift away from
aerobic FAO towards anaerobic glycolysis seems necessary to assist in sufficient production of ATP.
During ischemia reperfusion, a shift towards glucose oxidation seems favourable to prevent
uncoupling between glycolysis and glucose oxidation. This uncoupling may lead to the accumulation
of protons and lactate, which in turn lowers cardiac energetic efficiency because more energy is
used to reduce the levels of these harmful metabolites. In the non-ischemic failing heart,
particularly under pressure overload conditions, it seems that preserved energetics, rather than
increased glucose oxidation and/or decreased FAO per se, is key for restoring cardiac function. A
shift towards glucose oxidation may be a rescue mechanism under stressed conditions; however,
the increase in glucose utilization then needs to be sufficient to maintain the required energy
levels. On the other hand, the shift towards glucose oxidation may not be necessary and energetic
compensation through high FAO can also be a successful strategy to protect the heart from failure.
Therefore, the efficacy of FAO inhibition in the treatment of non-ischemic heart failure is as of yet
not evident.
Future research should focus on elucidating when and how alterations in FA and glucose
metabolism affect heart failure development and progression. Data which provide a comprehensive
view of cardiac metabolic adaptations during different stages of the development and progression
of heart failure are currently lacking, while the availability of these data would be invaluable to give
insight into cause-and-effect relationships between cardiac metabolic and functional changes in the
failing heart. Therefore, non-invasive in vivo techniques, such as magnetic resonance imaging
39
Chapter 3
(MRI), magnetic resonance spectroscopy (MRS), and positron emission tomography (PET), that
allow longitudinal measurements of cardiac function and metabolism during the progression of
heart failure within the same animal, are crucial for advancement in this field.
Figure 3.2. Myocardial substrate contribution and potential therapeutic targets in (A) ischemic and (B)
non-ischemic heart failure. The relative contribution of the major metabolic pathways (fatty acid
oxidation (FAO), glucose oxidation, and glycolysis) to myocardial energy production changes during
heart failure development and progression. Therapeutic targets should aim at providing sufficient energy
production. The probability of success of therapeutic strategies largely depends on the type and stage of
heart failure.
40
Altered fatty acid metabolism – review
3.7. Acknowledgement
This work was supported by the Netherlands Organisation for Scientific Research (NWO) (grant
numbers 700.58.421 VIDI to J.J.P., 916.14.050 VENI to M.N.).
3.8. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Aasum E, Hafstad AD, Severson DL, and Larsen TS. Age-dependent changes in metabolism,
contractile function, and ischemic sensitivity in hearts from db/db mice. Diabetes 52: 434-441, 2003.
Abdurrachim D, Ciapaite J, van Weeghel M, Wessels B, Nicolay K, Houten SM, and Prompers JJ.
In vivo cardiac 1H MRS, 31P MRS, and MRI in mouse models of increased fatty acid oxidation with and
without myocardial lipid accumulation. In: Proc 21st Ann Meeting ISMRM p. 117, 2013.
Abdurrachim D, Ciapaite J, Wessels B, Nabben M, Luiken JJ, Nicolay K, and Prompers JJ. Cardiac
diastolic dysfunction in high-fat diet fed mice is associated with lipotoxicity without impairment of cardiac
energetics in vivo. Biochim Biophys Acta 1841: 1525-1537, 2014.
Aksentijevic D, McAndrew DJ, Karlstadt A, Zervou S, Sebag-Montefiore L, Cross R, Douglas G,
Regitz-Zagrosek V, Lopaschuk GD, Neubauer S, and Lygate CA. Cardiac dysfunction and periweaning mortality in malonyl-coenzyme A decarboxylase (MCD) knockout mice as a consequence of
restricting substrate plasticity. J Mol Cell Cardiol 75C: 76-87, 2014.
Allard MF, Schonekess BO, Henning SL, English DR, and Lopaschuk GD. Contribution of oxidative
metabolism and glycolysis to ATP production in hypertrophied hearts. Am J Physiol 267: H742-750, 1994.
An D, Pulinilkunnil T, Qi D, Ghosh S, Abrahani A, and Rodrigues B. The metabolic "switch" AMPK
regulates cardiac heparin-releasable lipoprotein lipase. Am J Physiol Endocrinol Metab 288: E246-253,
2005.
Arany Z, He H, Lin J, Hoyer K, Handschin C, Toka O, Ahmad F, Matsui T, Chin S, Wu PH, Rybkin,
II, Shelton JM, Manieri M, Cinti S, Schoen FJ, Bassel-Duby R, Rosenzweig A, Ingwall JS, and
Spiegelman BM. Transcriptional coactivator PGC-1 alpha controls the energy state and contractile
function of cardiac muscle. Cell Metab 1: 259-271, 2005.
Arany Z, Novikov M, Chin S, Ma Y, Rosenzweig A, and Spiegelman BM. Transverse aortic
constriction leads to accelerated heart failure in mice lacking PPAR-gamma coactivator 1alpha. Proc Natl
Acad Sci U S A 103: 10086-10091, 2006.
Augustus A, Yagyu H, Haemmerle G, Bensadoun A, Vikramadithyan RK, Park SY, Kim JK,
Zechner R, and Goldberg IJ. Cardiac-specific knock-out of lipoprotein lipase alters plasma lipoprotein
triglyceride metabolism and cardiac gene expression. J Biol Chem 279: 25050-25057, 2004.
Augustus AS, Buchanan J, Park TS, Hirata K, Noh HL, Sun J, Homma S, D'Armiento J, Abel ED,
and Goldberg IJ. Loss of lipoprotein lipase-derived fatty acids leads to increased cardiac glucose
metabolism and heart dysfunction. J Biol Chem 281: 8716-8723, 2006.
Bakermans AJ, Dodd MS, Nicolay K, Prompers JJ, Tyler DJ, and Houten SM. Myocardial energy
shortage and unmet anaplerotic needs in the fasted long-chain acyl-CoA dehydrogenase knockout mouse.
Cardiovasc Res 100: 441-449, 2013.
Bakermans AJ, Geraedts TR, van Weeghel M, Denis S, Joao Ferraz M, Aerts JM, Aten J, Nicolay
K, Houten SM, and Prompers JJ. Fasting-induced myocardial lipid accumulation in long-chain acyl-CoA
dehydrogenase knockout mice is accompanied by impaired left ventricular function. Circ Cardiovasc
Imaging 4: 558-565, 2011.
Banke NH, Wende AR, Leone TC, O'Donnell JM, Abel ED, Kelly DP, and Lewandowski ED.
Preferential oxidation of triacylglyceride-derived fatty acids in heart is augmented by the nuclear receptor
PPARalpha. Circ Res 107: 233-241, 2010.
Barger PM, and Kelly DP. PPAR signaling in the control of cardiac energy metabolism. Trends Cardiovasc
Med 10: 238-245, 2000.
Boudina S, and Abel ED. Mitochondrial uncoupling: a key contributor to reduced cardiac efficiency in
diabetes. Physiology (Bethesda) 21: 250-258, 2006.
Boudina S, Sena S, O'Neill BT, Tathireddy P, Young ME, and Abel ED. Reduced mitochondrial
oxidative capacity and increased mitochondrial uncoupling impair myocardial energetics in obesity.
Circulation 112: 2686-2695, 2005.
Brainard RE, Watson LJ, Demartino AM, Brittian KR, Readnower RD, Boakye AA, Zhang D,
Hoetker JD, Bhatnagar A, Baba SP, and Jones SP. High fat feeding in mice is insufficient to induce
cardiac dysfunction and does not exacerbate heart failure. PLoS ONE 8: e83174, 2013.
Calligaris SD, Lecanda M, Solis F, Ezquer M, Gutierrez J, Brandan E, Leiva A, Sobrevia L, and
Conget P. Mice long-term high-fat diet feeding recapitulates human cardiovascular alterations: an animal
model to study the early phases of diabetic cardiomyopathy. PLoS ONE 8: e60931, 2013.
Carley AN, Bi J, Wang X, Banke NH, Dyck JR, O'Donnell JM, and Lewandowski ED. Multiphasic
triacylglycerol dynamics in the intact heart during acute in vivo overexpression of CD36. J Lipid Res 54:
97-106, 2013.
Chegary M, Brinke H, Ruiter JP, Wijburg FA, Stoll MS, Minkler PE, van Weeghel M, Schulz H,
Hoppel CL, Wanders RJ, and Houten SM. Mitochondrial long chain fatty acid beta-oxidation in man and
mouse. Biochim Biophys Acta 1791: 806-815, 2009.
41
Chapter 3
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
42
Chiu HC, Kovacs A, Blanton RM, Han X, Courtois M, Weinheimer CJ, Yamada KA, Brunet S, Xu H,
Nerbonne JM, Welch MJ, Fettig NM, Sharp TL, Sambandam N, Olson KM, Ory DS, and Schaffer
JE. Transgenic expression of fatty acid transport protein 1 in the heart causes lipotoxic cardiomyopathy.
Circ Res 96: 225-233, 2005.
Chiu HC, Kovacs A, Ford DA, Hsu FF, Garcia R, Herrero P, Saffitz JE, and Schaffer JE. A novel
mouse model of lipotoxic cardiomyopathy. J Clin Invest 107: 813-822, 2001.
Christe ME, and Rodgers RL. Altered glucose and fatty acid oxidation in hearts of the spontaneously
hypertensive rat. J Mol Cell Cardiol 26: 1371-1375, 1994.
Christian B, El Alaoui-Talibi Z, Moravec M, and Moravec J. Palmitate oxidation by the mitochondria
from volume-overloaded rat hearts. Mol Cell Biochem 180: 117-128, 1998.
Cole MA, Murray AJ, Cochlin LE, Heather LC, McAleese S, Knight NS, Sutton E, Jamil AA, Parassol
N, and Clarke K. A high fat diet increases mitochondrial fatty acid oxidation and uncoupling to decrease
efficiency in rat heart. Basic Res Cardiol 106: 447-457, 2011.
Depre C, Vanoverschelde JL, and Taegtmeyer H. Glucose for the heart. Circulation 99: 578-588,
1999.
Drosatos K, and Schulze PC. Cardiac lipotoxicity: molecular pathways and therapeutic implications. Curr
Heart Fail Rep 10: 109-121, 2013.
Duerr GD, Heinemann JC, Arnoldi V, Feist A, Kley J, Ghanem A, Welz A, and Dewald O.
Cardiomyocyte specific peroxisome proliferator-activated receptor-alpha overexpression leads to
irreversible damage in ischemic murine heart. Life Sci 2014.
Dyck JR, Hopkins TA, Bonnet S, Michelakis ED, Young ME, Watanabe M, Kawase Y, Jishage K,
and Lopaschuk GD. Absence of malonyl coenzyme A decarboxylase in mice increases cardiac glucose
oxidation and protects the heart from ischemic injury. Circulation 114: 1721-1728, 2006.
Dyck JR, Kudo N, Barr AJ, Davies SP, Hardie DG, and Lopaschuk GD. Phosphorylation control of
cardiac acetyl-CoA carboxylase by cAMP-dependent protein kinase and 5'-AMP activated protein kinase.
Eur J Biochem 262: 184-190, 1999.
Ellis JM, Mentock SM, Depetrillo MA, Koves TR, Sen S, Watkins SM, Muoio DM, Cline GW,
Taegtmeyer H, Shulman GI, Willis MS, and Coleman RA. Mouse cardiac acyl coenzyme a synthetase
1 deficiency impairs Fatty Acid oxidation and induces cardiac hypertrophy. Mol Cell Biol 31: 1252-1262,
2011.
Exil VJ, Roberts RL, Sims H, McLaughlin JE, Malkin RA, Gardner CD, Ni G, Rottman JN, and
Strauss AW. Very-long-chain acyl-coenzyme a dehydrogenase deficiency in mice. Circ Res 93: 448-455,
2003.
Fillmore N, Mori J, and Lopaschuk GD. Mitochondrial fatty acid oxidation alterations in heart failure,
ischaemic heart disease and diabetic cardiomyopathy. Br J Pharmacol 171: 2080-2090, 2014.
Finck BN, Lehman JJ, Leone TC, Welch MJ, Bennett MJ, Kovacs A, Han X, Gross RW, Kozak R,
Lopaschuk GD, and Kelly DP. The cardiac phenotype induced by PPARalpha overexpression mimics that
caused by diabetes mellitus. J Clin Invest 109: 121-130, 2002.
Folmes CD, Sowah D, Clanachan AS, and Lopaschuk GD. High rates of residual fatty acid oxidation
during mild ischemia decrease cardiac work and efficiency. J Mol Cell Cardiol 47: 142-148, 2009.
Ge F, Hu C, Hyodo E, Arai K, Zhou S, Lobdell Iv H, Walewski JL, Homma S, and Berk PD.
Cardiomyocyte triglyceride accumulation and reduced ventricular function in mice with obesity reflect
increased long chain Fatty Acid uptake and de novo Fatty Acid synthesis. Journal of Obesity 2012:
205648, 2012.
Gelinas R, Thompson-Legault J, Bouchard B, Daneault C, Mansour A, Gillis MA, Charron G,
Gavino V, Labarthe F, and Des Rosiers C. Prolonged QT interval and lipid alterations beyond betaoxidation in very long-chain acyl-CoA dehydrogenase null mouse hearts. Am J Physiol Heart Circ Physiol
301: H813-823, 2011.
Glatz JF, Luiken JJ, and Bonen A. Membrane fatty acid transporters as regulators of lipid metabolism:
implications for metabolic disease. Physiol Rev 90: 367-417, 2010.
Glenn DJ, Wang F, Nishimoto M, Cruz MC, Uchida Y, Holleran WM, Zhang Y, Yeghiazarians Y, and
Gardner DG. A murine model of isolated cardiac steatosis leads to cardiomyopathy. Hypertension 57:
216-222, 2011.
Goldberg IJ, Eckel RH, and Abumrad NA. Regulation of fatty acid uptake into tissues: lipoprotein
lipase- and CD36-mediated pathways. J Lipid Res 50 Suppl: S86-90, 2009.
Goldberg IJ, Trent CM, and Schulze PC. Lipid metabolism and toxicity in the heart. Cell Metab 15: 805812, 2012.
Greer JJ, Ware DP, and Lefer DJ. Myocardial infarction and heart failure in the db/db diabetic mouse.
Am J Physiol Heart Circ Physiol 290: H146-153, 2006.
Haemmerle G, Lass A, Zimmermann R, Gorkiewicz G, Meyer C, Rozman J, Heldmaier G, Maier R,
Theussl C, Eder S, Kratky D, Wagner EF, Klingenspor M, Hoefler G, and Zechner R. Defective
lipolysis and altered energy metabolism in mice lacking adipose triglyceride lipase. Science 312: 734-737,
2006.
Haemmerle G, Moustafa T, Woelkart G, Buttner S, Schmidt A, van de Weijer T, Hesselink M,
Jaeger D, Kienesberger PC, Zierler K, Schreiber R, Eichmann T, Kolb D, Kotzbeck P, Schweiger
M, Kumari M, Eder S, Schoiswohl G, Wongsiriroj N, Pollak NM, Radner FP, Preiss-Landl K, Kolbe
T, Rulicke T, Pieske B, Trauner M, Lass A, Zimmermann R, Hoefler G, Cinti S, Kershaw EE,
Schrauwen P, Madeo F, Mayer B, and Zechner R. ATGL-mediated fat catabolism regulates cardiac
mitochondrial function via PPAR-alpha and PGC-1. Nat Med 17: 1076-1085, 2011.
Hafstad AD, Khalid AM, Hagve M, Lund T, Larsen TS, Severson DL, Clarke K, Berge RK, and
Aasum E. Cardiac peroxisome proliferator-activated receptor-alpha activation causes increased fatty acid
oxidation, reducing efficiency and post-ischaemic functional loss. Cardiovasc Res 83: 519-526, 2009.
Altered fatty acid metabolism – review
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
Haynie KR, Vandanmagsar B, Wicks SE, Zhang J, and Mynatt RL. Inhibition of carnitine
palymitoyltransferase1b induces cardiac hypertrophy and mortality in mice. Diabetes Obes Metab 16: 757760, 2014.
He L, Kim T, Long Q, Liu J, Wang P, Zhou Y, Ding Y, Prasain J, Wood PA, and Yang Q. Carnitine
palmitoyltransferase-1b deficiency aggravates pressure overload-induced cardiac hypertrophy caused by
lipotoxicity. Circulation 126: 1705-1716, 2012.
Heather LC, Pates KM, Atherton HJ, Cole MA, Ball DR, Evans RD, Glatz JF, Luiken JJ, Griffin JL,
and Clarke K. Differential translocation of the fatty acid transporter, FAT/CD36, and the glucose
transporter, GLUT4, coordinates changes in cardiac substrate metabolism during ischemia and reperfusion.
Circ Heart Fail 6: 1058-1066, 2013.
Hue L, and Taegtmeyer H. The Randle cycle revisited: a new head for an old hat. Am J Physiol
Endocrinol Metab 297: E578-591, 2009.
Irie H, Krukenkamp IB, Brinkmann JF, Gaudette GR, Saltman AE, Jou W, Glatz JF, Abumrad NA,
and Ibrahimi A. Myocardial recovery from ischemia is impaired in CD36-null mice and restored by
myocyte CD36 expression or medium-chain fatty acids. Proc Natl Acad Sci U S A 100: 6819-6824, 2003.
Jastroch M, Divakaruni AS, Mookerjee S, Treberg JR, and Brand MD. Mitochondrial proton and
electron leaks. Essays Biochem 47: 53-67, 2010.
Jeckel KM, Miller KE, Chicco AJ, Chapman PL, Mulligan CM, Falcone PH, Miller ML, Pagliassotti
MJ, and Frye MA. The role of dietary fatty acids in predicting myocardial structure in fat-fed rats. Lipids
Health Dis 10: 92, 2011.
Ji S, You Y, Kerner J, Hoppel CL, Schoeb TR, Chick WS, Hamm DA, Sharer JD, and Wood PA.
Homozygous carnitine palmitoyltransferase 1b (muscle isoform) deficiency is lethal in the mouse. Mol
Genet Metab 93: 314-322, 2008.
Khan RS, Lin Y, Hu Y, Son NH, Bharadwaj KG, Palacios C, Chokshi A, Ji R, Yu S, Homma S,
Schulze PC, Tian R, and Goldberg IJ. Rescue of heart lipoprotein lipase-knockout mice confirms a role
for triglyceride in optimal heart metabolism and function. Am J Physiol Endocrinol Metab 305: E13391347, 2013.
Kienesberger PC, Pulinilkunnil T, Nagendran J, and Dyck JR. Myocardial triacylglycerol metabolism.
J Mol Cell Cardiol 55: 101-110, 2013.
Kienesberger PC, Pulinilkunnil T, Nagendran J, Young ME, Bogner-Strauss JG, Hackl H, Khadour
R, Heydari E, Haemmerle G, Zechner R, Kershaw EE, and Dyck JR. Early structural and metabolic
cardiac remodelling in response to inducible adipose triglyceride lipase ablation. Cardiovasc Res 99: 442451, 2013.
Kienesberger PC, Pulinilkunnil T, Sung MM, Nagendran J, Haemmerle G, Kershaw EE, Young ME,
Light PE, Oudit GY, Zechner R, and Dyck JR. Myocardial ATGL overexpression decreases the reliance
on fatty acid oxidation and protects against pressure overload-induced cardiac dysfunction. Mol Cell Biol
32: 740-750, 2012.
Kolwicz SC, Jr., Olson DP, Marney LC, Garcia-Menendez L, Synovec RE, and Tian R. Cardiacspecific deletion of acetyl CoA carboxylase 2 prevents metabolic remodeling during pressure-overload
hypertrophy. Circ Res 111: 728-738, 2012.
Kuang M, Febbraio M, Wagg C, Lopaschuk GD, and Dyck JR. Fatty acid translocase/CD36 deficiency
does not energetically or functionally compromise hearts before or after ischemia. Circulation 109: 15501557, 2004.
Labinskyy V, Bellomo M, Chandler MP, Young ME, Lionetti V, Qanud K, Bigazzi F, Sampietro T,
Stanley WC, and Recchia FA. Chronic activation of peroxisome proliferator-activated receptor-alpha
with fenofibrate prevents alterations in cardiac metabolic phenotype without changing the onset of
decompensation in pacing-induced heart failure. J Pharmacol Exp Ther 321: 165-171, 2007.
Lee L, Horowitz J, and Frenneaux M. Metabolic manipulation in ischaemic heart disease, a novel
approach to treatment. Eur Heart J 25: 634-641, 2004.
Lehman JJ, Boudina S, Banke NH, Sambandam N, Han X, Young DM, Leone TC, Gross RW,
Lewandowski ED, Abel ED, and Kelly DP. The transcriptional coactivator PGC-1alpha is essential for
maximal and efficient cardiac mitochondrial fatty acid oxidation and lipid homeostasis. Am J Physiol Heart
Circ Physiol 295: H185-196, 2008.
Lei B, Lionetti V, Young ME, Chandler MP, d'Agostino C, Kang E, Altarejos M, Matsuo K, Hintze
TH, Stanley WC, and Recchia FA. Paradoxical downregulation of the glucose oxidation pathway despite
enhanced flux in severe heart failure. J Mol Cell Cardiol 36: 567-576, 2004.
Liao R, Jain M, Cui L, D’Agostino J, Aiello F, Luptak I, Ngoy S, Mortensen RM, and Tian R.
Cardiac-Specific Overexpression of GLUT1 Prevents the Development of Heart Failure Attributable to
Pressure Overload in Mice. Circulation 106: 2125-2131, 2002.
Liu L, Shi X, Bharadwaj KG, Ikeda S, Yamashita H, Yagyu H, Schaffer JE, Yu YH, and Goldberg
IJ. DGAT1 expression increases heart triglyceride content but ameliorates lipotoxicity. J Biol Chem 284:
36312-36323, 2009.
Liu Q, Docherty JC, Rendell JC, Clanachan AS, and Lopaschuk GD. High levels of fatty acids delay
the recovery of intracellular pH and cardiac efficiency in post-ischemic hearts by inhibiting glucose
oxidation. J Am Coll Cardiol 39: 718-725, 2002.
Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, and Stanley WC. Myocardial fatty acid metabolism
in health and disease. Physiol Rev 90: 207-258, 2010.
Louwe MC, van der Hoorn JW, van den Berg SA, Jukema JW, Romijn JA, Willems van Dijk K,
Rensen PC, Smit JW, and Steendijk P. Gender-dependent effects of high-fat lard diet on cardiac
function in C57Bl/6J mice. Appl Physiol Nutr Metab 37: 214-224, 2012.
Luiken JJ. Sarcolemmal fatty acid uptake vs. mitochondrial beta-oxidation as target to regress cardiac
insulin resistance. Appl Physiol Nutr Metab 34: 473-480, 2009.
43
Chapter 3
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
44
Luiken JJ, Niessen HE, Coort SL, Hoebers N, Coumans WA, Schwenk RW, Bonen A, and Glatz JF.
Etomoxir-induced partial carnitine palmitoyltransferase-I (CPT-I) inhibition in vivo does not alter cardiac
long-chain fatty acid uptake and oxidation rates. Biochem J 419: 447-455, 2009.
Luiken JJ, van Nieuwenhoven FA, America G, van der Vusse GJ, and Glatz JF. Uptake and
metabolism of palmitate by isolated cardiac myocytes from adult rats: involvement of sarcolemmal
proteins. J Lipid Res 38: 745-758, 1997.
Luptak I, Balschi JA, Xing Y, Leone TC, Kelly DP, and Tian R. Decreased Contractile and Metabolic
Reserve in Peroxisome Proliferator–Activated Receptor-α–Null Hearts Can Be Rescued by Increasing
Glucose Transport and Utilization. Circulation 112: 2339-2346, 2005.
Masoud WG, Ussher JR, Wang W, Jaswal JS, Wagg CS, Dyck JR, Lygate CA, Neubauer S,
Clanachan AS, and Lopaschuk GD. Failing mouse hearts utilize energy inefficiently and benefit from
improved coupling of glycolysis and glucose oxidation. Cardiovasc Res 101: 30-38, 2014.
Nakajima H, Ishida T, Satomi-Kobayashi S, Mori K, Hara T, Sasaki N, Yasuda T, Toh R, Tanaka H,
Kawai H, and Hirata K. Endothelial lipase modulates pressure overload-induced heart failure through
alternative pathway for fatty acid uptake. Hypertension 61: 1002-1007, 2013.
Neely JR, Liedtke AJ, Whitmer JT, and Rovetto MJ. Relationship between coronary flow and
adenosine triphosphate production from glycolysis and oxidative metabolism. Recent Adv Stud Cardiac
Struct Metab 8: 301-321, 1975.
Neglia D, De Caterina A, Marraccini P, Natali A, Ciardetti M, Vecoli C, Gastaldelli A, Ciociaro D,
Pellegrini P, Testa R, Menichetti L, L'Abbate A, Stanley WC, and Recchia FA. Impaired myocardial
metabolic reserve and substrate selection flexibility during stress in patients with idiopathic dilated
cardiomyopathy. Am J Physiol Heart Circ Physiol 293: H3270-3278, 2007.
Noh HL, Okajima K, Molkentin JD, Homma S, and Goldberg IJ. Acute lipoprotein lipase deletion in
adult mice leads to dyslipidemia and cardiac dysfunction. Am J Physiol Endocrinol Metab 291: E755-760,
2006.
Nyman LR, Cox KB, Hoppel CL, Kerner J, Barnoski BL, Hamm DA, Tian L, Schoeb TR, and Wood
PA. Homozygous carnitine palmitoyltransferase 1a (liver isoform) deficiency is lethal in the mouse. Mol
Genet Metab 86: 179-187, 2005.
Ouwens DM, Boer C, Fodor M, de Galan P, Heine RJ, Maassen JA, and Diamant M. Cardiac
dysfunction induced by high-fat diet is associated with altered myocardial insulin signalling in rats.
Diabetologia 48: 1229-1237, 2005.
Ouwens DM, Diamant M, Fodor M, Habets DD, Pelsers MM, El Hasnaoui M, Dang ZC, van den
Brom CE, Vlasblom R, Rietdijk A, Boer C, Coort SL, Glatz JF, and Luiken JJ. Cardiac contractile
dysfunction in insulin-resistant rats fed a high-fat diet is associated with elevated CD36-mediated fatty
acid uptake and esterification. Diabetologia 50: 1938-1948, 2007.
Paul DS, Grevengoed TJ, Pascual F, Ellis JM, Willis MS, and Coleman RA. Deficiency of cardiac AcylCoA synthetase-1 induces diastolic dysfunction, but pathologic hypertrophy is reversed by rapamycin.
Biochim Biophys Acta 1841: 880-887, 2014.
Pereira RO, Wende AR, Crum A, Hunter D, Olsen CD, Rawlings T, Riehle C, Ward WF, and Abel
ED. Maintaining PGC-1alpha expression following pressure overload-induced cardiac hypertrophy
preserves angiogenesis but not contractile or mitochondrial function. FASEB J 28: 3691-3702, 2014.
Pereira RO, Wende AR, Olsen C, Soto J, Rawlings T, Zhu Y, Anderson SM, and Abel ED. Inducible
overexpression of GLUT1 prevents mitochondrial dysfunction and attenuates structural remodeling in
pressure overload but does not prevent left ventricular dysfunction. J Am Heart Assoc 2: e000301, 2013.
Pereira RO, Wende AR, Olsen C, Soto J, Rawlings T, Zhu Y, Riehle C, and Abel ED. GLUT1
deficiency in cardiomyocytes does not accelerate the transition from compensated hypertrophy to heart
failure. J Mol Cell Cardiol 2014.
Pulinilkunnil T, Kienesberger PC, Nagendran J, Sharma N, Young ME, and Dyck JR. Cardiacspecific adipose triglyceride lipase overexpression protects from cardiac steatosis and dilated
cardiomyopathy following diet-induced obesity. Int J Obes 38: 205-215, 2014.
Pulinilkunnil T, Kienesberger PC, Nagendran J, Waller TJ, Young ME, Kershaw EE, Korbutt G,
Haemmerle G, Zechner R, and Dyck JR. Myocardial adipose triglyceride lipase overexpression protects
diabetic mice from the development of lipotoxic cardiomyopathy. Diabetes 62: 1464-1477, 2013.
Semenza GL, Roth PH, Fang HM, and Wang GL. Transcriptional regulation of genes encoding glycolytic
enzymes by hypoxia-inducible factor 1. J Biol Chem 269: 23757-23763, 1994.
Soyal S, Krempler F, Oberkofler H, and Patsch W. PGC-1alpha: a potent transcriptional cofactor
involved in the pathogenesis of type 2 diabetes. Diabetologia 49: 1477-1488, 2006.
Stanley WC, Dabkowski ER, Ribeiro RF, Jr., and O'Connell KA. Dietary fat and heart failure: moving
from lipotoxicity to lipoprotection. Circ Res 110: 764-776, 2012.
Stanley WC, Recchia FA, and Lopaschuk GD. Myocardial Substrate Metabolism in the Normal and
Failing Heart. Physiol Rev 85: 1093-1129, 2005.
Steinbusch LK, Luiken JJ, Vlasblom R, Chabowski A, Hoebers NT, Coumans WA, Vroegrijk IO,
Voshol PJ, Ouwens DM, Glatz JF, and Diamant M. Absence of fatty acid transporter CD36 protects
against Western-type diet-related cardiac dysfunction following pressure overload in mice. Am J Physiol
Endocrinol Metab 301: E618-627, 2011.
Suzuki J, Shen WJ, Nelson BD, Patel S, Veerkamp JH, Selwood SP, Murphy GM, Jr., Reaven E,
and Kraemer FB. Absence of cardiac lipid accumulation in transgenic mice with heart-specific HSL
overexpression. Am J Physiol Endocrinol Metab 281: E857-866, 2001.
Taegtmeyer H, and Overturf ML. Effects of moderate hypertension on cardiac function and metabolism
in the rabbit. Hypertension 11: 416-426, 1988.
Tian R, and Abel ED. Responses of GLUT4-deficient hearts to ischemia underscore the importance of
glycolysis. Circulation 103: 2961-2966, 2001.
Altered fatty acid metabolism – review
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
Tucci S, Flogel U, Hermann S, Sturm M, Schafers M, and Spiekerkoetter U. Development and
pathomechanisms of cardiomyopathy in very long-chain acyl-CoA dehydrogenase deficient (VLCAD(-/-))
mice. Biochim Biophys Acta 1842: 677-685, 2014.
Tucci S, Herebian D, Sturm M, Seibt A, and Spiekerkoetter U. Tissue-specific strategies of the verylong chain acyl-CoA dehydrogenase-deficient (VLCAD-/-) mouse to compensate a defective fatty acid
beta-oxidation. PLoS ONE 7: e45429, 2012.
Ueno M, Suzuki J, Zenimaru Y, Takahashi S, Koizumi T, Noriki S, Yamaguchi O, Otsu K, Shen WJ,
Kraemer FB, and Miyamori I. Cardiac overexpression of hormone-sensitive lipase inhibits myocardial
steatosis and fibrosis in streptozotocin diabetic mice. Am J Physiol Endocrinol Metab 294: E1109-1118,
2008.
Ussher JR, Koves TR, Jaswal JS, Zhang L, Ilkayeva O, Dyck JR, Muoio DM, and Lopaschuk GD.
Insulin-stimulated cardiac glucose oxidation is increased in high-fat diet-induced obese mice lacking
malonyl CoA decarboxylase. Diabetes 58: 1766-1775, 2009.
Ussher JR, Wang W, Gandhi M, Keung W, Samokhvalov V, Oka T, Wagg CS, Jaswal JS, Harris RA,
Clanachan AS, Dyck JR, and Lopaschuk GD. Stimulation of glucose oxidation protects against acute
myocardial infarction and reperfusion injury. Cardiovasc Res 94: 359-369, 2012.
van der Vusse GJ, van Bilsen M, and Glatz JF. Cardiac fatty acid uptake and transport in health and
disease. Cardiovasc Res 45: 279-293, 2000.
Wang H, and Eckel RH. Lipoprotein lipase: from gene to obesity. Am J Physiol Endocrinol Metab 297:
E271-288, 2009.
Wende AR, and Abel ED. Lipotoxicity in the heart. Biochim Biophys Acta 1801: 311-319, 2010.
Werdich AA, Baudenbacher F, Dzhura I, Jeyakumar LH, Kannankeril PJ, Fleischer S, LeGrone A,
Milatovic D, Aschner M, Strauss AW, Anderson ME, and Exil VJ. Polymorphic ventricular tachycardia
and abnormal Ca2+ handling in very-long-chain acyl-CoA dehydrogenase null mice. Am J Physiol Heart
Circ Physiol 292: H2202-2211, 2007.
Wolkart G, Schrammel A, Dorffel K, Haemmerle G, Zechner R, and Mayer B. Cardiac dysfunction in
adipose triglyceride lipase deficiency: treatment with a PPARalpha agonist. Br J Pharmacol 165: 380-389,
2012.
Xiong D, He H, James J, Tokunaga C, Powers C, Huang Y, Osinska H, Towbin JA, Purevjav E,
Balschi JA, Javadov S, McGowan FX, Jr., Strauss AW, and Khuchua Z. Cardiac-specific VLCAD
deficiency induces dilated cardiomyopathy and cold intolerance. Am J Physiol Heart Circ Physiol 306:
H326-338, 2014.
Yan J, Young ME, Cui L, Lopaschuk GD, Liao R, and Tian R. Increased glucose uptake and oxidation
in mouse hearts prevent high fatty acid oxidation but cause cardiac dysfunction in diet-induced obesity.
Circulation 119: 2818-2828, 2009.
Zhabyeyev P, Gandhi M, Mori J, Basu R, Kassiri Z, Clanachan A, Lopaschuk GD, and Oudit GY.
Pressure-overload-induced heart failure induces a selective reduction in glucose oxidation at physiological
afterload. Cardiovasc Res 97: 676-685, 2013.
Zhang L, Jaswal JS, Ussher JR, Sankaralingam S, Wagg C, Zaugg M, and Lopaschuk GD. Cardiac
insulin-resistance and decreased mitochondrial energy production precede the development of systolic
heart failure after pressure-overload hypertrophy. Circ Heart Fail 6: 1039-1048, 2013.
Zhong M, Alonso CE, Taegtmeyer H, and Kundu BK. Quantitative PET imaging detects early metabolic
remodeling in a mouse model of pressure-overload left ventricular hypertrophy in vivo. J Nucl Med 54:
609-615, 2013.
45
Chapter 4
High frame-rate retrospectively triggered cine MRI for
assessment of murine diastolic function
Bram Coolen*, Desiree Abdurrachim*, Abdallah G. Motaal, Klaas Nicolay, Jeanine J. Prompers,
Gustav J. Strijkers
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven,
The Netherlands
*both authors contributed equally
Published in Magnetic Resonance in Medicine 69: 648-656, 2013.
DOI: 10.1002/mrm.24287. Epub 2012 Apr 19.
Chapter 4
Abstract
To assess left ventricular (LV) diastolic function in mice with cine MRI, a high frame rate (> 60
frames per cardiac cycle) is required. For conventional ECG-triggered cine MRI, the frame rate is
equal to 1/TR. However, TR cannot be lowered at will to increase the frame rate because of
gradient hardware, spatial resolution and signal-to noise limitations. To overcome these
limitations associated with ECG-triggered cine MRI, in this paper we introduce a retrospectively
triggered cine MRI protocol capable of producing high-resolution high frame-rate cine MRI of the
mouse heart for addressing LV diastolic function. Simulations were performed to investigate the
influence of MRI sequence parameters and the k-space filling trajectory in relation to the desired
number of frames per cardiac cycle. An optimized protocol was applied in vivo and compared
with ECG-triggered cine for which a high-frame rate could only be achieved by several
interleaved acquisitions. Retrospective high frame-rate cine MRI proved superior to the
interleaved ECG-triggered protocols. High spatial-resolution cine movies with frames rates up to
80 frames per cardiac cycle were obtained in 25 min. Analysis of LV filling rate curves allowed
accurate determination of early and late filling rates and revealed subtle impairments in LV
diastolic function of diabetic mice in comparison with non-diabetic mice.
Keywords
mouse cardiac MRI, cine, diastolic function, retrospective triggering
48
High frame rate cine MRI of murine diastolic function
4.1. Introduction
The cardiac cycle is generally divided in a systolic and diastolic phase. In the diastolic phase,
filling of the left ventricle with blood occurs via early LV relaxation and late atrial contraction. E
and A denote the peak filling rates in these two distinct diastolic phases. In a healthy compliant
heart the E/A ratio is around 1.5, whereas in a less compliant heart a decrease in the E/A ratio
is observed, which clinically marks the onset of diastolic dysfunction (6, 16). Diabetic
cardiomyopathy has a high prevalence in type 2 diabetes patients and is characterized by
diastolic dysfunction (1, 7). The E/A ratio is therefore an important readout when studying
experimental models of diabetic cardiomyopathy and when evaluating emerging treatment
strategies for diastolic dysfunction.
The traditional way to measure the E/A ratio in humans and small animals is by Doppler
echocardiography (4, 5, 16, 18). Recently, Stuckey et al. demonstrated that also cardiac cine
MRI can be used to measure the E/A ratio in small rodents (20). For accurate definition of the E
and A filling phases a high cine MRI frame rate was required, providing approximately 60
frames per cardiac cycle.
For traditional ECG-triggered cine MRI the frame rate is equal to 1/TR. Reducing TR to achieve a
high frame rate is however subject to experimental limitations. The excitation pulse and echo
readout can be shortened by sacrificing spatial resolution or by increasing the gradient strength
and receiver bandwidth at the cost of a reduced signal-to-noise (SNR) ratio. However, gradient
strength and gradient duty cycle are limited by gradient hardware restrictions. Additionally, a
short TR reduces SNR and contrast-to-noise between blood and myocardium due to saturation
of available magnetization. As a remedy, one can interleave multiple ECG-triggered cine
acquisitions with moderately long TR to obtain an effectively shorter TR.
As an alternative approach to overcome the above limitations associated with ECG-triggered
cine MRI, we here introduce a retrospective triggering protocol capable of producing highresolution high frame-rate cine movies of the mouse heart for assessing LV diastolic function.
The retrospectively triggered sequence (9) records k-lines with constant TR asynchronously with
the cardiac and respiratory motion (Fig. 4.1A). From a navigator echo cardiac triggering and
respiratory gating are derived, which are used to retrospectively bin k-lines in the appropriate
cine frame and discard the k-lines acquired during respiratory motion. The frame rate for a
retrospectively triggered cine movie is not dictated by 1/TR, but can be chosen at will as long as
sufficient k-lines are measured to completely fill the k-spaces of all cardiac frames.
Because of the asynchronous nature of the retrospectively triggered cine MRI sequence, filling
of k-space is a stochastic process. We therefore performed simulations to optimize the sequence
parameters and the k-space-filling trajectory in relation to the desired effective number of
averages and number of frames per cardiac cycle. Subsequently, the optimized protocol was
applied in vivo in healthy mice and compared with interleaved ECG-triggered cine MRI. Finally,
the retrospective triggering protocol was applied to measure diastolic function of diabetic and
non-diabetic mice.
49
Chapter 4
4.2. Materials and Methods
4.2.1. Simulations
Figure 4.1. Retrospectively triggered sampling of cardiac frames. K-Lines are acquired
asynchronously with the periodic cardiac motion. The vertical lines represent reconstructions with
(1) a low frame rate for which TR < L/NF and (2) a high frame for which TR > L/NF. For the latter,
a potential problem occurs because some cardiac frames may receive no hits. Simulations in this
paper show that the asynchrony between sampling and the heart cycle is sufficient to assure
efficient filling of all cardiac frames, even for very high frame rates. Different k-space acquisition
trajectories were tested: (A) k-space averaged, (B) k-line averaged, and (C) random sampling.
Numerical
simulations of the
retrospectively triggered
acquisitions were
performed in
Mathematica 7.0 (Wolfram Research, Champaign, IL, USA). Fig. 4.1 schematically illustrates the
retrospectively triggered cine acquisition with the 3 investigated k-space sampling trajectories:
a linear scheme in which k-space is filled in a ‘zigzag’ phase encoding pattern repeated a
number of times (‘k-space averaged’, Fig. 4.1A); a linear scheme in which first individual k-lines
are repeated a number of times, before acquisition of the next phase encoding step (‘k-line
averaged’, Fig. 4.1B); and a random k-space filling trajectory (‘random’, Fig. 4.1C). The number
of phase encoding steps was NP and NR refers to the number of times the total k-space was
measured (number of repetitions). The simulated total acquisition time therefore amounted to
NP x NR x TR.
Over the duration of the simulated acquisition time, the length of the cardiac cycle was assumed
Gaussian distributed around a mean value L and with a standard deviation δ in order to address
the effects of variations in the duration of the R-R interval. The Gaussian distribution was
experimentally
determined
from
experiments
with
healthy
C57BL/6
mice
(Fig.
4.2A).
Respiration was not included in the simulations as this merely involves discarding a certain
fraction of k-lines (typically 30%) acquired during respiratory movement.
50
High frame rate cine MRI of murine diastolic function
From the simulated data, each cardiac cycle was divided in the desired number of cardiac
frames (NF), after which all k-lines were binned for assignment to the different cardiac frames.
Because of the binning and asynchrony between the cardiac cycle and TR, not all k-lines were
averaged an equal number of times. The effective number of averages NA eff for a cine movie
was therefore defined as the minimal number of averages that was reached for 99% of all kline/cardiac frame combinations. This high threshold was chosen because a full sampling of kspace for each cardiac frame was considered to be important for the accuracy of diastolic
function parameters. NAmax refers to the maximum averages that would be obtained when all
acquisitions were evenly distributed over all k-spaces of all cardiac frames. NAmax is therefore
equal to NR/NF. Each simulation was performed 10 times, from which the average NAeff was
calculated.
4.2.2. MRI
MRI measurements were performed with a 9.4 T Bruker animal scanner (Bruker BioSpin GmbH,
Ettlingen, Germany). The gradient system allowed a maximum gradient strength of 740 mT/m
with a slew rate of 6600 T/m/s. A 4-element (2x2) phased array surface receive coil for mouse
cardiac MRI was used in combination with a 72-mm-diameter volume transmit coil (Bruker
BioSpin).
4.2.2.1. Systolic function
Systolic function measurements were performed using a 2D ECG- and respiratory-triggered
FLASH sequence with the following parameters: TR = 7.0 ms, TE = 1.8 ms, number of cardiac
frames = 12-15, FA = 15o, NP  Nfreq = 192  192, FOV = 3  3 cm2, slice thickness = 1 mm, NA
= 6, scan time  3 min. The sequence was repeated for 5-6 slices in LV short-axis orientation
from base to apex and in 2- and 4-chamber long-axis orientations.
4.2.2.2. Diastolic function with ECG triggering
The same ECG-triggered sequence described above for systolic function measurements was also
used for diastolic function measurements in a single midventricular slice. To achieve the higher
frame rate necessary for accurate delineation of the diastolic filling phases, the sequence was
interleaved multiple times. Interleaving was done in two ways. In method I, four consecutive
separate scans were made with trigger delays of 0, ¼TR, ½TR, and ¾TR, respectively, which
were subsequently interleaved to obtain an effective frame duration of 1/4TR. In method II, the
same delays were immediately applied for consecutive ECG-triggers, resulting in one single cine
movie containing all cardiac frames. The effective frame duration of both interleaved
acquisitions was 1.75 ms. ECG-triggering was performed every other R-wave to assure that the
full diastolic phase was covered. Eight averages were used resulting in a total acquisition time of
∼20 min.
4.2.2.3. Diastolic function with retrospective triggering
For diastolic function measurements with retrospective triggering, the k-line averaging method
was used. The sequence was a single-slice FLASH sequence with an in-slice navigator, as
described by Heijman et al. (9). Sequence parameters were TR = 4.7 ms, TE = 2.35 ms, FA =
15o, RF pulse = 1 ms Gaussian, NP  Nfreq = 128  192, parallel imaging acceleration (PI) = 1.5,
FOV = 3  3 cm2, slice thickness = 1 mm, NR = 2500, scan time = 25 min.
51
Chapter 4
Reconstruction of the retrospectively triggered cine movies was done off-line using homebuilt
software in Matlab 8.1 (The Mathworks, Natick, MA, USA). All navigator echo signals were
concatenated and interpolated with a spline function. Subsequently, a local maximum detection
algorithm was applied to determine the start of each cardiac cycle. Next, all cardiac cycles were
divided into NF cardiac frames, after which the k-lines were assigned to their corresponding
cardiac frames. K-lines that were acquired during respiratory motion (typically 30% of the
acquisition time) were omitted. No sharing or interpolation of k-lines between frames was used.
Image reconstruction was performed using the GRAPPA reconstruction algorithm (8).
4.2.3. Animal handling
A group of n=4 healthy C57BL/6 mice (Charles River, Maastricht, Netherlands) was used to
compare the retrospectively triggered and the ECG-triggered acquisition protocols. The
retrospectively triggered method was applied to longitudinally measure diastolic function of nondiabetic (db/+, n=6) and diabetic (db/db, n=6) C57BL/Ks mice (Harlan Laboratories, Venray,
The Netherlands) at 7, 13, and 19 weeks of age. Mice were housed under controlled conditions.
Both db/+ and db/db mice were put on a diet with 18% of the calories from fat (2018 Teklad
Global 18% Protein Rodent Diet, Harlan, Bicester, UK). Before MRI, anesthesia was induced with
2–3% isoflurane in 0.4 L/min medical air and maintained during MRI with 1–2% isoflurane. Mice
were positioned in prone position on the surface coil and placed in the MRI scanner.
Temperature was maintained at 36–37°C with a heating pad and monitored with a rectal probe.
Respiration frequency was kept at 60–80 bpm, monitored using a pressure balloon. For
prospectively triggered MRI, the ECG signal was obtained by putting EEG paste on the front
paws and taping them on copper ECG-electrodes, which were connected to an ECG gating
system (Small Animal Instruments Inc., Stony Brook, NY). Mice were sacrificed following the
last MRI measurement. The local institutional animal care committee (Maastricht University
Maastricht, Netherlands) reviewed and approved all experimental procedures.
4.2.4. Data analysis
4.2.4.1. Systolic function
LV volumes were determined by segmenting the epicardium and endocardium (excluding
papillary muscles) in end-diastolic and end-systolic frames in all short-axis slices using CAAS
MRV FARM software (Pie Medical Imaging, Maastricht, Netherlands). From the segmentations,
end-systolic volume, end-diastolic volume, ejection fraction, stroke volume, and LV mass index
(LVM = LV mass normalized by tibia length) were calculated.
4.2.4.2. Diastolic function
LV volume–time curves were created by semiautomatic segmentation of the LV endocardium
(excluding papillary muscles) using Segment (version 1.8 R1145, http://segment.heiberg.se)
(11). LV volume–time curves were normalized to the end-diastolic volume and smoothed over a
running average of three data points. From the derivative of the LV volume–time curves —the
filling rate curve— the following parameters were calculated using MATLAB 8.1: early peak
filling rate (E), late peak filling rate (A), E/A ratio, the contribution of the early filling to the enddiastolic volume (E-cont), and peak ejection rate.
52
High frame rate cine MRI of murine diastolic function
4.2.5. Statistics
All data are presented as mean ± standard deviation. Statistical analysis was performed with
SPSS 17.0 (SPSS Inc.) using analysis of variance for repeated measures with age (7, 13, and 19
weeks) as the within-subject factor and group (db/+ and db/db) as the between-subject factor.
In the case where the interaction between group and time was significant, the effects of group
and time were analyzed separately using one-way analysis of variance with Bonferroni post hoc
tests and Student's t-tests, respectively. The level of significance was set at α = 0.05.
4.3. Results
4.3.1. Simulations
Fig. 4.2A shows a representative distribution of the duration of the cardiac cycle L, determined
in a healthy C57BL/6 mouse from the navigators of a 2-min cardiac cine MR measurement. The
mean cardiac cycle duration measured in the whole healthy C57BL/6 mouse group was L = 115
± 4 ms. For all simulations L was set to 115 ms. The standard deviation of the cardiac cycle was
δ = 1.1 ± 0.2 ms. We used a value of δ = 1 ms in the simulations and compared this with the
case without variations in the heart rate (δ = 0 ms).
NAeff was calculated as function of NAmax for two cases: (1) considering only the number of
cardiac frames that are sampled (Fig. 4.2B), and (2) taking into account the cardiac frames as
well as sorting of k-lines in the correct phase encoding steps in k-space (Fig. 4.2C and Fig.
4.2D). Fig. 4.2B demonstrates that with respect to sampling of the different cardiac frames,
NAeff was equal to NAmax, even for high frame rates (NF = 80) where TR ≫ L/NF (situation
indicated with 2 in Fig. 4.1A). A homogeneous sampling for all cardiac frames was reached,
except for the situation when TR was an integer multiple of L and there was zero variation in the
cardiac cycle duration (TR = 5.0 ms, δ = 0 ms). For the latter, the synchrony of TR and L
prevents sampling of k-lines in certain cardiac frames, resulting in NAeff = 0. The simulations
indicate that experimentally this problem is very unlikely to occur, because even small
variations in the cardiac cycle (TR = 5.0 ms, δ = 1 ms) restore efficient sampling of all cardiac
frames with NAeff = NAmax.
Fig. 4.2C addresses the different k-space filling trajectories. This puts an additional restraint on
the sampling of cardiac frames, because not only a sufficient number of k-lines should be
acquired for every cardiac frame, but also these k-lines should be distributed equally over kspace such that all phase encoding steps are acquired at least once. Therefore, NAeff is always
lower than NAmax and it takes a number of averages for NAeff to become larger than zero. The
dashed line with unity slope in Fig. 4.2C indicates NAeff = NAmax, i.e., the hypothetical case when
all k-lines would be equally distributed to all phase encoding steps in all cardiac frames. The
closer the simulations are to this dashed line, the more efficient the sampling is. Our simulations
indicate that the k-line averaged filling scheme is more efficient compared with k-space
averaged sampling, although differences are not large and become smaller for higher frame
rates. Random sampling of k-space was not better than k-space averaged sampling (not
shown). Fig. 4.2D shows that the use of parallel imaging to reduce the number of k-lines
53
Chapter 4
needed for image reconstruction has almost no effect on the sampling efficiency, independent of
the k-space filling trajectory.
Figure 4.2. Simulations and in vivo results of the sampling efficiency of retrospective triggering cine
MRI. (A) Histogram of the duration of the cardiac cycle measured in a healthy mouse during 2 min cine
MRI. For the entire group, the mean duration of the cardiac cycle was L = 115 ± 4 ms, while the mean δ
was 1.1 ± 0.2. (B) Sampling of cardiac frames, depicting NAeff as function of NA for different
combinations of TR, NF and δ, with NP = 128. (C) Sampling of k-space, depicting NAeff as function of NA
for k-line averaged and k-space averaged trajectories and different combinations of TR and NF, with δ =
1 ms and NP = 128. (D) NAeff as function of NA for k-line averaged and k-line averaged trajectories and
different parallel imaging acceleration factors, with TR = 4.7 ms, NF = 80, and δ = 1 ms (parallel
imaging = 1.0 for NP = 192). The solid circles in (B)–(D) represent the parameter settings used for the
in vivo measurements. (E) In vivo sampling efficiency (mean of control group). Standard deviations in all
panels were small and are not shown.
54
High frame rate cine MRI of murine diastolic function
From the simulations, we conclude that small variations in the duration of the cardiac cycle lead
to sufficient asynchrony such that all cardiac frames and phase encoding steps are efficiently
sampled, even for high frame rates when TR >> L/NF. The k-line averaged sampling trajectory
resulted in the most homogeneous sampling of k-lines and cardiac frames. However, as there
was no major efficiency benefit for any of the three k-space trajectories, we chose the ‘k-space
averaging’ trajectory for in vivo measurements, because in this scheme averaging of k-lines is
distributed over the entire length of the acquisition, which decreases image artifacts from blood
flow, residual respiratory motion and heart-rate variations. Because also no significant
differences were seen between different TR values, the minimal reachable TR of 4.7 ms was
used in vivo.
The sampling efficiency in vivo was determined from the retrospectively triggered cine images
of non-diabetic db/+ mice at 7 weeks of age (Fig. 4.2E). Compared with the simulations, a
remarkably similar behavior was observed: initially, a certain number of averages is required
before NAeff > 0, after which NAeff increases linearly as a function of NAmax. The k-space
sampling efficiency in vivo is somewhat lower as compared with the simulation results (NAeff =
14.5 vs. 17.9 with NAmax = 30), which agrees with the observation that a homogeneous
sampling of all cardiac frames was not fully reached (open circles).
Figure 4.3. Comparison of ECG-triggered and retrospectively triggered high frame rate
cine MRI. To the left, a representative frame from the cine movie is shown. The white
line is the segmentation of the LV endocardium. The complete movies can be found in
the Supporting Information. To the right, the LV volume–time curves and ejection and
filling rate curves are shown. Three methods were compared: (A) ECG-triggered method
I, (B) ECG-triggered method II, and (C) retrospective triggering.
55
Chapter 4
4.3.2. Diastolic function: ECG triggering versus retrospective triggering
A representative comparison between the two prospectively ECG-triggered methods and the
retrospectively triggered high frame rate cine MRI (NF = 80) of healthy mouse heart is
presented in Fig. 4.3. Corresponding cine movies are provided in the Supporting Information
online. ECG-triggered method I resulted in periodic variations in the segmented LV volume–time
curves, likely due to a drift in heart rate or changes in the ECG quality during the course of the
four consecutive scans (Fig. 4.3A). Large oscillations were clearly visible in the filling rate curve,
in which no clear E and A filling phases could be distinguished in the diastolic phase. The
prospectively ECG-triggered method II performed better (Fig. 4.3B) resulting in a smoother LV
volume–time curve and smoother filling rate curve, in which the E and A phases could be
recognized.
The retrospectively triggered cine acquisitions (Fig. 4.3C) resulted in the highest quality LV
volume–time and filling rate curves from which E and A phases and peak rates could be
accurately determined. In the retrospectively triggered images, notably less blood-flood artifacts
were visible than for the ECG-triggered methods, particularly during the rapid E filling phases in
diastole, which facilitated segmentation of the endocardium.
Figure 4.4. LV volume-time and filling rate curves for increasing number of cardiac frames
reconstructed from the same retrospectively triggered cine MRI acquisition. The number of
frames were NF = 20 (■), 50 (), 70 (▲), and 90 (●). Panels A and B represent two different
mice. Frame duration for NF = 90 was 1.45 ms for mouse 1 and 1.17 ms for mouse 2.
56
High frame rate cine MRI of murine diastolic function
4.3.3. Diastolic function: frame fate
Fig. 4.4 shows LV volume and filling rate curves for different frame rates, reconstructed from a
single retrospectively triggered acquisition in a healthy mouse. With 20 frames/cardiac cycle,
which commonly suffices to assess systolic functional parameters, E and A diastolic phases
could not be recognized. Moreover, LV filling was underestimated at low frame rates due to a
smoothing effect of the running-average filter applied to the volume–time curves. The filter was
needed to compensate for small fluctuations in the volume–time curves, which were introduced
by inaccuracies in the semiautomatic LV segmentations. With increasing frame rate, more
details in the LV volume–time and filling rate curves became apparent, although at 50
frames/cardiac cycle, E and A were still underestimated. The curves for frame rates of 70 and
90 frames/cardiac cycle were similar, with E and A peak filling rates converging to a stable
value. For frame rates above 90, signal-to-noise became insufficient for accurate delineation of
the LV endocardium. Although we observed that for some mice the E and A phases could
already be resolved using 50 frames/cardiac cycle (Fig. 4.4B), a conservatively high frame rate
of 80 frames/cardiac cycle was used to assess the diastolic function of diabetic and non-diabetic
mice to ensure that E and A phases were accurately determined in all animals.
4.3.4. Diastolic function: diabetic and non-diabetic mice
The results on heart morphology and systolic cardiac function parameters for non-diabetic db/+
and diabetic db/db mice are presented in Table 4.1. Ejection fraction, peak ejection rate, enddiastolic volume, and end-systolic volume were similar between groups at all ages. Stroke
volume tended to be higher in db/db mice compared with db/+ mice (P=0.064), independent of
age. The LV mass index was higher in db/db mice compared with db/+ mice (P=0.002). No
significant difference in cardiac cycle duration L was observed between groups (db/+: 123 ± 5
ms and db/db: 128 ± 5 ms).
Table 4.1. Systolic LV function parameters and LV mass index at age 7, 13 and 19 weeks for non-diabetic
db/+ and diabetic db/db mice.
7 weeks
EF [%]
19 weeks
db/+
db/db
db/+
db/db
db/+
db/db
72.4±10.3
77.2±7.3
72.7±8.0
74.6±8.3
73.3±2.7
75.6±1.9
PER [μl/ms]
23.5±4.4
24.9±4.2
24.8±2.8
26.1±4.0
21.4±2.0
22.9±5.9
EDV [μl]
46.9±7.1
55.0±6.3
48.3±6.4
55.8±11.1
51.5±6.5
52.9±8.7
ESV [μl]
13.3±6.2
12.8±5.2
13.3±4.4
13.7±4.0
13.7±1.2
13.0±2.6
]#
SV [μl
LVM [mg]*
#
13 weeks
33.6±4.9
42.2±3.6
35.1±5.6
41.8±10.7
37.9±5.8
40.0±6.4
3.20±0.21
3.56±0.21
3.28±0.39
3.92±0.53
3.03±0.19
4.00±0.36
Trend for a general genotype effect (P=0.064)
*General genotype effect (P<0.05)
57
Chapter 4
Fig. 4.5 presents the parameters for diastolic function. In diabetic db/db mice, E/A tended to be
lower when compared with non-diabetic db/+ mice (P=0.091). This could be attributed to a
lower E (P=0.06) rather than a higher A (P=0.162). No significant changes for E, A, and E/A
were observed between 7 and 19 weeks of age. The contribution of early peak filling to the
diastolic filling (E-cont) was significantly lower in db/db mice compared with db/+ mice
(P=0.046), also independent of age.
Figure 4.5. LV diastolic function parameters of non-diabetic db/+
(open bars) and diabetic db/db (solid bars) mice at 7, 13, and 19 weeks
of age. #P=0.091 vs. db/+, ##P=0.06 vs. db/+, *P<0.05 vs. db/+.
4.4. Discussion
Cine MRI has been applied previously in humans (12) and mice (20) for diastolic function
measurements. However, the novelty of our protocol for measuring murine diastolic function is
that it does not rely on the most powerful gradient hardware to achieve high frame rates nor
does it compromise between spatial and temporal resolution. Instead, a high frame rate is
achieved by a retrospectively triggered acquisition, which can be reconstructed at any desired
frame rate after completion of the experiment. The highest feasible frame rate is primarily
limited by SNR and thus by the number of averages and the total acquisition time. However,
variations in the length of the cardiac cycle decrease the temporal resolution, since k-lines for a
specific cardiac frame are recorded over the entire duration of the experiment. Fig. 4.2A
therefore suggests that the best achievable temporal resolution in our experiments would be
1 ms, which corresponds to approximately 115 frames per cardiac cycle.
58
High frame rate cine MRI of murine diastolic function
The retrospectively triggered cine acquisition is based on stochastic sampling of the cardiac
frames. Simulations revealed under which conditions k-lines were distributed evenly across kspaces and cardiac frames. For the in vivo protocol, using a ‘zigzag’ linear k-space filling
trajectory (‘k-space averaging’), TR = 4.7 ms, and 80 cardiac frames, the effective number of
averages nearly reached 50% of the maximum available averages within the total acquisition
time. For diastolic function measurements it is important that the full diastolic phase is sampled.
Therefore an ECG-triggered sequence can only be started every other R-wave, also reducing its
sampling efficiency to 50%. Thus, although for ECG triggering filling of cardiac frames and kspaces is fully predictable, its efficiency for diastolic function measurements is at best equal to a
retrospectively triggered sequence. Although the current protocol would benefit from a shorter
total measurement time, in this study we did not investigate the effect of decreasing NR on the
diastolic function parameters, because this would only be of value when also considering the use
of spatio-temporal acceleration techniques (15, 22).
The high frame-rate retrospective triggering cine protocol was experimentally compared with
ECG triggering for the ability to assess diastolic function in a group of healthy mice. Our
gradient hardware did not permit a TR smaller than approximately 4 ms, without sacrificing
spatial resolution below acceptable levels. Therefore the ECG-triggered sequences were
interleaved to achieve an effective shorter frame duration. For the ECG-triggered interleaved
acquisitions, method II, in which frame delays were immediately applied for consecutive ECGtriggers, resulted in the smoothest LV volume-time and filling rate curves. Retrospective
triggering, however, outperformed ECG triggering, due to less motion and blood-flow related
artifacts in the former during the diastolic phase, which facilitated segmentation of the LV
endocardium. We believe that motion and blood-flow artifacts are suppressed in the
retrospectively triggered acquisitions because of the asynchronous nature of the acquisition
scheme, which greatly reduces coherence between the acquisition and the motion. A blackblood acquisition, e.g. by including a saturation slice above the heart (3), could completely
suppress blood flow related artifacts and improve LV segmentation. Incorporation of a
saturation slice comes at the cost of increased TR and total scan time. The sampling efficiency,
however, almost remains equal as shown by simulations comparing TR = 4.7 and TR = 8.0 ms.
Finally, the retrospectively triggered acquisition is in steady-state over the complete time-span
of the measurement, avoiding the oscillations in signal intensity that occur due to periodic
pauses in the ECG-triggered sequence when waiting for a next trigger.
In this study, retrospectively triggered data was reconstructed using a frame rate of 80
frames/cardiac cycle (approximately 700 frames/s for a cardiac cycle of 115 ms). This number
approaches the temporal resolution that can be achieved by echocardiography, which is about
700-1000 frames/s (17), although in echocardiography no averaging over time is performed.
The minimal number of frames required to resolve the E and A filling rate peaks may depend on
the heart rate and may therefore vary between individual mice (Fig. 4.4). In humans, the
diastasis period, i.e. the period between E and A peaks, contributes mostly to variations in heart
rate (21). If the same were true for mice, E and A peaks would be more easily resolved in mice
with a lower heart rate assuming an equal number of frames. On the other hand, a similar
temporal resolution might still be required for accurate estimations of the absolute E and A
filling rate values. In this study we chose a fixed number of 80 frames per cardiac cycle, which
59
Chapter 4
sufficed to resolve the E and A peaks in all mice. Alternatively, one could fix the time resolution
(e.g. 1.5 ms/frame) and adjust the number of frames to fill the length of the cardiac cycle.
Different methods to obtain cardiac or respiratory navigator MRI signals have been described.
Pencil-beam navigators are commonly used to assess respiratory motion (19), whereas cardiac
and respiratory motion may be retrieved from the peak echo intensity in a radial acquisition
(10) or from ROI correlations in co-acquired low-resolution images (13). Alternatively, (vector-)
ECG signals may be recorded to retrospectively assign k-lines to cardiac frames during image
reconstruction (14). Although these approaches offer good alternatives for the retrospective
triggering method employed in this study, the latter approach is less suitable, since continuous
excitation with short TR and fast gradient switching renders the mouse ECG unusable.
Recently, it was demonstrated that tracking of the atrioventricular junction motion in 4-chamber
cine provides an attractive alternative to assess diastolic function (2).
Stuckey et al. (20) have previously measured diastolic dysfunction in 12 week-old db/db mice
using a high frame-rate ECG-triggered sequence with a TR of 2.4 ms, reporting lower E/A ratios
as compared with non-diabetic littermates (db/db: E/A = 0.74 ± 0.11 and db/+: E/A =
1.06 ± 0.06). In contrast, measurements by Daniels et al. (5) using echocardiography revealed
no significant differences between 18 week-old diabetic and non-diabetic mice (db/db: E/A =
1.4 ± 0.2 and db/+: E/A = 1.5 ± 0.1). In this study, we observed a trend towards a lower E/A
ratio in diabetic db/db mice compared with non-diabetic db/+ mice (db/db: E/A = 1.51 ±
0.08 and db/+: E/A = 1.29 ± 0.08, P=0.091), independent of age. The E/A ratios measured in
this study are higher than the values reported by Stuckey et al. (20), but are in agreement with
echocardiographic findings (5).
The lower E/A ratio in diabetic db/db mice could be assigned to a lower E rather than a higher A
peak filling rate. In addition, the contribution of early LV relaxation to diastolic filling (E-cont)
was slightly, but significantly, lower in db/db mice compared with db/+ mice. In addition to the
subtle impairments in diastolic function, db/db mice had a 20% higher LVM, indicating LV
hypertrophy. Cardiac hypertrophy is generally associated with reduced LV compliance (11),
which could explain the reduced diastolic capacity of the db/db mice. Systolic functional
parameters, i.e. ejection fraction (EF) and peak ejection rate (PER), were not affected in the
db/db mice. Together, the observations in diabetic db/db mice point towards early stage
diabetic cardiomyopathy (7), which does not progress between 7 and 19 weeks of age.
In summary, we introduced a high frame-rate, high spatial resolution retrospectively triggered
cine MRI protocol for measuring diastolic function in mice to overcome limitations associated
with traditional high frame-rate ECG-triggered cine MRI. The optimized protocol was put to the
test in the assessment of diastolic function in non-diabetic and diabetic mice, which revealed
subtle alterations in diastolic function parameters of diabetic animals.
60
High frame rate cine MRI of murine diastolic function
4.5. Acknowledgements
This work is supported by VIDI grant (number: 700.58.421) from the Netherlands Organization
for Scientific Research (NWO), VIDI grant (number: 07952) from the Dutch Technology
Foundation STW, Applied Science Division of NWO and the Technology Program of the Ministry
of Economic Affairs, and the Institute for Imaging Science and Technology Eindhoven (IST/e).
4.6. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Bell DSH. Diabetic cardiomyopathy. Diabetes Care 26: 2949-2951, 2003.
Chung S, Breton E, and Axel L. A preliminary assessment of diastolic dysfunction with normal
ejection fraction with cine MRI of the atrioventricular junction motion. In: Proc 19th Ann Meeting ISMRM
p. 3367, 2011.
Coolen BF, Geelen T, Paulis LEM, Nauerth A, Nicolay K, and Strijkers GJ. Three-dimensional T1
mapping of the mouse heart using variable flip angle steady-state MR imaging. NMR Biomed 24: 154162, 2011.
Cosson S, and Kevorkian J. Left ventricular diastolic dysfunction: an early sign of diabetic
cardiomyopathy? Diabetes Metab 29: 455-466, 2003.
Daniels A, van Bilsen M, Janssen BJA, Brouns AE, Cleutjens JPM, Roemen THM, Schaart G,
van der Velden J, van der Vusse GJ, and van Nieuwenhoven FA. Impaired cardiac functional
reserve in type 2 diabetic db/db mice is associated with metabolic, but not structural, remodelling.
Acta Physiol 200: 11-22, 2010.
Diamant M, Lamb HJ, Groeneveld Y, Endert EL, Smit JWA, Bax JJ, Romijn JA, de Roos A, and
Radder JK. Diastolic dysfunction is associated with altered myocardial metabolism in asymptomatic
normotensive patients with well-controlled type 2 diabetes mellitus. J Am Coll Cardiol 42: 328-335,
2003.
Fang ZY, Prins JB, and Marwick TH. Diabetic cardiomyopathy: evidence, mechanisms, and
therapeutic implications. Endocr Rev 25: 543-567, 2004.
Griswold MA, Jakob PM, Heidemann RM, Nittka M, Jellus V, Wang JM, Kiefer B, and Haase A.
Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47: 1202-1210,
2002.
Heijman E, de Graaf WL, Niessen P, Nauerth A, van Eys G, de Graaf L, Nicolay K, and
Strijkers GJ. Comparison between prospective and retrospective triggering for mouse cardiac MRI.
NMR Biomed 20: 439-447, 2007.
Hiba B, Richard N, Thibault H, and Janier M. Cardiac and respiratory self-gated cine MRI in the
mouse: comparison between radial and rectilinear techniques at 7T. Magn Reson Med 58: 745-753,
2007.
Klabunde RE. Cardiac function. In: Cardiovascular Physiology Concepts. Lippincott Williams &
Wilkins, 2011, p. 59-90.
Krishnamurthy R, Pednekar A, Cheong B, and Muthupillai R. High temporal resolution SSFP Cine
MRI for estimation of left ventricular diastolic parameters. J Magn Reson Imaging 31: 872-880, 2010.
Larson AC, White RD, Laub G, McVeigh ER, Li DB, and Simonetti OP. Self-gated cardiac cine
MRI. Magn Reson Med 51: 93-102, 2004.
Lenz GW, Haacke EM, and White RD. Retrospective cardiac gating - a review of technical aspects
and future directions. Magn Reson Imaging 7: 445-455, 1989.
Makowski M, Jansen C, Webb I, Chiribiri A, Nagel E, Botnar R, Kozerke S, and Plein S. Firstpass contrast-enhanced myocardial perfusion MRI in mice on a 3T clinical MR scanner. Magn Reson
Med 64: 1592-1598, 2010.
Oh JK, Park SJ, and Nagueh SF. Established and novel clinical applications of diastolic function
assessment by echocardiography. Circ Cardiovasc Imag 4: 444-455, 2011.
Ram R, Mickelsen DM, Theodoropoulos C, and Blaxall BC. New approaches in small animal
echocardiography: imaging the sounds of silence. Am J Physiol Heart Circ Physiol 301: H1765-H1780,
2011.
Semeniuk LM, Kryski AJ, and Severson DL. Echocardiographic assessment of cardiac function in
diabetic db/db and transgenic db/db-hGLUT4 mice. Am J Physiol Heart Circ Physiol 283: H976-H982,
2002.
Stehning C, Bornert P, Nehrke K, Eggers H, and Stuber M. Free-breathing whole-heart coronary
MRA with 3D radial SSFP and self-navigated image reconstruction. Magn Reson Med 54: 476-480,
2005.
Stuckey DJ, Carr CA, Tyler DJ, Aasum E, and Clarke K. Novel MRI method to detect altered left
ventricular ejection and filling patterns in rodent models of disease. Magn Reson Med 60: 582-587,
2008.
Wang Y, Vidan E, and Bergman GW. Cardiac motion of coronary arteries: variability in the rest
period and implications for coronary MR angiography. Radiology 213: 751-758, 1999.
Wech T, Lemke A, Medway D, Stork L, Lygate C, Neubauer S, Kostler H, and Schneider JE.
Accelerating Cine-MR imaging in mouse hearts using compressed sensing. J Magn Reson Imaging 34:
1072-1079, 2011.
61
Chapter 5
In vivo quantification of mouse myocardial metabolite
content and T1 relaxation using 1H MRS
Partly based on:
In vivo proton T1 relaxation times of mouse myocardial metabolites at 9.4 T
Adrianus J. Bakermans1, Desiree Abdurrachim1, Tom R. Geraedts1, Sander M. Houten2, Klaas
Nicolay1, and Jeanine J. Prompers1
1
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven,
The Netherlands
2
Laboratory Genetic Metabolic Diseases, Department of Clinical Chemistry, and Department of Pediatrics,
Emma Children’s Hospital, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands
Published in Magnetic Resonance in Medicine 2014.
DOI: 10.1002/mrm.25340. Epub 2014 Jun 24.
Chapter 5
Abstract
Proton magnetic resonance spectroscopy (1H MRS) for quantitative in vivo assessment of mouse
myocardial metabolism requires accurate acquisition timing to minimize motion artifacts, as well
as knowledge of metabolite T1 relaxation times to correct for T1-dependent partial saturation
effects. In this study, mouse myocardial water and metabolite T1 relaxation time constants were
quantified with 1H MRS. Furthermore, the utility of the method was demonstrated by quantifying
myocardial lipid content in diabetic mice before and after treatment with the anti-diabetic drug
pioglitazone.
Cardiac-triggered and respiratory-gated PRESS-localized
1
H MRS was employed at 9.4 T to
acquire signal from a 4 µL voxel in the mouse interventricular septum in wild type, non-diabetic
db/+, and diabetic db/db mice. Steady-state of magnetization was maintained with dummy
scans during respiratory gates. Signal stability was assessed via standard deviations (SD) of
zero-order phases and amplitudes of water spectra. Saturation-recovery experiments were
performed to determine T1 values in wild type mice. Using the calculated T1s, myocardial lipid
content of non-diabetic db/+, diabetic db/db, and diabetic db/db mice after pioglitazone
treatment was quantified via corrections for partial saturation effects.
Phase SD did not vary for different repetition times (TR), and was 13.1° ± 4.5°. Maximal
amplitude SD was 14.2% ± 5.1% at TR = 500 ms. Myocardial T1 values (mean ± SD) were
quantified for water (1.71 ± 0.25 s), taurine (2.18 ± 0.62 s), trimethylamine from cholinecontaining compounds and carnitine (1.67 ± 0.25 s), creatine-methyl (1.34 ± 0.19 s),
triglyceride-methylene (0.60 ± 0.15 s), and triglyceride-methyl (0.90 ± 0.17 s) protons.
Myocardial lipid content, as determined from the triglyceride-methylene resonance, was 70%
higher in diabetic compared with non-diabetic mice, but was normalized to the levels of nondiabetic mice after pioglitazone treatment.
This work provides in vivo quantifications of proton T1 values for mouse myocardial water and
metabolites, and its application for the quantification of myocardial lipids in diabetic mice.
Keywords
1
H MRS, PRESS, saturation-recovery, metabolite T1 relaxation, myocardial metabolism, diabetic
mouse heart
64
Quantification of myocardial metabolites in vivo using 1H MRS
5.1. Introduction
Myocardial metabolism has become an important area of preclinical research, as metabolic
derangements have been implicated in the pathogenesis of many forms of heart disease,
including diabetic cardiomyopathy (17, 18). The accumulation of lipids in the diabetic heart has
been thought to play a role in the development of diabetic cardiomyopathy through the effect of
lipotoxicity (40). Myocardial lipid accumulation may induce myocardial fibrosis, which can
contribute to left ventricular stiffness and diastolic dysfunction (14, 25, 28). Indeed, data from
diabetic patients show that increased myocardial lipid content is an independent predictor of
cardiac diastolic dysfunction (32), which occurs prior to the development of systolic dysfunction
(26).
A wealth of mouse models is available to investigate the pathophysiology of diabetes and heart
disease (27). Mouse models are invaluable for investigations at genetic and molecular levels,
assessment of novel pharmaceutical interventions, as well as study designs with well-controlled
conditions. To this end, longitudinal investigations of the in vivo mouse heart have assumed a
major role in cardiac research. Proton magnetic resonance spectroscopy ( 1H MRS) is a powerful,
noninvasive technique to assess myocardial content of metabolites such as lipids (13, 31, 35,
37) and creatine (6, 29, 39) in vivo, and can be combined with MRI to quantify heart
morphology and function. Whereas MRI has become the gold standard for quantification of
cardiac function in mice (15), the application of 1H MRS for a quantitative in vivo assessment of
mouse myocardial metabolism has not been widely established yet (16, 24).
For MRS to be of quantitative value, signal acquisition needs to be accurately localized to the
region of interest, while minimizing potential contamination with signal originating from
surrounding tissue. To achieve a localized signal acquisition devoid of motion artifacts in cardiac
applications, synchronization of the MR sequence with the cardiac cycle is required. This can be
realized using the electrocardiogram (ECG) R-wave to prospectively trigger acquisitions. In
humans, respiratory effects can be eliminated by instructing the subjects to hold their breath
during acquisition (31). Alternatively, a navigator signal can be used to exclude signal
acquisitions during respiration, allowing the subjects to breathe freely (37). For anesthetized
small animal experiments, a breath-hold approach would require intubation, which is not
favorable due to its invasive nature (41). Instead, both ECG and respiratory signals, typically
monitored to assess the degree of anesthesia, can be used to orchestrate cardiac-triggered and
respiratory-gated acquisitions (7).
Prospective gating will lead to variability in the effective repetition time (TR). Moreover, to
reduce experimental time, spectra are often not acquired under fully-relaxed conditions (2, 22,
33), which induces metabolite-specific T1-dependent partial saturation effects. These saturation
effects will yield complex modulations of signal amplitudes if steady-state conditions are
violated by variation in TR due to gating, consequently hampering a quantitative evaluation of
the spectra. Hence, maintaining steady-state of magnetization when measuring at TR<5×T1
(33), as well as correcting for partial saturation effects are required for quantitative analysis of
the spectra (5). These requirements, combined with the small heart size (~100 mg) and the
high heart rate (>400 beats per minute) compared to humans (~300 g, ~80 beats per minute),
impose major challenges for cardiac application of in vivo quantitative 1H MRS in the mouse.
65
Chapter 5
Approaches to perform mouse cardiac MRI and MRS under steady-state conditions have been
developed (7) and applied (33) previously. However, to include corrections for partial saturation
effects in a quantitative analysis of tissue metabolite content, T1 relaxation time constants of
the metabolites under investigation need to be known. Here, we present quantitative in vivo
measurements of proton T1 values of mouse myocardial water and metabolites at 9.4 T.
Saturation-recovery experiments were performed with single-voxel localized
1
H MRS. A
physiologically-derived acquisition timing strategy was employed, ensuring accurate cardiac
triggering and respiratory gating, whilst maintaining steady-state of magnetization through
dummy pulses during respiration (1, 2, 33). Localized 1H MRS was then applied to determine
the modulation of myocardial lipid levels in diabetic mice after treatment with the anti-diabetic
drug pioglitazone. Piaglitazone has been shown to improve diabetes by increasing lipid synthesis
in adipose tissue (42). Therefore, we hypothesized that pioglitazone also reduces ectopic lipid
accumulation in the heart.
5.2. Methods
5.2.1. Animals
All procedures were reviewed and approved by the Animal Ethics Committee of Maastricht
University (Maastricht, The Netherlands). A group of male C57BL/6 wild type mice (n=10;
Charles River, Maastricht, The Netherlands) was used for the quantification of myocardial proton
T1 relaxation time. Myocardial lipid content was quantified in female C57BL/Ks db/+ nondiabetic mice (n=8) and C57BL/Ks db/db diabetic mice (n=9), at 7 and 13 weeks of age. After
the measurements at 7 weeks of age, a subset of diabetic mice (n=4) was treated with
pioglitazone admixed in chow (0.17 mg/g chow; Bio Services, Uden, The Netherlands) for 6
weeks, resulted in an average daily dose of 37.4 ± 3.1 mg/kg body weight. Prior to MR
measurements, the animals were anesthetized with 2-3% isoflurane in 0.4 L/min medical air,
and positioned supine in a support cradle. During the MR measurements, anesthesia was
maintained via a continuous flow of 0.4 L/min of medical air with 1-1.5% isoflurane through a
nose cone. Body temperature was maintained at 37.0 ± 0.5 °C with a warm water flow
integrated in the setup, and monitored with a rectal fiber optic probe. To allow for cardiac
triggering of MR acquisitions and monitoring of heart rate during anesthesia, conductive paste
(Ten20, Weaver and Company, Aurora, CO, USA) was applied to the front paws, which were
then fixed with medical tape onto copper-plated ECG electrodes integrated in the anesthesia
mask. A pneumatic pillow (Graseby, Watford, UK) was taped onto the abdomen to obtain a
respiratory signal for monitoring and gating. Vital signs were monitored and used to derive
cardiac triggers and respiratory gates (Fig. 5.1A) for MR acquisitions with the SA Monitoring and
Gating System (Model 1015, Small Animal Instruments, Stony Brook, NY, USA).
5.2.2. MR protocol
The setup was inserted into a horizontal-bore 9.4 T (400 MHz for
1
H) magnet (Magnex
Scientific, Oxon, UK), which was interfaced to a Bruker Avance III console (Bruker Biospin,
Ettlingen, Germany), and controlled by the ParaVision 5.0 software package (Bruker Biospin).
The system was equipped with a gradient set capable of delivering 740 mT/m at maximum
66
Quantification of myocardial metabolites in vivo using 1H MRS
amplitude and a slew rate of 6600 T/m/s. A quadrature-driven birdcage coil (Ø 35 mm, Bruker
Biospin) was used for RF transmission and signal reception.
Figure 5.1. PRESS localized 1H MRS. (a) Timing strategy for single-voxel localized 1H MRS of the in
vivo mouse heart. Electrocardiogram (ECG) and respiratory signal (Resp) are used to derive a trigger
signal (Trig) synchronizing the acquisition scheme (Acq) with the cardiac cycle during the respiratory
plateau. A trigger delay Δ shifts the water suppression module (WS) and the PRESS-localized
acquisition (P) towards the diastolic phases of two consecutive cardiac cycles. If no trigger is detected
during a <30 ms window for trigger detection (*), e.g., during respiratory gates, a dummy sequence
(D) is performed to maintain steady-state of magnetization. (b) Pulse sequence diagram for PRESS.
Crusher gradients are depicted in grey. TE, echo time; RF, radiofrequency transmission and reception;
Gx, Gy, Gz, orthogonal gradient directions. Position of the 1 × 2 × 2 mm3 voxel in the interventricular
septum, in (c) a 4-chamber long-axis view and (d) an equatorial short-axis view.
After global shimming and radiofrequency (RF) pulse calibration, a segmented, prospectively
cardiac-triggered and respiratory-gated FLASH sequence was used to collect cine MRI series of
an equatorial left-ventricular (LV) short-axis slice, an apical and a basal slice contiguous to the
equatorial slice, and 4- and 2-chamber long-axis slices (34). Imaging parameters were: field of
view = 30 × 30 mm2; matrix = 192 × 192; TE = 1.8 ms; TR = 7 ms; flip angle = 15°; number
of averages (NA) = 6; 16-18 frames per cardiac cycle after ECG R-wave upslope detection.
Using the cine MR images for anatomical reference, a 1 × 2 × 2 mm3 voxel for localized 1H MRS
was positioned in the diastolic interventricular septum (Fig. 5.1C and Fig. 5.1D) to avoid
potential contamination of the signal with pericardial lipids (33). Spectra were acquired using a
cardiac-triggered and respiratory-gated single-voxel point resolved spectroscopy (PRESS) (4)
sequence (Fig. 5.1B). To maintain a steady-state of magnetization, dummy scans were
performed during respiratory gates (7), allowing a window of <30 ms for trigger detection (2).
67
Chapter 5
As such, the effective TR could be kept nearly constant. Localized acquisition was timed at
~75% of the cardiac cycle in the diastolic phase, typically requiring a trigger delay (Δ) of 80100 ms after ECG R-wave upslope detection (Fig. 5.1A). PRESS pulse parameters were: 0.4 ms
Hermite-shaped 90° excitation pulse (bandwidth = 32.9 ppm); 0.9 ms Mao-type 180°
refocusing pulses (bandwidth = 15.5 ppm). TE was minimized to 9.1 ms to reduce signal loss
due to T2 relaxation as well as additional signal loss due to motion. Volume-selective shimming
of the region of interest was performed by positioning a 4 μL shim box in the interventricular
septum using the same PRESS sequence, while manually adjusting the shim settings to
minimize the water line width to ~35 Hz. Suppression of the dominant water signal was
achieved by preceding the localization sequence with a cardiac-triggered and respiratory-gated
chemical shift selective (CHESS) water suppression module (20) consisting of three frequencyselective 21.6 ms Hermite-shaped RF excitation pulses (bandwidth = 250 Hz), each followed by
dephasing gradients in orthogonal directions. Total CHESS module duration was 90.8 ms, which
fitted just within one cardiac cycle (Fig. 5.1A), minimizing the delay between water suppression
during one cycle and signal acquisition in the next, thus optimizing water suppression efficiency
(33).
5.2.2.1. PRESS saturation-recovery experiments in healthy mice
To determine mouse myocardial water and metabolite T1 relaxation time constants, PRESSlocalized saturation-recovery experiments were performed in the diastolic interventricular
septum. Water spectra were acquired with the CHESS module enabled but the RF pulses turned
off at TR = 500, 800, 1000, 1500, 2000, 3000, 4000, and 6000 ms; on-resonance on water.
The acquisition was repeated 32 times (number of repetition NR = 32), with NA = 1 for each
spectrum. Water-suppressed spectra were acquired in the same voxel at TR = 500, 800, 1000,
1500, 2000, 3000, 4000, and 6000 ms; on-resonance on the methyl protons of creatine
(Cr-CH3), 1.71 ppm upfield of water (2.99 ppm); number of averages = 512-128. To reach
steady-state conditions, dummy scans (2 for TR = 6000 ms to 16 for TR = 500 ms) were
applied prior to acquisition of the first spectra.
5.2.2.2. PRESS experiments in diabetic mice
In non-diabetic and diabetic mice, the water and water-suppressed spectra were acquired at TR
= 2000 ms (NR = 32, NA = 1 for the water spectra; NR = 1, NA = 256 for the watersuppressed spectra). Other parameters were kept the same as in the wild type mice.
5.2.3. Data analysis
Spectra were fitted in the time domain using AMARES in jMRUI (38). Individual water peaks
(4.70 ppm) were fitted to Lorentzian line shapes to obtain line widths, frequencies, zero-order
phases, and amplitudes. For each TR, the summed water spectrum was used as a reference for
initial phasing the corresponding water-suppressed spectrum. The prominent triglyceridemethylene (TG-CH2-) peak was set at 1.28 ppm as an internal chemical shift reference in the
water-suppressed spectra. Other peaks were assigned according to literature reports (19, 36).
To accommodate potential line broadening effects due to averaging of signals with small
frequency differences, metabolite peaks were fitted to Gaussian line shapes with equal zeroorder phase.
68
Quantification of myocardial metabolites in vivo using 1H MRS
5.2.3.1. Signal stability
In the wild type group, per mouse and per TR, the series of 32 water spectra was used to
assess signal stability via the standard deviations (SD) of signal zero-order phase and amplitude
(23, 33). Zero-order phase SD is reported in degrees (°), and amplitude SD is given as a
percentage of the mean amplitude. Because series of water spectra were acquired in multiple
mice, we report the mean ± SD of the SDs per TR.
5.2.3.2. Proton T1 relaxation time and metabolite content in healthy mice
Peak amplitudes (M) of water from the average of 32 spectra, taurine (Tau; 3.39 ppm),
trimethylamine from choline-containing compounds and carnitine (TMA; 3.21 ppm), Cr-CH3, TGCH2-, and TG-methyl (TG-CH3; 0.84 ppm) from the saturation-recovery experiments were used
as input for fitting mono-exponential functions to estimate longitudinal relaxation rates (R1),
fully-relaxed magnetization magnitudes (M0), and potential deviations (δ) from TR due to the
delay for trigger detection via M = M0 {1 - exp[-(TR + δ) R1]}. Repeatability of the saturationrecovery experiments was assessed via the repeatability coefficient of myocardial water R 1
values, defined as twice the SD of the difference in R 1 (ΔR1) between two separate
measurements in the same mice within a 1-3 week interval (n=5). To quantify the myocardial
metabolite content per mouse, metabolite M0 was divided by the corresponding water M0, thus
reporting metabolite content as a percentage of the fully-relaxed water signal.
5.2.3.3. Myocardial lipid content in diabetic mice
Myocardial lipid content was quantified from the prominent TG-CH2- peak expressed as TG-CH2M0 normalized to water M0 (corrected for partial saturation effects). TG-CH2- M0 and water M0
were calculated using the equation in section B above, with M is the acquired signal magnitude,
TR = 2000 ms, and T1 is 600 ms, as determined from the fitting of the saturation-recovery
experiments for TG-CH2- in wild type mice.
5.3. Results
5.3.1. Signal stability
Using cardiac-triggered and respiratory-gated PRESS-localized 1H-MRS acquisitions, we obtained
water spectra and water-suppressed spectra in the in vivo mouse myocardium at physiological
heart rates (541 ± 33 beats per minute). Water line width did not vary for different TR, and was
27.9 ± 6.9 Hz when averaged over all TR and all mice, indicating good magnetic field
homogeneity in the 4 µL voxels of interest. Likewise, fluctuations of the water frequency were
small (SD: 0.0082 ± 0.0062 ppm). Signal stability in terms of zero-order phase fluctuations and
amplitude variations of the water signal are displayed for each TR in Fig. 5.2. Zero-order phase
SD did not vary for different TR (Fig. 5.2A), and was 13.1° ± 4.5° on average. Amplitude SD
was 14.2% ± 5.1% at TR = 500 ms, which slightly decreased with increasing TR, resulting in an
amplitude SD of 6.6% ± 3.0% at TR = 6000 ms (P<0.05, two-sided paired t-test; Fig. 5.2B).
Combined, these data demonstrate good signal stability and an adequate maintenance of
steady-state conditions.
69
Chapter 5
Figure 5.2. Water signal stability for cardiac-triggered and respiratory-gated PRESSlocalized 1H MRS of the in vivo mouse heart, with steady-state maintenance during
respiratory gates. Mean zero-order phase fluctuations (A) and amplitude variations
(B) were determined for each TR, using 32 water spectra per mouse (n=10) per TR.
Data are presented as mean SDs, with error bars representing the SD of the SDs.
5.3.2. T1 relaxation time constants
With the capability to maintain steady-state of magnetization, we performed saturationrecovery experiments to estimate in vivo proton T1 relaxation time constants of mouse
myocardial water and metabolites at 9.4 T. A saturation-recovery series of water-suppressed
spectra acquired in an individual mouse is shown in Fig. 5.3. Nine resonance peaks could be
distinguished and are annotated in the spectrum acquired at TR = 6000 ms (Fig. 5.3). Mean
saturation-recovery curves for water, Cr-CH3, and TG-CH2- are plotted in Fig. 5.4. For water and
five metabolite peaks (Tau, TMA from choline-containing compounds and carnitine, Cr-CH3, TGCH2-, and TG-CH3), T1 relaxation time constants could be determined in five or more animals
(Table 5.1). Deviations (δ) from TR, as estimated from the offset on the time axis of the
saturation-recovery curve, were small: 23.8 ± 18.5 ms when averaged over all curves for all
metabolites in all mice, which is in agreement with the <30 ms window allowed for trigger
detection. The repeatability coefficient for myocardial water R 1 was 0.18 s-1, with the mean ΔR1
close to zero (0.0078 ± 0.089 s-1), indicating absence of systematic measurement errors.
Knowledge of water as well as metabolite T1 relaxation time constants allowed for the correction
of partial saturation effects to obtain quantitative values of mouse myocardial metabolite
content. Metabolite tissue levels, expressed as a percentage of the water signal, are collected in
Table 5.1.
70
Quantification of myocardial metabolites in vivo using 1H MRS
Figure 5.3. Saturation-recovery series of water-suppressed spectra acquired in the in vivo mouse
heart with PRESS-localized 1H MRS. The prominent triglyceride-methylene (TG-CH2-) peak was used as
a chemical shift reference at 1.28 ppm. Tau, taurine (3.39 ppm); TMA, trimethylamine (3.21 ppm); CrCH3, creatine-methyl (2.99 ppm); TG-CH3, triglyceride-methyl (0.84 ppm). Annotations of numbered
peaks: 1, TG-CH=CH-CH2-CH=CH- (2.72 ppm); 2, TG-CαH2COO (2.21 ppm); 3, TG-CH2-CH=CH-CH2(1.99 ppm); 4, TG-CβH2CH2COO (1.57 ppm).
Figure 5.4. Mean saturation-recovery curves for in vivo mouse (A) myocardial water (n=10), (B)
creatine-methyl (Cr-CH3; n=5), and (C) triglyceride-methylene (TG-CH2-; n=8) protons obtained with
PRESS-localized acquisitions at 9.4 T. Per mouse, the acquired water, Cr-CH3, and TG-CH2- signal
amplitudes at each repetition time were normalized to the corresponding M0 of water, Cr-CH3 and TGCH2-, respectively. For comparison, signal amplitudes are expressed as a percentage of water M 0. Error
bars represent SD.
71
Chapter 5
Table 5.1. In vivo water and metabolite T1 relaxation time constants and tissue metabolite content for
mouse myocardium determined with PRESS-localized saturation-recovery 1H MRS experiments at 9.4 T.
Chemical shift
T1 relaxation time constant
Metabolite contenta
(ppm)
(s)
(% of water)
Water (n=10)
4.70
1.71
±
0.25
Tau (n=8)
3.39
2.18
±
0.62
0.099
±
0.026
TMA (n=8)
3.21
1.67
±
0.25
0.108
±
0.040
Cr-CH3 (n=5)
2.99
1.34
±
0.19
0.056
±
0.037
TG-CH2- (n=8)
1.28
0.60
±
0.15
0.477
±
0.114
TG-CH3 (n=6)
0.84
0.90
±
0.17
0.098
±
0.038
a
Corrected for partial saturation effects. Data are expressed as mean ± SD. Tau, taurine; TMA,
trimethylamine from choline-containing compounds and carnitine; Cr-CH3, creatine-methyl; TG-CH2-,
triglyceride-methylene; TG-CH3, triglyceride-methyl. TMA, trimethylamine (3.21 ppm); Cr-CH3, creatinemethyl (2.99 ppm); TG-CH3, triglyceride-methyl (0.84 ppm). Annotations of numbered peaks: 1, TGCH=CH-CH2-CH=CH- (2.72 ppm); 2, TG-CαH2COO (2.21 ppm); 3, TG-CH2-CH=CH-CH2- (1.99 ppm); 4, TGCβH2CH2COO (1.57 ppm).
5.3.3. Myocardial lipid content in diabetic mice
As a realistic test of the usefulness of the developed technique, we applied the technique to
measure myocardial lipid content in diabetic mice at baseline, and after 6 weeks of pioglitazone
treatment. At 7 weeks, myocardial lipid (TG-CH2-) levels were significantly higher in diabetic
db/db mice compared with non-diabetic db/+ mice (Fig. 5.5). This was also the case at 13
weeks for untreated db/db mice, while pioglitazone treatment normalized myocardial lipid levels
to the levels of db/+ mice (Fig. 5.5).
Figure 5.5. Myocardial TG-CH2- levels as a percentage of water
signal, corrected for partial saturation effects (white: db/+ (n=8),
black: non-treated db/db (n=5), hatched: db/db (n=4) before
pioglitazone treatment (7 weeks of age) and after pioglitazone
treatment (13 weeks of age), *P<0.05, **P<0.01, ***P < 0.001).
72
Quantification of myocardial metabolites in vivo using 1H MRS
5.4. Discussion and Conclusions
Besides an accurate localization of signal acquisition, a steady-state of magnetization is required
for 1H MRS to be a quantitative tool for the in vivo assessment of myocardial metabolism (33).
In addition, a correction for partial saturation effects when measuring at TR<5×T1 requires
knowledge of T1 relaxation time constants (5). With the capability to maintain steady-state of
magnetization (1, 2), we performed quantitative in vivo assessments of proton T1 values for
myocardial metabolites in the mouse heart. Prior assessments of metabolite T1 values at 9.4 T
were limited to the rat brain (10, 11). Notably, the current results for metabolite T1 values in
the mouse myocardium are consistent with those obtained in rat brain, with a relatively long T 1
relaxation time constant for Tau (this work: 2.2 ± 0.6 s vs. brain: 2.3 ± 0.4 s (11) and 2.6 ±
0.2 s (10)), and short T1 values for lipid-associated peaks (TG-CH2- in this work: 0.60 ± 0.15 s
vs. brain: 0.58 ± 0.02 s (11) and 0.51 ± 0.07 s (10)). The T1 value for mouse myocardial water
reported here (1.71 ± 0.25 s) is in close agreement with a previous study that used MRI to map
T1 values (1.76 ± 0.17 s) of mouse myocardial water at 9.4 T (8). Furthermore, the
repeatability coefficient of water R1 values achieved in the current study (0.18 s-1) is similar to
those reported for MRI T1 relaxometry (0.14 - 0.21 s-1 (8)).
In previous reports, in vivo assessment of myocardial metabolite content using 1H MRS under
steady-state conditions was shown to correlate strongly with standard biochemical assays on ex
vivo tissue, both for Cr-CH3 (30) and triglycerides (2, 3). In contrast, not adhering to steadystate conditions if TR<5×T1 can be a source of discrepancy between
1
H MRS-based
quantifications and biochemical assays (22). Our data indicate that at 9.4 T, a long TR (~10 s)
is needed to assess the levels of water, Tau, TMA and Cr-CH3 under fully relaxed conditions,
whereas fully relaxed signals of myocardial TG resonances can be obtained at considerably
shorter TR (~5 s). Measuring at fully relaxed conditions may therefore be an alternative
approach to quantitatively assess myocardial levels of metabolites with short T 1, e.g., lipids,
within an acceptable experimental time without implementing a physiologically-derived timing
strategy for dummy scans as used here and elsewhere (33).
For absolute quantification of metabolite concentrations using 1H MRS, knowledge of the number
of protons in the compound of interest is required. As the composition of TG can vary, e.g., due
to dietary intake or pathophysiology, absolute quantification of in vivo lipid concentrations is not
straightforward. Moreover, tissue water content must be determined for the water signal to be
used as a quantification reference (35). In addition, to correct for signal loss due to T2
relaxation during TE, metabolite T2 relaxation time constants need to be known. In stationary
tissue, metabolite T2 values are typically assessed by acquiring series of spectra at multiple TEs.
In the mouse myocardium, this approach would lead to an underestimation of T 2 values, due to
additional cardiac motion-induced signal loss at longer TEs. Therefore, assessment of
myocardial metabolite T2 values requires an alternative approach, in which T2 weighting is
performed prior to signal localization. Recently, a nonselective T 2 preparation module (MLEV)
was introduced to map mouse myocardial water T2 values with MRI (9). Incorporating this MLEV
preparation in combination with a water suppression module to precede
1
H MRS signal
localization could be the next step towards absolute quantification of mouse myocardial
metabolite concentrations.
73
Chapter 5
Using cardiac-triggered and respiratory-gated PRESS-localized 1H MRS, we demonstrated the
quantification of myocardial lipid content in diabetic mice, which was corrected for partial
saturation using the T1 relaxation time constants. Myocardial lipid content was higher in diabetic
compared with non-diabetic mice, but was normalized after 6 weeks of pioglitazone treatment.
Pioglitazone is an anti-diabetic drug, which action involves redistribution of ectopic lipid from
non-adipose tissues into adipose tissue. The mechanism has been thought to involve adipocyte
differentiation to increase the capacity of adipose tissue to store lipids (12, 21, 42). This results
in lower plasma lipid levels, and hence, may lead to reduced myocardial lipid uptake and
accumulation (12, 21).
In conclusion, this work reports a quantitative in vivo assessment of mouse myocardial proton
T1 relaxation time constants for water, Tau, TMA, Cr-H3, TG-CH2-, and TG-CH3, and its
application for the quantification of myocardial lipids in diabetic mice. Using saturation-recovery
experiments with single-voxel localized 1H MRS, we obtained results that are in good agreement
with published T1 values for rat brain at 9.4 T. With the anticipated increase of interest in in vivo
cardiac phenotyping of mouse models of metabolic derangements (24), the data presented here
will be of value for future quantitative 1H MRS studies and provide a basis towards absolute
quantification of metabolite concentrations in the in vivo mouse heart.
5.5. Acknowledgement
This work was supported by VIDI grants (project number 700.58.421 and 016.086.336 to J.J.P.
and S.M.H., respectively) from the Netherlands Organisation for Scientific Research (NWO).
5.6. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
74
Bakermans AJ, Abdurrachim D, Nicolay K, and Prompers JJ. 1H-MRS in the in vivo mouse heart:
a comparison between PRESS and STEAM. In: Proc 21st Ann Meeting ISMRM p. 2011, 2013.
Bakermans AJ, Geraedts TR, van Weeghel M, Denis S, Joao Ferraz M, Aerts JM, Aten J,
Nicolay K, Houten SM, and Prompers JJ. Fasting-induced myocardial lipid accumulation in longchain acyl-CoA dehydrogenase knockout mice is accompanied by impaired left ventricular function. Circ
Cardiovasc Imaging 4: 558-565, 2011.
Bakermans AJ, van Weeghel M, Denis S, Nicolay K, Prompers JJ, and Houten SM. Carnitine
supplementation attenuates myocardial lipid accumulation in long-chain acyl-CoA dehydrogenase
knockout mice. J Inherit Metab Dis 36: 973-981, 2013.
Bottomley PA. Spatial localization in NMR spectroscopy in vivo. Ann N Y Acad Sci 508: 333-348,
1987.
Bottomley PA. The trouble with spectroscopy papers. Radiology 181: 344-350, 1991.
Bottomley PA, and Weiss RG. Non-invasive magnetic-resonance detection of creatine depletion in
non-viable infarcted myocardium. Lancet 351: 714-718, 1998.
Cassidy PJ, Schneider JE, Grieve SM, Lygate C, Neubauer S, and Clarke K. Assessment of
motion gating strategies for mouse magnetic resonance at high magnetic fields. J Magn Reson Imaging
19: 229-237, 2004.
Coolen BF, Geelen T, Paulis LE, Nauerth A, Nicolay K, and Strijkers GJ. Three-dimensional T1
mapping of the mouse heart using variable flip angle steady-state MR imaging. NMR Biomed 24: 154162, 2011.
Coolen BF, Simonis FF, Geelen T, Moonen RP, Arslan F, Paulis LE, Nicolay K, and Strijkers GJ.
Quantitative T1 mapping of the mouse heart by segmented MLEV phase-cycled T preparation. Magn
Reson Med 2013.
Cudalbu C, Mlynarik V, Xin L, and Gruetter R. Comparison of T1 relaxation times of the
neurochemical profile in rat brain at 9.4 tesla and 14.1 tesla. Magn Reson Med 62: 862-867, 2009.
de Graaf RA, Brown PB, McIntyre S, Nixon TW, Behar KL, and Rothman DL. High magnetic field
water and metabolite proton T1 and T2 relaxation in rat brain in vivo. Magn Reson Med 56: 386-394,
2006.
de Souza CJ, Eckhardt M, Gagen K, Dong M, Chen W, Laurent D, and Burkey BF. Effects of
pioglitazone on adipose tissue remodeling within the setting of obesity and insulin resistance. Diabetes
50: 1863-1871, 2001.
Quantification of myocardial metabolites in vivo using 1H MRS
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
den Hollander JA, Evanochko WT, and Pohost GM. Observation of cardiac lipids in humans by
localized 1H magnetic resonance spectroscopic imaging. Magn Reson Med 32: 175-180, 1994.
Ellims AH, Shaw JA, Stub D, Iles LM, Hare JL, Slavin GS, Kaye DM, and Taylor AJ. Diffuse
myocardial fibrosis evaluated by post-contrast t1 mapping correlates with left ventricular stiffness. J
Am Coll Cardiol 63: 1112-1118, 2014.
Epstein FH. MR in mouse models of cardiac disease. NMR Biomed 20: 238-255, 2007.
Faller KM, Lygate CA, Neubauer S, and Schneider JE. 1H-MR spectroscopy for analysis of cardiac
lipid and creatine metabolism. Heart Fail Rev 2012.
Fang ZY, Prins JB, and Marwick TH. Diabetic cardiomyopathy: Evidence, mechanisms, and
therapeutic implications. Endocr Rev 25: 543-567, 2004.
Fillmore N, Mori J, and Lopaschuk GD. Mitochondrial Fatty Acid Oxidation Alterations in Heart
Failure, Ischemic Heart Disease, and Diabetic Cardiomyopathy. Br J Pharmacol 2013.
Griffin JL, Williams HJ, Sang E, and Nicholson JK. Abnormal lipid profile of dystrophic cardiac
tissue as demonstrated by one- and two-dimensional magic-angle spinning 1H NMR spectroscopy.
Magn Reson Med 46: 249-255, 2001.
Haase A, Frahm J, Hanicke W, and Matthaei D. 1H NMR chemical shift selective (CHESS) imaging.
Phys Med Biol 30: 341-344, 1985.
Hallakou S, Doare L, Foufelle F, Kergoat M, Guerre-Millo M, Berthault MF, Dugail I, Morin J,
Auwerx J, and Ferre P. Pioglitazone induces in vivo adipocyte differentiation in the obese Zucker
fa/fa rat. Diabetes 46: 1393-1399, 1997.
Hankiewicz JH, Banke NH, Farjah M, and Lewandowski ED. Early impairment of transmural
principal strains in the left ventricular wall after short-term, high-fat feeding of mice predisposed to
cardiac steatosis. Circ Cardiovasc Imaging 3: 710-717, 2010.
Katz-Brull R, and Lenkinski RE. Frame-by-frame PRESS 1H-MRS of the brain at 3 T: the effects of
physiological motion. Magn Reson Med 51: 184-187, 2004.
Kober. Recent developments in small animal cardiovascular MRI. Curr Cardiovasc Imaging Rep 2014.
Li CJ, Lv L, Li H, and Yu DM. Cardiac fibrosis and dysfunction in experimental diabetic
cardiomyopathy are ameliorated by alpha-lipoic acid. Cardiovasc Diabetol 11: 73, 2012.
McGavock JM, Lingvay I, Zib I, Tillery T, Salas N, Unger R, Levine BD, Raskin P, Victor RG,
and Szczepaniak LS. Cardiac steatosis in diabetes mellitus: a 1H-magnetic resonance spectroscopy
study. Circulation 116: 1170-1175, 2007.
Molkentin JD, and Robbins J. With great power comes great responsibility: using mouse genetics to
study cardiac hypertrophy and failure. J Mol Cell Cardiol 46: 130-136, 2009.
Mori J, Patel VB, Abo Alrob O, Basu R, Altamimi T, Desaulniers J, Wagg CS, Kassiri Z,
Lopaschuk GD, and Oudit GY. Angiotensin 1-7 ameliorates diabetic cardiomyopathy and diastolic
dysfunction in db/db mice by reducing lipotoxicity and inflammation. Circ Heart Fail 7: 327-339, 2014.
Nakae I, Mitsunami K, Omura T, Yabe T, Tsutamoto T, Matsuo S, Takahashi M, Morikawa S,
Inubushi T, Nakamura Y, Kinoshita M, and Horie M. Proton magnetic resonance spectroscopy can
detect creatine depletion associated with the progression of heart failure in cardiomyopathy. J Am Coll
Cardiol 42: 1587-1593, 2003.
Phillips D, Ten Hove M, Schneider JE, Wu CO, Sebag-Montefiore L, Aponte AM, Lygate CA,
Wallis J, Clarke K, Watkins H, Balaban RS, and Neubauer S. Mice over-expressing the myocardial
creatine transporter develop progressive heart failure and show decreased glycolytic capacity. J Mol
Cell Cardiol 48: 582-590, 2010.
Rial B, Robson MD, Neubauer S, and Schneider JE. Rapid quantification of myocardial lipid content
in humans using single breath-hold 1H MRS at 3 Tesla. Magn Reson Med 66: 619-624, 2011.
Rijzewijk LJ, van der Meer RW, Smit JW, Diamant M, Bax JJ, Hammer S, Romijn JA, de Roos
A, and Lamb HJ. Myocardial steatosis is an independent predictor of diastolic dysfunction in type 2
diabetes mellitus. J Am Coll Cardiol 52: 1793-1799, 2008.
Schneider JE, Tyler DJ, ten Hove M, Sang AE, Cassidy PJ, Fischer A, Wallis J, SebagMontefiore LM, Watkins H, Isbrandt D, Clarke K, and Neubauer S. In vivo cardiac 1H-MRS in the
mouse. Magn Reson Med 52: 1029-1035, 2004.
Schneider JE, Wiesmann F, Lygate CA, and Neubauer S. How to perform an accurate assessment
of cardiac function in mice using high-resolution magnetic resonance imaging. J Cardiovasc Magn
Reson 8: 693-701, 2006.
Szczepaniak LS, Dobbins RL, Metzger GJ, Sartoni-D'Ambrosia G, Arbique D, Vongpatanasin
W, Unger R, and Victor RG. Myocardial triglycerides and systolic function in humans: in vivo
evaluation by localized proton spectroscopy and cardiac imaging. Magn Reson Med 49: 417-423, 2003.
Ugurbil K, Petein M, Maidan R, Michurski S, Cohn JN, and From AH. High resolution proton NMR
studies of perfused rat hearts. FEBS Lett 167: 73-78, 1984.
van der Meer RW, Doornbos J, Kozerke S, Schar M, Bax JJ, Hammer S, Smit JW, Romijn JA,
Diamant M, Rijzewijk LJ, de Roos A, and Lamb HJ. Metabolic imaging of myocardial triglyceride
content: reproducibility of 1H MR spectroscopy with respiratory navigator gating in volunteers.
Radiology 245: 251-257, 2007.
Vanhamme L, van den Boogaart A, and Van Huffel S. Improved method for accurate and efficient
quantification of MRS data with use of prior knowledge. J Magn Reson 129: 35-43, 1997.
Weiss K, Martini N, Boesiger P, and Kozerke S. Metabolic MR imaging of regional triglyceride and
creatine content in the human heart. Magn Reson Med 68: 1696-1704, 2012.
Wende AR, and Abel ED. Lipotoxicity in the heart. Biochim Biophys Acta 1801: 311-319, 2010.
Whalen M, and Shapiro JI. Controlled ventilation during NMR spectroscopic studies: hemodynamic
and biochemical consequences. Magn Reson Imaging 9: 229-234, 1991.
75
Chapter 5
42.
76
Yang KJ, Noh JR, Kim YH, Gang GT, Hwang JH, Yang SJ, Yeom YI, and Lee CH. Differential
modulatory effects of rosiglitazone and pioglitazone on white adipose tissue in db/db mice. Life Sci 87:
405-410, 2010.
Chapter 6
In vivo
31
P MR spectroscopy of the mouse heart using
respiratory-gated and cardiac-triggered 3D ISIS
Adrianus J. Bakermans1,2*, Desiree Abdurrachim1*, Bastiaan J. van Nierop1, Inge van der
Kroon1, Gustav J. Strijkers1,3, Sander M. Houten4, Klaas Nicolay1, and Jeanine J. Prompers1
1
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven,
The Netherlands
2
Department of Radiology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands
3
Biomedical Engineering and Physics, Academic Medical Center, University of Amsterdam, Amsterdam, The
Netherlands
4
Laboratory Genetic Metabolic Diseases, Department of Clinical Chemistry, and Department of Pediatrics,
Emma Children’s Hospital, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands
*both authors contributed equally
Submitted
Chapter 6
Abstract
Phosphorous-31 magnetic resonance spectroscopy (31P MRS) provides a unique non-invasive
window on the myocardial energy status. Mouse models of human cardiac disease are widely
used in preclinical studies, but application of
31
P MRS in the in vivo mouse heart has been
limited. The small-sized, fast-beating mouse heart imposes challenges regarding localized signal
acquisition devoid of contamination with signal originating from surrounding tissues. Here, we
report the implementation of 3D Image Selected In vivo Spectroscopy (ISIS) for single-voxel
localized
31
P MRS of the in vivo mouse heart at 9.4 T.
A respiratory-gated, cardiac-triggered 3D ISIS sequence was used, whilst maintaining a steady
state of magnetization with dummy excitations during respiratory gates to ensure a constant
repetition time. Cardiac
31
P MR spectra were acquired in vivo in the left ventricle of healthy mice
and of transverse aortic constricted (TAC) mice. Localization and potential signal contamination
were assessed with
31
P MRS experiments in the anterior myocardial wall, liver, skeletal muscle
and blood.
The myocardial energy status, assessed via the phosphocreatine (PCr) to ATP ratio, was ~25%
lower in TAC mice compared to controls (0.76 ± 0.13 vs. 1.00 ± 0.15; P<0.01). Contamination
of the spectra with signal from ATP in the blood was marginal (~4%). Localization with 1D ISIS
resulted in two-fold higher PCr/ATP ratio than measured with 3D ISIS, which could be attributed
to high PCr levels of chest skeletal muscle that contaminate the signal measured with 1D ISIS.
Absence of the PCr resonance in
31
P MR spectra acquired in the liver confirmed accurate
localization with 3D ISIS.
Respiratory-gated and cardiac-triggered 3D ISIS allows for the non-invasive detection of a
disturbed myocardial energy homeostasis with
31
P MRS in the in vivo mouse, and can be
combined with MRI measurements of cardiac function in the same experimental session.
Keywords
energy metabolism, heart failure, ISIS, mouse,
78
31
P MRS, transverse aortic constriction (TAC)
Mouse cardiac
31
P MR spectroscopy
6.1. Introduction
Mouse models are widely used in preclinical studies on the pathogenesis of cardiomyopathies.
Disturbed myocardial energy metabolism has been identified as an important contributor to the
development of cardiomyopathy (33). Assessment of the myocardial energy status is therefore
instrumental in characterizing disease progression or treatment response. The high-energy
phosphates adenosine 5’-triphosphate (ATP) and phosphocreatine (PCr) are essential for
providing energy for cellular processes such as sarcomere contraction in cardiomyocytes, and
for energy transport and buffering. The inherent instability of high-energy phosphates
compromises accurate assessment of the myocardial energy status with traditional biochemical
techniques, which require disruptive or terminal biopsies, precluding longitudinal in vivo
investigations. Many magnetic resonance imaging (MRI) and, to a lesser extent, spectroscopy
(MRS) methods have been introduced to study the in vivo mouse heart noninvasively (9, 11,
22). Application of these methods allows for longitudinal studies of disease progression and an
assessment of the effects of therapeutic strategies.
Phosphorous-31 MRS (31P MRS) is the only method that provides a non-invasive window on in
vivo high-energy phosphates (33). Localized signal acquisition is essential to restrict the
obtained spectrum to the heart, excluding signal from nearby liver tissue or chest skeletal
muscle. Localization methods for
(CSI) approaches.
31
31
P MRS include single-voxel as well as chemical shift imaging
P CSI allows for an assessment of regional differences in myocardial energy
status (46), but is susceptible to intervoxel signal contamination due to Fourier bleeding (20).
In contrast, single-voxel localization with 3D Image Selected In vivo Spectroscopy (ISIS) (36)
leads to a better defined voxel shape (10), but voxel size is typically much larger compared to
CSI, and commonly includes the whole left ventricle (LV) (35). Nowadays, both CSI and ISIS
approaches are applied for human cardiac
Localized
31
31
P MRS (5).
P MRS of the in vivo mouse heart is very challenging due the small organ size (~100
mg), the high heart rate (> 400 beats/min), and the intrinsically low sensitivity of
1
6.65% of the proton ( H) sensitivity). Methods for in vivo 1D and 2D
31
31
P MRS (only
P CSI of the mouse heart
were initially demonstrated in healthy mice (8) and in a transgenic mouse model for
cardiomyopathy (12). These experiments were performed at a constant TR, which is essential
for a reliable signal quantification. Remarkably, none of these methods used cardiac triggering
or respiratory gating to account for physiological motion of the tissue of interest. Together with
the effects of Fourier bleeding in CSI methods, tissue displacement could lead to contamination
of the spectrum with signal from the liver, blood, and/or chest skeletal muscle.
One early study describes the application of cardiac-triggered 3D ISIS for in vivo
31
P MRS of the
mouse heart (35), which until now has not been pursued by other research groups. While
cardiac triggering was used in those experiments to synchronize the acquisitions with the
cardiac cycle, no measures were taken to ensure a constant TR (35). A constant TR is important
for accurate localization when applying ISIS under partially saturated conditions (i.e., if
TR<5×T1), as variations in TR can lead to signal contamination (21, 28) as well as modulations
of signal amplitudes due to T1-dependent partial saturation effects. These initial experiments
were performed at 2.35 T and required a lengthy experimental time of almost 3 hours. At higher
79
Chapter 6
field strengths, scan time can possibly be reduced to acceptable values while maintaining
sufficient SNR for signal quantification.
Here, we report the implementation of 3D ISIS for single-voxel localized
31
P MRS of the in vivo
mouse heart at 9.4 T. Because 3D ISIS is a multi-shot localization method and hence
particularly sensitive to motion artifacts, we employed both respiratory gating and cardiac
triggering whilst maintaining a steady state of magnetization with dummy excitations during
respiratory gates to ensure a constant TR (2). To demonstrate suitability for cardiac
applications, the method was applied to a well-characterized mouse model of heart failure (31,
37). Cardiac functional parameters were quantified using MR images obtained during the same
experimental session, which could typically be completed well within 2 hours.
6.2. Methods
6.2.1. Animals
Male C57BL/6 mice (n=8; body weight = 26.2 ± 2.6 g) underwent transverse aortic constriction
(TAC) surgery as described previously (44). Anesthesia was induced using 2-3% isoflurane in
0.4 L/min flow of 1:1 O2:medical air, after which mice were intubated for mechanical
ventilation. Analgesia was provided using buprenorphine (0.1 mg/kg subcutaneous). An incision
was made above the first intercostal space to gain access to the aortic arch. The aorta was tied
off together with a 27 G needle between the innominate artery and the left common carotid
artery with a 6-0 silk suture. Subsequent needle removal left an aortic stenosis, inducing LV
pressure-overload. The chest was closed and the intubation tube was removed to allow full
recovery. Seven weeks after surgery, MR data were acquired as described below. Healthy mice
(n=9; body weight = 24.4 ± 2.0 g) served as controls. Following the MR measurements,
anesthetized mice were sacrificed by exsanguination. Blood was collected in EDTA tubes for ex
vivo analysis with
31
P MRS. All procedures were approved by the Animal Ethics Committee of
Maastricht University (Maastricht, The Netherlands).
6.2.2. MR protocol
Mice were anesthetized with 2-3% isoflurane in 0.4 L/min flow of medical air and positioned
prone in a purpose-built support cradle above a custom-built, actively decoupled, two-turn
31
P
surface coil (Ø 15 mm) for signal reception. Anesthesia was maintained with 1.2-1.6%
isoflurane in a continuous flow of 0.4 L/min medical air. The front paws were taped onto goldcoated ECG electrodes integrated in the anesthesia mask. A respiratory balloon was positioned
underneath the lower abdomen. Vital signs were monitored and used for MR gating and
triggering by the SA Monitoring and Gating System 1025 (Small Animal Instruments, Stony
Brook, NY, USA). Mouse body temperature was maintained using a heating pad with integrated
warm water flow, and monitored with an external abdominal fiber optic probe. The setup was
inserted into a horizontal-bore 9.4 T magnet (Magnex Scientific, Oxon, UK), interfaced to a
Bruker Avance III console (Bruker Biospin MRI, Ettlingen, Germany), and controlled by the
ParaVision 5.0 software package (Bruker Biospin). The system was equipped with a 740 mT/m
gradient set, and a volume coil (Ø 54 mm) composed of a quadrature 1H birdcage resonator and
80
Mouse cardiac
a linear
31
31
P MR spectroscopy
P birdcage resonator (RAPID Biomedical, Rimpar, Germany), used for 1H MRI and
shimming, and for radiofrequency transmission for
31
P MRS, respectively.
Scout 1H MR images were acquired to confirm positioning of the heart within the sensitive area
of the
31
P surface coil. A segmented, prospectively cardiac-triggered, respiratory-gated fast low-
angle shot sequence was used to acquire cine 1H MR image series of 16-18 frames per cardiac
cycle. Four 1-mm LV short-axis slices were complemented with 4- and 2-chamber long-axis
views, and used for quantification of LV function and morphology as well as for anatomical
reference during 3D ISIS voxel planning for localized
31
P MRS. Imaging parameters were: field
2
of view 30 × 30 mm ; matrix = 128 × 128; TE = 1.8 ms; TR = 7 ms; flip angle = 15°; number
of averages (NA) = 4. Total acquisition time was ≈ 20 minutes.
Subsequently, an 11 × 11 × 11 mm3 cubic voxel in the sensitive area of the surface coil was
shimmed manually by minimizing the
1
H2O line width, acquired with a respiratory-gated,
cardiac-triggered point resolved spectroscopy sequence (4). Calibration of the
31
P sinc excitation
pulse (pulse length = 1.2 ms; bandwidth = 32.0 ppm) was performed by varying pulse power to
achieve maximal signal from a spherical phantom (Ø 5 mm; 15 M phosphoric acid) positioned
underneath the
31
P surface coil. After removal of the phantom, unlocalized pulse-acquire
31
P MR
spectra were obtained from a subset of animals (n=5 per group) to assess metabolite T1 values
using conventional saturation-recovery experiments. Parameters were: 1.2 ms 90° sinc
excitation pulse; bandwidth = 32.0 ppm; γ-ATP on resonance; TR = 500, 1000, 2000, 4000,
6000, and 15000 ms; NA = 1024 - 32.
Next, a respiratory-gated, cardiac-triggered 3D ISIS sequence was used for localized cardiac
31
P
MRS. Dependent on heart size, a voxel (control: 174 ± 9 μL, TAC: 326 ± 43 μL) was positioned
to enclose the end-diastolic LV, carefully excluding the liver and chest skeletal muscle (Fig. 6.1A
and Fig. 6.1B). 3D ISIS parameters were: TR ≈ 2 seconds; NA = 768 (96 3D ISIS cycles)
preceded by 1 dummy cycle; 6.25 ms 180° adiabatic hyperbolic secant inversion pulses
(bandwidth = 37.5 ppm); 1.2 ms 90° sinc excitation pulse (bandwidth = 32.0 ppm); γ-ATP on
resonance; 2048 acquisition points. Triggering was timed at ECG R-wave upslope detection.
Respiratory gating causes fluctuations in effective TR, leading to variations in longitudinal
magnetization between subsequent acquisitions. If longitudinal magnetization is not equal at the
start of all eight acquisitions within one 3D ISIS cycle, cancellation of unwanted signals in the
addition/subtraction scheme is incomplete. Thus, when measuring at TR<5×T 1, a constant TR is
required to minimize signal contamination. Moreover, measuring at constant TR prevents
complex modulations of signal amplitudes that hamper corrections for partial saturation effects.
Therefore, we performed unlocalized dummy excitation pulses during respiratory gates to
achieve an essentially constant TR of ≈ 2 seconds. Acquisition time was < 40 minutes. In a
subset of healthy mice (n=6), 3D ISIS was also performed at TR ≈ 15 seconds; NA = 192 (24
cycles). These measurements were used to verify the partial saturation correction factor
obtained with unlocalized saturation-recovery experiments.
To compare 3D ISIS localization with 1D localization, we acquired spectra using 1D ISIS of the
anterior myocardial wall in a slice essentially parallel to the surface coil (Fig. 6.2A) in a subset of
mice (n=6). Acquisition parameters were kept similar to the 3D ISIS approach, except for NA =
384 (192 1D ISIS cycles) and slice thickness = 1 mm. Furthermore, to obtain spectra solely
81
Chapter 6
from the myocardium and to rule out any contamination with signal from LV cavity blood, 3D
ISIS was performed with a small voxel (1 × 3 × 3 mm 3) positioned within the anterior
myocardial wall (Fig. 6.2B). Acquisition time was ≈ 2.5 hours with NA = 3072 (384 cycles).
Additionally, spectra from liver and hind limb skeletal muscle (n=3) were acquired in vivo with
respiratory-gated and cardiac-triggered 3D ISIS (Fig. 6.3) to evaluate the performance of 3D
ISIS localization with respect to in vivo
31
P MRS of these tissues as reported in the literature
(19, 27).
Finally, to investigate the contribution of blood metabolites to the cardiac
31
P MR spectra
acquired in vivo, spectra of fresh blood were measured ex vivo using a pulse-acquire sequence.
A vial (n=6) with ~1 mL of blood in EDTA was positioned just over the surface coil and
maintained at 37°C by a heating pad. Parameters were: 1.2 ms 90° sinc excitation pulse; γ-ATP
on resonance; TR = 2000 ms; NA = 512.
6.2.3. Image analysis
LV cavity and myocardial wall volumes were quantified by semiautomatic segmentation of the
cine images (PIE Medical Imaging, Maastricht, The Netherlands) as described previously (17),
yielding LV end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV), ejection
fraction (EF), and LV myocardial mass.
6.2.4.
31
P MRS data analysis
Fitting of the metabolite signals to Lorentzian line shapes was performed in the time domain
using AMARES in jMRUI (45). The PCr resonance at 0.00 ppm was used as an internal chemical
shift reference. The ATP resonances at -2.48 ppm (γ; doublet), -7.52 ppm (α; doublet) and
-16.26 (β; triplet) were fitted with equal amplitudes and line widths within each multiplet, and a
J-coupling constant of 17 Hz. The γ-ATP line widths (LWγ-ATP) were constrained relative to the
PCr line width (LWPCr) according to an empirically determined relation: LWγ-ATP = LWPCr + 14.85
Hz (n=63; r = 0.78; P<0.001). Blood 2,3-diphosphoglycerate (2,3-DPG) resonances obscured
the inorganic phosphate (Pi) resonance. Therefore, these signals were fitted with two peaks:
one for 2,3-DPG5.4 ppm and Pi at 5.4 ppm, and one for 2,3-DPG6.3 ppm at 6.3 ppm.
Saturation-recovery curves of PCr, γ-ATP and α-ATP were fitted by a mono-exponential function
to estimate the corresponding longitudinal relaxation rate constants R 1. Mean R1 values were
used to determine metabolite T1 values via R1 = 1/T1. The in vivo myocardial energy status was
expressed as the PCr/γ-ATP ratio, corrected for partial saturation.
Because blood contains ATP, but no PCr (18), signal contamination from blood in in vivo cardiac
31
P MR spectra may lead to underestimation of the myocardial PCr/γ-ATP ratio. Therefore, we
evaluated whether the PCr/γ-ATP ratio in the current work could be affected by blood
contamination. The [γ-ATP/2,3-DPG6.3
ppm]blood
ratio in spectra acquired in fresh blood was
determined as a measure for ATP content in blood. The contribution of resonances from blood
metabolites to in vivo spectra was assessed by quantifying the [2,3-DPG6.3 ppm/γ-ATP]LV ratio in
the cardiac 3D ISIS spectra. Next, the relative contribution of signal from ATP in the blood to
82
Mouse cardiac
31
P MR spectroscopy
the ATP signal obtained with in vivo 3D ISIS was calculated as: [γ-ATP/2,3-DPG6.3
ppm]blood
×
[2,3-DPG6.3 ppm/γ-ATP]LV × 100%.
6.2.5. Statistical analyses
Data are reported as mean ± SD. The statistical significance of differences was analyzed using
two-sided paired or unpaired t-tests, as appropriate. The level of significance was set at P<0.05.
Figure 6.1. End-diastolic left ventricular (LV) MR images obtained from a control mouse (A) and a
mouse with a transverse aortic constriction (TAC) (B) The constriction is indicated by the arrow. Dilated
hypertrophic cardiomyopathy is evidenced by increased LV wall thickness and LV cavity volume in the
TAC mouse. Rectangles indicate the voxels selected for localized 31P-MRS with 3D ISIS. Panels C and D
display 31P MR spectra acquired in vivo with 3D ISIS in a healthy mouse heart and a TAC heart,
respectively. Myocardial PCr/γ-ATP, corrected for partial saturation, was lower in TAC mice (n=8)
compared to healthy controls (n=9) (E) Data are mean ± SD. **P<0.01. α-, β-, γ-ATP, α-, β-, and γphosphate groups in ATP; 2,3-DPG, 2,3-diphosphoglycerate; PCr, phosphocreatine; Pi, inorganic
phosphate.
83
Chapter 6
Figure 6.2. Comparisons of 1D ISIS and 3D ISIS localization for cardiac 31P-MRS. Geometrical
localization for (A) 1D ISIS in the anterior myocardial wall (slice thickness = 1 mm), indicated in
transversal and sagittal reference images. Circle and ellipses indicate the position of the 31P surface
coil. 3D ISIS voxel localization in (B) the anterior myocardial wall (1 × 3 × 3 mm3) and (C) enclosing
the whole LV (6 × 6 × 6 mm3). The corresponding 31P MR spectra from (D) 1D ISIS (NA = 384), (E)
3D ISIS in the myocardium (NA = 3072), and (F) 3D ISIS of the whole LV (NA = 384). All spectra
were acquired in the same mouse. (G) PCr/γ-ATP ratio assessed via 1D ISIS (as in A) and 3D ISIS of
the whole LV (as in C). **P<0.01 (two-sided paired t-test, n=6). α-, β-, γ-ATP, α-, β-, and γphosphate groups in ATP; 2,3-DPG, 2,3-diphosphoglycerate; PCr, phosphocreatine; Pi, inorganic
phosphate. SA: short axis, 4-ch LA: 4-chamber long-axis, 2-ch LA: 2-chamber long-axis.
84
Mouse cardiac
31
P MR spectroscopy
6.3. Results
6.3.1. MRI: LV hypertrophy in TAC mice
We assessed in vivo LV morphology and function from cine MR images to confirm the
hypertrophic phenotype and impaired cardiac performance in TAC mice (Table 6.1, Fig. 6.1A
and Fig. 6.1B). LV mass was 95% higher in TAC mice compared to healthy mice (P<0.001),
indicating LV hypertrophy in TAC mice. Concomitantly, EDV (P<0.001) and ESV (P<0.001) were
higher in TAC mice compared to controls. This translated in a lower SV (-33%, P<0.001) and EF
(-65%, P<0.001) in TAC mice. Combined, these data illustrate the development of dilated
hypertrophic cardiomyopathy with severe systolic dysfunction after 7 weeks of LV pressureoverload in TAC mice.
Table 1. LV morphology and functional parameters obtained with MRI in control mice and TAC mice.
Control (n=9)
TAC (n=8)
P
495 ± 50
539 ± 31
N.S.
90.0 ± 14.9
176 ± 19.1
***
3.7 ± 0.6
6.0 ± 0.9
***
End-diastolic volume (μL)
63.7 ± 10.8
123 ± 26.1
***
End-systolic volume (μL)
20.8 ± 5.4
94.1 ± 24.8
***
Stroke volume (μL)
43.0 ± 7.3
28.7 ± 6.6
***
Ejection fraction (%)
67.6 ± 5.6
23.9 ± 5.2
***
Heart rate (beats per minute)
LV mass (mg)
LV mass/body weight (mg/g)
Data are expressed as mean ± SD.
6.3.2.
Typical
31
***
P<0.001, N.S.: not significant.
P MRS: in vivo myocardial energy status
31
P MR spectra acquired with 3D ISIS in a healthy mouse heart and a TAC heart are
shown in Fig. 6.1C and Fig. 6.1D. Resonances of PCr, and ATP are indicated. Inorganic
phosphate (Pi, ~5 ppm) was obscured by 2,3-DPG arising from blood. Spectral line width for PCr
was 0.37 ± 0.15 ppm.
Conventional pulse-acquire
31
P MR saturation-recovery experiments of the mouse chest were
performed in order to estimate the high-energy phosphate metabolite T1 relaxation times at 9.4
T. T1 values did not differ between control mice and TAC mice, and were 2.54 ± 0.41 s for PCr,
1.45 ± 0.25 s for γ-ATP, and 1.09 ± 0.31 s for α-ATP. Given a TR of 2 seconds, this resulted in
a partial saturation correction factor for PCr/γ-ATP of 1.37 in 3D ISIS experiments. In healthy
mice (n=6), 3D ISIS was also performed under fully relaxed conditions. These localized
acquisitions yielded a partial saturation correction factor for myocardial PCr/γ-ATP of 1.38 ±
0.28 for spectra acquired at TR = 2 seconds, which is in good agreement with the value derived
from the unlocalized pulse-acquire saturation-recovery experiments. Myocardial PCr/γ-ATP,
corrected for partial saturation, was lower in TAC mice compared to healthy controls (0.76 ±
0.13 vs. 1.00 ± 0.15; P<0.01, Fig. 6.1E), which is indicative of a compromised myocardial
energy homeostasis in TAC mice.
85
Chapter 6
Localization of a slice containing anterior myocardial wall with 1D ISIS (Fig. 6.2A) systematically
resulted in higher PCr/γ-ATP ratios than those obtained with 3D ISIS of the whole LV (2.12 ±
0.61 vs. 1.08 ± 0.25; P<0.01, Fig. 6.2G). Skeletal muscle tissue surrounding the anterior
myocardial wall within the sensitive area of the surface coil (Fig. 6.2A) likely contributed to the
higher PCr/γ-ATP ratio observed with 1D ISIS. This was corroborated by spectra obtained with a
very small 9 µL 3D ISIS voxel positioned in the anterior myocardial wall (Fig. 6.2B), which
qualitatively confirm the PCr/γ-ATP ratio of ~1 for healthy myocardium (Fig. 6.2E).
6.3.3.
31
P MRS of liver, skeletal muscle, and blood
Figure 6.3. Voxel positioning for 3D ISIS in (A) the liver (5 × 5 × 5 mm3) and in (B) hind limb
skeletal muscle (4 × 4 × 4 mm3) in the mouse, and the corresponding 31P MR spectra for (C)
the liver (NA = 768) and (D) skeletal muscle (NA = 384). Note the prominent resonance of PCr
detected in skeletal muscle, whereas it is absent in liver tissue. α-, β-, γ-ATP, α-, β-, and γphosphate groups in ATP; PCr, phosphocreatine; Pi, inorganic phosphate.
Using the respiratory-gated and cardiac-triggered 3D ISIS approach, we acquired
31
P MR
spectra from in vivo mouse liver and hind limb skeletal muscle. Absence of PCr in liver tissue
was confirmed (Fig. 6.3C), illustrating that essentially no signal from PCr in skeletal muscle
contaminated the spectra that were obtained from the liver with 3D ISIS. Spectra obtained from
hind limb skeletal muscle yielded a PCr/γ-ATP ratio of 3.59 ± 0.58 (Fig. 6.3D), which is typical
for healthy mouse skeletal muscle in resting conditions (21).
In spectra obtained from fresh blood (Fig. 6.4), the blood ATP content was estimated via the [γATP/2,3-DPG6.3 ppm]blood ratio, which was 0.22 ± 0.14. Contribution of signal from metabolites in
the blood to the cardiac spectra, estimated via [2,3-DPG6.3 ppm/γ-ATP]LV, was similar for controls
and TAC mice, and was 0.18 ± 0.10. The relative contribution of signal from ATP in the blood to
the ATP signal in cardiac 3D ISIS spectra was therefore approximately 4%. These results show
86
Mouse cardiac
31
P MR spectroscopy
that contamination of the 3D ISIS spectra with signal from ATP in LV blood is marginal, and only
minimally affects the myocardial PCr/γ-ATP ratio.
Figure 6.4. 31P MR spectrum obtained from fresh blood. 2,3-DPG, 2,3-diphosphoglycerate;
PDE, phosphodiesters; α-, β-, γ-ATP, α-, β-, and γ-phosphate groups in ATP.
6.4. Discussion
We developed a noninvasive method to study the in vivo myocardial energy status in healthy
and diseased mice using 3D ISIS-localized
31
P MRS. Acquisitions were respiratory-gated and
cardiac-triggered, whilst maintaining steady-state conditions using dummy excitations to ensure
accurate localization. Using a 1H/31P MR setup, cardiac
cine
1
31
P MRS with 3D ISIS was combined with
H MRI to assess morphological, functional, and metabolic parameters in a single
experimental session of less than 2 hours. With this approach, we identified a reduced
myocardial energy status, evidenced by a ~25% lower PCr/γ-ATP ratio, accompanied by
hypertrophic growth and severely impaired myocardial function in a surgical mouse model of
heart failure. These pathophysiological changes are consistent with previous studies in patients
(25, 34) and mice (14, 31).
The small heart size and the high heart rate in mice require adequate strategies to avoid signal
contamination while maximizing sensitivity for in vivo cardiac
31
P MRS. A surface coil has
superior sensitivity over a volume resonator, and is therefore beneficial for
31
P MR signal
reception. However, chest skeletal muscle is located in the most sensitive area of the surface
coil, and can contribute considerably to the received signal (5). Therefore, localized acquisitions
are essential to restrict the obtained spectra to the heart. If CSI techniques are used for
localized signal acquisition (8, 12, 14, 31), Fourier bleeding can be an important contributor to
signal contamination (20), which is exacerbated in a surface coil setup. Typically, single-voxel
localization techniques provide better-defined signal localization than CSI methods (10). Here,
we minimized contamination of the spectra with signal from tissues other than the myocardium
by using the single-voxel approach of respiratory-gated and cardiac-triggered 3D ISIS.
87
Chapter 6
Myocardial PCr/γ-ATP ratio in healthy mice in this study was 1.00 ± 0.15. This value is on the
lower range of the normal PCr/ATP values reported in literature for humans (overall mean: 1.72
± 0.26) (5), which may be explained by species differences. It has been suggested that mice
have lower cardiac energy reserve than larger species (38). In humans, heart rate and cardiac
work can increase up to three-fold upon dobutamine stress (24) or during exercise (23),
whereas in healthy mice, increases in heart rate of only ~39% upon dobutamine infusion (32)
and 40% to 90% upon running exercise (1, 30) have been reported.
Interestingly, the PCr/γ-ATP in our healthy mice was also lower than the PCr/ATP values
reported in other in vivo mouse studies (8, 12, 35). It has been recognized that many aspects
of localized
31
P MRS acquisition and quantification can contribute to the variability in the
PCr/ATP values reported in literature (5), and that comparisons between laboratories should be
made with the differences in the methods used in mind (6, 26). The three main causes for
apparent discrepancies in literature reports are laboratory-dependent differences in 1)
correction for partial saturation, 2) contamination of the spectra by signal from the liver or
skeletal muscle tissue, and 3) contaminating blood in the LV cavity (5). Below, each of these
issues are addressed for the current study.
Partial saturation correction
For the correction of partial saturation effects, we used the metabolite T1 values as determined
by unlocalized pulse-acquire saturation-recovery experiments. Because the T1 values in chest
skeletal muscle could be different from those in cardiac muscle (43), we validated the correction
factor in healthy mice by acquiring 3D ISIS-localized spectra from the LV myocardium under
fully relaxed conditions. Indeed, the partial saturation correction factor derived from unlocalized
saturation-recovery experiments was in agreement with measurements localized to the heart.
This observation validates the assumption (7) that in the healthy mouse the T1(PCr)/T1(γ-ATP)
ratio is essentially the same for chest skeletal muscle and myocardium at 9.4 T (12).
Minimizing signal contamination
Because 3D ISIS requires multiple acquisitions for signal localization, the method is particularly
sensitive to motion artifacts and consequential contamination from tissues surrounding the
heart such as liver and chest skeletal muscle. Additional contamination (‘T1-smearing’) can be
introduced by differences in longitudinal magnetization between subsequent acquisitions within
one ISIS cycle due to imperfect flip angle of the excitation pulse combined with a TR<5×T1 (21,
28). Similar effects occur when TR is not constant between acquisitions, while measuring at
TR<5×T1. Previous studies in rodents (35) and in humans (13, 39, 42) using 3D ISIS for cardiac
applications did not measure at constant TR for a steady state of magnetization, nor at fully
relaxed conditions. The 3D ISIS sequence presented here uses a respiratory-gated and cardiactriggered timing strategy to ensure localized inversion of the LV signal at identical cardiac and
respiratory phases for all acquisitions, in combination with dummy excitations during respiratory
gates to maintain a constant TR. This approach may find applications beyond investigations of in
vivo myocardial energy status in the mouse. Indeed, localized
benefit from the strategy proposed here (40).
88
31
P MRS of the liver (27) will also
Mouse cardiac
31
P MR spectroscopy
Moreover, we demonstrated the effect of potential contamination of the spectra with signal from
chest skeletal muscle by comparing 3D ISIS of the whole LV with 1D ISIS of a slice containing
the anterior myocardial wall, similar to 1D CSI approaches reported previously (8, 15).
Localization within the selected slice is realized by surface coil positioning, which may introduce
signal from the surrounding tissue to the resultant spectrum. Indeed, the PCr/γ-ATP ratio
obtained with 1D ISIS was consistent with values observed with 1D CSI (8), but two-fold higher
than the myocardial PCr/γ-ATP ratio measured with 3D ISIS.
Blood contamination
Blood in the LV cavity contains ATP, but no PCr. If signal from ATP in the blood contributes to
the signal acquired from the voxel of interest, the myocardial PCr/ATP ratio may be
underestimated (16). A correction factor can be applied to account for signal contribution from
ATP in the blood to the spectra, which depends on both the amount of ATP in the blood and the
amount of blood contributing to the spectrum. We showed that the amount of ATP signal from
fresh blood, in terms of [γ-ATP/2,3-DPG6.3 ppm]blood, was approximately 20%, which is consistent
with previous reports of
31
P MRS studies of blood in humans (18) and mice (29, 35). The
contribution of signal from metabolites in blood to the in vivo 3D ISIS spectra of the whole LV,
estimated via [2,3-DPG6.3
ppm/γ-ATP]LV,
was less than 20%, and was not different between
healthy and TAC mice. The high velocity of flowing blood during acquisition may attenuate the
peaks from blood metabolites in spectra obtained in vivo (47). Based on these observations, we
estimated that the contribution of signal from ATP in the blood to the ATP signal in in vivo 3D
ISIS spectra was only 4%, and concluded that correcting for blood contamination was not
necessary in this study. Moreover, spectra obtained with 3D ISIS selecting a small voxel
confined to the anterior myocardial wall were similar to those obtained from the whole LV,
confirming minimal contribution of signal from blood to the latter.
Study limitations
With our current approach, it is not possible to detect similar decreases in both PCr and ATP
concentrations, which would render the PCr/ATP ratio insensitive to alterations in myocardial
energy status. Absolute quantification of metabolite concentration might therefore be more
sensitive to changes in cardiac energy metabolism. Indeed, it has been shown that in pressureoverload mice, both myocardial PCr and ATP concentrations decrease, illustrating the added
value of absolute quantification over the PCr/ATP ratio (15). In addition, absolute values would
allow for a more straightforward comparison of results obtained with different methods at
different research sites. Nonetheless, the PCr/ATP ratio has been used to identify perturbations
in myocardial energy homeostasis in various mouse models of (metabolic) diseases (3, 31, 41).
Furthermore, a reduced PCr/ATP ratio has been shown to be an important indicator of disease
severity and can be of predictive value for disease progression (31).
A drawback of using single-voxel localized
31
P MRS of the entire LV is that myocardial energy
status cannot be assessed at a regional scale, which could be of importance in investigations of
myocardial ischemia. Spatially encoded MRS techniques, such as 2D
31
P CSI (12), would allow
mapping of myocardial energetics to assess differences in cardiac energy status between
infarcted and remote regions after a myocardial ischemic insult. However,
31
P MRS mapping of
89
Chapter 6
myocardial energetics demonstrating regional differences in high energy phosphate content in
mouse models of myocardial ischemia has not been reported yet. To date, many preclinical
animal studies focus on pathologies that have a global effect on the heart, such as aortic
stenosis, diabetes, and inborn errors of metabolism. For these investigations, the current 3D
ISIS approach for localized
31
P MRS can be a valuable addition to the toolbox of mouse cardiac
MR methods (9).
Even at the high magnetic field strength of 9.4 T, we were not able to unambiguously detect
and quantify myocardial Pi, because the signal was obscured by signal from 2,3-DPG in the
blood. Potentially, reliable detection of Pi would provide additional metabolic insights, as
decreased PCr levels may be mirrored by increased Pi levels. Additionally, the chemical shift of
Pi could potentially be used to quantify in vivo myocardial pH. Notably, even in the human
heart, Pi is often undetectable, suggesting that myocardial Pi may only be partially MR visible
(5).
In conclusion, the present work describes a noninvasive approach to assess myocardial energy
status in the in vivo mouse using single-voxel 3D ISIS-localized
31
P MRS. The method
encompasses a respiratory-gated, cardiac-triggered 3D ISIS sequence with dummy excitations
during respiratory gates, to ensure a well-defined localization. This method is able to identify
differences in high-energy phosphate metabolism between the healthy mouse heart and a
widely used model for heart failure, the TAC mouse. Furthermore, the 3D ISIS-localized spectra
can be obtained within 40 minutes, leaving room for measurements of cardiac function with MRI
during the same experimental session. We anticipate that localized
31
P MRS will provide valuable
contributions to preclinical investigations of cardiac disease progression and therapeutic
intervention efficacy.
6.5. Acknowledgements
We thank Larry de Graaf and Tom R. Geraedts for dedicated hardware design, Esther C.M.
Kneepkens for contributions to pulse sequence design, and Leonie B.P. Niesen for biotechnical
assistance. This work was supported by the Center for Translational Molecular Medicine, project
TRIUMPH (grant number 01C-103) with funding from the Dutch Heart Foundation. S.M.H. and
J.J.P. are
supported
by VIDI grants (project
numbers 016.086.336
and
700.58.421,
respectively) from the Netherlands Organisation for Scientific Research (NWO).
6.6. References
1.
2.
3.
4.
5.
6.
90
Adlam D, De Bono JP, Danson EJ, Zhang MH, Casadei B, Paterson DJ, and Channon KM.
Telemetric analysis of haemodynamic regulation during voluntary exercise training in mouse models.
Experimental Physiology 96: 1118-1128, 2011.
Bakermans AJ, Abdurrachim D, Geraedts TR, Houten SM, Nicolay K, and Prompers JJ. In vivo
proton T1 relaxation times of mouse myocardial metabolites at 9.4 T. Magn Reson Med doi:
10.1002/mrm.25340, 2014.
Bakermans AJ, Dodd MS, Nicolay K, Prompers JJ, Tyler DJ, and Houten SM. Myocardial energy
shortage and unmet anaplerotic needs in the fasted long-chain acyl-CoA dehydrogenase knockout
mouse. Cardiovasc Res 100: 441-449, 2013.
Bakermans AJ, Geraedts TR, van Weeghel M, Denis S, Joao Ferraz M, Aerts JM, Aten J,
Nicolay K, Houten SM, and Prompers JJ. Fasting-induced myocardial lipid accumulation in longchain acyl-CoA dehydrogenase knockout mice is accompanied by impaired left ventricular function. Circ
Cardiovasc Imaging 4: 558-565, 2011.
Bottomley PA. NMR Spectroscopy of the Human Heart. In: eMagRes. John Wiley & Sons, Ltd, 2007.
Bottomley PA. The trouble with spectroscopy papers. Radiology 181: 344-350, 1991.
Mouse cardiac
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
31
P MR spectroscopy
Bottomley PA, Hardy CJ, and Weiss RG. Correcting human heart 31P NMR spectra for partial
saturation. Evidence that saturation factors for PCr/ATP are homogeneous in normal and disease
states. J Magn Reson 95: 341-355, 1991.
Chacko VP, Aresta F, Chacko SM, and Weiss RG. MRI/MRS assessment of in vivo murine cardiac
metabolism, morphology, and function at physiological heart rates. Am J Physiol Heart Circ Physiol
279: H2218-2224, 2000.
Coolen BF, Paulis LE, Geelen T, Nicolay K, and Strijkers GJ. Contrast-enhanced MRI of murine
myocardial infarction - part II. NMR Biomed 25: 969-984, 2012.
De Graaf RA. In vivo NMR spectroscopy: principles and techniques. John Wiley & Sons, 2008.
Epstein FH. MR in mouse models of cardiac disease. NMR Biomed 20: 238-255, 2007.
Flogel U, Jacoby C, Godecke A, and Schrader J. In vivo 2D mapping of impaired murine cardiac
energetics in NO-induced heart failure. Magn Reson Med 57: 50-58, 2007.
Fragasso G, De Cobelli F, Spoladore R, Esposito A, Salerno A, Calori G, Montanaro C, Maranta
F, Lattuada G, Margonato A, Del Maschio A, and Perseghin G. Resting cardiac energy metabolism
is inversely associated with heart rate in healthy young adult men. Am Heart J 162: 136-141, 2011.
Gupta A, Chacko VP, Schar M, Akki A, and Weiss RG. Impaired ATP kinetics in failing in vivo
mouse heart. Circ Cardiovasc Imaging 4: 42-50, 2011.
Gupta A, Chacko VP, and Weiss RG. Abnormal energetics and ATP depletion in pressure-overload
mouse hearts: in vivo high-energy phosphate concentration measures by noninvasive magnetic
resonance. Am J Physiol Heart Circ Physiol 297: H59-64, 2009.
Hardy CJ, Weiss RG, Bottomley PA, and Gerstenblith G. Altered myocardial high-energy
phosphate metabolites in patients with dilated cardiomyopathy. Am Heart J 122: 795-801, 1991.
Heijman E, Aben JP, Penners C, Niessen P, Guillaume R, van Eys G, Nicolay K, and Strijkers
GJ. Evaluation of manual and automatic segmentation of the mouse heart from CINE MR images. J
Magn Reson Imaging 27: 86-93, 2008.
Horn M, Neubauer S, Bomhard M, Kadgien M, Schnackerz K, and Ertl G. 31P-NMR spectroscopy
of human blood and serum: first results from volunteers and patients with congestive heart failure,
diabetes mellitus and hyperlipidaemia. MAGMA 1: 55-60, 1993.
Kan HE, Veltien A, Arnts H, Nabuurs CI, Luijten B, de Haan A, Wieringa B, and Heerschap A.
Gated dynamic 31P MRS shows reduced contractile phosphocreatine breakdown in mice deficient in
cytosolic creatine kinase and adenylate kinase. NMR Biomed 22: 523-531, 2009.
Keevil SF. Spatial localization in nuclear magnetic resonance spectroscopy. Phys Med Biol 51: R579636, 2006.
Keevil SF, Porter DA, and Smith MA. Experimental characterization of the ISIS technique for
volume selected NMR spectroscopy. NMR Biomed 5: 200-208, 1992.
Kober. Recent developments in small animal cardiovascular MRI. Curr Cardiovasc Imaging Rep 2014.
La Gerche A, Claessen G, Van de Bruaene A, Pattyn N, Van Cleemput J, Gewillig M, Bogaert J,
Dymarkowski S, Claus P, and Heidbuchel H. Cardiac MRI: a new gold standard for ventricular
volume quantification during high-intensity exercise. Circ Cardiovasc Imaging 6: 329-338, 2013.
Lamb HJ, Beyerbacht HP, Ouwerkerk R, Doornbos J, Pluim BM, van der Wall EE, van der
Laarse A, and de Roos A. Metabolic response of normal human myocardium to high-dose atropinedobutamine stress studied by 31P-MRS. Circulation 96: 2969-2977, 1997.
Lamb HJ, Beyerbacht HP, van der Laarse A, Stoel BC, Doornbos J, van der Wall EE, and de
Roos A. Diastolic dysfunction in hypertensive heart disease is associated with altered myocardial
metabolism. Circulation 99: 2261-2267, 1999.
Lamb HJ, Doornbos J, den Hollander JA, Luyten PR, Beyerbacht HP, van der Wall EE, and de
Roos A. Reproducibility of human cardiac 31P-NMR spectroscopy. NMR Biomed 9: 217-227, 1996.
Landis CS, Yamanouchi K, Zhou H, Mohan S, Roy-Chowdhury N, Shafritz DA, Koretsky A,
Roy-Chowdhury J, Hetherington HP, and Guha C. Noninvasive evaluation of liver repopulation by
transplanted hepatocytes using 31P MRS imaging in mice. Hepatology 44: 1250-1258, 2006.
Lawry TJ, Karczmar GS, Weiner MW, and Matson GB. Computer simulation of MRS localization
techniques: an analysis of ISIS. Magn Reson Med 9: 299-314, 1989.
Lee J, Hu Q, Nakamura Y, Wang X, Zhang X, Zhu X, Chen W, Yang Q, and Zhang J. Open-chest
31
P magnetic resonance spectroscopy of mouse heart at 4.7 Tesla. J Magn Reson Imaging 24: 12691276, 2006.
Lujan HL, Janbaih H, Feng HZ, Jin JP, and DiCarlo SE. Ventricular function during exercise in mice
and rats. Am J Physiol Regul Integr Comp Physiol 302: R68-74, 2012.
Maslov MY, Chacko VP, Stuber M, Moens AL, Kass DA, Champion HC, and Weiss RG. Altered
high-energy phosphate metabolism predicts contractile dysfunction and subsequent ventricular
remodeling in pressure-overload hypertrophy mice. Am J Physiol Heart Circ Physiol 292: H387-391,
2007.
Naumova AV, Weiss RG, and Chacko VP. Regulation of murine myocardial energy metabolism
during adrenergic stress studied by in vivo 31P NMR spectroscopy. Am J Physiol Heart Circ Physiol 285:
H1976-1979, 2003.
Neubauer S. The failing heart--an engine out of fuel. N Engl J Med 356: 1140-1151, 2007.
Neubauer S, Horn M, Pabst T, Harre K, Stromer H, Bertsch G, Sandstede J, Ertl G, Hahn D,
and Kochsiek K. Cardiac high-energy phosphate metabolism in patients with aortic valve disease
assessed by 31P-magnetic resonance spectroscopy. J Investig Med 45: 453-462, 1997.
Omerovic E, Basetti M, Bollano E, Bohlooly M, Törnell J, Isgaard J, Hjalmarson Å, Soussi B,
and Waagstein F. In vivo metabolic imaging of cardiac bioenergetics in transgenic mice. Biochem
Biophys Res Commun 271: 222-228, 2000.
91
Chapter 6
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
92
Ordidge R, Connelly A, and Lohman J. Image-selected in Vivo spectroscopy (ISIS). A new
technique for spatially selective NMR spectroscopy. J Magn Reson 66: 283-294, 1986.
Rockman HA, Ross RS, Harris AN, Knowlton KU, Steinhelper ME, Field LJ, Ross J, Jr., and
Chien KR. Segregation of atrial-specific and inducible expression of an atrial natriuretic factor
transgene in an in vivo murine model of cardiac hypertrophy. Proc Natl Acad Sci U S A 88: 8277-8281,
1991.
Saupe KW, Spindler M, Tian R, and Ingwall JS. Impaired cardiac energetics in mice lacking
muscle-specific isoenzymes of creatine kinase. Circ Res 82: 898-907, 1998.
Schaefer S, Gober J, Valenza M, Karczmar GS, Matson GB, Camacho SA, Botvinick EH, Massie
B, and Weiner MW. Nuclear magnetic resonance imaging-guided phosphorus-31 spectroscopy of the
human heart. J Am Coll Cardiol 12: 1449-1455, 1988.
Solanky BS, Sanchez-Canon GJ, Cobbold JF, Taylor-Robinson SD, Bell JD, Scudamore CL,
Ross E, Holder JC, So PW, and Cox IJ. Metabolic profiling of the rat liver after chronic ingestion of
alpha-naphthylisothiocyanate using in vivo and ex vivo magnetic resonance spectroscopy. Toxicol Sci
126: 306-316, 2012.
Tucci S, Flogel U, Hermann S, Sturm M, Schafers M, and Spiekerkoetter U. Development and
pathomechanisms of cardiomyopathy in very long-chain acyl-CoA dehydrogenase deficient (VLCAD(-/)) mice. Biochim Biophys Acta 1842: 677-685, 2014.
van der Meer RW, Hammer S, Smit JW, Frolich M, Bax JJ, Diamant M, Rijzewijk LJ, de Roos
A, Romijn JA, and Lamb HJ. Short-term caloric restriction induces accumulation of myocardial
triglycerides and decreases left ventricular diastolic function in healthy subjects. Diabetes 56: 28492853, 2007.
van Dobbenburgh JO, Lekkerkerk C, van Echteld CJ, and de Beer R. Saturation correction in
human cardiac 31P MR spectroscopy at 1.5 T. NMR Biomed 7: 218-224, 1994.
van Nierop BJ, van Assen HC, van Deel ED, Niesen LB, Duncker DJ, Strijkers GJ, and Nicolay
K. Phenotyping of left and right ventricular function in mouse models of compensated hypertrophy and
heart failure with cardiac MRI. PLoS ONE 8: e55424, 2013.
Vanhamme L, van den Boogaart A, and Van Huffel S. Improved method for accurate and efficient
quantification of MRS data with use of prior knowledge. J Magn Reson 129: 35-43, 1997.
Weiss RG, Bottomley PA, Hardy CJ, and Gerstenblith G. Regional myocardial metabolism of highenergy phosphates during isometric exercise in patients with coronary artery disease. N Engl J Med
323: 1593-1600, 1990.
Zhang J, Duncker DJ, Xu Y, Zhang Y, Path G, Merkle H, Hendrich K, From AH, Bache RJ, and
Ugurbil K. Transmural bioenergetic responses of normal myocardium to high workstates. Am J Physiol
268: H1891-1905, 1995.
Chapter 7
Cardiac diastolic dysfunction in high-fat diet fed mice
is associated with lipotoxicity without impairment of
cardiac energetics in vivo
Desiree Abdurrachim1*, Jolita Ciapaite1,2*, Bart Wessels1, Miranda Nabben1, Joost J.F.P. Luiken3,
Klaas Nicolay1, Jeanine J. Prompers1
1
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven,
The Netherlands
2
Department of Pediatrics, Center for Liver, Digestive and Metabolic Diseases, University of Groningen,
University Medical Center Groningen, Groningen, The Netherlands
3
Department of Molecular Genetics, Cardiovascular Research Institute Maastricht (CARIM), Maastricht
University, Maastricht, The Netherlands
*both authors contributed equally
Published in Biochimica et Biophysica Acta Lipids and Lipid Metabolism 1841(10):1525-1537,
2014. DOI: 10.1016/j.bbalip.2014.07.016.
Chapter 7
Abstract
Obesity is often associated with abnormalities in cardiac morphology and function. This study
tested the hypothesis that obesity-related cardiomyopathy is caused by impaired cardiac
energetics. In a mouse model of high-fat diet (HFD)-induced obesity, we applied in vivo cardiac
31
P magnetic resonance spectroscopy (MRS) and magnetic resonance imaging (MRI) to
investigate cardiac energy status and function, respectively. The measurements were
complemented by ex vivo determination of oxygen consumption in isolated cardiac mitochondria,
the expression of proteins involved in energy metabolism, and markers of oxidative stress and
calcium homeostasis. We also assessed whether HFD induced myocardial lipid accumulation
using in vivo 1H MRS, and if this was associated with apoptosis and fibrosis. Twenty weeks of
HFD feeding resulted in early stage cardiomyopathy, as indicated by diastolic dysfunction and
increased left ventricular mass, without any effects on systolic function. In vivo cardiac
phosphocreatine-to-ATP ratio and ex vivo oxygen consumption in isolated cardiac mitochondria
were not reduced after HFD feeding, suggesting that the diastolic dysfunction was not caused by
impaired cardiac energetics. HFD feeding promoted mitochondrial adaptations for increased
utilization of fatty acids, which was however not sufficient to prevent the accumulation of
myocardial lipids and lipid intermediates. Myocardial lipid accumulation was associated with
oxidative stress and fibrosis, but not apoptosis. Furthermore, HFD feeding strongly reduced the
phosphorylation of phospholamban, a prominent regulator of cardiac calcium homeostasis and
contractility. In conclusion, HFD-induced early stage cardiomyopathy in mice is associated with
lipotoxicity-associated oxidative stress, fibrosis, and disturbed calcium homeostasis, rather than
impaired cardiac energetics.
Keywords
diet-induced obesity, cardiomyopathy, cardiac energetics, lipotoxicity, calcium homeostasis
94
Cardiac energetics and function in diet-induced obesity
7.1. Introduction
Over-nutrition
and
consumption
of
high-fat
diets
are
associated
with
obesity-related
cardiomyopathy and increased risk for heart failure (36, 48). In obesity, the elevated plasma
free fatty acid (FFA) concentration induces an increase in myocardial fatty acid (FA) uptake and
oxidation (63). Despite the increase in FA oxidation, the excessive availability of FA may lead to
an imbalance between FA uptake and oxidation, resulting in an increased deposition of
potentially toxic lipids in the heart (18, 42). Myocardial lipid accumulation has been implicated
in the development of cardiomyopathy through a number of pathways, including lipid-induced
apoptosis (77) and fibrosis (40, 67).
Upregulation of FA oxidation may also result in a disturbance in cardiac energetics. Compared to
glucose oxidation, stoichiometrically, FA oxidation produces 10% less ATP per oxygen consumed
(i.e. 10% decrease in cardiac efficiency). However, up to 30% lower cardiac efficiency has been
observed as a consequence of increased FA oxidation, suggesting that also other mechanisms
than the inefficiency of FA oxidation affect the cardiac efficiency (25). The increase in the supply
of reducing equivalents (i.e. NADH and FADH2) to the electron transport chain (ETC), without a
parallel increase in the oxidative phosphorylation (OXPHOS) capacity, might result in the loss of
electrons from the ETC and subsequently, the generation of reactive oxygen species (ROS) (60).
In turn, increased ROS may promote mitochondrial uncoupling (8, 9, 15, 19, 52), as a
mechanism to reduce the electrochemical proton gradient required for ROS formation, which will
however result in lower ATP generation. ROS may induce mitochondrial defects, such as shown
by lower expression of OXPHOS proteins in obese and diabetic mice (10, 27). Increased ROS
production might also stress sarco/endoplasmic reticulum, which could lead to disturbed calcium
homeostasis and thereby impair cardiac muscle contraction/relaxation (39). Although in vitro
data suggest that cardiac energy metabolism is disturbed in obesity, in vivo evaluation of
cardiac energetics is currently lacking.
In animal models, cardiac energetics is usually assessed ex vivo in isolated perfused heart
setups by measuring cardiac power and oxygen consumption to calculate cardiac energy
efficiency (33, 50). Reduced cardiac efficiency has been observed in rodents with diet-induced
obesity (19, 75), obese ob/ob mice (10, 13, 50), and diabetic db/db mice (11, 33, 50).
Unfortunately, the ex vivo setup does not entirely mimic the complexity of the in vivo situation
as the ex vivo heart is usually perfused with glucose and FA at a constant concentration, while
in obesity the in vivo heart is exposed to a hyperlipidemic and, possibly, a hyperinsulinemic
environment, which affects the choice of substrates. In vivo cardiac energetics can be assessed
using
31
P magnetic resonance spectroscopy (MRS), in which the PCr/ATP ratio is used as a
measure of cardiac energy status (34, 53). The implementation of
31
P MRS in vivo in the mouse
heart has proven very challenging, because of its small size (~2000 times smaller than the
human heart) and high heart rates (~10 times faster than the human heart). In a number of
recent studies, in vivo cardiac
31
P MRS has been applied in mouse models of heart failure,
demonstrating a robust association between heart failure and cardiac energy deficiency (31, 32).
However, in vivo cardiac PCr/ATP data in obesity-related cardiomyopathy is currently not
available.
95
Chapter 7
In this study, we applied cardiac
31
P MRS in a mouse model of diet-induced obesity to
investigate whether obesity-related cardiomyopathy is associated with an impairment in cardiac
energetics in vivo. For this purpose, C57BL/6J mice were fed a high-fat diet (HFD) for 20 weeks.
Cardiac systolic and diastolic function were examined using cinematic magnetic resonance
imaging (MRI). The
31
P MRS measurements of in vivo cardiac energy status were complemented
by ex vivo determinations of oxygen consumption in isolated cardiac mitochondria, the
expression or activity of various proteins involved in energy metabolism, as well as markers of
oxidative stress and calcium homeostasis. Additionally, we assessed whether HFD feeding
induced myocardial lipid accumulation using in vivo localized 1H MRS and if this was associated
with apoptosis and fibrosis.
7.2. Materials and Methods
7.2.1. Animals
Male C57BL/6J mice were purchased from Charles River (Elsene, Belgium). The animals were
housed under controlled temperature (23°C) and humidity (50%) with a 12:12-h dark-light
cycle and were given ad libitum access to food and water. Starting from the age of 12 weeks,
the animals were divided into two groups (n=8 per group) and fed either low fat diet (LFD; 10
kcal% (4.3 g%) palm oil-based fat, 70 kcal% (67.3 g%) carbohydrate, 20 kcal% (19.2 g%)
protein; based on OpenSource Diets No. D12450B, Research Diet Services, Wijk bij Duurstede,
The Netherlands) or HFD (45 kcal% (24 g%) palm oil-based fat, 35 kcal% (41 g%)
carbohydrate, 20 kcal% (24 g%) protein; based on OpenSource Diets No. D12451, Research
Diet Services, Wijk bij Duurstede, The Netherlands). The animals received the diet for 20 weeks.
The animals then underwent MR measurements in the fed state, after which they were
sacrificed within one week, during which they were maintained on the same diet. The heart was
excised and part of it was used immediately for the isolation of mitochondria. The other part
was either snap-frozen in liquid nitrogen and stored at −80C until further biochemical analyses,
or immersed in 4% paraformaldehyde (PFA) in PBS at 4C overnight and embedded in paraffin
the following day. Animal handling procedures and experimental protocols conformed to and
were approved by the Animal Experimental Committee of Maastricht University (The
Netherlands).
7.2.2. MR measurements
All MR measurements were performed on a 9.4 T horizontal bore MR scanner (Bruker,
Germany). MRI and 1H MRS were performed using a 35-mm quadrature birdcage coil (Bruker,
Germany) for both signal reception and transmission. For
1
quadrature H and linear
31
31
P MRS, a 54-mm double tuned
P birdcage coil (Rapid Biomedical, Germany) was used for signal
transmission, while a 15-mm diameter, home-built, actively decoupled, two-turn
31
P surface coil
was used for signal reception. Before the experiments, the animals were sedated in a chamber
with 3% isoflurane in medical air at a flow rate of 0.4 L/min. During the measurements, the
anesthesia was maintained at 1-2% isoflurane through a customized anesthesia mask.
Temperature was maintained at 36-37°C with a heating pad. Rectal temperature, ECG signal,
and breathing rate were monitored throughout the measurements. All measurements were
performed with respiratory gating and cardiac triggering.
96
Cardiac energetics and function in diet-induced obesity
7.2.2.1. Cardiac function measurement using cardiac cine MRI
Cardiac systolic and diastolic function were measured as described previously (20). For systolic
function measurement, cine movies from the beating heart (15-18 frames/cardiac cycle) were
acquired using prospectively cardiac-triggered gradient echo imaging of 5-6 contiguous short
axis and 2 long axis slices (slice thickness: 1 mm). The imaging parameters were as follows:
repetition time: 7 ms, echo time: 1.8 ms, flip angle: 15°, matrix: 192 x 192, and field of view:
30 x 30 mm2. For diastolic function, cardiac movies of only the mid-ventricular slice were
acquired using retrospectively-triggered gradient echo imaging, allowing data reconstruction
with a much higher temporal resolution (50-60 frames/cardiac cycle). The parameters used
were as follows: repetition time: 4.7 ms, echo time: 2.35 ms, flip angle: 15°, matrix: 128 x 128,
and field of view: 30 x 30 mm2, effective time resolution: 2 ms. Image segmentation and data
analysis were performed using CAAS MRV 2.0 (Pie Medical, Maastricht, The Netherlands) or
Segment (version 1.8 R1145, http://segment.heiberg.se).
7.2.2.2. Cardiac energy status measurements using
Cardiac
31
31
P MRS
P MR spectra were acquired using the image selected in vivo spectroscopy (ISIS)
sequence, on a voxel of typically ~6 x 6 x 6 mm3 covering the left ventricle, at the end of the
diastolic phase, as described previously (4). The parameters were as follows: repetition time:
~2 s, 1.2 ms 90º sinc-shaped excitation pulse (bandwidth: 32.0 ppm), 6.25 ms 180º adiabatic
hyperbolic secant inversion pulses (bandwidth: 37.5 ppm), 96 ISIS cycles (768 scans), and γATP on resonance. The 90º sinc-shaped excitation pulse was calibrated during the in vivo scan,
by performing a series of single pulse measurements with varying pulse power on a 5-mm
diameter glass phantom containing 15 M phosphoric acid, which was positioned underneath the
surface coil. Data fitting and analysis were performed using AMARES in jMRUI (66), as described
as in (4). The γ-ATP line widths (LWγ-ATP) were constrained relative to the PCr line width (LWPCr)
according to an empirically determined relation: LWγ-ATP = LWPCr + 14.8 Hz (n=63 data sets,
R=0.78, P<0.001). As a measure of cardiac energy status, the ratio of PCr to γ-ATP was
determined. The PCr/ATP ratio was corrected for partial T1 saturation using correction factors of
1.75 and 1.31 for PCr and γ-ATP, respectively, which were determined by analyzing ISIS
spectra acquired at repetition times of 2 s and 15 s (n=19 data sets from the current and
previous studies). The contamination of the ATP from the blood to the spectra was shown to be
less than 4%, and was considered negligible (5).
7.2.2.3. Myocardial lipid measurements using 1H MRS
1
H MR spectra were acquired from a 1 x 2 x 2 mm3 voxel positioned in the interventricular
septum during the diastolic phase of the cardiac cycle, using point resolved spectroscopy
(PRESS) with chemical shift selective (CHESS) water suppression, as described in (3). The
parameters were as follows: repetition time: ~2 s, echo time: 9.1 ms, 0.41 ms 90° Hermiteshaped excitation pulse, 0.9 ms 180° Mao-type refocusing pulses, and 256 averages. The
spectra were processed and analyzed using AMARES in the jMRUI software package (66). All
metabolites (taurine, choline/carnitine, creatine, and seven peaks from lipids) were fitted to
Gaussian line shapes. Myocardial metabolite levels were then calculated from the metabolite
signal relative to the unsuppressed water signal measured from the same voxel.
97
Chapter 7
7.2.3. Isolation of mitochondria
Following excision, the heart was rinsed in ice cold 0.9% KCl, quickly minced with scissors in 0.5
ml of medium containing 160 mM KCl, 10 mM NaCl, 20 mM Tris, 5 mM EGTA and 0.05 mg/ml
bacterial proteinase type XXIV (pH 7.7), and incubated for ~1 min on ice. The mixture was then
homogenized using a Potter-Elvehjem homogenizer. Mitochondria were isolated through a
differential centrifugation procedure as described in (51) and resuspended in buffer containing
180 mM KCl, 20 mM Tris, 3 mM EGTA and 1 mg/ml bovine serum albumin (pH 7.4). Protein
content was determined using a BCA protein assay kit (Pierce, Thermo Fisher Scientific Inc.,
Rockfort, IL, USA).
7.2.4. High-resolution respirometry
A 2-channel high-resolution Oroboros oxygraph-2k (Oroboros, Innsbruck, Austria) was used to
measure oxygen consumption rates at 37°C. Isolated mitochondria (0.15 mg/ml) were
incubated in 1 ml of assay medium (110 mM KCl, 20 mM Tris, 2.3 mM MgCl2, 5 mM KH2PO4 and
1 mg/ml bovine serum albumin, pH 7.2) with i) 5 mM pyruvate plus 5mM malate, or ii) 25 µM
palmitoyl-L-carnitine plus 2 mM malate as the oxidizable substrates. The maximal O2
consumption rate in a coupled state of oxidative phosphorylation (state 3) was measured after
the addition of 0.1 mg/ml hexokinase, 12.5 mM glucose and 1 mM ATP. The basal O2
consumption rate (state 4) was measured after blocking ATP synthesis with 1.25 µM
carboxyatractyloside (CAT). The uncoupled O2 consumption rate (state U) was determined after
the addition of 1 µM carbonyl cyanide 3-chlorophenyl hydrazone (CCCP). Data acquisition and
analysis were performed using Oxygraph-2k-Datlab 4.3.1.15 software (Oroboros, Innsbruck,
Austria).
7.2.5. Citrate synthase activity
Tissues were weighed and 20% (w/v) homogenates were prepared in ice cold PBS, pH 7.4.
Homogenates were sonicated for 30 s in the pulse mode (pulse duration 1 s, interval between
the pulses 1 s, power input 10 W) on ice. Half of each sample was used for the determination of
acylcarnitine concentrations (see below). The remainder of the sample was centrifuged at 1000
g for 10 min at 4 C. Protein content in the supernatant was measured using BCA protein assay
kit (Pierce, Thermo Fisher Scientific Inc., Rockford, IL, USA). Citrate synthase activity was
determined spectrophotometrically as described in (65).
7.2.6. Determination of ATP, ADP, and AMP concentration
Tissues were weighed, grinded in liquid nitrogen and homogenized in 0.5 M perchloric acid at 4
C, followed by centrifugation for 10 min at 14,000 g, 4C. The supernatant was neutralized with
2 M KOH to ~pH 7 and centrifuged for 10 min at 14,000 g, 4C. Nucleotides in 100 µl of
supernatant were separated with HPLC as described in (55). The concentrations in the eluate
were calculated from ATP, ADP and AMP (Sigma-Aldrich, Zwijndrecht, The Netherlands)
standard curves and expressed as nmol/mg wet tissue.
98
Cardiac energetics and function in diet-induced obesity
7. 2.7. Quantification of acylcarnitines
Sonicated 20% homogenates were centrifuged for 10 min at 14,000 g, 4C. Next, 10 µl of the
supernatant was mixed with 100 µl acetonitrile and 100 µl of methanol–water (80:20 v/v)
containing internal standards [1,1,1-N-methyl-2H3]-L-carnitine, [2H3]acetyl-L-carnitine, [3,3,32
H3]propionyl-L-carnitine,
[8,8,8-2H3]octanoyl-L-carnitine,
[10,10,10-2H3]decanoyl-L-carnitine
and [16,16,16-2H3]hexadecanoyl-L-carnitine (VU Medical Center, Amsterdam, The Netherlands),
vortexed and centrifuged for 10 min at 14,000 g at 4C. Concentrations of acylcarnitines were
measured in the supernatant with an API 3000 LC-MS/MS equipped with a Turbo ion spray
source (Applied Biosystems/MDS Sciex, Ontario, Canada) as described in (22).
7.2.8. Biochemical determination of intramyocelullar lipids
Intramyocellular lipids were determined in cardiac muscle homogenates as described previously
(1). In short, samples containing 400 µg of protein were used for intracellular lipid extraction in
methanol/chloroform, and an internal standard and water were added. Afterwards thin-layer
chromatography
was
used
to
separate
lipids.
Bands
were
resolved
with
a
hexane/diethylether/propanol (87:10:3) resolving solution. Triacylglycerol (TAG), diacylglycerol
(DAG), monoacylglycerol (MAG), and cholesterol bands were detected with a Molecular Imager
(ChemiDoc XRS, BioRad) and analyzed with Quantity One® (BioRad).
7.2.9. Immunoblotting
Total cardiac protein extracts were prepared, resolved with SDS-PAGE and transferred to
polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA) as described in (71). After
blocking with TBS containing 0.1% Tween 20 and 5% skim milk powder for 1 h at room
temperature, the membranes were incubated overnight at 4C with one of the following
antibodies: rabbit polyclonal anti-peroxisome proliferator-activated receptor  coactivator 1
(PGC1) (1:500), rabbit anti-caspase 3 (1:500), rabbit polyclonal anti-collagen α1 Type I
(1:1000), rabbit anti-cardiac troponin T (1:500), goat anti-sarcoplasmic reticulum calcium
ATPase2 (SERCA2) (1:200), rabbit anti-phospholamban (1:200), rabbit anti-phospho Thr17
phospholamban (1:200) (all antibodies were from Santa Cruz Biotechnology, Santa Cruz, CA,
USA), rabbit anti-uncoupling protein 3 (UCP3; 1:500; Sigma-Aldrich), rabbit polyclonal anticaspase 9 (1:1000; Stressgen Biotechnologies Corporation, San Diego, CA, USA), rabbit
polyclonal
anti-Cu/Zn
Corporation),
rabbit
superoxide
dismutase
polyclonal
anti-Mn
(SOD;
1:1000;
superoxide
Stressgen
dismutase
Biotechnologies
(1:1000;
Stressgen
Biotechnologies Corporation), or MitoProfile® Total OXPHOS Rodent WB Antibody Cocktail
(1:2000, MitoSciences, Eugene, OR, USA) containing mouse monoclonal antibodies against
Complex I, II, III, IV, and V subunits. Membranes were washed 3× for 5 min with TBS
containing 0.1% Tween 20 and incubated with a corresponding horse-radish peroxidaseconjugated secondary antibody for 1 h at room temperature. After the final wash of 3× for 5
min with TBS containing 0.1% Tween 20 and 1× 5 min with TBS, the immunocomplexes were
detected using SuperSignal West Dura Extended Duration Substrate (Pierce, Thermo Fisher
Scientific Inc., Rockford, IL, USA), visualized using ChemiDoc™ XRS+ imaging system and
quantified using Image Lab™ analysis software version 3.0 (Bio-Rad Laboratories Inc., Hercules,
CA, USA). All data were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
99
Chapter 7
expression level and expressed relative to the LFD-fed controls. Immunoblot detection of
protein carbonylation in the total cardiac protein extracts was done using the OxyBlot™ Protein
Oxidation Detection Kit (Chemicon/Millipore, Billerica, MA, USA) following recommendations of
the manufacturer.
7.2.10. Histology
The paraffin-embedded hearts were sliced in 5 μm-thick sections. The collagen was stained with
Picrosirius red (Direct Red 80, Sigma-Aldrich, Zwijndrecht, The Netherlands) according to
standard
histological
procedures.
To
detect
apoptotic
cells,
terminal
deoxynucleotidyl
transferase dUTP nick end labeling (TUNEL) staining was performed using Apoptag peroxidase in
situ
apoptosis
detection
kit
(Millipore
Corporation,
Billerica,
MA,
USA)
according
to
manufacturer’s protocol. The stained sections were digitalized with a Pannoramic MIDI scanner
(3DHISTECH, Budapest, Hungary). Fibrotic area, represented by red area in Sirius Red staining,
was quantified from 15 random locations in 2-4 heart slices (magnification 40 x), using a macro
in ImageJ 1.47v (NIH, USA). TUNEL-positive cells were counted from 5 random locations (40 x
magnification, about 200-300 cells per location) in a slice, and quantified as a percentage of the
total number of cells detected using ‘analyze particles’ in imageJ 1.47v (NIH, USA).
7.2.11. Plasma FFA, triglyceride, insulin, and glucose determination
Blood samples were collected in EDTA coated tubes, at 20 weeks of feeding after 4 h fasting and
at sacrifice (fed conditions). Plasma was obtained by centrifuging the blood at 1000 g for 10 min
and stored at -80°C until plasma insulin, FFA, and triglyceride (TG) determination. Fasting
plasma glucose levels were measured immediately after blood collection in one drop of blood
using a Glucose-201 glucose meter (HemoCue, Ängelholm, Sweden). Plasma insulin levels were
determined using a mouse insulin determination kit (Mercodia, Uppsala, Sweden). Plasma TG
levels were determined using a TG determination kit (Sigma-Aldrich, Zwijndrecht, The
Netherlands). Plasma FFA levels were determined using a NEFA-HR kit (Wako Chemicals, Neuss,
Germany).
7.2.12. Statistical analysis
All data are presented as means ± standard deviation. Statistical analysis was performed with
SPSS 17.0 (SPSS Inc.) using independent-samples Student’s t-test. Multivariate ANOVA was
applied to analyze different lipid peaks in the cardiac 1H MR spectra. The level of significance
was set at P<0.05.
7.3. Results
7.3.1. Mouse characteristics
After 20 weeks of feeding, body weight of HFD mice was 50% higher compared with LFD mice
(P<0.001, suppl Fig. 7.1A). The average caloric intake over the 20 weeks of feeding tended to
be higher for HFD mice than for LFD mice (18%, P=0.052, suppl Fig. 7.1B). Fasting plasma
glucose levels were higher in HFD compared with LFD mice (P=0.018; Table 7.1). Fasting
100
Cardiac energetics and function in diet-induced obesity
plasma insulin tended to be higher in HFD compared with LFD mice (P=0.082; Table 7.1).
Plasma TG and FFA levels, determined in the fed state, were about 1.5-fold higher (P<0.05) in
HFD compared with LFD mice, indicating hyperlipidemia in HFD mice (Table 7.1).
Suppl. Figure 7.1. Animal characteristics. (A) Body weight and (B) caloric intake during the
experiment. Dashed lines in panel B indicate means of the caloric intake over the 20 weeks of
diet. Data are means ± SD (n=8 per diet group). Open symbols: LFD, closed symbols: HFD.
#
P<0.01 and *P<0.001 vs. LFD.
Table 7.1. Plasma insulin, glucose, total TG, and FFA concentrations.
LFD
HFD
Insulin (fasted, ng/mL)
0.87 ± 0.50
2.25 ± 1.02#
Glucose (fasted, mM)
9.03 ± 0.99
11.23 ± 1.97*
Total TG (fasted, mM)
0.63 ± 0.10
0.62 ± 0.07
Total TG (fed, mM)
0.59 ± 0.20
0.92 ± 0.30*
FFA (fed, mM)
0.39 ± 0.09
0.63 ± 0.17**
Data are means ± SD (n=8 per diet group). #P=0.08, *P<0.05, **P<0.01 vs. LFD.
TG: triglycerides, FFA: free fatty acids.
7.3.2. HFD feeding induced increased left ventricular mass and diastolic dysfunction
Ejection fraction, end-diastolic volume, end-systolic volume, and stroke volume were all similar
between LFD and HFD mice (P=0.16, 0.56, 0.73, 0.23, respectively; Table 7.2). However, left
ventricular (LV) mass and LV mass normalized to tibia length were 13-16% higher in HFD mice
compared with LFD mice (P<0.05; Table 7.2 and Fig. 7.1A). MRI with high temporal resolution
(time resolution: 2 ms) was performed on the midventricular slice to determine cardiac diastolic
function parameters. From the images, the LV cavity volume-time curve (Fig. 7.1B) and the
time derivative of the LV cavity volume-time curve (Fig. 7.1C) were generated to analyze the
early and late diastolic filling phases. Early peak filling rates (E) were 15% lower (P<0.05) in
HFD mice compared with LFD mice (Table 7.2 and Fig. 7.1C), which indicates an impairment in
the active relaxation phase of the left ventricle. The late peak filling rate (A), i.e. the rate of
diastolic filling due to atrial contraction, did not differ significantly between LFD and HFD mice
101
Chapter 7
(P=0.11; Table 7.2 and Fig. 7.1C). In both LFD and HFD mice, the early filling phase
contributed about 60% of the total diastolic filling (P=0.85; Table 7.2).
Figure 7.1. In vivo cardiac function measured using MRI. (A) Two-chamber long axis, 4-chamber
long axis, and short axis cardiac MR images of a LFD and a HFD mouse. The larger heart size in the
HFD compared with the LFD mouse can be appreciated. (B) The LV cavity volume-time curve and (C)
the time derivative of LV cavity volume-time curve, showing the peak ejection rate (PER), early peak
filling rate (E), and late peak filling rate (A). Open symbols: LFD, closed symbols: HFD.
102
Cardiac energetics and function in diet-induced obesity
Table 7.2. MRI parameters of cardiac function.
LFD
HFD
72.3 ± 3.6
74.7 ± 2.7
EDV (μL)
71.6 ± 14.5
75.4 ± 11.0
ESV (μL)
20.2 ± 6.4
19.2 ± 4.5
SV (μL)
51.4 ± 8.2
56.2 ± 7.0
PER (%EDV/ms)
2.3 ± 0.2
2.1 ± 0.3
E (%EDV/ms)
2.5 ± 0.2
2.1 ± 0.3*
EF (%)
A (%EDV/ms)
E contribution (%EDV)
LV mass (mg)
LV mass/tibia length (mg/mm)
1.6 ± 0.2
1.3 ± 0.4
60.4 ± 5.3
61.0 ± 5.4
101.8 ± 13.5
114.9 ± 7.8*
5.6 ± 0.7
6.5 ± 0.5*
Data are means ± SD (n=8 per diet group, except for PER, E, A, and E contribution n=7).
*P<0.05 vs. LFD. EF: ejection fraction, EDV: end-diastolic volume, ESV: end-systolic
volume, SV: stroke volume, PER: peak ejection rate, E: early peak filling rate, A: late
peak filling rate.
Figure 7.2. Cardiac energy status measured using 31P MRS. (A) Twochamber long axis, 4-chamber long axis, and short axis MR images from
a LFD mouse showing the positioning of the 6 x 6 x 6 mm3 voxel used
for 31P MRS in white. (B) Typical example of an in vivo cardiac 31P MR
spectrum. (C) PCr/γ-ATP ratio. (D) ATP, ADP, and AMP concentrations,
as determined biochemically. Data are means ± SD (n=8 per diet
group). White bar: LFD, black bar: HFD.
103
Chapter 7
7.3.3. HFD feeding did not alter cardiac energy status
Cardiac energy status in vivo was determined using
7.2A). A typical
31
31
P MRS localized in the left ventricle (Fig.
P MR spectrum contains a phosphocreatine (PCr) peak originating from
myocardium, a 2,3-diphosphoglycerate (DPG) peak from blood in the LV lumen (obscuring the
inorganic phosphate (Pi) peak from myocardium), and γ-, α-, β-ATP peaks from mainly
myocardium with a minor contamination from blood (Fig. 7.2B). PCr/ γ-ATP ratios were similar
for HFD and LFD mice (P=0.73; Fig. 7.2C), indicating that 20 weeks of HFD feeding did not
affect cardiac energy status in vivo. ATP, ADP, and AMP concentrations, determined
biochemically, were similar between HFD and LFD mice (P=0.27, P=0.85, and P=0.73,
respectively; Fig. 7.2D).
7.3.4. HFD feeding increased mitochondrial FA oxidation capacity, without altering
mitochondrial protein levels
Oxygen consumption rates when ETC is coupled with ATP synthase (coupled state, state 3) and
in the uncoupled state (state U) were similar in isolated heart mitochondria from LFD and HFD
mice, when pyruvate plus malate was used as the oxidizable substrate (P=0.38 and P=0.57,
respectively; Fig. 7.3A). This indicates that glucose oxidation capacity did not change upon HFD
feeding. When palmitoyl-L-carnitine plus malate was used as the substrate, state 3 and state U
oxygen consumption rates in isolated HFD heart mitochondria were 1.3- (P<0.01) and 1.4-fold
(P<0.05) higher than those in isolated LFD heart mitochondria, respectively (Fig. 7.3B),
indicating an increased FA oxidation capacity in the HFD group. The oxygen consumption rates
in state 4 were similar in HFD and LFD groups, for both pyruvate plus malate and palmitoyl-Lcarnitine plus malate as substrates (P=0.07 and P=0.85, respectively; Fig. 7.3), indicating that
HFD feeding did not affect the leakiness of the inner mitochondrial membrane to protons. In
agreement, the expression of mitochondrial uncoupling protein 3 (UCP3) was similar between
LFD and HFD hearts (P=0.90; Table 7.3).
Figure 7.3. Oxygen consumption rates in isolated cardiac mitochondria. For (A) pyruvate plus
malate, and (B) palmitoyl-L-carnitine plus malate as the substrate. Data are means ± SD (n=8 per
diet group). White bars: LFD, black bars: HFD. *P<0.05 and **P<0.01 vs. LFD.
104
Cardiac energetics and function in diet-induced obesity
The activity of citrate synthase, a mitochondrial marker enzyme, was similar in cardiac muscle
from both diet groups (LFD: 2.43 ± 0.16 vs HFD: 2.43 ± 0.19 μmol/min/mg protein, n=8 per
group, P=0.92), suggesting that mitochondrial density was not affected by HFD feeding. In
addition, the expression of peroxisome proliferator-activated receptor γ cofactor-1α (PGC1α),
which regulates mitochondrial biogenesis, was not altered in HFD hearts (P=0.75; Table 7.3).
Furthermore, the expression of all mitochondrial oxidative phosphorylation complexes was
similar between LFD and HFD hearts (P=0.83, 0.65, 0.40, 0.80, 0.09 for Complexes I, II, III, IV,
and V, respectively; Table 7.3).
Table 7.3. Cardiac protein expression.
LFD
HFD
UCP3
1.00 ± 0.10
1.01 ± 0.06
PGC1α
1.00 ± 0.08
0.98 ± 0.05
OXPHOS Complex I
1.00 ± 0.11
1.01 ± 0.09
OXPHOS Complex II
1.00 ± 0.11
1.04 ± 0.07
OXPHOS Complex III
1.00 ± 0.09
1.05 ± 0.02
OXPHOS Complex IV
1.00 ± 0.08
0.99 ± 0.03
OXPHOS Complex V
1.00 ± 0.09
1.15 ± 0.07
Data are means ± SD (n=3 per diet group). All data were normalized to glyceraldehyde
3-phosphate dehydrogenase (GAPDH) expression level and expressed relative to the
LFD-fed controls. UCP3: uncoupling protein 3, PGC1α: peroxisome proliferator activator
γ coactivator-1α, OXPHOS: oxidative phosphorylation.
7.3.5. HFD feeding induced accumulation of myocardial lipids and lipid intermediates
Myocardial lipid content was measured using 1H MRS from a single voxel located in the septum
(Fig. 7.4A). Seven peaks associated with lipids are visible in the spectra (Fig. 7.4B and Fig.
7.4C). After 20 weeks of feeding, the peak area from the most prominent lipid signal from
methylene protons (peak 9; CH2) was 1.6-fold higher (P<0.001) in HFD compared with LFD
mice (Fig. 7.4D). In addition, peak areas of the lipid signals from allylic methylene protons
(peak 7; CH2-CH=CH-CH2), and -methylene protons (peak 8; CβH2CH2COO) were significantly
higher (P<0.01) in HFD compared with LFD mice (Fig. 7.4D). The peak areas from other lipid
resonances and those from creatine, choline/carnitine, and taurine were similar in both diet
groups (Fig. 7.4D).
In agreement with the higher myocardial lipid content in HFD mice in vivo, ex vivo biochemical
analysis also showed 1.6-fold higher TAG levels in HFD mouse hearts compared with LFD mouse
hearts (P<0.001, Fig. 7.5A). The higher myocardial TAG content in HFD mice was accompanied
by elevated levels of DAG (1.4-fold, P=0.006), MAG (1.7-fold, P<0.001), and cholesterol (1.6fold, P=0.004), indicating increased lipotoxicity in the HFD heart (Fig. 7.5A). Furthermore, HFD
feeding resulted in the accumulation of acylcarnitines in the heart (Fig. 7.5B). As a consequence,
the content of free carnitine (C0) in the heart was 40% lower (P<0.001) in HFD compared with
LFD mice (Fig. 7.5B). This suggests stimulation of FA oxidation by HFD feeding. The observation
that the content of acetylcarnitine (C2) was 23% higher (P<0.001) in the hearts of HFD-fed
mice may suggest that the increase in mitochondrial -oxidation capacity was not matched by
the TCA cycle capacity to utilize acetyl-coenzyme A (CoA) produced by the former.
105
Chapter 7
Figure 7.4. Myocardial lipid content measured using in vivo 1H MRS. (A) Short axis and 4-chamber
long axis MR images showing the positioning of the 1 x 2 x 2 mm3 voxel used for 1H MRS in black.
Representative in vivo 1H MR spectra from (B) a LFD-fed mouse and (C) a HFD-fed mouse. (D)
Metabolite signal amplitude expressed as a percentage of the water signal. Data are means ± SD
(n=8 per diet group). White bars: LFD, black bars: HFD. Bold: resonating protons for the
corresponding triglyceride (TG) peaks. **P<0.01 and ***P<0.001 vs. LFD.
Figure 7.5. Cardiac lipid intermediates. (A) Triacylglycerols (TAG), diacylglycerols (DAG),
monoacylglycerols (MAG), and cholesterol, and (B) cardiac acylcarnitine profiles. Data are means ±
SD (n=6-8 per diet group). White bars: LFD, black bars: HFD. *P<0.05, **P<0.01 and ***P<0.001
vs. LFD.
106
Cardiac energetics and function in diet-induced obesity
7.3.6. HFD feeding increased the production of reactive oxygen species
To investigate the effect of HFD feeding on the production of reactive oxygen species (ROS), we
measured the expression of cytosolic superoxide dismutase (Cu/ZnSOD) and mitochondrial
superoxide dismutase (MnSOD), two enzymes involved in the scavenging of superoxide anion
Figure 7.6. Markers for the production of ROS. Expression of antioxidant enzymes (A) Cu/Zn
SOD (n=3 per diet group) and (B) Mn SOD (n=6 per diet group), normalized to GAPDH. (C)
Representative immunoblot images and (D) the quantification of protein oxidation using
OxyBlot™ (n=3 per diet group). SOD: superoxide dismutase, GAPDH: glyceraldehyde 3phosphate dehydrogenase, NC: negative control, DNPH: 2,4-dinitrophenylhydrazine. Data are
means ± SD. White bars: LFD, black bars: HFD. *P<0.05, **P<0.01, ***P<0.001 vs. LFD.
107
Chapter 7
radicals produced mostly during mitochondrial respiration. The expression of these enzymes
was higher (Cu/ZnSOD: 16%, P<0.05; MnSOD: 20%, P<0.001) in HFD compared with LFD
mice (Fig. 7.6A and Fig. 7.6B). In agreement, we observed an increase in protein oxidation in
HFD compared with LFD hearts (Fig. 7.6C and Fig. 7.6D).
7.3.7. HFD feeding and apoptosis
The expression of caspases involved in the initiation of apoptosis, procaspase-9 and cleaved
(mature) caspase-9, was 37% (P<0.05) and 32% (P<0.01) higher, respectively, in hearts of
HFD mice compared with LFD mice (Fig. 7.7A and Fig. 7.7B). However, the expression level of
executing caspase, procaspase-3, which is processed by caspase-9 into mature caspase-3, was
similar between HFD and LFD groups (Fig. 7.7C). The mature caspase-3 was not detectable in
both groups (data not shown). Accordingly, the number of TUNEL-positive cells was not
different between TUNEL-stained LFD and HFD heart slices (P=0.56; Fig. 7.7D and Fig. 7.7E).
Figure 7.7. Apoptosis markers. Expression of (A) procaspase-9, (B) caspase-9, and (C)
procaspase-3, normalized to GAPDH (n=3 per diet group). (D) Visualization of TUNEL
staining, in which TUNEL-positive cells appear brown, while TUNEL-negative cells appear
purple. Black arrows indicate TUNEL-positive cells. (E) The quantification of TUNEL-positive
cells (LFD n=3, HFD n=4). GAPDH: glyceraldehyde 3-phosphate dehydrogenase, TUNEL:
terminal deoxynucleotidyl transferase dUTP nick end labeling. Data are means ± SD. White
bars: LFD, black bars: HFD. *P<0.05, **P<0.01 vs. LFD.
108
Cardiac energetics and function in diet-induced obesity
7.3.8. HFD feeding induced myocardial fibrosis
The expression of procollagen type I was similar in both diet groups (Fig. 7.8A). The expression
of mature collagen type I, the main collagen type expressed in the heart, was 40% increased
(P<0.01) in response to HFD feeding (Fig. 7.8B). This increase in collagen content indicates
excess formation of fibrous tissue, which was confirmed by a higher fraction of Sirius Redenhanced areas in HFD compared with LFD heart slices (P=0.027; Fig. 7.8C and Fig. 7.8D),
indicating the presence of increased fibrosis in HFD hearts.
Figure 7.8. Fibrosis markers. Expression of (A) procollagen I and (B) collagen I, normalized
to GAPDH (n=3 per diet group). (C) Representative images of Sirius Red staining from 3
different mice per diet group, and (D) the quantification of Sirius Red-enhanced areas (LFD
n=3, HFD n=4). Collagen appears red on yellow cytoplasm background. GAPDH:
glyceraldehyde 3-phosphate dehydrogenase. Data are means ± SD. White bars: LFD, black
bars: HFD. *P<0.05 vs. LFD.
7.3.9. HFD feeding affected cardiac calcium handling and contractile apparatus
Increased ROS might stress sarco/endoplasmic reticulum, which could disturb calcium handling
and the contractile apparatus. The expression of sarcoplasmic reticulum calcium ATPase2
(SERCA2) did not change in hearts of HFD compared with LFD mice (P=0.84; Fig. 7.9A);
however, there was a 60% decrease in the phosphorylation of phospholamban (PLB) (P=0.018;
Fig. 7.9B), a regulatory protein of SERCA2. HFD feeding also decreased the expression of
contractile protein troponin T (cTNT; 19%, P=0.020; Fig. 7.9C).
109
Chapter 7
Figure 7.9. Calcium handling and contractile protein. Expression of (A) SERCA2 (n=3 per
diet group), (B) phosphorylated PLBThr17 (n=3 per diet group), and (C) troponin T (n=6 per
diet group). The expression of SERCA2 and troponin T were normalized to GAPDH.
Phosphorylated PLBThr17 was expressed relative to total PLB. SERCA2: sarcoplasmic
reticulum calcium ATPase2, GAPDH: glyceraldehyde 3-phosphate dehydrogenase,
pPLBThr17: phosphorylated phospholamban, PLB: phospholamban. Data are means ± SD.
White bars: LFD, black bars: HFD. *P<0.05, **P<0.01 vs. LFD.
7.4. Discussion
Obesity is often associated with abnormalities in cardiac morphology and function. In this study,
we tested the hypothesis that obesity-related cardiomyopathy is associated with impaired
cardiac energy metabolism. In mice fed a HFD for 20 weeks, we evaluated cardiac function, and
in vivo and ex vivo parameters of cardiac energy metabolism. In addition, we assessed whether
HFD feeding induced myocardial lipid accumulation. Twenty weeks of HFD feeding resulted in
diastolic dysfunction and increased LV mass, without changes in systolic function. In vivo
cardiac PCr/ATP ratio and ex vivo oxygen consumption in isolated cardiac mitochondria were not
reduced after HFD feeding, suggesting that the diastolic dysfunction was not caused by impaired
cardiac energetics. HFD feeding promoted mitochondrial adaptations for increased utilization of
FA, which was however not sufficient to prevent the accumulation of myocardial lipids and lipid
intermediates. In addition, HFD feeding increased oxidative stress markers and collagen
110
Cardiac energetics and function in diet-induced obesity
deposition, and reduced phosphorylation of phospholamban in the heart. Our data thus show
that HFD-induced early stage cardiomyopathy in mice is associated with lipotoxicity-associated
oxidative stress, fibrosis, and disturbed calcium homeostasis, rather than impaired cardiac
energetics.
7.4.1. Effect of HFD feeding on cardiac function
We showed that 20 weeks of HFD feeding was associated with early stage cardiomyopathy (24),
marked by an increased LV mass and lower early peak filling rates, indicating impaired diastolic
function, while systolic function was normal. This supports diet-induced obesity as a model of
diastolic dysfunction. Similar to our findings, in a previous study, 20 weeks of HFD feeding in
mice did not affect systolic function, although no data on diastolic function was reported (76). In
contrast, another study in mice reported lower systolic function (27, 45) and minor diastolic
dysfunction (45) after 12 weeks of HFD feeding. In Wistar rats, posterior wall thickness
increased and systolic function decreased after 8 weeks of HFD feeding (57). In Sprague Dawley
rats fed a HFD for 32 weeks, LV hypertrophy was observed without any cardiac dysfunction (35).
These different results might be caused by different experimental conditions: differences in
types of animal models, differences in macronutrient composition of the diet and duration of the
experiment, as well as differences in the sensitivity of the measurement techniques.
7.4.2. The role of cardiac energetics in HFD-induced cardiac dysfunction
Continuous repetition of the cardiac contraction-relaxation cycle requires permanent supply of
energy in the form of ATP. Therefore, an impairment of mitochondrial ATP production may
underlie HFD-induced diastolic dysfunction. However, in vivo cardiac
31
P MRS showed that 20
weeks of HFD feeding did not affect the cardiac PCr/ATP ratio, a measure of in vivo cardiac
energy status. To our knowledge, we are the first to measure the cardiac PCr/ATP ratio in vivo
in a mouse model of HFD-induced obesity. The finding that HFD feeding had no effect on the
cardiac PCr/ATP ratio in vivo indicates that in this condition ATP production still meets the ATP
demand. This was also supported by unaltered ATP, ADP, and AMP concentrations in hearts of
HFD compared with LFD mice, as determined biochemically. These data are in agreement with
observations in perfused heart setups, where unaffected PCr and ATP concentrations have been
reported in mouse hearts after 20 weeks of HFD feeding (76).
The PCr/ATP ratio measured in this study (0.85 ± 0.15 in LFD mice) is lower compared to
literature values of healthy mice (17). The values are also generally lower than values reported
for humans (38, 54), which is likely due to methodological differences (7). It has been
recognized that many aspects of localized
31
P-MRS acquisition and quantification can contribute
to differences in PCr/ATP values (7). Although the method we used (cardiac-triggered,
respiratory-gated 3D ISIS) may underestimate the PCr/ATP values, our method has been shown
to be sensitive to detect a decrease in PCr/ATP in the hearts of fasted long-chain acyl-CoA
dehydrogenase (LCAD) knockout mice compared with controls (4).
In vitro assessment of mitochondrial function revealed that HFD feeding increased oxidative
phosphorylation capacity with the FA -oxidation substrate palmitoyl-L-carnitine, while oxidative
phosphorylation capacity for pyruvate was unaffected, showing that cardiac metabolism adapted
111
Chapter 7
to the dietary lipid oversupply. This is in agreement with previous findings in perfused heart
studies demonstrating that HFD feeding shifts cardiac substrate use from glucose to FA (61, 70,
76), while myocardial energetics is maintained (70, 76). A shift in cardiac substrate use from
glucose to FA has been associated with impaired cardiac function (43, 63). HFD feeding had no
effect on the basal (state 4) oxygen consumption rate in isolated cardiac mitochondria or on the
expression of cardiac UCP3, suggesting unaltered ion permeability of the inner mitochondrial
membrane/mitochondrial uncoupling. In contrast, a previous study in rats fed with HFD showed
increased mitochondrial uncoupling, which was associated with reduced cardiac energetics (19).
The maintenance of in vivo cardiac energetics upon HFD feeding in the present study was also
supported by unaltered cardiac citrate synthase activity, a mitochondrial marker enzyme,
unaltered expression of PGC1α, a key regulator of mitochondrial biogenesis, and unaltered
expression levels of OXPHOS complexes. Studies with other obesity and diabetes mouse models,
such as leptin-deficient (ob/ob) mice and leptin receptor-deficient (db/db) mice, showed
decreased expression of oxidative phosphorylation proteins in association with impaired cardiac
energetics and function (10, 21, 27), suggesting that genetic obesity has more profound
detrimental effects on cardiac energy metabolism compared to HFD-induced obesity.
Taken together, our data show that HFD feeding has a subtle effect on cardiac energy
metabolism, manifesting in the activation of signal transduction pathways that shift cardiac
metabolism toward increased use of FA as a fuel, rather than a gross impairment of energy
producing function.
7.4.3. The role of lipotoxicity in HFD-induced cardiac dysfunction
The storage of lipids in cardiomyocytes, which occurs when cardiac FA uptake exceeds the rate
of oxidation, is a protective mechanism against the formation of intermediate lipid metabolites,
such as diacylglycerols, ceramides, and long chain fatty acyl-CoA esters (12, 69). However,
prolonged myocardial lipid overload could contribute to cardiac dysfunction through increased
ROS production and oxidative stress (41), apoptosis (28, 77), fibrosis (40, 67), and induction of
endoplasmic reticulum (ER) stress (74). We showed that 20 weeks of HFD feeding resulted in a
strong accumulation of myocardial neutral lipids as well as lipid intermediates, such as
acylcarnitines and DAG. DAG may interfere with insulin signaling (6), which supports the
reduction in cardiac insulin sensitivity as observed in mice as early as after 1.5 weeks of HFD
feeding (59). The accumulation of myocardial lipids suggests that FA supply and uptake exceed
the rate of oxidation, despite the increase in mitochondrial FA oxidation capacity (64). In this
respect, we have earlier observed in HFD-fed rats that the main cardiac FA transporter CD36
relocates from intracellular membrane compartments to the sarcolemma (58). This CD36
relocation occurs in the absence of changes in expression of this transporter, and causes a
chronically increased influx of FA into the heart, followed by insulin resistance, and finally
contractile dysfunction (29). Furthermore, the accumulation of cardiac acylcarnitines indicates
that the increased FA oxidation in the HFD mouse hearts is associated with elevated incomplete
FA oxidation (37), which can be explained by the lack of a concomitant increase in OXPHOS
capacity as shown by the unchanged glucose oxidation capacity and protein expression of
OXPHOS complexes. The increase in FA oxidation without a concomitant increase in OXPHOS
capacity may result in the loss of electrons from the ETC and subsequently, excessive
112
Cardiac energetics and function in diet-induced obesity
stimulation of ROS production. In the present study, we observed upregulation of antioxidant
enzymes Cu/ZnSOD and MnSOD and increased protein oxidation in response to HFD feeding,
which indeed suggests increased production of ROS.
7.4.3.1. Lipotoxicity and apoptosis
Increased ROS production may lead to the activation of the apoptosis pathway (16, 28), as well
as lower cardiac efficiency due to the induction of mitochondrial uncoupling, a mechanism to
reduce ROS formation at the expense of ATP production (8). In the present study, HFD feeding
had no effect on the coupling of cardiac mitochondria but it indeed induced activation of the
apoptosis pathway, as indicated by upregulation of procaspase-9 as well as increased formation
of the mature caspase-9. However, we did not observe increased processing of procaspase-3 to
its mature form, which would ultimately lead to the execution of apoptosis. Supporting this data,
TUNEL staining did not show differences in the amount of apoptotic cells in LFD and HFD heart
slices. This is not surprising, since the processing of procaspase-3 to its active form catalyzed by
caspase-9 requires additional mitochondrial signals, such as the release of cytochrome c from
mitochondria (14) and a drop in cytosolic ATP level (73) due to the loss of mitochondrial
integrity and impairment of mitochondrial function. We showed that ATP concentration was
similar in LFD and HFD hearts, and the integrity of cardiac mitochondria was not affected by
HFD as indicated by unaltered oxygen consumption rate in state 4, implying that the extent of
damage did not yet lead to the actual execution of apoptosis.
7.4.3.2. Lipotoxicity and fibrosis
Excessive production of toxic lipid intermediates such as DAG and ROS stimulates the
production of pro-collagens and collagens, through the activation of protein kinases PKC-β or
PKA (2). Collagens, primarily collagen type I, make up the extracellular matrix (ECM) of the
heart (2). The increase in ECM, a hallmark of fibrosis, has been shown to increase myocardial
stiffness (72). The stiffening of myocardium reduces LV compliance (i.e. the change in LV
pressure needed to allow the LV filling), and therefore the ability of the muscle to relax (i.e.
reduced LV relaxation/diastolic function). Indeed, it has been shown that collagen content was
increased in rats with pre-diabetes or upon HFD feeding, together with reduced LV diastolic
function (40, 67). In addition, treatment using anti-diabetic drug pioglitazone was shown to
normalize LV diastolic function, which was associated with reduced collagen content (67). In
human studies, the serum levels of breakdown products of procollagen I and collagen I (i.e.
carboxy-terminal propeptide of procollagen I (PICP) and carboxy-terminal telopeptide of
collagen I (CITP)) were shown to be higher in relation to the severity of diastolic dysfunction
(49), providing evidence for the relevance of fibrosis in the development of diastolic dysfunction.
In the present study, we found higher expression of collagen I levels in HFD compared with LFD
mouse hearts. This indication of excess formation of fibrous connective tissue was supported by
an increase in the fibrotic area in Sirius Red-stained HFD heart slices. Therefore, our data
suggest that myocardial fibrosis plays a key role in HFD-induced impairment of cardiac function
in mice.
7.4.3.3. Lipotoxicity and calcium homeostasis
Finally, the increase in lipotoxicity-induced oxidative stress may impair calcium homeostasis
(30, 68), which may also play a role in HFD-induced diastolic dysfunction (23), since diastolic
dysfunction is characterized by a prolonged relaxation time (62). Cardiac muscle relaxation after
113
Chapter 7
contraction is induced in response to the lowering of the sarcoplasmic calcium concentration due
to active re-uptake of calcium into the sarcoplasmic reticulum, catalyzed by sarcoplasmic
reticulum calcium ATPases (SERCA). SERCA activity is crucially regulated by the phosphorylation
of PLB (47) and a reduction in PLB phosphorylation may be associated with lipotoxicity-induced
ER stress as shown in a previous study in HFD mice (68). In the unphosphorylated state, PLB
binds to SERCA, inhibiting calcium pump activity. The phosphorylation of PLB releases this
inhibition, which has been shown to result in an increase in calcium transport up to 4-fold or
greater (62). PLB levels were shown to be linearly correlated with the rates of contraction and
relaxation in isolated perfused hearts (46) and in hearts of intact mice (44), suggesting that the
remarkable 60% decrease in PLB phosphorylation in the HFD mouse hearts may have
contributed to the diastolic dysfunction observed in the present study. Supporting our results, a
recent study reported decreased PLB phosphorylation, impaired calcium regulation, and
contractile dysfunction in cardiomyocytes of HFD-fed mice (68). In addition, we also observed
reduced cardiac troponin T (cTNT) in HFD mouse hearts. cTNT is essential in regulating calcium
sensitivity and magnitude of force production during the contraction. Cardiac troponin T
mutation or deletion has been associated with diastolic dysfunction in mice (26, 56), and a
decrease in cTNT expression has been observed in response to HFD feeding in mice (21). Taken
together, our data show that disturbed calcium regulation may contribute to the observed
diastolic dysfunction.
7.5. Conclusion
In conclusion, 20 weeks of HFD feeding led to cardiac dysfunction in mice as indicated by
increased LV mass and reduced diastolic function. HFD feeding induced mitochondrial
adaptations promoting the utilization of FA. However, HFD-induced cardiac dysfunction was not
caused by abnormalities in cardiac energetics, but rather was associated with lipotoxicityinduced myocardial oxidative stress, fibrosis, and disturbed calcium homeostasis.
7.6. Acknowledgements
We thank Leonie Niesen and David Veraart for their assistance in animal handling, Albert
Gerding for the assistance with cardiac acylcarnitine quantification, Will Coumans for the
assistance with biochemical determination of intramyocellular lipids, and Abdallah Motaal for the
reconstruction of high temporal resolution cine MRI data.
D.A., B.W. and J.J.P. are supported by a VIDI grant (project number 700.58.421) from the
Netherlands Organisation for Scientific Research (NWO). J.C. is supported by the NWO-funded
Groningen Systems Biology Center for Energy Metabolism and Ageing. This work was supported
by the Dutch Technology Foundation STW, the Applied Science Division of NWO, and the
Technology Program of the Ministry of Economic Affairs (grant number 10191). M.N. is
supported by a VENI grant (project number 916.14.050) from NWO.
114
Cardiac energetics and function in diet-induced obesity
7.7. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Alkhateeb H, Chabowski A, Glatz JF, Luiken JF, and Bonen A. Two phases of palmitate-induced
insulin resistance in skeletal muscle: impaired GLUT4 translocation is followed by a reduced GLUT4
intrinsic activity. Am J Physiol Endocrinol Metab 293: E783-793, 2007.
Asbun J, and Villarreal FJ. The pathogenesis of myocardial fibrosis in the setting of diabetic
cardiomyopathy. J Am Coll Cardiol 47: 693-700, 2006.
Bakermans AJ, Abdurrachim D, Geraedts TR, Houten SM, Nicolay K, and Prompers JJ. In vivo
proton T1 relaxation times of mouse myocardial metabolites at 9.4 T. Magn Reson Med doi:
10.1002/mrm.25340, 2014.
Bakermans AJ, Dodd MS, Nicolay K, Prompers JJ, Tyler DJ, and Houten SM. Myocardial energy
shortage and unmet anaplerotic needs in the fasted long-chain acyl-CoA dehydrogenase knockout
mouse. Cardiovasc Res 100: 441-449, 2013.
Bakermans AJ, van Nierop BJ, Abdurrachim D, Nicolay K, and Prompers JJ. In vivo cardiac 31PMRS in a mouse model of heart failure. In: Proc 20th Ann Meeting ISMRM p. 1115, 2012.
Boden G, and Shulman GI. Free fatty acids in obesity and type 2 diabetes: defining their role in the
development of insulin resistance and beta-cell dysfunction. Eur J Clin Invest 32: 14-23, 2002.
Bottomley PA. NMR Spectroscopy of the Human Heart. In: eMagRes. John Wiley & Sons, Ltd, 2007.
Boudina S, and Abel ED. Mitochondrial uncoupling: a key contributor to reduced cardiac efficiency in
diabetes. Physiology (Bethesda) 21: 250-258, 2006.
Boudina S, Bugger H, Sena S, O'Neill BT, Zaha VG, Ilkun O, Wright JJ, Mazumder PK,
Palfreyman E, Tidwell TJ, Theobald H, Khalimonchuk O, Wayment B, Sheng X, Rodnick KJ,
Centini R, Chen D, Litwin SE, Weimer BE, and Abel ED. Contribution of impaired myocardial
insulin signaling to mitochondrial dysfunction and oxidative stress in the heart. Circulation 119: 12721283, 2009.
Boudina S, Sena S, O'Neill BT, Tathireddy P, Young ME, and Abel ED. Reduced mitochondrial
oxidative capacity and increased mitochondrial uncoupling impair myocardial energetics in obesity.
Circulation 112: 2686-2695, 2005.
Boudina S, Sena S, Theobald H, Sheng X, Wright JJ, Hu XX, Aziz S, Johnson JI, Bugger H,
Zaha VG, and Abel ED. Mitochondrial energetics in the heart in obesity-related diabetes: direct
evidence for increased uncoupled respiration and activation of uncoupling proteins. Diabetes 56: 24572466 2007.
Brindley DN, Kok BP, Kienesberger PC, Lehner R, and Dyck JR. Shedding light on the enigma of
myocardial lipotoxicity: the involvement of known and putative regulators of fatty acid storage and
mobilization. Am J Physiol Endocrinol Metab 298: E897-908, 2010.
Buchanan J, Mazumder PK, Hu P, Chakrabarti G, Roberts MW, Yun UJ, Cooksey RC, Litwin SE,
and Abel ED. Reduced cardiac efficiency and altered substrate metabolism precedes the onset of
hyperglycemia and contractile dysfunction in two mouse models of insulin resistance and obesity.
Endocrinology 146: 5341-5349, 2005.
Budihardjo I, Oliver H, Lutter M, Luo X, and Wang X. Biochemical pathways of caspase activation
during apoptosis. Annu Rev Cell Dev Biol 15: 269-290, 1999.
Bugger H, and Abel ED. Mitochondria in the diabetic heart. Cardiovasc Res 88: 229-240, 2010.
Cai L, Li W, Wang G, Guo L, Jiang Y, and Kang YJ. Hyperglycemia-induced apoptosis in mouse
myocardium: mitochondrial cytochrome C-mediated caspase-3 activation pathway. Diabetes 51: 19381948, 2002.
Chacko VP, Aresta F, Chacko SM, and Weiss RG. MRI/MRS assessment of in vivo murine cardiac
metabolism, morphology, and function at physiological heart rates. Am J Physiol Heart Circ Physiol 279:
H2218-2224, 2000.
Chiu H-C, Kovacs A, Ford DA, Hsu F-F, Garcia R, Herrero P, Saffitz JE, and Schaffer JE. A novel
mouse model of lipotoxic cardiomyopathy. J Clin Invest 107: 813-822, 2001.
Cole MA, Murray AJ, Cochlin LE, Heather LC, McAleese S, Knight NS, Sutton E, Jamil AA,
Parassol N, and Clarke K. A high fat diet increases mitochondrial fatty acid oxidation and uncoupling
to decrease efficiency in rat heart. Basic Res Cardiol 106: 447-457, 2011.
Coolen BF, Abdurrachim D, Motaal AG, Nicolay K, Prompers JJ, and Strijkers GJ. High frame
rate retrospectively triggered Cine MRI for assessment of murine diastolic function. Magn Reson Med
69: 648-656, 2013.
Cruz-Topete D, List EO, Okada S, Kelder B, and Kopchick JJ. Proteomic changes in the heart of
diet-induced pre-diabetic mice. J Proteomics 74: 716-727, 2011.
Derks TG, Boer TS, van Assen A, Bos T, Ruiter J, Waterham HR, Niezen-Koning KE, Wanders
RJ, Rondeel JM, Loeber JG, Ten Kate LP, Smit GP, and Reijngoud DJ. Neonatal screening for
medium-chain acyl-CoA dehydrogenase (MCAD) deficiency in The Netherlands: the importance of
enzyme analysis to ascertain true MCAD deficiency. J Inherit Metab Dis 31: 88-96, 2008.
Fang CX, Dong F, Thomas DP, Ma H, He L, and Ren J. Hypertrophic cardiomyopathy in high-fat
diet-induced obesity: role of suppression of forkhead transcription factor and atrophy gene
transcription. Am J Physiol Heart Circ Physiol 295: H1206-H1215, 2008.
Fang ZY, Prins JB, and Marwick TH. Diabetic cardiomyopathy: Evidence, mechanisms, and
therapeutic implications. Endocr Rev 25: 543-567 2004.
Fillmore N, and Lopaschuk GD. Targeting mitochondrial oxidative metabolism as an approach to
treat heart failure. Biochim Biophys Acta 1833: 857-865, 2013.
115
Chapter 7
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
116
Frey N, Franz WM, Gloeckner K, Degenhardt M, Muller M, Muller O, Merz H, and Katus HA.
Transgenic rat hearts expressing a human cardiac troponin T deletion reveal diastolic dysfunction and
ventricular arrhythmias. Cardiovasc Res 47: 254-264, 2000.
Ge F, Hu C, Hyodo E, Arai K, Zhou S, Lobdell Iv H, Walewski JL, Homma S, and Berk PD.
Cardiomyocyte triglyceride accumulation and reduced ventricular function in mice with obesity reflect
increased long chain fatty acid uptake and de novo fatty acid synthesis. J Obes 2012: 205648, 2012.
Ghosh S, and Rodrigues B. Cardiac cell death in early diabetes and its modulation by dietary fatty
acids. Biochim Biophys Acta 1761: 1148-1162, 2006.
Glatz JF, Luiken JJ, and Bonen A. Membrane fatty acid transporters as regulators of lipid
metabolism: implications for metabolic disease. Physiol Rev 90: 367-417, 2010.
Guo R, Ma H, Gao F, Zhong L, and Ren J. Metallothionein alleviates oxidative stress-induced
endoplasmic reticulum stress and myocardial dysfunction. J Mol Cell Cardiol 47: 228-237, 2009.
Gupta A, Chacko VP, Schar M, Akki A, and Weiss RG. Impaired ATP kinetics in failing in vivo
mouse heart. Circulation Cardiovascular Imaging 4: 42-50, 2011.
Gupta A, Chacko VP, and Weiss RG. Abnormal energetics and ATP depletion in pressure-overload
mouse hearts: in vivo high-energy phosphate concentration measures by noninvasive magnetic
resonance. Am J Physiol Heart Circ Physiol 297: H59-64, 2009.
How OJ, Aasum E, Severson DL, Chan WY, Essop MF, and Larsen TS. Increased myocardial
oxygen consumption reduces cardiac efficiency in diabetic mice. Diabetes 55: 466-473, 2006.
Hudsmith LE, and Neubauer S. Magnetic resonance spectroscopy in myocardial disease. JACC
Cardiovasc Imaging 2: 87-96, 2009.
Jeckel KM, Miller KE, Chicco AJ, Chapman PL, Mulligan CM, Falcone PH, Miller ML,
Pagliassotti MJ, and Frye MA. The role of dietary fatty acids in predicting myocardial structure in
fat-fed rats. Lipids Health Dis 10: 92, 2011.
Kenchaiah S, Evans JC, Levy D, Wilson PW, Benjamin EJ, Larson MG, Kannel WB, and Vasan
RS. Obesity and the risk of heart failure. N Engl J Med 347: 305-313, 2002.
Koves TR, Ussher JR, Noland RC, Slentz D, Mosedale M, Ilkayeva O, Bain J, Stevens R, Dyck
JR, Newgard CB, Lopaschuk GD, and Muoio DM. Mitochondrial overload and incomplete fatty acid
oxidation contribute to skeletal muscle insulin resistance. Cell Metab 7: 45-56, 2008.
Lamb HJ, Doornbos J, den Hollander JA, Luyten PR, Beyerbacht HP, van der Wall EE, and de
Roos A. Reproducibility of human cardiac 31P-NMR spectroscopy. NMR Biomed 9: 217-227, 1996.
Lebeche D, Davidoff AJ, and Hajjar RJ. Interplay between impaired calcium regulation and insulin
signaling abnormalities in diabetic cardiomyopathy. Nat Clin Pract Cardiovasc Med 5: 715-724, 2008.
Leopoldo AS, Sugizaki MM, Lima-Leopoldo AP, do Nascimento AF, Luvizotto Rde A, de
Campos DH, Okoshi K, Dal Pai-Silva M, Padovani CR, and Cicogna AC. Cardiac remodeling in a
rat model of diet-induced obesity. Can J Cardiol 26: 423-429, 2010.
Li YJ, Wang PH, Chen C, Zou MH, and Wang DW. Improvement of mechanical heart function by
trimetazidine in db/db mice. Acta Pharmacol Sin 31: 560-569, 2010.
Lopaschuk GD, Folmes CD, and Stanley WC. Cardiac energy metabolism in obesity. Circ Res 101:
335-347, 2007.
Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, and Stanley WC. Myocardial fatty acid
metabolism in health and disease. Physiol Rev 90: 207-258, 2010.
Lorenz JN, and Kranias EG. Regulatory effects of phospholamban on cardiac function in intact mice.
Am J Physiol 273: H2826-2831, 1997.
Louwe MC, van der Hoorn JW, van den Berg SA, Jukema JW, Romijn JA, Willems van Dijk K,
Rensen PC, Smit JW, and Steendijk P. Gender-dependent effects of high-fat lard diet on cardiac
function in C57Bl/6J mice. Appl Physiol Nutr Metab 37: 214-224, 2012.
Luo W, Wolska BM, Grupp IL, Harrer JM, Haghighi K, Ferguson DG, Slack JP, Grupp G,
Doetschman T, Solaro RJ, and Kranias EG. Phospholamban gene dosage effects in the mammalian
heart. Circ Res 78: 839-847, 1996.
MacLennan DH, and Kranias EG. Phospholamban: a crucial regulator of cardiac contractility. Nat
Rev Mol Cell Biol 4: 566-577, 2003.
Mandavia CH, Pulakat L, DeMarco V, and Sowers JR. Over-nutrition and metabolic
cardiomyopathy. Metabolism 61: 1205-1210, 2012.
Martos R, Baugh J, Ledwidge M, O'Loughlin C, Conlon C, Patle A, Donnelly SC, and McDonald
K. Diastolic heart failure: evidence of increased myocardial collagen turnover linked to diastolic
dysfunction. Circulation 115: 888-895, 2007.
Mazumder PK, O'Neill BT, Roberts MW, Buchanan J, Yun UJ, Cooksey RC, Boudina S, and
Abel ED. Impaired cardiac efficiency and increased fatty acid oxidation in insulin-resistant ob/ob
mouse hearts. Diabetes 53: 2366-2374, 2004.
Mildaziene V, Baniene R, Nauciene Z, Marcinkeviciute A, Morkuniene R, Borutaite V,
Kholodenko B, and Brown GC. Ca2+ stimulates both the respiratory and phosphorylation
subsystems in rat heart mitochondria. Biochem J 320 ( Pt 1): 329-334, 1996.
Murray AJ, Anderson RE, Watson GC, Radda GK, and Clarke K. Uncoupling proteins in human
heart. Lancet 364: 1786-1788, 2004.
Neubauer S. The failing heart--an engine out of fuel. N Engl J Med 356: 1140-1151, 2007.
Neubauer S, Krahe T, Schindler R, Horn M, Hillenbrand H, Entzeroth C, Mader H, Kromer EP,
Riegger GA, Lackner K, and et al. 31P magnetic resonance spectroscopy in dilated cardiomyopathy
and coronary artery disease. Altered cardiac high-energy phosphate metabolism in heart failure.
Circulation 86: 1810-1818, 1992.
Cardiac energetics and function in diet-induced obesity
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
Noack H, Kunz WS, and Augustin W. Evaluation of a procedure for the simultaneous determination
of oxidized and reduced pyridine nucleotides and adenylates in organic phenol extracts from
mitochondria. Anal Biochem 202: 162-165, 1992.
Oberst L, Zhao G, Park JT, Brugada R, Michael LH, Entman ML, Roberts R, and Marian AJ.
Dominant-negative effect of a mutant cardiac troponin T on cardiac structure and function in
transgenic mice. J Clin Invest 102: 1498-1505, 1998.
Ouwens DM, and Diamant M. Myocardial insulin action and the contribution of insulin resistance to
the pathogenesis of diabetic cardiomyopathy. Arch Physiol Biochem 113: 76-86, 2007.
Ouwens DM, Diamant M, Fodor M, Habets DD, Pelsers MM, El Hasnaoui M, Dang ZC, van den
Brom CE, Vlasblom R, Rietdijk A, Boer C, Coort SL, Glatz JF, and Luiken JJ. Cardiac contractile
dysfunction in insulin-resistant rats fed a high-fat diet is associated with elevated CD36-mediated fatty
acid uptake and esterification. Diabetologia 50: 1938-1948, 2007.
Park SY, Cho YR, Kim HJ, Higashimori T, Danton C, Lee MK, Dey A, Rothermel B, Kim YB,
Kalinowski A, Russell KS, and Kim JK. Unraveling the temporal pattern of diet-induced insulin
resistance in individual organs and cardiac dysfunction in C57BL/6 mice. Diabetes 54: 3530-3540,
2005.
Patti ME, and Corvera S. The role of mitochondria in the pathogenesis of type 2 diabetes. Endocr
Rev 31: 364-395, 2010.
Pellieux C, Montessuit C, Papageorgiou I, Pedrazzini T, and Lerch R. Differential effects of highfat diet on myocardial lipid metabolism in failing and nonfailing hearts with angiotensin II-mediated
cardiac remodeling in mice. Am J Physiol Heart Circ Physiol 302: H1795-1805, 2012.
Periasamy M, and Janssen PM. Molecular basis of diastolic dysfunction. Heart Fail Ciln 4: 13-21,
2008.
Rider OJ, Cox P, Tyler D, Clarke K, and Neubauer S. Myocardial substrate metabolism in obesity.
Int J Obes 37: 972-979, 2013.
Schaffer JE. Lipotoxicity: when tissues overeat. Curr Opin Lipidol 14: 281, 2003.
Srere PA, Brazil H, and Gonen L. The citrate condensing enzyme of pigeon breast muscle and moth
flight muscle. Acta chem scand 17: S129-S134, 1963.
Stefan D, Cesare FD, Andrasescu A, Popa E, Lazariev A, Vescovo E, Strbak O, Williams S,
Starcuk Z, Cabanas M, van Ormondt D, and Graveron-Demilly D. Quantitation of magnetic
resonance spectroscopy signals: the jMRUI software package. Meas Sci Technol 20: 104035, 2009.
Tsuji T, Mizushige K, Noma T, Murakami K, Ohmori K, Miyatake A, and Kohno M. Pioglitazone
improves left ventricular diastolic function and decreases collagen accumulation in prediabetic stage of
a type II diabetic rat. J Cardiovasc Pharmacol 38: 868, 2001.
Turdi S, Hu N, and Ren J. Tauroursodeoxycholic acid mitigates high fat diet-induced cardiomyocyte
contractile and intracellular Ca2+ anomalies. PLoS ONE 8: e63615, 2013.
Unger RH, and Orci L. Lipoapoptosis: its mechanism and its diseases. Biochim Biophys Acta 1585:
202-212, 2002.
Ussher JR, Koves TR, Jaswal JS, Zhang L, Ilkayeva O, Dyck JR, Muoio DM, and Lopaschuk GD.
Insulin-stimulated cardiac glucose oxidation is increased in high-fat diet-induced obese mice lacking
malonyl CoA decarboxylase. Diabetes 58: 1766-1775, 2009.
van den Broek NM, Ciapaite J, De Feyter HM, Houten SM, Wanders RJ, Jeneson JA, Nicolay K,
and Prompers JJ. Increased mitochondrial content rescues in vivo muscle oxidative capacity in longterm high-fat-diet-fed rats. FASEB J 24: 1354-1364, 2010.
van Heerebeek L, Franssen CP, Hamdani N, Verheugt FW, Somsen GA, and Paulus WJ.
Molecular and cellular basis for diastolic dysfunction. Curr Heart Fail Rep 9: 293-302, 2012.
Vander Heiden MG, Chandel NS, Schumacker PT, and Thompson CB. Bcl-xL prevents cell death
following growth factor withdrawal by facilitating mitochondrial ATP/ADP exchange. Mol Cell 3: 159167, 1999.
Wende AR, and Abel ED. Lipotoxicity in the heart. Biochim Biophys Acta 1801: 311-319, 2010.
Wright JJ, Kim J, Buchanan J, Boudina S, Sena S, Bakirtzi K, Ilkun O, Theobald HA, Cooksey
RC, Kandror KV, and Abel ED. Mechanisms for increased myocardial fatty acid utilization following
short-term high-fat feeding. Cardiovasc Res 82: 351-360, 2009.
Yan J, Young ME, Cui L, Lopaschuk GD, Liao R, and Tian R. Increased glucose uptake and
oxidation in mouse hearts prevent high fatty acid oxidation but cause cardiac dysfunction in dietinduced obesity. Circulation 119: 2818-2828, 2009.
Zhou YT, Grayburn P, Karim A, Shimabukuro M, Higa M, Baetens D, Orci L, and Unger RH.
Lipotoxic heart disease in obese rats: implications for human obesity. Proc Natl Acad Sci U S A 97:
1784-1789, 2000.
117
Chapter 8
Reduced Cpt1b sensitivity to malonyl-CoA is associated
with disturbed cardiac energy metabolism and reduced
cardiac function in a Cpt1bE3A mouse model
Desiree Abdurrachim1*, Michel van Weeghel2*, Miranda Nabben1, Riekelt H. Houtkooper2,
Rianne Nederlof4, Klaas Nicolay1, Johan Auwerx5, Christine Des Rosiers6, Coert J. Zuurbier4,
Ronald J. Wanders2,3, Jeanine J. Prompers1, Sander M. Houten2,3
1
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, The
Netherlands
2
Laboratory Genetic Metabolic Diseases, Departments of Clinical Chemistry, Academic Medical Center,
Amsterdam, The Netherlands
3
Department of Pediatrics, Emma Children’s Hospital, Academic Medical Center, Amsterdam, The
Netherlands
4
Laboratory of Intensive Care and Anesthesiology, Department of Anesthesiology, Academic Medical Center,
Amsterdam, The Netherlands
5
Laboratory of Integrative and Systems Physiology, Ecole Polytechnique Fédérale de Lausanne, Switzerland
6
Department of Nutrition, Université de Montréal, Montréal, Quebec, Canada
*both authors contributed equally
Chapter 8
Abstract
Increased cardiac fatty acid oxidation (FAO) in diabetes and obesity has been associated with
cardiac dysfunction. Carnitine palmitoyltransferase 1b (CPT1b) is a component of the carnitine
shuttle, which is essential for fatty acid (FA) import into mitochondria. CPT1b activity is inhibited
by malonyl-CoA and is considered rate-limiting in FAO. The control of malonyl-CoA on CPT1b
has recently emerged as a potential contributor to abnormal cardiac FAO; however, the
implication of reduced CPT1b sensitivity to malonyl-CoA on cardiac function and metabolic
homeostasis is currently not known. We generated a knockin mouse model expressing a
Cpt1bE3A mutant enzyme, which has reduced sensitivity to malonyl-CoA. Cardiac function and
metabolism of Cpt1bE3A mice were assessed in isolated cardiac mitochondria, in perfused hearts,
and in vivo using magnetic resonance imaging (MRI) and spectroscopy (MRS).
Increased FAO flux was observed in the perfused Cpt1bE3A/E3A hearts, demonstrating reduced
malonyl-CoA control of Cpt1b and that Cpt1b is rate limiting for FAO. The increased FA flux was
paralleled by decreased cardiac energy status, increased left ventricular mass, and reduced
cardiac function in Cpt1bE3A/E3A mice in vivo. In isolated heart mitochondria of Cpt1bE3A/E3A mice,
oxidative capacity was reduced for glucose-derived substrates, and to a greater extent, for FAderived substrates, suggesting a downregulation in oxidative phosphorylation and FA βoxidation pathways. Furthermore, reduced Cpt1 activity and Cpt1b protein expression were
observed in Cpt1bE3A/E3A hearts. This study demonstrates the importance of malonyl-CoA control
on Cpt1b for cardiac function and metabolic homeostasis. Reduced Cpt1b sensitivity to malonylCoA increases cardiac FAO, and is associated with impaired cardiac energetics and cardiac
dysfunction.
Keywords
Cpt1b, malonyl-CoA levels, fatty acid oxidation, cardiac energetics, cardiac function
120
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.1. Introduction
Alterations in cardiac metabolism have been implicated in the development of cardiomyopathy.
In the healthy heart, 60-70% of the energy is provided through fatty acid (FA) oxidation, while
the remaining 30-40% is contributed mostly by glucose oxidation (1, 5, 28, 45). In obesity and
diabetes, cardiac substrate preference shifts towards almost exclusive use of FAs. The increased
FAO, together with the loss of flexibility to switch between glucose and FA substrates, is
associated with a reduced cardiac function of the diabetic heart (14).
The reciprocal relationship between glucose oxidation and FAO is known as the glucose-fatty
acid, or Randle cycle. Several mechanisms contribute to the regulation of glucose oxidation and
FAO (38, 39). High levels of glucose and insulin promote the carboxylation of glucose oxidationderived acetyl-CoA by acetyl-CoA carboxylase-2 (ACC2) to form malonyl-CoA, a potent inhibitor
of carnitine palmitoyl transferase-1 (CPT1) (32). CPT1 performs the first step in the transport of
FA into mitochondria and its activity is considered a rate-limiting step in FAO (32). It has been
demonstrated in isolated rat hepatocytes that the activity of Cpt1a, the liver isoform of Cpt1,
strongly determines FAO flux (10). However, the control of Cpt1b, the heart and skeletal muscle
isoform of Cpt1, on FAO flux has not been determined, and moreover, has been questioned
recently. Compared with Cpt1a, Cpt1b has a higher affinity for malonyl-CoA, but the cardiac
malonyl-CoA concentration greatly exceeds the concentration of malonyl-CoA needed to inhibit
Cpt1b activity by 50% (IC50), indicating that Cpt1b activity would be totally repressed (12).
Furthermore, reducing Cpt1b activity by 44% using an irreversible inhibitor did not affect FAO
rates in isolated cardiomyocytes (30). There are, however, also strong arguments in favor of a
high control of Cpt1b. Acc2 knockout mice have lower malonyl-CoA levels, accompanied by an
increase in cardiac FAO, proving flux control at the level of Cpt1b (13, 25).
In diabetes and heart disease, the modulation of malonyl-CoA levels might play a potential role
in the abnormal (cardiac) FA metabolism (18). Low malonyl-CoA levels due to increased activity
and expression of malonyl-CoA decarboxylase (MCD) seem to contribute to the high FAO rates
in diabetes (41). Moreover, low malonyl-CoA levels have also been shown to be associated with
the increase in FAO during ischemia reperfusion (26). Apart from changes in malonyl-CoA levels,
a decrease in malonyl-CoA control of FAO has also been suggested to contribute to the
increased FAO rates observed in obesity and diabetes (18). In this case, Cpt1 sensitivity to
malonyl-CoA may be of particular importance. Indeed, it has been shown that Cpt1a sensitivity
to malonyl-CoA was reduced in diabetic liver (37). However, it is not known whether reduced
Cpt1b sensitivity to malonyl-CoA contributes to increased FAO in the heart.
To this end, we generated a novel Cpt1bE3A knockin mouse, based on the decreased malonylCoA sensitivity of the E3A mutation previously shown in Cpt1a (15, 43). Indeed, here we show
that malonyl-CoA sensitivity of Cpt1b was decreased in heart and skeletal muscle tissue of
Cpt1bE3A mice. We investigated 1) whether reduced Cpt1b sensitivity to malonyl-CoA increases
cardiac FAO flux, and more importantly, 2) whether the reduced control of malonyl-CoA on
Cpt1b activity has a direct impact on cardiac function and metabolic homeostasis.
121
Chapter 8
8.2. Materials and Methods
8.2.1. Materials
Human serum albumin (HSA), ʟ-carnitine, malonyl-CoA, bicinchoninic acid (BCA) were obtained
from Sigma-Aldrich. Complete mini protease inhibitor cocktail tablets were obtained from Roche.
The cell culture medium DMEM, nutrient mixture (25 mM HEPES + ʟ-glutamine) and the trypsinEDTA solution were acquired from Gibco. 1-butanol and acetylchloride were obtained from
Merck. Acetonitrile (ACN) gradient grade was obtained from Biosolve. The [ 2H3]-C3, [2H3]-C8
and [2H3]-C16 acylcarnitine internal standards were obtained from Dr. H.J. ten Brink (Vrije
University Medical Center, Amsterdam, The Netherlands).
8.2.2. Site-directed mutagenesis
Site-directed mutagenesis was performed using the Quickchange II-E Site-directed mutagenesis
kit from Stratagene. Primers for Cpt1a and Cpt1b E3A were designed using the Stratagene
website. Cpt1a was amplified from pGEM-T-Cpt1a plasmid using the following primers: a
forward primer containing the E3A mutation 5'-CACTCAAGATGGCAGCGGCTCACCAAGCTGT-3'
and a reverse primer 5'-ACAGCTTGGTGAGCCGCTGCCATCTTGAGTG-3'. Cpt1b was amplified
from pGEM-T-Cpt1b plasmid using the following primers: a forward primer containing the E3A
mutation
5'-CCAGGATGGCGGCAGCACACCAGGC-3'
and
a
reverse
primer
5'-
GCCTGGTGTGCTGCCGCCATCCTGG-3'.
8.2.3. Mouse Cpt1a and Cpt1b plasmid expression in S. cerevisiae
Cpt1a, Cpt1b, Cpt1aE3A and Cpt1bE3A were expressed in the Saccharomyces cerevisiae strain
BY4742 MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0 ΔCAT2 (ATCC number 4016728 from Invitrogen)
using the pYES vector. The cells were grown in glucose medium (20g/L glucose, 6.7 g/L yeast
nitrogen base (without amino acids), and amino acids (0.3 g/L Leu, 0.2 g/L Try, 0.2 g/L His, 0.3
g/L Lys) for 24h at 28°C in a gyro shaker, harvested by centrifugation and transferred to
galactose medium (20 g/L galactose, 6.7 g/L yeast nitrogen base, amino acids and 1 g/L yeast
extract). The cells from overnight cultures were harvested again by centrifugation and
protoplasts were prepared using the lytic enzyme zymolyase, as described by Spaan et al. (44).
The resulting protoplasts were centrifuged for 5 min at 700g and the obtained pellets were
stored at -80°C until usage for enzyme activity measurements. Protein concentration in the
homogenates was determined using BCA solution and HSA as standard.
8.2.4. Overexpression in HEK293 cells
Approximately 20.000 HEK293 cells were seeded in a 6 wells plate and cultured o/n in DMEM
with glutamine, 10% fetal bovine serum (Gibco, Carlsbad, CA, USA), and 1% mixture of
penicillin, streptomycin, fungizone (Gibco). Cells were incubated in a CO2 incubator (5% CO2) at
37°C. The cells were transfected using lipofectamine 2000 (Invitrogen) with 3.6 µg POG44 and
either 0.4 µg pcDNA5-EV, pcDNA5-Cpt1bWT or pcDNA5-Cpt1bE3A in optimem. After 5-6 hours,
the transfection medium was removed and replaced by fresh DMEM culture medium.
122
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.2.5. Cpt1b constitutive knockin generation
The Cpt1b constitutive knockin mouse (c.8_9delinsCT, coding effect Glu3Ala (E3A)) was
generated on a C57BL/6N background by Artemis Taconic (Cologne, Germany) (Fig. 1A-B). The
positive selection marker (Puromycin resistance) was flanked by F3 sites and inserted within
intron 7, into a region with minimal sequence conservation among species, in order to minimize
the risk of interfering with elements important for the regulation of transcription or splicing of
the Cpt1b gene. The targeting vector was generated using BAC clones from the C57BL/6J RPCI23 BAC library and transfected into Artemis Taconic C57BL/6N Tac ES cell lines. Homologous
recombinant
clones were
isolated
using
positive
(Puromycin
resistance)
and
negative
(Thymidine kinase) selections. The constitutive KI allele was left over after Flp-mediated
removal of the selection marker. This allele expressed the mutated Cpt1b E3A protein. The
remaining recombination site will be located in non-conserved region of the genome. The
Cpt1bE3A colony was maintained by crossing with C57BL/6N (Charles River) mice.
8.2.6. Animals
Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mice were generated via heterozygous breeding pairs.
The mice were housed at 21 ± 1 °C, 40-50% humidity, and a 12h light-dark cycle, with ad
libitum access to water and a standard rodent diet. Four different sets of mice were used for the
following experiments: (1) intraperitoneal glucose tolerance test (IPGTT) and biochemical tissue
analysis (Cpt1bWT/WT n=6, Cpt1bWT/E3A n=5, and Cpt1bE3A/E3A n=7) at 8-15 weeks of age, (2)
perfused heart setup (Cpt1bWT/WT n=5 and Cpt1bE3A/E3A n=5) at 25 weeks of age, (3) in vivo MRI
and MRS measurements (Cpt1bWT/WT n=10; homebred or from Charles River, and Cpt1bE3A/E3A
n=10) at 12-13 weeks of age, and (4) respiration measurement in isolated cardiac mitochondria
(Cpt1bWT/WT n=10 from Charles River, and Cpt1bE3A/E3A n=10) at 18 weeks of age. All
experiments other than MRI and MRS were approved by the institutional review board for
animal experiments at the Academic Medical Center, Amsterdam (The Netherlands) or were
carried out according to national Swiss and EU ethical guidelines and approved by the local
animal experimentation committee of the Canton de Vaud (Switzerland) under license #2469.
MRI and MRS experimental protocols were approved by the Animal Experiment Committee of
Maastricht University (The Netherlands).
8.2.7. IPGTT and refeeding
At 8 weeks of age, 50 µl of blood was collected from the vena saphena for the measurements of
glucose, acylcarnitines, and ketones in the fed state. Intraperitoneal glucose tolerance test
(IPGTT) was determined at 9 and 12 weeks of age. At 9 weeks of age, mice were fasted for 4h,
followed by an intraperitoneal (i.p.) injection of 2 g/kg glucose (20% D-glucose) at t=0 minutes.
Glucose level was measured via a tail cut at time points, t=0, 15, 30, 60, 90 and 120 minutes.
At 12 weeks of age, mice were fasted overnight, followed by an i.p. injection of 2g/kg glucose
(20% D-glucose) at T=0 minutes. Glucose level was measured via a tail cut at time points, t=0,
15, 30, 60, 90 and 120 minutes. At 15 weeks of age, mice were weighed, placed in a clean cage
without food but with access to water, and fasted overnight. This was followed by refeeding the
mice for 4h, where after, the mice were anesthetized with an intraperitoneal injection of 100
mg/kg pentobarbital. Anesthetized mice were euthanized by exsanguination from the vena cava
123
Chapter 8
inferior. The heart, liver, BAT, WAT, and muscles were rapidly excised, weighed, and processed
for biochemical analysis.
8.2.8. Acylcarnitine analysis
Cardiac acylcarnitine levels were determined in freeze-dried tissue specimens using tandem
mass spectrometry as described previously (48). In short, acylcarnitines were extracted using
80% acetonitrile (ACN), dried, propylated, dissolved in ACN, and stored at -20°C until analysis.
Semiquantitative determination of the formed acylcarnitines in the samples was done using
internal standards (650 µM d3-C0, 5 µM d3-C3, 2 µM d3-C6, 2 µM d3-C8, 2 µM d3-C10, and 2
µM d3-C16 acylcarnitine).
8.2.9. Immunoblotting
The polyclonal antibody against Cpt1b (HPA029583) was obtained from Sigma-Aldrich. The
polyclonal antibody against UCP1 was obtained from Merck Millipore (662045, USA). Secondary
antibodies goat anti-rabbit IrD cw800 were from Li-Cor Biosciences (Lincoln, NE, USA), and
immunoblot images were obtained using the Odyssey infrared imaging system (Li-Cor
Biosciences).
8.2.10. In vitro Cpt1 activity measurements
Mouse tissues were homogenized using the ultra-turrax T10 (IKA-Werke Germany) in PBS +
protease inhibitors, sonicated 3 x 10s (8W) and diluted to 1 mg/mL. The Cpt1 activity in these
tissues was measured in vitro after incubating homogenates (final protein concentration of 10
µg/mL) with 25 μM [U-13C]-palmitoyl-CoA (C16:0-CoA), 0.5 mM ʟ-carnitine, 1 mg/ml BSA, 150
mM KCL, 25 mM Tris (pH 7.4), 10 mM KPi (pH 7.4) and 2 mM EDTA at 37°C, for 10 min. The
activity of Cpt1 was evaluated in the presence or absence of its specific inhibitor malonyl-CoA
(0-256 µM) or after the addition of 0.1% triton X-100.
The protoplasts of the yeast strains expressing Cpt1 proteins (5 µg/mL final concentration
suspended in PBS + protease inhibitors) were sonicated 3 x 10s (8W) and used for Cpt1 activity
measurements to determine the IC50 for malonyl-CoA (0-1 mM) and the affinities for the
substrates, [U-13C]-C16:0-CoA (0-40 µM) and ʟ-carnitine (0-1.25 mM). The Km for carnitine and
the K0.5 for palmitoyl-CoA were determined by plotting reaction rate against concentration, and
using nonlinear regression of the Michaelis-Menten equation. The IC50 for malonyl-CoA was
determined using 1/V Dixon plot equation.
The Cpt1 activity was measured as the synthesis rate of [U-13C]-C16:0-carnitine, which was
analyzed and quantified by ESI-MS/MS (47). A mixture of acylcarnitines internal standard (50
pmol [2H3]-propionyl-carnitine, 50 pmol [2H3]-octanoyl-carnitine and 25 pmol [2H3]-palmitoylcarnitine) was added to each sample. The samples were analyzed in duplicate. The peak height
ratio of the formed [U-13C]-C16:0-carnitine to the peak height of the internal standard [2H3]palmitoyl-carnitine was determined using MassLynx NT software version 4.1 (Waters–Micromass,
Manchester, UK). This ratio was used to calculate the Cpt1 activity.
124
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.2.11. Perfused heart setup
Mice were heparinized (15 IU) and anesthetized with pentobarbital (80 mg/kg). After
tracheotomy, mice were mechanically ventilated. Heart cannulation was performed in situ and
perfusion was started before excision of the heart. Hearts were Langendorff perfused at a
constant flow (initial perfusion pressure 80 mmHg) at 37°C with Krebs-Henselheit solution
containing 118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, 0.5 mM
EDTA, 2.5 mM CaCl2, 5.5 mM glucose, 0.3 mM palmitate bound to 3% albumin, 0.5 mM
glutamine, 1.0 mM lactate, 0.1 mM pyruvate, 0.05 mM ʟ-carnitine, and 100 mU/L insulin gassed
with 95% O2/5% CO2. Hearts were perfused with either [U-13C6]-glucose or [U-13C16]-palmitate
(n=5 per group). A cannula was pierced through the apex to drain thebesian venous effluent
from the left ventricular lumen. End-diastolic pressure was set at ~2-6 mmHg using a waterfilled polyethylene balloon inserted in the left ventricle. After stabilization hearts were perfused
for 20 min after which venous oxygen tension was measured. The hearts were then freeze
clamped and stored at -80°C until further analysis. Cardiac effluent and influent was also
sampled and stored at -80°C until further analysis.
8.2.12. Determination of citric acid cycle intermediates
The isotope distribution of citric acid cycle intermediates was determined in heart tissue freeze
clamped after Langendorff perfusion. Heart tissue (~100 mg wet weight) was homogenized in 1
mL 8% (w/v) sulfosalicylic acid and 100 μL of internal standard (0.2 mM 2-phenylbutyric acid)
using an Ultra-Turrax and sonication or the Tissuelyser (Qiagen). Suspension was centrifuged
(10 minutes, 12000 g), after which the pH of the supernatant was adjusted to 7 by adding 30%
NaOH dropwise. The sample was then split in two. In half of the extract the keto groups were
directly oximated by addition of 50 μL 5M hydroxylamine hydrochloride and inducation at 60°C
for 30 minutes. The other half of the extract was first reduced by 50 μL 200 mM NaBH4, and
incubated at room temperature for 30 minutes. After inactivation and neutralization using
hydrochloric acid and KOH/MOPS, respectively, the sample was further treated with citrate lyase
and hydroxylamine, to convert citrate into oximated oxaloacetate (750 μL of 500 mM
triethylamine pH 7.4, 100 mM MgSO4, 50 mM EDTA, 10 μL 5M hydroxylamine hydrochloride and
2.5U citrate lyase). The mixture was sonicated and incubated for 5 minutes at 37°C after which
the sample was acidified by adding 100 μL saturated sulfosalicylic acid and centrifuged (10
minutes, 1200 g). Next, both samples were reacidified by adding 100 μL 37% HCl, and
saturated with salt by adding ~0.2 g NaCl. The organic acids were extracted twice by adding 3
mL ethylacetate. Both ethylacetate extracts were pooled and dried. The residue was derivatized
by addition of 75 μL N-tert-butyldimethylsilyl-N-methyltrifluoroacetamide, and incubation at
80°C for 60 minutes. Within 24 hours, the sample (1 μL) was analyzed by gas chromatographymass spectrometry (GC-MS). Peak intensities were integrated, corrected for natural abundance,
and used for the calculation of contribution of glucose and palmitate to acetyl-CoA formation for
citrate synthesis (7, 8).
8.2.13. Magnetic resonance imaging (MRI) and spectroscopy (MRS) setup
MRI and MRS measurements were performed on a 9.4 T horizontal bore MR scanner (Bruker,
Germany), during two separate sessions with 2-4 days interval between the two sessions. One
125
Chapter 8
session was for
1
H MRS and MRI, while the other was for
measurements assigned randomly. For
1
31
P MRS, with the order of the
H MRS and MRI, a 35-mm quadrature birdcage coil
(Bruker, Germany) was used for signal transmission and reception. For
1
31
P MRS, a 53.8-mm
31
double tuned H quadrature/ P linear birdcage coil (Rapid Biomedical, Germany) was used for
signal transmission and a 15-mm diameter home-built
31
P surface coil was used for signal
reception. During the MR measurements, the animals were anaesthetized through a customized
mask with 1-2% isoflurane in medical air at a flow rate of 0.4 L/min. Body temperature was
controlled at 36-37°C with a heating pad. Rectal temperature, ECG signal, and breathing rate
were monitored throughout the measurement period. All measurements were acquired using
cardiac-triggering and respiratory-gating.
8.2.13.1. Myocardial TG measurements using 1H MRS
1
H MRS was performed using the point resolved spectroscopy (PRESS) sequence, with chemical
shift selective (CHESS) water suppression, as described previously (3). Briefly, 1H MR spectra
were acquired from a 1x2x2 mm3 voxel positioned in the interventricular septum during the
diastolic phase of the cardiac cycle. The parameters were as follows: repetition time: ~2 s, echo
time: 9.1 ms, 0.41 ms 90° Hermite-shaped excitation pulse, 0.9 ms 180° Mao-type refocusing
pulses. The spectra were processed and analyzed using AMARES (49) in the jMRUI software
package (35, 46). The unsuppressed water peak was fitted to a Lorentzian line shape, and
metabolite peaks (taurine, choline/carnitine, creatine, and 7 peaks from triglycerides) were
fitted to Gaussian line shape. The metabolite levels were then calculated from the metabolite
signal relative to the unsuppressed water peak.
8.2.13.2. Cardiac energy status measurements using
31
31
P MRS
P MRS was performed in a voxel of typically ~6x6x6 mm3 covering the left ventricle, using the
image selected in vivo spectroscopy (ISIS) sequence, as described previously (2). The
parameters were as follows: repetition time: ~2 s, 1.2 ms 90º sinc-shaped excitation pulse
(bandwidth: 32.0 ppm), 6.25 ms 180º adiabatic hyperbolic secant inversion pulses (bandwidth:
37.5 ppm), 96 ISIS cycles (768 scans), γ-ATP on resonance. Phosphocreatine (PCr) and γ-ATP
peaks were fitted to Lorentzian line shapes with a linewidth (LW) constraint of LW-γ-ATP = LWPCr + 14.8 Hz, based on the intrinsic linewidth difference between γ-ATP and PCr as determined
from 60 cardiac spectra from our laboratory. As a measure of cardiac energy status, the ratio of
the PCr to γ-ATP peak (PCr/ATP) was determined and corrected for T1 partial saturation
(correction factor for PCr = 1.75 and for ATP = 1.31, determined from sixteen cardiac
31
P
spectra with repetition times of 2 and 15 s). The PCr/ATP ratio was not corrected for the
contribution of blood to the spectra, as this was shown to be negligible (4).
8.2.13.3. Cardiac function measurement using cinematic MRI
Cardiac movies (~15 frames/cardiac cycle) were acquired using gradient echo imaging in 5-6
contiguous short axis and 2 long axis slices with 1-mm thickness. The imaging parameters were
as follows: repetition time: 7 ms, echo time: 1.8 ms, flip angle: 15°, matrix: 192x192, field of
view: 30x30 mm2, frame resolution: 7 ms. To determine diastolic function parameters, cardiac
movies (~50-60 frames/cardiac cycle) of the mid-ventricular slice were acquired using
retrospectively triggered gradient echo imaging, allowing high temporal resolution, as described
previously (9). The parameters used were as follows: repetition time: 4.7 ms, echo time: 2.35
ms, flip angle: 15°, matrix: 128x128, field of view: 30x30 mm 2, frame resolution: 2 ms. The
126
Cardiac function and metabolism in Cpt1bE3A knockin mice
images were segmented semi-automatically using CAAS MRV 2.0 (Pie Medical, Maastricht, The
Netherlands) or Segment (version 1.8 R1145, http://segment.heiberg.se), after which the
systolic and diastolic functional parameters were calculated.
8.2.14. Isolation of cardiac mitochondria
Following excision, the heart was rinsed in ice cold 0.9% KCl, quickly minced with scissors in 0.5
ml of medium containing 160 mM KCl, 10 mM NaCl, 20 mM Tris, 5 mM EGTA and 0.05 mg/ml
bacterial proteinase type XXIV (pH 7.7), and incubated for ~1 min on ice. The mixture was then
homogenized using a Potter-Elvehjem homogenizer. Mitochondria were isolated through a
differential centrifugation procedure as described in (34) and and resuspended in buffer
containing 180 mM KCl, 20 mM Tris, 3 mM EGTA and 1 mg/ml bovine serum albumin (pH 7.4).
Protein content was determined using a BCA protein assay kit (Pierce, Thermo Fisher Scientific
Inc., Rockfort, IL, USA).
8.2.15. High-resolution respirometry
A 2-channel high-resolution Oroboros oxygraph-2k (Oroboros, Innsbruck, Austria) was used to
measure oxygen consumption rates at 30°C. Isolated mitochondria (0.15 mg/ml) were
incubated in 1 ml of assay medium (0.5 mM EGTA, 3 mM MgCl 2, 60 mM K-lactobionate, 20 mM
taurine, 10 mM KH2PO4, 20 mM HEPES, 110 mM sucrose, and 1 mg/ml bovine serum
albumin, pH 7.1 at 30°C) with i) 5 mM pyruvate plus 5mM malate, ii) 25 µM palmitoyl-CoA plus
2.5 mM ʟ-carnitine plus 2.5 mM malate, or iii) 25 µM palmitoyl-ʟ-carnitine plus 2.5 mM malate
as the oxidizable substrates. The maximal O2 consumption rate in a coupled state of oxidative
phosphorylation (state 3) was measured after addition of 2 mM ADP. The basal O2 consumption
rate (state 4) was measured after blocking ATP synthesis with 1.25 µM carboxyatractyloside
(CAT). The uncoupled O2 consumption rate (state U) was determined after addition of 1 µM
carbonyl cyanide 3-chlorophenyl hydrazone (CCCP). Data acquisition and analysis were
performed using Oxygraph-2k-Datlab 4.3.1.15 software (Oroboros, Innsbruck, Austria).
8.2.16. Statistical analysis
All data are presented as means ± standard deviation. Statistical analysis was performed using
GraphPad Prism 5 (GraphPad, La Jolla, CA, USA) or SPSS 17.0 (SPSS Inc). Differences were
evaluated using a 2-sided Student’s t-test or a one-way analysis of variance (ANOVA) with
Bonferroni-corrected post hoc tests. Statistical significance was set at P<0.05.
127
Chapter 8
Figure 8.1. Generation and characterization of Cpt1bE3A knockin mice. (A) Targeting strategy for the
generation of Cpt1bE3A knockin mice. Organization of the Cpt1b gene on chromosome 22. (B) Analysis of
the E3A mutation in cDNA isolated from Cpt1bWT/E3A hearts. Displayed is one representative
electropherogram of 3 Cpt1bWT/E3A mice analyzed. (C) Malonyl-CoA sensitivity of Cpt activity in heart,
quadriceps and liver of Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mice. White bars: Cpt1bWT/WT, grey bars:
Cpt1bWT/E3A, black bars: Cpt1bE3A/E3A. Data are mean ± SD (Cpt1bWT/WT n=6, Cpt1bWT/E3A n=5, Cpt1bE3A/E3A
n=7). *P<0.05, **P<0.01, ***P<0.001.
128
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.3. Results
8.3.1. Activity and kinetic parameters for WT and mutant E3A Cpt1a and Cpt1b
The Glu3Ala point mutation (E3A) is known to decrease malonyl-CoA sensitivity in rat Cpt1a
(43). In order to determine whether the E3A point mutation also affects malonyl-CoA sensitivity
of Cpt1b, we expressed mouse Cpt1a and Cpt1b and the E3A mutants for both isoforms in S.
cerevisiae and measured their kinetic properties in homogenates. We determined the IC 50
values for malonyl-CoA for the different enzymes (Table 8.1). Cpt1a and Cpt1b differ in their
malonyl-CoA sensitivity (IC50= 116 µM vs. 0.29 µM, respectively), which is already well
described (33, 40). When we compared mutant Cpt1E3A with their WT counterparts there was a
17-fold and 123-fold decrease in malonyl-CoA sensitivity for Cpt1a and Cpt1b, respectively. The
K0.5 and Km for palmitoyl-CoA and carnitine, respectively, as well as the specific activity were
unchanged in the Cpt1E3A mutants when compared to their WT counterparts (Table 8.1). Thus,
like for Cpt1aE3A, the Cpt1bE3A mutant has a decreased malonyl-CoA sensitivity and retains
normal Cpt1 activity and kinetics.
Table 8.1. Comparison of the kinetic parameters of recombinant Cpt1aWT, Cpt1bWT, Cpt1aE3A and
Cpt1bE3A expressed in S. cerevisiae.
Construct
Specific activity
IC50 malonyl-CoA
K0.5 palmitoyl-CoA
Km carnitine
(nmol/min∙mg protein)
(µM)
(µM)
(µM)
WT
0.76 ± 0.07
116
5.64 ± 1.70
45.09 ± 9.93
E3A
0.39
1957
5.91 ± 1.40
44.77 ± 9.31
WT
Cpt1b
0.59 ± 0.10
0.29
8.41 ± 1.78
6925 ± 2770
Cpt1bE3A
0.47 ± 0.02
41.53
5.59 ± 2.17
15783 ± 10716
Cpt1a
Cpt1a
8.3.2. Generation and characterization of the Cpt1bE3A mutation in the Cpt1bE3A mouse
The E3A mutation (c.8_9delinsCT) was introduced into the Cpt1b gene using homologous
recombination in C57BL/6N ES cells (Fig. 8.1A). Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mice
were born according to Mendelian ratios (28%, 47% and 25% (n=60)) and pups appeared
healthy. To verify equal expression of the Cpt1bWT and Cpt1bE3A allele, we sequenced Cpt1b
cDNA derived from Cpt1bWT/E3A mouse hearts, which revealed that both alleles are expressed
equally (Fig. 8.1B). To establish whether Cpt1bE3A mice exhibited decreased malonyl-CoA
sensitivity in heart and skeletal muscle, we performed Cpt1 activity measurements using a
range of malonyl-CoA concentrations in Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mouse tissue
extracts. Cpt activity in Cpt1bE3A/E3A mouse heart and quadriceps muscle had a pronounced
decrease in malonyl-CoA sensitivity (Fig. 8.1C). In heterozygous Cpt1bWT/E3A mice, Cpt activity
was only slightly less malonyl-CoA sensitive as compared to Cpt1bWT/WT mice, which indicates
that the E3A mutation is not dominant. As expected, the malonyl-CoA sensitivity of Cpt activity
in liver was not changed as this activity is catalyzed by Cpt1a (Fig. 8.1C). Thus, as for the
Cpt1bE3A mutation in yeast, we observed decreased malonyl-CoA sensitivity in the heart and
quadriceps muscle of the Cpt1bE3A/E3A mouse.
129
Chapter 8
8.3.3. Physiological characteristics of Cpt1bE3A mouse
To characterize the physiological effects of the Cpt1b E3A mutation, we compared plasma
metabolites of Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mice. Upon 4 hours refeeding after an
overnight fast, blood levels of glucose, pyruvate, lactate, ketone bodies, free fatty acids,
glycerol and triglycerides were not different between Cpt1b WT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A
mice (Fig. 8.2A). The plasma amino acid profile was also not different (data not shown).
Furthermore, glucose levels in fed and overnight fasted mice were also unchanged in
Cpt1bWT/E3A and Cpt1bE3A/E3A mice when compared to Cpt1bWT/WT mice (Fig. 8.2B, at time 0).
Next, we determined glucose tolerance in all three genotypes in the fed and fasted state. In
both conditions, the glucose tolerance was not affected in Cpt1b WT/E3A and Cpt1bE3A/E3A mice
when compared to Cpt1bWT/WT mice (Fig. 8.2B). Body weight was similar between all groups (Fig.
8.2C).
Figure 8.2. Physiological characteristics of Cpt1bE3A knockin mice. (A) Plasma levels of pyruvate,
lactate, hydroxybutyrate, free fatty acids, glycerol, trigycerides, and glucose determined at 4 hours
refeeding after overnight fasting. White bars: Cpt1bWT/WT, grey bars: Cpt1bWT/E3A, black bars:
Cpt1bE3A/E3A. (B) Glucose tolerance test under fed and fasted conditions, (C) body weight. Data are
mean ± SD (Cpt1bWT/WT n=6, Cpt1bWT/E3A n=5, Cpt1bE3A/E3A n=7). Black dotted lines: Cpt1bWT/WT, grey
solid lines: Cpt1bWT/E3A, black solid lines: Cpt1bE3A/E3A.
130
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.3.4. Cardiac acylcarnitine profile
We determined acylcarnitine profiles in the hearts of Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A
mice using LC-MS/MS analysis. Upon 4 hours refeeding after an overnight fast, cardiac free
carnitine was lower in Cpt1bE3A/E3A hearts compared with Cpt1bWT/WT hearts (P=0.041; Fig. 8.3).
In parallel, there was a trend for general genotype-dependent effect on long-chain acylcarnitine
levels (ANOVA P=0.059; Fig. 8.3).
Figure 8.3. Cardiac acylcarnitine profiles. Measured at 4 hours refeeding after overnight fasting.
There is a trend for a general genotype-dependent effect for long-chain acylcarnitines (P=0.059). Data
are mean ± SD (Cpt1bWT/WT n=6, Cpt1bWT/E3A n=5, Cpt1bE3A/E3A n=7). White bars: Cpt1bWT/WT, grey bars:
Cpt1bWT/E3A, black bars: Cpt1bE3A/E3A. *P<0.05.
8.3.5. Ex vivo cardiac substrate use
To determine whether the decreased malonyl-CoA sensitivity in Cpt1bE3A/E3A hearts results into
an increased FAO flux, we perfused hearts from Cpt1b WT/WT and Cpt1bE3A/E3A mice with media
containing either [U-13C]-glucose or [U-13C]-palmitate and measured the relative contribution of
glucose and palmitate to acetyl-CoA formation for citrate synthesis in the TCA cycle (Fig. 8.4A).
In Cpt1bE3A/E3A hearts, palmitate contributed to 63% of the acetyl-CoA formation, compared with
only 24% in Cpt1bWT/WT hearts (P<0.001). The contribution of glucose to the acetyl-CoA
formation was similar in Cpt1bE3A/E3A and Cpt1bWT/WT hearts (11% vs. 15%; P=0.53).
Furthermore, the production of pyruvate and lactate from exogenous glucose, indicating
cytosolic glycolysis, was also similar in Cpt1b E3A/E3A and Cpt1bWT/WT hearts (P=0.66; Fig. 8.4B).
The unchanged mRNA expression levels of Glut4 and Pdk4 supported the unaffected glucose
131
Chapter 8
metabolism (P=0.57 and P=0.26, respectively; Fig. 8.4C and Fig. 8.4D). Furthermore, the
mRNA expression of Pgc1α was similar between groups (P=0.80, Fig. 8.4E).
Figure 8.4. Ex vivo cardiac substrate use. (A) Relative contribution of exogenous FA, glucose, and
other sources to the production of acetyl-CoA for citrate synthesis in the TCA cycle, and (B) lactate
and pyruvate production from exogenous glucose, as measured in perfused heart setup (Cpt1bWT/WT
n=3-5, Cpt1bE3A/E3A n=5 per labeled substrates). Relative mRNA expression of (C) Glut4, (D) Pdk4, and
(E) Pgc1α (n=5 per group). Data are mean ± SD. White bars: Cpt1b WT/WT, grey bars: Cpt1bWT/E3A,
black bars: Cpt1bE3A/E3A. ***P<0.001.
8.3.6. Ex vivo cardiac function
Cardiac functional parameters were calculated from the perfused heart experiments (Table 8.2).
Left-ventricular end-diastolic pressure (LVEDP) and developed left-ventricular pressure (DLVP)
were similar between Cpt1bE3A/E3A and Cpt1bWT/WT hearts. However, heart rates were higher in
Cpt1bE3A/E3A than in Cpt1bWT/WT hearts, which resulted in higher flow, rate pressure product
(RPP), and cardiac oxygen consumption (Table 8.2).
8.3.7. In vivo myocardial lipid content
Myocardial lipid content was measured in vivo in the myocardial septum using 1H MRS (Fig.
8.5A). In vivo myocardial lipid levels, as determined from all lipid-associated peaks in the 1H MR
spectra (peaks 1; 5-10 in Fig. 8.5B), were similar between Cpt1bE3A/E3A and Cpt1bWT/WT mice (Fig.
8.5C).
132
Cardiac function and metabolism in Cpt1bE3A knockin mice
Table 8.2. Cardiac functional parameters of Cpt1bWT/WT and Cpt1bE3A/E3A mice as measured ex vivo using
perfused heart setups.
Parameters
Cpt1bWT/WT
Cpt1bE3A/E3A
P
Flow (ml/min)
2.0 ± 0.4
2.9 ± 1.0
0.14
Left-ventricular end-diastolic pressure (LVEDP; mmHg)
2.3 ± 0.7
2.3 ± 2.9
0.950
Developed left-ventricular pressure (DLVP; mmHg)
96.3 ± 16
98.4 ± 17
0.78
Heart rate (HR; beats/min)
320 ± 30
409 ± 31
<0.001
30616 ± 4899
40352 ± 8462
0.006
1.05 ± 0.2
1.28 ± 1.28
0.095
Rate pressure product (RPP; mmHg∙beats/min)
Cardiac oxygen consumption (μmol/min)
RPP = HR x RPP. Data are mean ± SD (Cpt1b
WT/WT
n=11, Cpt1b
E3A/E3A
n=10).
Figure 8.5. Myocardial metabolites content measured using 1H MRS. (A) Voxel localization (in white)
for 1H MRS shown in short axis and 4-chamber long axis views, (B) representative cardiac 1H MR
spectrum, (C) concentration of metabolites as percentage of the water signal measured in the same
voxel (peak 1: -CH=CH-, peak 2: taurine, peak 3: choline/carnitine, peak 4: creatine, peak 5:
-CH=CH-CH2-CH=CH-, peak 6: -CαH2-COO-, peak 7: -CH2-CH=CH-CH2-, peak 8: -CβH2-CH2-COO-,
peak 9: (-CH2-)n, peak 10: CH3). Peak 1 and peaks 5-10 are from lipids. Data are mean ± SD
(Cpt1bWT/WT n=9, Cpt1bE3A/E3A n=8). White bars: Cpt1bWT/WT, black bars: Cpt1bE3A/E3A.
8.3.8. In vivo cardiac energy status
We employed
31
P MRS to assess the in vivo cardiac energy status. A typical cardiac
31
P MR
spectrum contains a phosphocreatine (PCr) peak originating from myocardium, a 2,3diphosphoglycerate (DPG) peak from blood in the left ventricular lumen (which obscures the
inorganic phosphate (Pi) peak from myocardium), and γ-ATP, α-ATP, β-ATP peaks from mainly
myocardium with a minor contamination from blood (Fig. 8.6A and Fig. 8.6B). The in vivo PCr to
ATP ratio (PCr/ATP ratio), which is a measure of in vivo cardiac energy status, was 23% lower in
Cpt1bE3A/E3A mice compared with Cpt1bWT/WT mice (P<0.05, Fig. 8.6C).
133
Chapter 8
Figure 8.6. Cardiac energy status measured using 31P MRS. (A) Voxel localization (in white)
for 31P MRS shown in short axis, 2-chamber long axis, and 4-chamber long axis views of the
heart, (B) representative cardiac 31P MR spectrum, (C) PCr/ATP ratio. DPG: diphosphoglycerate, Pi: inorganic phosphate, PCr: phosphocreatine, ATP: adenosine triphosphate. Data
are mean ± SD (Cpt1bWT/WT n=10, Cpt1bE3A/E3A n=9). White bars: Cpt1bWT/WT, black bars:
Cpt1bE3A/E3A. *P<0.05.
Table 8.3. Cardiac function of Cpt1bWT/WT and Cpt1bE3A/E3A mice as measured in vivo using MRI.
Parameter
Heart rate (beats/minute)
Cpt1bE3A/E3A
P
550 ± 63
542 ± 55
0.839
Ejection fraction (%)
67.8 ± 4.7
65.0 ± 3.7
0.155
End-diastolic volume (EDV; μL)
60.7 ± 5.7
68.5 ± 10.5
0.058
End-systolic volume (ESV; μL)
19.6 ± 3.8
24.1 ± 5.2
0.040
Stroke volume (SV; μL)
41.1 ± 4.3
44.4 ± 6.5
0.205
Peak ejection rate (PER; %EDV/ms)
2.3 ± 0.4
1.9 ± 0.3
0.048
Early peak filling rate (E; %EDV/ms)
2.4 ± 0.3
2.0 ± 0.3
0.017
Late peak filling rate (A; %EDV/ms)
1.2 ± 0.4
0.9 ± 0.2
0.228
Left ventricular mass (LV mass; mg)
94.6 ± 7.7
104.2 ± 10.7
0.034
3.4 ± 0.2
3.7 ± 0.3
0.048
LV mass/body weight (mg/g)
Data are means ± SD (n=8-10 per group).
134
Cpt1bWT/WT
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.3.9. In vivo cardiac function
In Cpt1bE3A/E3A mice, end-diastolic volume and end-systolic volume were both increased when
compared with Cpt1bWT/WT mice, which was paralleled by ~10% increase in LV mass and LV
mass/body weight ratio (Table 8.3). In addition to the enlarged heart, peak ejection rate (PER)
and early peak filling rate (E) were ~17% lower in Cpt1bE3A/E3A mice compared with Cpt1bWT/WT
mice, indicating some degree of both systolic and diastolic dysfunction. Ejection fraction, stroke
volume, and late peak filling rate (A) were similar between Cpt1b WT/WT and Cpt1bE3A/E3A mice
(Table 8.3).
Figure 8.7. Oxygen consumption rates in isolated heart mitochondria. State 3, state 4, and state U
respiration when (A) pyruvate plus malate, or (B) palmitoyl-CoA plus carnitine plus malate, or (C)
palmitoyl-carnitine plus malate were used as substrates. The ratio of oxygen consumption rates of (D)
palmitoyl-CoA/pyruvate, (E) palmitoyl-carnitine/pyruvate, and (F) palmitoyl-CoA/palmitoyl-carnitine
(Cpt1bWT/WT n=9, Cpt1bE3A/E3A n=8). Data are mean ± SD. White bars: Cpt1bWT/WT, grey bar: Cpt1bWT/E3A,
black bars: Cpt1bE3A/E3A. *P<0.05, **P<0.01, ***P<0.001.
135
Chapter 8
8.3.10. Respiration measurements in isolated heart mitochondria
Despite the marked increase in FAO flux in perfused hearts, interestingly, overall oxidative
capacity for both glucose and FA substrates was reduced in isolated Cpt1b E3A/E3A heart
mitochondria (Fig. 8.7). When pyruvate plus malate were used as substrates, oxygen
consumption rates in the coupled state (i.e. when the electron transfer chain is coupled with
ATP production; state 3) were ~20% lower in Cpt1b E3A/E3A than in Cpt1bWT/WT isolated heart
mitochondria (P=0.017; Fig. 8.7A). The maximal consumption rates (i.e. in the uncoupled state;
state U) were also ~20% lower in Cpt1bE3A/E3A than in Cpt1bWT/WT isolated heart mitochondria
(P=0.011; Fig. 8.7A). For palmitoyl-CoA plus carnitine plus malate, state 3 and state U
respiration were ~30-35% lower in Cpt1bE3A/E3A than in Cpt1bWT/WT isolated heart mitochondria
(P<0.001; Fig. 8.7B). Similarly, when palmitoyl-carnitine plus malate were used as substrates,
oxygen consumption rates in state 3 and state U were reduced by ~40% in Cpt1b E3A/E3A
compared with Cpt1bWT/WT heart mitochondria (P<0.001; Fig. 8.7C). Even when normalized to
state 3 and state U respiration for pyruvate plus malate, state 3 and state U respiration for
these fatty acid substrates were lower in Cpt1bE3A/E3A compared with Cpt1bWT/WT heart
mitochondria (P<0.001; Fig. 8.7D and Fig. 8.7E). State 3 and state U respiration for palmitoylCoA when normalized to palmitoyl-carnitine were ~1, and similar between Cpt1b WT/WT and
Cpt1bE3A/E3A heart mitochondria (P=0.30 and P=0.63, respectively; Fig. 8.7F), indicating that
Cpt1b activity is not limiting respiration in Cpt1bE3A/E3A heart mitochondria.
State 4 respiration was similar in Cpt1bE3A/E3A and Cpt1bWT/WT heart mitochondria for pyruvate
plus malate (Fig. 8.7A). For palmitoyl-CoA plus carnitine plus malate, and palmitoyl-carnitine
plus malate, state 4 respiration was ~25% lower in Cpt1b E3A/E3A than in Cpt1bWT/WT heart
mitochondria (P=0.039 and P=0.032, respectively; Fig. 8.7B and Fig. 8.7C). However, these
differences disappeared when normalized to the state 4 respiration for pyruvate plus malate (Fig.
8.7D and Fig. 8.7E).
8.3.11. Cpt1 activity is decreased in mutant Cpt1bE3A/E3A mice due to lower Cpt1b
protein content
In our Cpt1 activity assays, we noted an unexpected decrease in Cpt1 activity in the hearts and
quadriceps muscle of Cpt1bE3A/E3A mice (34% and 67% lower vs. Cpt1bWT/WT, P<0.001; Fig. 8.8A
and Fig. 8.8B). This decrease in Cpt1 activity in Cpt1b E3A/E3A mice appeared to be due to lower
Cpt1b protein levels (P<0.001 and P=0.002, respectively, vs. Cpt1bWT/WT; Fig. 8.8D and Fig.
8.8E), which is not explained by lower gene expression of mutant Cpt1b mRNA (P=0.73; Fig.
8.8F). These data show that Cpt1 activity and Cpt1b protein expression are decreased in parallel
with the increased malonyl-CoA insensitivity.
136
Cardiac function and metabolism in Cpt1bE3A knockin mice
Figure 8.8. Cpt activity and Cpt1b expression. (A) heart, (B) quadriceps, and (C) liver (in
pmol/min.mg) of Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mice (Cpt1bWT/WT n=6, Cpt1bWT/E3A n=5,
Cpt1bE3A/E3A n=7). Cpt2 activity is represented by the Cpt activity in the presence of triton X-100. Cpt1
activity is calculated by subtracting the Cpt2 activity from the total Cpt activity. Cpt1b protein
expression in (D) heart and (E) quadriceps of Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mice (n=3 per
group). (F) Relative mRNA expression levels of Cpt1bWT/WT, Cpt1bWT/E3A, and Cpt1bE3A/E3A mouse hearts
(n=5 per group). Protein expression is corrected for tubulin. Relative mRNA expression is corrected for
36B4. Data are means ± SD. White bars: Cpt1bWT/WT, grey bars: Cpt1bWT/E3A, black bars: Cpt1bE3A/E3A.
*P<0.05, **P<0.01, ***P<0.001.
8.3.12. The expression of mutant Cpt1bE3A is temperature sensitive
We investigated whether the decreased protein expression of Cpt1b E3A/E3A in heart and skeletal
muscle of our knockin model could be caused by decreased stability of the mutant protein. To
address this, we overexpressed mouse Cpt1bWT and Cpt1bE3A in yeast cells and grew them at
30°C and 37°C (Fig. 8.9A and Fig. 8.9B). We observed no differences in expression of Cpt1b WT
and Cpt1bE3A in yeast cells grown at 30°C and 37°C suggesting that the E3A Cpt1b protein is
equally stable when compared to the WT enzyme. However, because yeast cells differ from
mammalian cells, we performed a similar experiment in HEK293 cells again employing
conditions that are known to promote stability and folding of proteins. We compared expression
at 30°C or in medium supplemented with 5% glycerol with cells cultured under regular
conditions. In HEK293 cells grown at 30°C there was no difference in protein expression if we
137
Chapter 8
compared the Cpt1bWT with Cpt1bE3A (Fig. 8.9C). However, in HEK293 cells grown at 37°C, a
lower Cpt1bE3A protein level was observed when compared with Cpt1b WT (Fig. 8.9D).
Furthermore, expression of Cpt1bE3A was also increased when cells were incubated with 5%
glycerol (Fig. 8.9E). Overall, these data show that Cpt1bE3A protein is less stable than Cpt1bWT
protein in HEK293 cells, likely caused by a posttranscriptional mechanisms involving protein
folding.
Figure 8.9. Cpt1bE3A protein expression is temperature sensitive. Measurement of Cpt1b activity in
yeast grown at (A) 30°C and (B) 37°C overexpressing Cpt1bWT or Cpt1bE3A, respectively (n=2-3 per
group). Overexpression of EV, Cpt1bWT or Cpt1bE3A in HEK293 cells grown at (C) 30°C and (D) 37°C,
respectively, and (E) Overexpression of EV, Cpt1bWT or Cpt1bE3A in HEK293 cells grown with or without
the addition of 5% glycerol (n=3 per group). Protein expression is corrected for β-actin. Data are
mean ± SD. ***P<0.001.
138
Cardiac function and metabolism in Cpt1bE3A knockin mice
8.4. Discussion
Increased cardiac FAO in diabetes and obesity has been associated with cardiac dysfunction.
Based on the inhibitory properties of malonyl-CoA on Cpt1, Cpt1 is predicted to control FAO flux.
However, it is under debate whether the skeletal and cardiac muscle isoform of Cpt1, Cpt1b,
has a strong control over cardiac FAO flux. In this study, we generated a Cpt1bE3A knockin
mouse, in which Cpt1b has a reduced sensitivity to malonyl-CoA. We determined whether
reduced Cpt1b sensitivity to malonyl-CoA increases cardiac FAO flux, and more importantly,
whether the reduced control of malonyl-CoA on Cpt1b activity has a direct impact on cardiac
function and metabolic homeostasis.
There are two important findings in the present study. First, our data demonstrate the control of
Cpt1b over cardiac FAO flux. The mutant Cpt1bE3A/E3A had a decreased sensitivity to malonylCoA, and in a perfused heart setup, we observed a 2.6 fold increase in the contribution of FA to
acetyl-CoA formation in Cpt1bE3A/E3A hearts compared with Cpt1bWT/WT hearts. This shows that
reduced Cpt1b sensitivity to malonyl-CoA increases FAO flux and supports the hypothesis that
Cpt1b is rate limiting for FAO. Second, we show that the control of malonyl-CoA over Cpt1b is
crucial for metabolic homeostasis and maintenance of cardiac function. In Cpt1b E3A/E3A mice, the
increase in FAO was paralleled by an increased left ventricular mass and reduced in vivo cardiac
function. Cardiac PCr/ATP ratio, which is a measure of cardiac energy status in vivo, was
decreased, together with reduced ex vivo oxidative capacity of cardiac mitochondria in
Cpt1bE3A/E3A mice.
The Cpt1bE3A knockin mouse model is a novel mouse model for increased cardiac FAO, which is
based on reduced sensitivity of Cpt1b to malonyl-CoA. Other mouse models of increased or
decreased cardiac FAO are available, but these mouse models are based on the modulation of
either malonyl-CoA levels (i.e. Acc2-/- or Mcd-/- knockout mice) or Cpt1b levels (i.e. Cpt1b
deficient Cpt1b+/- mice; Cpt1b-/- mice are embryonically lethal) (11, 22, 24, 25). Increased FAO
in acetyl-CoA carboxylase2 knockout (Acc2-/-) mice is associated with reduced levels of malonylCoA due to the deletion of cardiac-specific Acc2 (25). Malonyl-CoA decarboxylase knockout
(Mcd-/-) mice have increased malonyl-CoA levels and stimulated glucose oxidation (via Randle
cycle upon inhibition of FAO) after ischemia reperfusion (11). Interestingly, modifying cardiac
metabolism in the Acc2-/-, Mcd-/-, or Cpt1b+/- mice did not alter cardiac function at rest (11, 22,
25), while modification of cardiac metabolism in the Cpt1bE3A/E3A mice resulted in cardiac
dysfunction. Furthermore, cardiac PCr/ATP ratio was maintained in the perfused hearts of Acc2-/mice (25), while the cardiac PCr/ATP ratio was reduced in vivo in the Cpt1bE3A/E3A mice,
indicating disturbed cardiac energy metabolism in the latter. Together, these data suggest that
not only the malonyl-CoA levels, but also the control of malonyl-CoA over Cpt1b is indeed
crucial to maintain cardiac metabolic homeostasis.
The lower cardiac PCr/ATP ratio in the Cpt1bE3A/E3A mice might be related with lower cardiac
efficiency (i.e. the amount of ATP produced per oxygen consumed), which is known to be
associated with increased FAO (17). FAO requires more oxygen to produce the same amount of
ATP as produced by glucose oxidation. Indeed, increased FAO flux in the perfused Cpt1bE3A/E3A
hearts was paralleled by a tendency of increased myocardial oxygen consumption. Furthermore,
the rate pressure product (RPP), which is also a good indicator of myocardial oxygen
139
Chapter 8
consumption (6, 19), was significantly higher in the perfused Cpt1b E3A/E3A hearts compared with
Cpt1bWT/WT hearts. In obese ob/ob and diabetic db/db mice, it has been shown that increased
myocardial oxygen consumption reduces cardiac efficiency (6, 23). With regard to the higher
heart rates of the perfused Cpt1bE3A/E3A hearts compared with the Cpt1bWT/WT hearts, this may
be related to the higher FAO flux in the Cpt1bE3A/E3A hearts. In ex vivo perfused heart setups,
increased FAO flux has been shown to be associated with increased anaplerosis, which was
associated with higher heart rates (27). The lower cardiac PCr/ATP ratio in the Cpt1bE3A/E3A
hearts might also be explained by the lower oxidative capacity measured in isolated heart
mitochondria. We speculate that this could result in decreased total cardiac substrate oxidation,
depriving the Cpt1bE3A/E3A heart of energy. As ATP is essential for cardiac work and ion
homeostasis, energy deprivation may lead to the reduced cardiac function observed in vivo in
the Cpt1bE3A/E3A mice. Indeed, the association between lower PCr/ATP ratio and reduced cardiac
function has previously been shown in heart failure both in mice (20, 21) and in patients (36).
In the Cpt1bE3A/E3A mouse hearts, the increased contribution of exogenous FA to substrate
oxidation was not accompanied by reduced contribution of exogenous glucose. Furthermore, the
production of lactate and pyruvate from exogenous glucose, which is an indicator of glycolytic
flux, was also unaffected. In a situation where total oxidative phosphorylation capacity is
reduced, such as in an advanced stage of heart failure or in myocardial ischemia, the heart
usually compensates for the reduction in energy production by increasing glycolysis (16). Even
though cardiac energetics was lower in the Cpt1b E3A/E3A mice, our data show that this does not
seem to induce an increase in glycolysis to provide more energy. The lack of increase in
glycolysis may not seem beneficial for the heart; however, this protects the hearts from
damaging accumulation of protons and lactate, which could occur when increased glycolysis is
not coupled with increased glucose oxidation (16, 29). In a very recent study, Mcd-/- mice were
shown to be able to increase the coupling between glycolysis and glucose oxidation during
myocardial ischemia, which reduces the degree of cardiac ischemia/reperfusion damage (31).
Respiration measurements in isolated cardiac mitochondria of the Cpt1bE3A/E3A mice showed
reduced mitochondrial oxidative capacity for both glucose- and FA-derived substrates, which
might suggest a compensatory mechanism to suppress the increased FAO in the Cpt1bE3A/E3A
mice. This attempt to lower the increased FAO is further supported by the observation that the
reduced ex vivo mitochondrial oxidative capacity in Cpt1bE3A/E3A mice was not compensated by a
higher number of mitochondria as the mRNA expression of Pgc1α, a marker of mitochondrial
biogenesis, was similar in Cpt1bWT/WT and Cpt1bE3A/E3A mice. On the other hand, rather than a
compensatory mechanism for the increased FAO flux, the reduced mitochondrial oxidative
capacity in the Cpt1bE3A/E3A mice could actually be the result of mitochondrial dysfunction
induced by the increased flux of FAs into the mitochondria. As mitochondrial oxidative capacity
of Cpt1bE3A/E3A mice was lower for both glucose- and FA-derived substrates, the defect in
mitochondria may be located in the TCA cycle or oxidative phosphorylation pathway. However, a
defect in the TCA cycle is unlikely to be the cause because acetylcarnitine levels were similar
between Cpt1bE3A/E3A and Cpt1bWT/WT mice. Limited TCA cycle capacity could result in excess
acetyl-CoA which cannot be oxidized by the TCA cycle, and would therefore be converted back
to acetylcarnitine (42). Thus, we may expect some defects in the OXPHOS complexes instead.
Uncoupling between ETC and ATP production might not be responsible for the lower
mitochondrial capacity, because state 3 and state U respiration were similarly suppressed in
140
Cardiac function and metabolism in Cpt1bE3A knockin mice
isolated Cpt1bE3A/E3A heart mitochondria. Furthermore, as mitochondrial oxidative capacity was
suppressed even more for the FA-derived substrates than for the glucose-derived substrate, one
may expect a downregulation in the FAO pathway. We observed reduced Cpt1 activity and
Cpt1b protein in the Cpt1bE3A/E3A mice, which may therefore explain the downregulation in the
FAO pathway. However, we unexpectedly observed similar degree of reduced respiration in the
isolated heart mitochondria of Cpt1bE3A/E3A mice for palmitoyl-CoA, a Cpt1b-dependent substrate,
and for palmitoyl-carnitine, a Cpt1b-independent substrate. These data suggest that instead of
Cpt1b, proteins downstream of Cpt1b may be responsible for the reduced capacity for FAO.
The lower Cpt1 activity and Cpt1b protein might demonstrate an adaptation at the level of
Cpt1b in the Cpt1bE3A/E3A mice, probably as a mechanism to cope with the increase in FAO flux.
The lower protein expression of Cpt1b in the heart was not due to lower expression of the Cpt1b
gene, suggesting that protein stability of Cpt1b E3A is affected. Indeed, overexpression of
Cpt1bE3A protein in HEK293 under conditions that promote protein stability i.e. culturing at 30°C
and in the presence of 5% glycerol, yielded more protein and suggests that stability and folding
of Cpt1bE3A is affected. Furthermore, a similar experiment in yeast did not show affected Cpt1
activity, suggesting that besides protein folding, other posttranscriptional mechanisms may also
be involved in the regulation of Cpt1b protein expression. It has been shown that
posttranslational protein modification mechanisms could indeed alter catalytic activity of Cpt1
(50).
In conclusion, our study with Cpt1bE3A/E3A mice demonstrates the importance of malonyl-CoA
control on Cpt1b for cardiac function and metabolic homeostasis. Reduced Cpt1b sensitivity to
malonyl-CoA increases cardiac FAO, and is associated with impaired cardiac energetics and
cardiac dysfunction.
8.5. Acknowledgements
D.A. and J.J.P. are supported by a VIDI grant (project number 700.58.421) from the
Netherlands Organisation for Scientific Research (NWO). R.H.H. is supported by a VENI grant
from ZonMw (number 91613050) and an AMC postdoctoral fellowship. M.N. is supported by a
VENI grant (project number 916.14.050) from NWO. This work was supported by the Dutch
Technology Foundation STW, Applied Science Division of NWO and the Technology Program of
the Ministry of Economic Affairs (grant number 10191). The authors thank the employees of the
Animal Research Institute Amsterdam and Central Animal Facilities Maastricht for technical
assistance.
8.6. References
1.
2.
3.
4.
Avogaro A, Nosadini R, Doria A, Fioretto P, Velussi M, Vigorito C, Sacca L, Toffolo G, Cobelli C,
Trevisan R, and et al. Myocardial metabolism in insulin-deficient diabetic humans without coronary
artery disease. Am J Physiol 258: E606-618, 1990.
Bakermans AJ, Dodd MS, Nicolay K, Prompers JJ, Tyler DJ, and Houten SM. Myocardial energy
shortage and unmet anaplerotic needs in the fasted long-chain acyl-CoA dehydrogenase knockout
mouse. Cardiovasc Res 100: 441-449, 2013.
Bakermans AJ, Geraedts TR, van Weeghel M, Denis S, Joao Ferraz M, Aerts JM, Aten J,
Nicolay K, Houten SM, and Prompers JJ. Fasting-induced myocardial lipid accumulation in longchain acyl-CoA dehydrogenase knockout mice is accompanied by impaired left ventricular function. Circ
Cardiovasc Imaging 4: 558-565, 2011.
Bakermans AJ, van Nierop BJ, Abdurrachim D, Nicolay K, and Prompers JJ. In vivo cardiac 31PMRS in a mouse model of heart failure In: Proc 20th Ann Meeting ISMRM p. 1115, 2012.
141
Chapter 8
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
142
Belke DD, Larsen TS, Gibbs EM, and Severson DL. Altered metabolism causes cardiac dysfunction
in perfused hearts from diabetic (db/db) mice. Am J Physiol Endocrinol Metab 279: E1104-1113, 2000.
Boudina S, Sena S, O'Neill BT, Tathireddy P, Young ME, and Abel ED. Reduced mitochondrial
oxidative capacity and increased mitochondrial uncoupling impair myocardial energetics in obesity.
Circulation 112: 2686-2695, 2005.
Comte B, Vincent G, Bouchard B, and Des Rosiers C. Probing the origin of acetyl-CoA and
oxaloacetate entering the citric acid cycle from the 13C labeling of citrate released by perfused rat
hearts. J Biol Chem 272: 26117-26124, 1997.
Comte B, Vincent G, Bouchard B, Jette M, Cordeau S, and Rosiers CD. A 13C mass isotopomer
study of anaplerotic pyruvate carboxylation in perfused rat hearts. J Biol Chem 272: 26125-26131,
1997.
Coolen BF, Abdurrachim D, Motaal AG, Nicolay K, Prompers JJ, and Strijkers GJ. High frame
rate retrospectively triggered Cine MRI for assessment of murine diastolic function. Magn Reson Med
69: 648-656, 2013.
Drynan L, Quant PA, and Zammit VA. Flux control exerted by mitochondrial outer membrane
carnitine palmitoyltransferase over beta-oxidation, ketogenesis and tricarboxylic acid cycle activity in
hepatocytes isolated from rats in different metabolic states. Biochem J 317 ( Pt 3): 791-795, 1996.
Dyck JR, Hopkins TA, Bonnet S, Michelakis ED, Young ME, Watanabe M, Kawase Y, Jishage K,
and Lopaschuk GD. Absence of malonyl coenzyme A decarboxylase in mice increases cardiac glucose
oxidation and protects the heart from ischemic injury. Circulation 114: 1721-1728, 2006.
Eaton S, Bartlett K, and Quant PA. Carnitine palmitoyl transferase I and the control of betaoxidation in heart mitochondria. Biochem Biophys Res Commun 285: 537-539, 2001.
Essop MF, Camp HS, Choi CS, Sharma S, Fryer RM, Reinhart GA, Guthrie PH, Bentebibel A, Gu
Z, Shulman GI, Taegtmeyer H, Wakil SJ, and Abu-Elheiga L. Reduced heart size and increased
myocardial fuel substrate oxidation in ACC2 mutant mice. Am J Physiol Heart Circ Physiol 295: H256265, 2008.
Fang ZY, Prins JB, and Marwick TH. Diabetic cardiomyopathy: evidence, mechanisms, and
therapeutic implications. Endocr Rev 25: 543-567, 2004.
Faye A, Borthwick K, Esnous C, Price NT, Gobin S, Jackson VN, Zammit VA, Girard J, and
Prip-Buus C. Demonstration of N- and C-terminal domain intramolecular interactions in rat liver
carnitine palmitoyltransferase 1 that determine its degree of malonyl-CoA sensitivity. Biochem J 387:
67-76, 2005.
Fillmore N, and Lopaschuk GD. Targeting mitochondrial oxidative metabolism as an approach to
treat heart failure. Biochim Biophys Acta 1833: 857-865, 2013.
Fillmore N, Mori J, and Lopaschuk GD. Mitochondrial fatty acid oxidation alterations in heart failure,
ischemic heart disease, and diabetic cardiomyopathy. Br J Pharmacol 2013.
Folmes CD, and Lopaschuk GD. Role of malonyl-CoA in heart disease and the hypothalamic control
of obesity. Cardiovasc Res 73: 278-287, 2007.
Gobel FL, Norstrom LA, Nelson RR, Jorgensen CR, and Wang Y. The rate-pressure product as an
index of myocardial oxygen consumption during exercise in patients with angina pectoris. Circulation
57: 549-556, 1978.
Gupta A, Chacko VP, Schar M, Akki A, and Weiss RG. Impaired ATP kinetics in failing in vivo
mouse heart. Circ Cardiovasc Imaging 4: 42-50, 2011.
Gupta A, Chacko VP, and Weiss RG. Abnormal energetics and ATP depletion in pressure-overload
mouse hearts: in vivo high-energy phosphate concentration measures by noninvasive magnetic
resonance. Am J Physiol Heart Circ Physiol 297: H59-64, 2009.
He L, Kim T, Long Q, Liu J, Wang P, Zhou Y, Ding Y, Prasain J, Wood PA, and Yang Q. Carnitine
palmitoyltransferase-1b deficiency aggravates pressure overload-induced cardiac hypertrophy caused
by lipotoxicity. Circulation 126: 1705-1716, 2012.
How OJ, Aasum E, Severson DL, Chan WY, Essop MF, and Larsen TS. Increased myocardial
oxygen consumption reduces cardiac efficiency in diabetic mice. Diabetes 55: 466-473, 2006.
Ji S, You Y, Kerner J, Hoppel CL, Schoeb TR, Chick WS, Hamm DA, Sharer JD, and Wood PA.
Homozygous carnitine palmitoyltransferase 1b (muscle isoform) deficiency is lethal in the mouse. Mol
Genet Metab 93: 314-322, 2008.
Kolwicz SC, Jr., Olson DP, Marney LC, Garcia-Menendez L, Synovec RE, and Tian R. Cardiacspecific deletion of acetyl CoA carboxylase 2 prevents metabolic remodeling during pressure-overload
hypertrophy. Circ Res 111: 728-738, 2012.
Kudo N, Barr AJ, Barr RL, Desai S, and Lopaschuk GD. High rates of fatty acid oxidation during
reperfusion of ischemic hearts are associated with a decrease in malonyl-CoA levels due to an increase
in 5'-AMP-activated protein kinase inhibition of acetyl-CoA carboxylase. J Biol Chem 270: 17513-17520,
1995.
Lauzier B, Vaillant F, Merlen C, Gelinas R, Bouchard B, Rivard ME, Labarthe F, Dolinsky VW,
Dyck JR, Allen BG, Chatham JC, and Des Rosiers C. Metabolic effects of glutamine on the heart:
anaplerosis versus the hexosamine biosynthetic pathway. J Mol Cell Cardiol 55: 92-100, 2013.
Lopaschuk GD. Metabolic abnormalities in the diabetic heart. Heart Fail Rev 7: 149-159, 2002.
Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, and Stanley WC. Myocardial fatty acid
metabolism in health and disease. Physiol Rev 90: 207-258, 2010.
Luiken JJ, Niessen HE, Coort SL, Hoebers N, Coumans WA, Schwenk RW, Bonen A, and Glatz
JF. Etomoxir-induced partial carnitine palmitoyltransferase-I (CPT-I) inhibition in vivo does not alter
cardiac long-chain fatty acid uptake and oxidation rates. Biochem J 419: 447-455, 2009.
Cardiac function and metabolism in Cpt1bE3A knockin mice
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
Masoud WG, Ussher JR, Wang W, Jaswal JS, Wagg CS, Dyck JR, Lygate CA, Neubauer S,
Clanachan AS, and Lopaschuk GD. Failing mouse hearts utilize energy inefficiently and benefit from
improved coupling of glycolysis and glucose oxidation. Cardiovasc Res 101: 30-38, 2014.
McGarry JD, Mannaerts GP, and Foster DW. A possible role for malonyl-CoA in the regulation of
hepatic fatty acid oxidation and ketogenesis. J Clin Invest 60: 265-270, 1977.
McGarry JD, Mills SE, Long CS, and Foster DW. Observations on the affinity for carnitine, and
malonyl-CoA sensitivity, of carnitine palmitoyltransferase I in animal and human tissues.
Demonstration of the presence of malonyl-CoA in non-hepatic tissues of the rat. Biochem J 214: 21-28,
1983.
Mildaziene V, Baniene R, Nauciene Z, Marcinkeviciute A, Morkuniene R, Borutaite V,
Kholodenko B, and Brown GC. Ca2+ stimulates both the respiratory and phosphorylation subsystems
in rat heart mitochondria. Biochem J 320 ( Pt 1): 329-334, 1996.
Naressi A, Couturier C, Devos JM, Janssen M, Mangeat C, de Beer R, and Graveron-Demilly D.
Java-based graphical user interface for the MRUI quantitation package. MAGMA 12: 141-152, 2001.
Neubauer S. The failing heart--an engine out of fuel. N Engl J Med 356: 1140-1151, 2007.
Park EA, Mynatt RL, Cook GA, and Kashfi K. Insulin regulates enzyme activity, malonyl-CoA
sensitivity and mRNA abundance of hepatic carnitine palmitoyltransferase-I. Biochem J 310 ( Pt 3):
853-858, 1995.
Randle PJ. Regulatory interactions between lipids and carbohydrates: the glucose fatty acid cycle
after 35 years. Diabetes Metab Rev 14: 263-283, 1998.
Randle PJ, Garland PB, Hales CN, and Newsholme EA. The glucose fatty-acid cycle. Its role in
insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1: 785-789, 1963.
Saggerson ED. Carnitine acyltransferase activities in rat liver and heart measured with palmitoyl-CoA
and octanoyl-CoA. Latency, effects of K+, bivalent metal ions and malonyl-CoA. Biochem J 202: 397405, 1982.
Sakamoto J, Barr RL, Kavanagh KM, and Lopaschuk GD. Contribution of malonyl-CoA
decarboxylase to the high fatty acid oxidation rates seen in the diabetic heart. Am J Physiol Heart Circ
Physiol 278: H1196-1204, 2000.
Schroeder MA, Atherton HJ, Dodd MS, Lee P, Cochlin LE, Radda GK, Clarke K, and Tyler DJ.
The cycling of acetyl-coenzyme A through acetylcarnitine buffers cardiac substrate supply: a
hyperpolarized 13C magnetic resonance study. Circ Cardiovasc Imaging 5: 201-209, 2012.
Shi J, Zhu H, Arvidson DN, and Woldegiorgis G. A single amino acid change (substitution of
glutamate 3 with alanine) in the N-terminal region of rat liver carnitine palmitoyltransferase I abolishes
malonyl-CoA inhibition and high affinity binding. J Biol Chem 274: 9421-9426, 1999.
Spaan AN, Ijlst L, van Roermund CW, Wijburg FA, Wanders RJ, and Waterham HR.
Identification of the human mitochondrial FAD transporter and its potential role in multiple acyl-CoA
dehydrogenase deficiency. Mol Genet Metab 86: 441-447, 2005.
Stanley WC, Lopaschuk GD, and McCormack JG. Regulation of energy substrate metabolism in the
diabetic heart. Cardiovasc Res 34: 25-33, 1997.
Stefan D, Cesare FD, Andrasescu A, Popa E, Lazariev A, Vescovo E, Strbak O, Williams S,
Starcuk Z, Cabanas M, van Ormondt D, and Graveron-Demilly D. Quantitation of magnetic
resonance spectroscopy signals: the jMRUI software package. Meas Sci Technol 20: 104035, 2009.
van Vlies N, Ruiter JP, Doolaard M, Wanders RJ, and Vaz FM. An improved enzyme assay for
carnitine palmitoyl transferase I in fibroblasts using tandem mass spectrometry. Mol Genet Metab 90:
24-29, 2007.
van Vlies N, Tian L, Overmars H, Bootsma AH, Kulik W, Wanders RJ, Wood PA, and Vaz FM.
Characterization of carnitine and fatty acid metabolism in the long-chain acyl-CoA dehydrogenasedeficient mouse. Biochem J 387: 185-193, 2005.
Vanhamme L, van den Boogaart A, and Van Huffel S. Improved method for accurate and efficient
quantification of MRS data with use of prior knowledge. J Magn Reson 129: 35-43, 1997.
Zammit VA. Carnitine palmitoyltransferase 1: central to cell function. IUBMB Life 60: 347-354, 2008.
143
Chapter 9
Cardiac metabolic adaptations in diabetic mice prevent
the heart from failing upon pressure overload, as
evidenced by a combined in vivo PET, MRI, and MRS
approach
Desiree Abdurrachim1*, Miranda Nabben1*, Verena Hoerr2, Michael T. Kuhlmann3,
Philipp Bovenkamp2, Michael Schäfers3, Klaas Nicolay1, Cornelius Faber2, Sven Hermann3,
Jeanine J. Prompers1
1
Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven,
The Netherlands
2
Department of Clinical Radiology, University Hospital of Münster, Münster, Germany
3
European Institute for Molecular Imaging-EIMI, Münster, Germany
*both authors contributed equally
Chapter 9
Abstract
Heart failure has been associated with altered myocardial substrate metabolism and impaired
cardiac energetics. However, it is currently not clear whether these alterations are cause or
consequence of heart failure progression. Furthermore, comorbidities such as diabetes may
influence the metabolic adaptations during heart failure. In this study, we aim to quantify the
cardiac metabolic changes during different stages of heart failure progression in non-diabetic
and diabetic mice in a longitudinal in vivo study design.
Transverse aortic constriction (TAC) surgery was performed in non-diabetic db/+ and diabetic
db/db mice to induce pressure overload heart failure. Magnetic resonance imaging (MRI),
1
magnetic resonance spectroscopy (MRS), H MRS, and
18
31
P
F-fluorodeoxyglucose-positron emission
tomography (18F-FDG-PET) were applied to measure cardiac function, energy status, lipid
content, and glucose uptake, respectively, at baseline and 1, 5, and 12 weeks post TAC. At 12
weeks post TAC, ex vivo mitochondrial function was determined in isolated heart mitochondria.
In non-diabetic mice, TAC induced progressive left ventricular (LV)
hypertrophy and
dysfunction, which correlated with myocardial FDG uptake. Myocardial FDG uptake was
increased at 1 and 5 weeks post TAC, but tended to decrease again at 12 weeks post TAC,
which was associated with a trend for lowered cardiac energy status. Interestingly, whereas
diabetic mice showed overall lower myocardial FDG uptake, lower cardiac energy status, and
higher myocardial lipid accumulation than non-diabetic mice at baseline, cardiac function of
diabetic mice was only mildly affected by TAC. Also in diabetic mice, myocardial FDG uptake
increased upon TAC, but it remained lower than in non-diabetic mice. Ex vivo mitochondrial
oxidative capacity for fatty acid substrates was lower in hearts of non-diabetic mice compared
with diabetic mice and was more severely reduced upon TAC in non-diabetic animals.
This study presents the first longitudinal in vivo data of cardiac metabolic and functional
adaptations during heart failure development in non-diabetic and diabetic mice. The progression
to heart failure was correlated with increased myocardial FDG uptake, which preceded the
decrease in cardiac energy status. The mild cardiac hypertrophy and dysfunction in diabetic
mice, together with lower myocardial glucose uptake upon TAC, suggests that maintaining fatty
acid oxidation may be beneficial for cardiac function and energetics in pressure overloadinduced heart failure.
Keywords
heart failure, metabolic adaptations, myocardial glucose uptake, cardiac energetics,
transverse aortic constriction, pressure overload
146
Cardiac metabolic adaptations in heart failure
9.1. Introduction
Heart failure is one of the leading causes of death (WHO, Sept 2012), with 1 in 9 deaths
attributable to heart failure in the US population (16). In 2010, heart failure was prevalent in an
estimated 5.1 million Americans over 20 years of age (16). It is projected that by 2030, the
number of heart failure patients in the US will increase by 46%, to more than 8 million patients
(24). Altered cardiac energy metabolism is proposed to be an important contributor to the
development of heart failure. The heart has sustained high energy demands in order to pump
sufficient blood through the body. In order to obtain this energy, the healthy adult heart
primarily relies on the oxidation of fatty acids (FA; 60-90%) and to a lesser extent glucose (1040%) (52). It is generally accepted that during the transition from cardiac hypertrophy to
failure, the heart shifts its substrate preference towards glucose, which is usually paralleled by a
concomitant suppression of FA oxidation (FAO) (3, 11, 27, 43, 44). The switch from FAO to
glucose utilization thus seems to precede the onset of cardiac dysfunction (61). Energetically,
the failing heart is characterized by an imbalance between energy supply and demand (21).
More importantly, a decreased cardiac energy status, i.e. a lower phosphocreatine-to-ATP ratio
(PCr/ATP), has been identified as a strong predictor of mortality in heart failure patients (40). It
is currently not clear, however, whether the metabolic switch towards glucose utilization is a
cause or a consequence of the reduced cardiac energetic status in heart failure.
The diabetic heart almost exclusively (90-100%) relies on FAO and is suggested to have lost the
flexibility to switch from FAO to glucose oxidation (42). In addition to these metabolic
adaptations, it has been shown that the diabetic heart has a decreased energy status (49) and
an impaired function (15). Furthermore, patients with diabetes have a 2-3 times higher risk for
cardiovascular diseases as compared to healthy subjects (26). These observations suggest that
increased FAO is potentially detrimental for the heart and that the shift from FAO towards
glucose oxidation upon heart failure may be an adaptive mechanism. Interestingly, however,
recent studies show that increased FAO can also be beneficial for restoring function in heart
failure. Upon pressure overload in acetyl-CoA carboxylase 2 (ACC2) knockout mice, high rates
of cardiac FAO were maintained, which was associated with protection against the development
of heart failure (27). Furthermore, maintained FAO without a switch to glucose oxidation was
shown to delay heart failure progression in dogs with ventricular pacing-induced heart failure
(29). This shows that an increased reliance on FAO is not necessarily detrimental, and suggests
that the development of heart failure in diabetes patients has other underlying causes. Possibly,
the excessive accumulation of myocardial lipid (intermediates) as found in diabetes causes
lipotoxicity-induced fibrosis and apoptosis, and hence alters myocardial function (10, 62).
Therapeutic interventions for heart failure treatment generally focus on the inhibition of FAO,
while this might thus not be the best approach. It is therefore important to investigate to what
extent changes in cardiac substrate metabolism and flexibility, lipid accumulation, and energy
status predict the longitudinal development of cardiac hypertrophy and subsequent failure. To
this end, an in vivo longitudinal study setting is crucial to investigate the time course of changes
in cardiac metabolism and energetics, in relation to the decline in cardiac function. Current
evidence on the role of cardiac metabolism in heart failure is largely based on ex vivo studies,
which do not allow such longitudinal observations. Furthermore, ex vivo setups might not be
147
Chapter 9
representative for the in vivo situation, where alterations in substrate availability, e.g. during
heart failure development or in diabetic conditions, affect substrate choice.
Here we aim to quantify the cardiac metabolic changes during different stages of heart failure
progression in non-diabetic and diabetic mice in a longitudinal in vivo study design. Heart failure
was induced by transverse aortic constriction (TAC) surgery, and non-invasive magnetic
resonance imaging (MRI),
31
P magnetic resonance spectroscopy (MRS),
1
H MRS, and
18
F-
18
fluorodeoxyglucose-positron emission tomography ( F-FDG-PET) were applied to measure
cardiac function, energy status, lipid content, and glucose uptake, respectively, at baseline and
1, 5, and 12 weeks post TAC.
9.2. Materials and methods
9.2.1. Animals
Male non-diabetic (db/+) and diabetic (db/db) C57BL/Ks mice were purchased from Charles
River (Elsene, Belgium). The animals were housed under controlled temperature (23°C) and
humidity (50%) with a 12:12-h dark-light cycle and had ad libitum access to food (5K52
LabDiet; 22 kcal% protein, 16 kcal% fat, 62 kcal% carbohydrates) and water. Three different
groups of animals were used: (1) for MRI and
31
P MRS in Eindhoven, The Netherlands (n=8 for
non-diabetic mice, n=10 for diabetic mice); (2) for MRI, 1H MRS, and PET in Münster, Germany
(n=12 for non-diabetic mice, n=14 for diabetic mice); (3) for baseline mitochondrial respiration
measurement (n=6 for non-diabetic mice, n=6 for diabetic mice). At 10 weeks of age, the
animals underwent baseline PET and/or MR measurements. At 11 weeks of age, the animals
underwent TAC surgery (needle size 27 G) to induce pressure overload heart failure, as
described previously (59). To follow the progression of heart failure, the PET and/or MR
measurements were then repeated at 1, 5, and 12 weeks post TAC surgery. The animals were
sacrificed after the measurements at 12 weeks post TAC. The heart was excised and part of it
was used immediately for isolation of mitochondria. Animal handling procedures and
experimental protocols conformed to and were approved by the Animal Experimental
Committees of Maastricht University (The Netherlands) and the University of Münster
(Germany).
9.2.2. MR measurements
MR measurements were performed on 9.4 T horizontal bore MR scanners (Bruker, Germany).
MRI and 1H MRS were performed using a 35-mm quadrature birdcage coil (Bruker, Germany or
Rapid Biomedical, Germany) for both signal reception and transmission. For
1
double tuned quadrature H and linear
31
31
P MRS, a 54-mm
P birdcage coil (Rapid Biomedical, Germany) was used
for signal transmission, while a 15-mm diameter, home-built, actively decoupled, two-turn
31
P
surface coil was used for signal reception. Before the experiments, the animals were sedated in
a chamber with 3% isoflurane in medical air at a flow rate of 0.4 L/min and were injected with
150 μL of saline solution to keep the animals hydrated. During the measurements, the
anesthesia was maintained at 1-2% isoflurane through a customized anesthesia mask.
Temperature was maintained at 36-37°C with a heating pad. Rectal temperature, ECG signal,
148
Cardiac metabolic adaptations in heart failure
and breathing rate were monitored throughout the measurements. All measurements were
performed with prospective or retrospective respiratory gating and cardiac triggering.
9.2.2.1. MR measurements: Cine MRI to measure cardiac function
Cardiac systolic and diastolic function were measured as described previously (12). For systolic
function measurement, cine movies from the beating heart (15-18 frames/cardiac cycle) were
acquired using prospectively cardiac-triggered gradient echo imaging of 5-6 contiguous short
axis and 2 long axis slices (slice thickness: 1 mm). The imaging parameters were as follows:
repetition time: 7 ms, echo time: 1.8 ms, flip angle: 15°, matrix: 192x192, field of view: 30x30
mm2, NA: 6. For diastolic function, cardiac movies of only the mid-ventricular slice were
acquired using retrospectively-triggered gradient echo imaging with Gaussian weighted
sampling for the phase encoding. The parameters used were as follows: repetition time: 4.7 ms,
echo time: 2.35 ms, flip angle: 15°, matrix: 128x128, field of view: 30x30 mm 2, NA: 250. Data
was reconstructed using compressed sensing (38), resulting in cine movies with an effective
time resolution of 2 ms (50-70 frames/cardiac cycle). Image segmentation and data analysis
were performed using CAAS MRV 2.0 (Pie Medical, Maastricht, The Netherlands) or Segment
(version 1.8 R1145, http://segment.heiberg.se).
9.2.2.2. MR measurements: Flow through right and left carotid arteries
To determine the severity of the TAC surgery, blood flow through the right carotid artery (RCA)
and the left carotid artery (LCA) were measured using a 2D FLASH sequence with a four-point
Hadamard encoding scheme for flow velocity encoding. The slice was positioned above the
bifurcation of the right subclavian artery, containing (semi) cross-sectional areas of the RCA and
LCA (Fig. 9.1A). The imaging parameters were as follows: repetition time: 15 ms, echo time:
2.6 ms, flip angle: 30°, matrix: 256x256, field of view: 30x30 mm2, slice thickness: 1 mm, NA:
10. Total scan time was 30-40 minutes. Data reconstruction was performed using MATLAB (The
Mathworks, Inc., Natick, USA). Magnitude images and phase images were extracted from the
complex image data. The magnitude image was used to filter the phase images and cutoff the
statistical phase noise in the low signal regions. Velocity vectors were constructed from the
three velocity components in three directions. The peak velocity was calculated from the length
of the velocity vectors inside the vessels.
9.2.2.3. MR measurements:
Cardiac
31
31
P MRS to measure cardiac energy status
P MR spectra were acquired using the image selected in vivo spectroscopy (ISIS)
sequence in a voxel of typically ~6x6x6 mm3 covering the left ventricle, at the end of the
diastolic phase, as described previously (5). The parameters were as follows: repetition time:
~2 s, 1.2 ms 90º sinc-shaped excitation pulse (bandwidth: 32.0 ppm), 6.25 ms 180º adiabatic
hyperbolic secant inversion pulses (bandwidth: 37.5 ppm), 96 ISIS cycles (768 scans), γ-ATP
on resonance. The 90º sinc-shaped excitation pulse was calibrated during the in vivo scan, by
performing a series of single pulse measurements with varying pulse power on a 5-mm
diameter glass phantom containing 15 M phosphoric acid, which was positioned underneath the
surface coil. Data fitting and analysis was performed using AMARES in jMRUI (54), as described
in (5). The γ-ATP line widths (LWγ-ATP) were constrained relative to the PCr line width (LWPCr)
according to an empirically determined relation: LWγ-ATP = LWPCr + 14.8 Hz (n=63 data sets,
R=0.78, P<0.001). As a measure of cardiac energy status, the ratio of PCr to γ-ATP was
determined. The PCr/ATP ratio was corrected for partial T1 saturation using correction factors of
149
Chapter 9
1.75 and 1.31 for PCr and γ-ATP, respectively, which were determined by analyzing ISIS
spectra acquired at repetition times of 2 s and 15 s (n=19 data sets from the current and
previous studies). The contamination of ATP from the blood to the spectra was shown to be less
than 4%, and was considered negligible (6).
9.2.2.4. MR measurements: 1H MRS to measure myocardial lipid content
1
H MR spectra were acquired from a 1x2x2 mm3 voxel positioned in the interventricular septum
during the diastolic phase of the cardiac cycle, using point resolved spectroscopy (PRESS) with
chemical shift selective (CHESS) water suppression, as described in (4). The parameters were
as follows: repetition time: ~2 s, echo time: 9.1 ms, 0.41 ms 90° Hermite-shaped excitation
pulse, 0.9 ms 180° Mao-type refocusing pulses, 256 averages. The spectra were processed and
analyzed using AMARES in the jMRUI software package (54). All metabolites (taurine,
choline/carnitine, creatine, and seven peaks from lipids) were fitted to Gaussian line shapes.
Myocardial metabolite levels were then calculated from the metabolite signal relative to the
unsuppressed water signal measured from the same voxel.
9.2.3.
18
F-fluorodeoxyglucose (FDG)-PET imaging to measure myocardial glucose
uptake
FDG-PET measurements were performed using a small-animal quadHIDAC PET (Oxford Positron
System, Oxford, UK). The measurements were performed after a 12-13 hour overnight fast. The
animals were anaesthetized using isoflurane. A tail vein catheter (26 G) was then inserted, from
which a drop of blood was collected to measure plasma glucose levels. Subsequently, FDG (~10
MBq in 100 μL) was injected and flushed with 150 μL of physiological saline. The acquisition was
performed one hour after the FDG injection, for 15 minutes. The animals were kept on a
warming plate throughout the procedure. Data was reconstructed into an image volume of
110×60×20 mm3 and a voxel size of 0.4×0.4×0.4 mm3, using a resolution recovery
reconstruction algorithm (46) leading to an effective resolution of 0.7 mm (full width at half
maximum) (46, 48). Quantification of segmental tracer uptake and volumes of the left ventricle
were performed using an automated 3D contour detection algorithm developed in-house and
validated against MRI (55).
9.2.4. Determination of fasting plasma glucose
At baseline, and at 1, 5, and 12 weeks post TAC, blood samples were collected after a 4 hour
fast. Fasting plasma glucose levels were measured immediately after blood collection in one
drop of blood using a Glucose-201 glucose meter (HemoCue, Ängelholm, Sweden).
9.2.5. Isolation of cardiac mitochondria
Following excision, the heart was rinsed in ice cold 0.9% KCl, quickly minced with scissors in 0.5
ml of medium containing 160 mM KCl, 10 mM NaCl, 20 mM Tris, 5 mM EGTA and 0.05 mg/ml
bacterial proteinase type XXIV (pH 7.7), and incubated for ~1 min on ice. The mixture was then
homogenized using a Potter-Elvehjem homogenizer. Mitochondria were isolated through a
differential centrifugation procedure as described in (37) and resuspended in buffer containing
180 mM KCl, 20 mM Tris, 3 mM EGTA and 1 mg/ml bovine serum albumin (pH 7.4). Protein
150
Cardiac metabolic adaptations in heart failure
content was determined using a BCA protein assay kit (Pierce, Thermo Fisher Scientific Inc.,
Rockfort, IL, USA).
9.2.6. High-resolution respirometry
A 2-channel high-resolution Oroboros oxygraph-2k (Oroboros, Innsbruck, Austria) was used to
measure oxygen consumption rates at 30°C. Isolated mitochondria (0.15 mg/ml) were
incubated in 1 ml of assay medium (0.5 mM EGTA, 3 mM MgCl 2, 60 mM K-lactobionate, 20 mM
taurine, 10 mM KH2PO4, 20 mM HEPES, 110 mM sucrose, and 1 mg/ml bovine serum
albumin, pH 7.1 at 30°C) with i) 5 mM pyruvate plus 5 mM malate or ii) 25 µM palmitoyl-CoA
plus 2.5 mM ʟ-carnitine plus 2.5 mM malate. The maximal O2 consumption rate in a coupled
state of oxidative phosphorylation (state 3) was measured after addition of 2 mM ADP. The
basal O2 consumption rate (state 4) was measured after blocking ATP synthesis with 1.25 µM
carboxyatractyloside (CAT). The uncoupled O2 consumption rate (state U) was determined after
addition of 1 µM carbonyl cyanide 3-chlorophenyl hydrazone (CCCP). Data acquisition and
analysis were performed using Oxygraph-2k-Datlab 4.3.1.15 software (Oroboros, Innsbruck,
Austria).
9.2.7. Statistics
Data are presented as means ± SD. Statistical significance of genotype (non-diabetic db/+ and
diabetic db/db mice) and time (baseline, 1, 5, and 12 weeks post TAC) effects were assessed by
applying a two-way Analysis of Variance (ANOVA) in the IBM SPSS 21.0 statistical package
(SPSS Inc., Chicago, IL, USA). In case of a significant effect of time, Bonferroni corrected posthoc tests were carried out in order to identify differences between different time points. In case
the interaction between genotype and time was significant, the differences were evaluated in
more detail by separately analyzing the effects of genotype and time using Bonferroni-corrected
two-sided unpaired t-tests and one-way ANOVA with Bonferroni corrected post-hoc tests,
respectively. Correlations between parameters were assessed using Pearson correlation.
Statistical significance was set at P<0.05.
9.3. Results
9.3.1. Diabetic mice had a higher mortality rate durling TAC surgery, but higher
survival rate after TAC surgery
TAC surgery was performed to induce pressure-overload cardiac hypertrophy and heart failure.
While all non-diabetic mice survived the surgery, the diabetic mice had ~30% mortality during
and directly after surgery. The higher mortality in diabetic mice is likely to be attributed to fat
tissue around and fragility of the aorta. After TAC surgery, however, the survival rate was
higher in diabetic mice compared with non-diabetic mice. While the survival rate was 100% for
both groups until 5 weeks post TAC, at 12 weeks post TAC, survival rates were 62.5% versus
83% for non-diabetic and diabetic mice, respectively.
151
Table 1. Animal characteristics
Parameters
Baseline
db/+
BW (g)
§
1 week post TAC
db/db
db/+
db/db
5 weeks post TAC
db/+
12 weeks post TAC
db/db
†
db/+
†
db/db
26.4 ± 2.0
46.9 ± 3.2
27.1 ± 1.1
47.7 ± 1.6
29.4 ± 1.0
52.8 ± 2.0
31.0 ± 2.2
46.8 ± 7.5
HW (mg)
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
262.8 ± 40.0
154.4 ± 16.8*
TL (mm)
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
18.6 ± 0.4
18.4 ± 1.0
HW/ TL (mg/mm)
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
13.1 ± 1.5
8.5 ± 1.2*
Dry/wet LW ratio (mg/mg)
Plasma glucose (mM)
n.a.
n.a.
n.a.
n.a.
8.3 ± 1.3
26.3 ± 6.7*
8.5 ± 0.5
21.4 ± 5.5*
n.a.
6.8 ± 0.7
#,‡
n.a.
3.9 ± 0.5
3.8 ± 0.6
31.8 ± 11.4*
8.4 ± 1.2^
29.1 ± 6.2*
BW: body weight, HW: heart weight, LW: lung weight, TL: tibia length. Data are means ± SD (n=5-8 for non-diabetic db/+ mice, n=5-10 for diabetic db/db mice). §P<0.05
for general genotype effect, †P<0.05 vs. baseline independent of genotype, *P<0.05 vs. non-diabetic db/+ mice at the same time point, #P<0.05 vs. baseline for the same
genotype, ‡P<0.05 vs. 1 wk post TAC for the same genotype, ^P<0.05 vs. 5 wk post TAC for the same genotype.
Cardiac metabolic adaptations in heart failure
Figure 9.1. Blood flow measurements in RCA and LCA upon TAC. (A) Long axis view of the heart,
showing the TAC banding located between RCA and LCA branches (data from a non-diabetic mouse).
The white line shows the orientation of the slice used for flow measurements, where the main
direction of RCA and LCA flow is almost perpendicular to the slice. (B) Representative flow velocity
curves at baseline and 1week post TAC for the (B) RCA and (C) LCA from a diabetic mouse. (D) RCA
peak velocity and (E) LCA peak velocity for non-diabetic and diabetic mice at baseline and 1 week
post TAC. Data are means ± SD (n=8-10 for non-diabetic mice, n=3-8 for diabetic mice). †P<0.05 vs.
baseline independent of genotype.
153
Chapter 9
9.3.2. Animal characteristics
Body weight of diabetic mice was two times higher than that of non-diabetic mice, independent
of time point (general genotype effect: P<0.001; Table 9.1). For both non-diabetic and diabetic
mice, body weight was increased at 5 weeks post TAC compared with baseline (P=0.024; Table
9.1). At 12 weeks post TAC, heart weight was lower in diabetic mice than in non-diabetic mice
(P<0.001), also after normalization to tibia length (P=0.001; Table 9.1). Tibia length was
similar between non-diabetic and diabetic mice (P=0.62; Table 9.1). The ratio of lung dry
weight to wet weight did not differ between non-diabetic and diabetic mice (P=0.86).
Plasma glucose levels were 3-5 times higher in diabetic than in non-diabetic mice at all time
points (P<0.05; Table 1). In non-diabetic mice, plasma glucose levels were decreased at 5
weeks post TAC (P=0.021 vs. baseline and P=0.008 vs. 1 wk post TAC; Table 9.1), but returned
to baseline levels at 12 weeks post TAC (P=1.000 vs. baseline and P=0.032 vs. 5 wk post TAC;
Table 9.1). In diabetic mice, plasma glucose levels did not change over time.
9.3.3. Carotid blood flow velocity was similar between non-diabetic and diabetic mice
at baseline and upon TAC
Measurements of blood flow velocities in the carotid arteries (Fig. 9.1A-C) allows assessing
variations in the severity of TAC surgery. At baseline, peak flow velocities were similar in the
RCA and LCA, both for non-diabetic and diabetic mice (Fig. 9.1D and Fig. 9.1E). In both nondiabetic and diabetic mice, TAC increased RCA and decreased LCA peak flow velocities 1 week
post TAC (P<0.001 and P=0.004, respectively; Fig. 9.1D and Fig. 9.1E). However, no difference
in RCA and LCA peak flow velocities was observed between non-diabetic and diabetic mice
(P=0.58 and P=0.16, respectively), indicating that the severity of the constriction was similar
among genotypes.
9.3.4. TAC affected cardiac morphology and function severely in non-diabetic mice,
but only mildly in diabetic mice
TAC resulted in a hypertrophic growth of the heart over time (Fig. 9.2A and Fig. 9.1B). In nondiabetic mice, LV mass progressively increased from a 37% increase with respect to baseline at
1 week post TAC (P=0.001 vs. baseline) to a 96% increase at 12 weeks post TAC (P<0.001 vs.
baseline; Fig. 9.2B). Interestingly, the increase in LV mass was less pronounced in diabetic mice
compared with non-diabetic mice (Fig. 9.2B). In diabetic mice, LV mass only significantly
increased after 5 weeks post TAC (44% increase with respect to baseline; P=0.001) and did not
further increase at 12 weeks post TAC.
In non-diabetic mice, the progressive hypertrophy upon TAC was accompanied by increases of
end-diastolic and end-systolic volumes (EDV and ESV) over time (Fig. 9.2C and Fig. 9.2D),
resulting in a reduced stroke volume (SV) at 1 week post TAC, which was maintained until 12
weeks post TAC (Fig. 9.2E). In contrast, in diabetic mice EDV and SV were not affected by TAC
(Fig. 9.2C and Fig 9.2E). In diabetic mice only ESV was increased at 5 weeks post TAC
(P<0.002 vs. baseline), but this was again normalized at 12 weeks post TAC (Fig. 9.2D).
154
Cardiac metabolic adaptations in heart failure
Figure 9.2. Cardiac mass and systolic function as measured using cine MRI. (A) Representative images
of 4-chamber long axis views (the white scale bar equals 5 mm), (B) left ventricular (LV) mass, (C) enddiastolic volume, (D) end-systolic volume, (E) stroke volume, and (F) ejection fraction at baseline, and
1, 5, and 12 weeks post TAC. Data are means ± SD (n=6-8 for non-diabetic mice, n=6 for diabetic
mice). *P<0.05 vs. non-diabetic mice at the same time point, #P<0.05 vs. baseline for the same
genotype, ‡P<0.05 vs. 1 week post TAC for the same genotype, ^P<0.05 vs. 5 weeks post TAC for the
same genotype.
155
Chapter 9
Figure 9.3. Cardiac ejection and filling rates as measured using high time resolution cine MRI. (A)
Peak ejection rates, and (B) peak filling rates at baseline, and 1, 5, and 12 weeks post TAC. Data are
means ± SD (n=4-6 for non-diabetic mice, n=3-4 for diabetic mice). *P<0.05 vs. non-diabetic mice at
the same time point, #P<0.05 vs. baseline for the same genotype, ‡P<0.05 vs. 1 week post TAC for
the same genotype, †P=0.078 vs. 1 week post TAC independent of genotype.
Similarly, ejection fraction (EF) was reduced to a higher degree in non-diabetic mice than in
diabetic mice (Fig. 9.2F). In non-diabetic mice, EF progressively decreased after TAC, from a
33% decrease with respect to baseline at 1 week post TAC (P<0.001 vs. baseline) to a 55%
decrease at 12 weeks post TAC (P<0.001 vs. baseline). In contrast, in diabetic mice EF was only
decreased by 9% compared with baseline at 1 week post TAC (P=0.027 vs. baseline), and was
not further decreased at 5 and 12 weeks post TAC. TAC progressively reduced the peak ejection
rate (PER) in non-diabetic mice, resulting in a 65% reduction with respect to baseline at 12
weeks post TAC (P<0.001 vs. baseline), while PER was not affected by TAC in diabetic mice
(Fig. 9.3A). The peak filling rate (PFR), as a measure of cardiac diastolic function, tended to be
reduced at 12 weeks post TAC independent of genotype (P=0.078 vs. 1 week post TAC; Fig.
9.3B).
9.3.5. Cardiac energy status tended to reduce in non-diabetic mice upon TAC, but was
maintained in diabetic mice despite the lower cardiac energy status at baseline
Cardiac
31
P MRS was used to quantify the myocardial PCr/ATP ratio as a measure of cardiac
energy status in vivo. Cardiac
31
P MR spectra contain resonances from myocardial PCr,
myocardial and blood γ-, α-, and β-ATP, and blood 2,3-diphosphoglyerate (DPG) obscuring the
myocardial inorganic phosphate peak (Fig. 9.4A and Fig. 9.4B). At baseline, the PCr/ATP was
41% lower in diabetic mice than in non-diabetic mice (Fig. 9.4C). Interestingly, the baseline
PCr/ATP ratio in diabetic mice correlated strongly with fasting plasma glucose (r=0. 855,
P=0.014; Fig. 9.4D). TAC did not affect the PCr/ATP ratio in diabetic mice (Fig. 9.4C). In nondiabetic mice, the PCr/ATP ratio was also unchanged at 1 and 5 weeks post TAC. However, in
non-diabetic mice the PCr/ATP ratio tended to decrease by ~25% at 12 weeks post TAC
(P=0.082 vs. baseline, P=0.077 vs. 5 weeks post TAC). Furthermore, in non-diabetic mice, we
observed a tendency for a positive correlation between the PCr/ATP ratio and EF (r=0.325,
156
Cardiac metabolic adaptations in heart failure
P=0.091; Fig. 9.4E) and a negative correlation between the PCr/ATP ratio and LV mass (r=0.341, P=0.076; Fig. 9.4F).
Figure 9.4. Cardiac energy status measured using 31P MRS. Representative cardiac 31P MR spectra of
(A) a non-diabetic mouse and (B) a diabetic mouse at 5 weeks post TAC from a voxel covering the left
ventricle. (C) PCr/ATP ratio at baseline, and 1, 5, and 12 weeks post TAC. Data are means ± SD (n=68 for non-diabetic mice, n=5-7 for diabetic mice). *P<0.05 vs. non-diabetic mice at the same time
point, $P=0.082 vs. baseline for the same genotype, ξP=0.077 vs. 5 weeks post TAC for the same
genotype. (D) Correlation between PCr/ATP ratio and fasting plasma glucose at baseline. Correlation
between (E) PCr/ATP ratio and ejection fraction, and (F) PCr/ATP ratio and left ventricular mass, for
non-diabetic mice.
157
Chapter 9
Figure 9.5. Myocardial lipid content. (A) Cardiac 1H MR spectra of a non-diabetic and a diabetic mouse
at baseline, showing peaks originating from taurine (tau), choline/carnitine (cho/car), creatine-CH3
(Cr-CH3), and 7 peaks associated with triglyceride (TG) protons (indicated in bold): (1) CH=CH-;
(2) -CH=CH-CH2-CH=CH-; (3) -CαH2COO-; (4) -CH2-CH=CH-CH2-; (5) -CβH2CH2COO-; (6) -CH2-;
(7) -CH3. (B) Quantification of the myocardial TG-CH2 signal for non-diabetic and diabetic mice at
baseline, and 1, and 12 weeks post TAC. The separator indicates data from separate experiments, and
statistical analysis was performed accordingly. Data are means ± SD (n=4-8 for non-diabetic mice, n=37 for diabetic mice). §P<0.05 for general genotype effect, *P<0.05 vs. non-diabetic mice at the same
time point.
9.3.6. Myocardial lipid content was higher in diabetic mice and not affected by TAC
Myocardial lipid content was measured using 1H MRS, by quantifying the TG-CH2 peak relative
to the water signal (Fig. 9.5A). Myocardial TG levels were ~50% higher in diabetic mice
compared with non-diabetic mice and were not affected by TAC (Fig. 9.5B).
9.3.7. Myocardial FDG uptake was higher in non-diabetic mice, but increased in both
non-diabetic and diabetic mice upon TAC
To measure myocardial glucose uptake, we performed
18
F-FDG-PET. At 75 minutes post
18
F-FDG
administration, FDG accumulation in the heart and bladder were observed (Fig. 9.6A). The FDG
accumulation in the heart was clearly observed in the myocardium (Fig. 9.6B). Overall, FDG
uptake was ~45% lower in diabetic mice than in non-diabetic mice (general genotype effect:
P=0.004). However, in both non-diabetic and diabetic mice, FDG uptake was increased at 1
week post TAC with respect to baseline (general time effect: P<0.001; Fig. 9.6C). For diabetic
mice, FDG-PET data at 5 and 12 weeks post TAC is not yet available. In non-diabetic mice, FDG
uptake at 5 weeks post TAC was 2.8-fold higher than at baseline (P<0.001). However, at 12
weeks post TAC, FDG uptake in non-diabetic mice tended to decrease with respect to 5 weeks
post TAC (P=0.094; Fig. 9.6C). Furthermore, in non-diabetic mice, myocardial FDG uptake was
negatively correlated with EF (r=-0.814, P<0.001; Fig. 9.6D) and positively correlated with LV
mass (r=0.804, P<0.001; Fig. 9.6E).
158
Cardiac metabolic adaptations in heart failure
Figure 9.6. Myocardial 18F-FDG uptake measured using PET. (A) Whole-body 18F-FDG PET image
(maximum intensity projection) from a non-diabetic mouse at baseline. Highest uptake is observed in
the heart and bladder. The spots in the head are from the Harderian glands. (B) 18F-FDG PET images
from 3 heart axis views of a non-diabetic mouse at baseline, and 1, 5, and 12 weeks post TAC. Color
intensity represents the level of 18F-FDG uptake, in the percentage of injected dose (%ID) per μL of
myocardial volume. (C) Quantification of myocardial 18F-FDG uptake at baseline, and 1, 5, and 12
weeks post TAC (experiments in diabetic mice are in progress). Data are means ± SD (n=3-8 for nondiabetic mice, n=3-8 for diabetic mice). Two-way ANOVA was performed for data at baseline and 1
week post TAC (§P<0.05 vs. non-diabetic mice independent of time, †P<0.05 vs. baseline independent
of genotype), while one-way ANOVA was performed separately on the non-diabetic mice (#P<0.05 vs.
baseline for the same genotype, $P=0.094 vs. 5 weeks post TAC for the same genotype). Correlation
between (D) myocardial FDG uptake and ejection fraction, (E) myocardial FDG uptake and left
ventricular mass (data from non-diabetic mice).
159
Chapter 9
Figure 9.7. Oxygen consumption rates in isolated heart mitochondria of non-diabetic and diabetic
mice at baseline and 12 weeks post TAC. State 3, state 4, and state U respiration when (A) pyruvate
plus malate or (B) palmitoyl-CoA plus carnitine plus malate were used as substrates. Data are means
± SD (n=4-6 per group). †P<0.05 vs. baseline independent of genotype, §P<0.05 vs. non-diabetic mice
independent of time, *P<0.05 vs. non-diabetic mice at the same time point, #P<0.05 vs. baseline for
the same genotype.
9.3.8. TAC lowers mitochondrial oxidative capacity for FA substrates to a higher
extent in hearts of non-diabetic mice than in diabetic mice
Ex vivo mitochondrial oxidative capacity was measured in isolated heart mitochondria of nondiabetic and diabetic mice at baseline and at 12 weeks post TAC. At baseline, mitochondrial
respiratory capacity for a glucose-derived substrate (i.e. pyruvate plus malate) was similar
between non-diabetic and diabetic mice (Fig. 9.7A). However, when using a FA substrate to fuel
mitochondrial respiration (i.e. palmitoyl-CoA plus carnitine plus malate) state 3 and state U
respiratory capacity were higher in diabetic mice than in non-diabetic mice (P<0.05; Fig. 9.7B).
Twelve weeks post TAC, state 3 and state U respiratory capacity with pyruvate were ~25%
lower compared with baseline, independent of genotype (general time effect P=0.001 for state
160
Cardiac metabolic adaptations in heart failure
3, P=0.013 for state U; Fig. 9.7A). TAC similarly resulted in a genotype-independent decrease in
state 3 respiratory capacity with palmitoyl-CoA (general time effect P<0.001; Fig. 9.7B), but
state 3 respiration with palmitoyl-CoA remained higher in diabetic mice than in non-diabetic
mice (P<0.05; Fig. 9.7B). Interestingly, while state U respiratory capacity of non-diabetic mice
for the FA substrate was 35% lower at 12 weeks post TAC compared with baseline (P=0.001),
state U respiratory capacity of diabetic mice with palmitoyl-CoA was not affected by TAC
(P=0.362; Fig. 9.7B). State 4 respiration was similar between non-diabetic and diabetic mice for
both substrates, and unaffected by TAC (Fig. 9.7A and Fig. 9.7B).
9.4. Discussion
The current in vivo study addressed to what extent changes in substrate use, lipid
accumulation, and energy status predict the longitudinal development of hypertrophy and
subsequent failure of the non-diabetic and diabetic heart. Evidence for the link between heart
failure and changes in cardiac metabolism largely originates from ex vivo studies using isolated
perfused hearts. These setups, however, cannot fully mimic the in vivo condition. The diabetic
heart, for example, is exposed to a hyperglycemic, hyperlipidemic, and hyperinsulinemic
environment, which will affect the choice of substrates (42). Furthermore, the type and
concentration of substrates may change during the development of heart failure, which is
difficult to address in ex vivo studies.
Using in vivo PET, MRS, and MRI techniques we measured myocardial FDG uptake, lipid content,
energy status, and function at baseline, and 1, 5, and 12 weeks after TAC-induced pressure
overload. We showed that in non-diabetic mice, pressure overload induces progressive LV
hypertrophy and systolic dysfunction, which strongly correlated with increased myocardial FDG
uptake. While myocardial FDG uptake was increased at 1 and 5 weeks post TAC, it tended to
decrease again at 12 weeks post TAC. This apparent reduction in myocardial FDG uptake was
associated with a decrease in cardiac energy status, which only occurred at 12 weeks post TAC.
Surprisingly, whereas diabetic mice showed overall lower myocardial FDG uptake, lower cardiac
energy status, and higher myocardial lipid accumulation than non-diabetic mice, cardiac
function of the diabetic mice was only mildly affected by pressure overload compared with nondiabetic mice. Additionally, we show that metabolic flexibility was not completely lost in the
hearts of diabetic mice.
9.4.1. Pressure overload-induced increase in myocardial FDG uptake precedes
changes in cardiac energy status in non-diabetic mice
Our study indicates that upon pressure overload the heart switches to an increased utilization of
glucose, and that this metabolic switch occurs already early in the development of heart failure.
The increase in myocardial FDG uptake was already present at 1 week post TAC, and an
increase in myocardial glucose uptake is usually accompanied by an increase in glycolysis
and/or glucose oxidation (43, 61). The increase in myocardial FDG uptake preceded the
reduction in cardiac energy status, which was observed as a trend at 12 weeks post TAC. A
reduced PCr/ATP ratio in the failing heart has been reported in both animals (18, 36) and
patients (22, 39-41). However, while the PCr/ATP ratio was greatly reduced in heart failure
161
Chapter 9
patients with coronary artery disease, the PCr/ATP ratio was maintained in non-ischemic heart
failure patients (60). Nonetheless, Maslov et al. showed that a reduced PCr/ATP ratio at 3 weeks
after TAC in mice was a predictor of cardiac dysfunction (36), which is in contrast with our
study. Our data show that cardiac dysfunction, which was already present at 1 week post TAC,
is unlikely to be explained by an impairment of cardiac energetics. The reduced cardiac energy
status observed at 12 weeks post TAC seems rather to be an effect of heart failure progression,
which may only become prominent when total mitochondrial oxidative metabolism has declined,
as indicated by the decrease in myocardial FDG uptake and lower ex vivo mitochondrial
oxidative capacity at 12 weeks post TAC.
The pattern of alterations in myocardial glucose uptake upon heart failure as observed in this
study, i.e. the increase in the early stage and the subsequent decrease at a later stage), is in
agreement with the current consensus based on collective findings in animals and humans
(reviewed in (39)). However, to our knowledge, our study provides the first longitudinal in vivo
data confirming these previous, single time point observations. So far, we are only aware of one
in vivo study that investigated myocardial FDG uptake upon TAC in mice longitudinally (61).
However, data from that study are limited to the early stage of heart failure development
(baseline, 1 day, and 7 days after TAC) (61). In the present study, we extended the time frame
to include also the more long-term effects of heart failure, and, moreover, we complemented
the data on myocardial FDG uptake with measurements of cardiac energetics, providing a more
comprehensive overview of the metabolic adaptations upon heart failure.
The alterations in cardiac substrate metabolism upon pressure overload may contribute to the
development of cardiac dysfunction. Through the Randle cycle (45), the increased glucose
utilization may suppress FAO, as observed in other studies of heart failure (27, 43, 44).
Importantly, reduced FAO was shown to correlate with the severity of cardiac dysfunction in
hypertensive heart disease (14). Suppression of FAO may lead to the accumulation of toxic lipid
intermediates (28), and although myocardial TG content in vivo was not changed upon TAC in
our mice, levels of potentially toxic diacylglycerols (DAG) and ceramides could have been
increased. Increased DAG and ceramide levels have been observed in previous studies of
pressure overload models and were associated with increased fibrosis, apoptosis, and oxidative
stress (43, 44, 56), the latter of which can eventually damage mitochondria. Reduced
mitochondrial oxidative capacity upon heart failure has been shown previously in isolated
mitochondria and permeabilized cardiac fibers of dogs (47, 50) and patients (31, 51). In
agreement, we observed a reduction in ex vivo mitochondrial oxidative capacity for both glucose
and FA substrates in hearts of non-diabetic mice upon 12 weeks of TAC.
9.4.2. Mild effect of pressure overload on LV mass and cardiac function in diabetic
mice, while cardiac energy status is maintained
Surprisingly, while the non-diabetic mice developed pressure overload-induced cardiac
dysfunction, cardiac function of diabetic mice was only mildly affected by TAC. The different
responses to TAC cannot be explained by differences in the surgery, leading to differences in the
severity of the constriction. TAC was shown to increase RCA blood flow velocity and decrease
LCA blood flow velocity as early as 1 day post TAC, which remained until 1 week post TAC (32).
Here, we showed that RCA and LCA blood flow velocities were similar at 1 week post TAC in
162
Cardiac metabolic adaptations in heart failure
diabetic and non-diabetic mice, indicating similar severity of constriction. The increase in LV
mass upon TAC in diabetic mice was much less pronounced than in non-diabetic mice (44%
versus 96% at 12 weeks post TAC), showing cardiac hypertrophy was less severe in the diabetic
mice. EF was only 9% reduced at 1 week post TAC in diabetic mice, and did not decrease
further during the time course of the study. The mild effect of TAC on cardiac function in
diabetic mice was accompanied by maintained cardiac energy status up to 12 weeks post TAC,
and a blunted reduction in ex vivo mitochondrial oxidative capacity for FA substrates compared
with non-diabetic mice. In agreement with our results, a recent study showed that diabetic mice
subjected to pressure overload do not exhibit an exacerbation of cardiac dysfunction compared
with non-diabetic controls and rather seem to be protected from the development of heart
failure (8).
Metabolically, the diabetic heart has been proposed to be inflexible to switch from FA to glucose
substrates (57). This was supported by an in vivo study in which glucose infusion after an 8hour fast did not alter the rates of FA or glucose utilization in the diabetic heart (42). However,
the concept of metabolic inflexibility of the diabetic heart has been a topic of debate (30). A
number of studies using ex vivo perfused heart setups demonstrated that FA and glucose
oxidation rates in the diabetic heart could be altered upon chronic treatment with lipid-lowering
drugs (9, 17, 25, 53) or glucose/insulin administration (7, 19, 20). Supporting these ex vivo
observations, we showed that the metabolic flexibility of the diabetic heart was not completely
lost. Myocardial FDG uptake was increased at one week post TAC compared to baseline,
although the absolute myocardial FDG uptake was still lower than in non-diabetic mice.
It is an interesting observation that the effect of TAC-induced pressure overload was milder in
diabetic mice compared with non-diabetic mice, especially because diabetes is a risk factor for
heart failure (26) and because cardiac energy status, which is a strong predictor of mortality in
heart failure patients (40), was already 41% lower in diabetic mice compared with non-diabetic
mice at baseline. Furthermore, diabetic hearts were previously associated with reduced resting
cardiac function (2, 7), although the cardiac function impairments were generally subtle (12,
13). Therefore, one might expect that the reduced energy status of the diabetic heart would be
exacerbated upon stress such as pressure overload. However, we showed that cardiac
energetics was maintained in diabetic mice after TAC. Supported by the data from the nondiabetic mice, our results suggest that cardiac energy status does not play a decisive role in the
development of heart failure in mice.
9.4.3. The switch towards glucose utilization during heart failure: (mal)adaptive?
Our data as well as other studies (27, 43, 44) show that during the development of heart failure
the heart switches towards increased glucose utilization. However, it is still not clear whether
this shift is adaptive or maladaptive. In the present study, we showed that myocardial FDG
uptake strongly correlates with the severity of cardiac dysfunction in non-diabetic mice. If the
increased myocardial FDG uptake was a compensatory response to the increased workload, our
data indicates that this compensatory mechanism does not seem to prevent the progression of
heart failure. In agreement, previous studies have shown that the hypertrophied heart seems to
be unable to increase glucose utilization to the levels required to maintain function (3, 43). As
FA substrates have a higher energy density compared with glucose (34), a substantial increase
163
Chapter 9
in glucose oxidation would be necessary to compensate for a decrease in FAO (58). It has been
demonstrated that further stimulation of glucose utilization by overexpression of the insulinindependent glucose transporter GLUT1 in mice could rescue the heart upon increased workload
(35) or upon pressure overload induced by constriction of the ascending aorta (33). On the
other hand, an intervention with propranolol preventing the switch towards glucose utilization in
mice upon TAC was also shown to improve cardiac function (61).
The trend for a reduced cardiac energy status in non-diabetic mice at 12 weeks post TAC,
together with reduced myocardial glucose uptake and mitochondrial oxidative capacity, may
indicate lower total rates of substrate oxidation at this time point. Maintenance of FAO may thus
be necessary to provide sufficient energy during later stages of HF. The importance of FAO in
maintaining cardiac function and energetics has been demonstrated in ACC2 knockout mice,
which have an increased FAO (27). Upon TAC-induced pressure overload, the high rates of
cardiac FAO in ACC2 knockout mice were maintained, while the heart was protected from failure
(27). Furthermore, it has been shown that reduced FAO through ablation of FAT/CD36 or
CPT1b, proteins involved in FA transport over the cellular and mitochondrial membranes,
respectively, results in exacerbated cardiac dysfunction and energetic impairment after TAC (23,
56). The observation that the development of cardiac hypertrophy and dysfunction upon TAC
was much less severe in diabetic mice compared with non-diabetic mice, and that cardiac
energetics was not further reduced in diabetic mice upon TAC, further supports the hypothesis
that FAO is important in maintaining cardiac function and energy status. Diabetic mice have
been shown to almost exclusively rely on FAO (1, 2, 7). Upon TAC, despite the increase in
myocardial glucose uptake, FAO may still be higher in diabetic mice compared with non-diabetic
mice, as also suggested by the higher mitochondrial oxidative capacity for FA substrates in
diabetic mice than in non-diabetic mice at 12 weeks post TAC. A recent study in glucose
transporter 1 (GLUT1) knockout mice demonstrated that maintained high FAO and the lack of a
shift towards glucose utilization after TAC did not accelerate the progression from hypertrophy
to heart failure, suggesting that the switch to glucose may not be a necessary mechanism to
cope with heart failure (44).
Taken together, this study provides the first longitudinal in vivo data of cardiac metabolic,
energetic, and functional adaptations during heart failure development in non-diabetic and
diabetic mice. Our data show that progression to heart failure was correlated with increased
myocardial glucose uptake, which preceded the decrease in cardiac energy status. The mild
cardiac hypertrophy and dysfunction in diabetic mice, together with lower myocardial glucose
uptake upon TAC, suggest that maintaining FAO may be beneficial for cardiac function and
energetics in heart failure.
9.5. Acknowledgements
We thank Leonie Niesen and Dirk Reinhardt for performing the TAC surgery, Leonie Niesen and
David Veraart for their assistance in animal handling, Sarah Koester and Roman Priebe for their
assistance in PET measurements, Nina Nagelmann for data analysis of systolic function, and
Laura Vergoossen for data reconstruction and analysis of diastolic function.
164
Cardiac metabolic adaptations in heart failure
D.A., and J.J.P. are supported by a VIDI grant (project number 700.58.421) from the
Netherlands Organisation for Scientific Research (NWO). D.A. is also supported by a travel grant
from Boehringer Ingelheim Fonds to carry out the experiments in Münster. M.N. is supported by
a VENI grant (project number 916.14.050) from NWO. This work was in part funded by the
Interdisciplinary Centre for Clinical Research (IZKF, core unit PIX), Münster, Germany.
9.6. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Aasum E, Belke DD, Severson DL, Riemersma RA, Cooper M, Andreassen M, and Larsen TS.
Cardiac function and metabolism in Type 2 diabetic mice after treatment with BM 17.0744, a novel
PPAR-alpha activator. Am J Physiol Heart Circ Physiol 283: H949-957, 2002.
Aasum E, Hafstad AD, Severson DL, and Larsen TS. Age-dependent changes in metabolism,
contractile function, and ischemic sensitivity in hearts from db/db mice. Diabetes 52: 434-441, 2003.
Allard MF, Schonekess BO, Henning SL, English DR, and Lopaschuk GD. Contribution of
oxidative metabolism and glycolysis to ATP production in hypertrophied hearts. Am J Physiol 267:
H742-750, 1994.
Bakermans AJ, Abdurrachim D, Geraedts TR, Houten SM, Nicolay K, and Prompers JJ. In vivo
proton T1 relaxation times of mouse myocardial metabolites at 9.4 T. Magn Reson Med 2014.
Bakermans AJ, Dodd MS, Nicolay K, Prompers JJ, Tyler DJ, and Houten SM. Myocardial energy
shortage and unmet anaplerotic needs in the fasted long-chain acyl-CoA dehydrogenase knockout
mouse. Cardiovasc Res 100: 441-449, 2013.
Bakermans AJ, van Nierop BJ, Abdurrachim D, Nicolay K, and Prompers JJ. In vivo cardiac 31PMRS in a mouse model of heart failure. In: Proc 20th Ann Meeting ISMRM p. 1115, 2012.
Belke DD, Larsen TS, Gibbs EM, and Severson DL. Altered metabolism causes cardiac dysfunction
in perfused hearts from diabetic (db/db) mice. Am J Physiol Endocrinol Metab 279: E1104-1113, 2000.
Brainard RE, Watson LJ, Demartino AM, Brittian KR, Readnower RD, Boakye AA, Zhang D,
Hoetker JD, Bhatnagar A, Baba SP, and Jones SP. High fat feeding in mice is insufficient to induce
cardiac dysfunction and does not exacerbate heart failure. PLoS ONE 8: e83174, 2013.
Carley AN, Semeniuk LM, Shimoni Y, Aasum E, Larsen TS, Berger JP, and Severson DL.
Treatment of type 2 diabetic db/db mice with a novel PPAR{gamma} agonist improves cardiac
metabolism but not contractile function. Am J Physiol Endocrinol Metab 286: E449-455, 2004.
Chiu HC, Kovacs A, Ford DA, Hsu FF, Garcia R, Herrero P, Saffitz JE, and Schaffer JE. A novel
mouse model of lipotoxic cardiomyopathy. J Clin Invest 107: 813-822, 2001.
Christe ME, and Rodgers RL. Altered glucose and fatty acid oxidation in hearts of the spontaneously
hypertensive rat. J Mol Cell Cardiol 26: 1371-1375, 1994.
Coolen BF, Abdurrachim D, Motaal AG, Nicolay K, Prompers JJ, and Strijkers GJ. High frame
rate retrospectively triggered Cine MRI for assessment of murine diastolic function. Magn Reson Med
69: 648-656, 2013.
Daniels A, van Bilsen M, Janssen BJA, Brouns AE, Cleutjens JPM, Roemen THM, Schaart G,
van der Velden J, van der Vusse GJ, and van Nieuwenhoven FA. Impaired cardiac functional
reserve in type 2 diabetic db/db mice is associated with metabolic, but not structural, remodelling.
Acta Physiol 200: 11-22, 2010.
de las Fuentes L, Herrero P, Peterson LR, Kelly DP, Gropler RJ, and Davila-Roman VG.
Myocardial fatty acid metabolism: independent predictor of left ventricular mass in hypertensive heart
disease. Hypertension 41: 83-87, 2003.
Diamant M, Lamb HJ, Groeneveld Y, Endert EL, Smit JW, Bax JJ, Romijn JA, de Roos A, and
Radder JK. Diastolic dysfunction is associated with altered myocardial metabolism in asymptomatic
normotensive patients with well-controlled type 2 diabetes mellitus. J Am Coll Cardiol 42: 328-335,
2003.
Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Blaha MJ, Dai S, Ford ES, Fox CS,
Franco S, Fullerton HJ, Gillespie C, Hailpern SM, Heit JA, Howard VJ, Huffman MD, Judd SE,
Kissela BM, Kittner SJ, Lackland DT, Lichtman JH, Lisabeth LD, Mackey RH, Magid DJ, Marcus
GM, Marelli A, Matchar DB, McGuire DK, Mohler ER, 3rd, Moy CS, Mussolino ME, Neumar RW,
Nichol G, Pandey DK, Paynter NP, Reeves MJ, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani
SS, Wong ND, Woo D, and Turner MB. Heart disease and stroke statistics--2014 update: a report
from the American Heart Association. Circulation 129: e28-e292, 2014.
Golfman LS, Wilson CR, Sharma S, Burgmaier M, Young ME, Guthrie PH, Van Arsdall M,
Adrogue JV, Brown KK, and Taegtmeyer H. Activation of PPAR{gamma} enhances myocardial
glucose oxidation and improves contractile function in isolated working hearts of ZDF rats. Am J
Physiol Endocrinol Metab 289: E328-336, 2005.
Gupta A, Chacko VP, and Weiss RG. Abnormal energetics and ATP depletion in pressure-overload
mouse hearts: in vivo high-energy phosphate concentration measures by noninvasive magnetic
resonance. Am J Physiol Heart Circ Physiol 297: H59-64, 2009.
Hafstad AD, Khalid AM, How O-J, Larsen TS, and Aasum E. Glucose and insulin improve cardiac
efficiency and postischemic functional recovery in perfused hearts from type 2 diabetic (db/db) mice.
Am J Physiol Endocrinol Metab 292: E1288-1294, 2007.
165
Chapter 9
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
166
Hafstad AD, Solevag GH, Severson DL, Larsen TS, and Aasum E. Perfused hearts from Type 2
diabetic (db/db) mice show metabolic responsiveness to insulin. Am J Physiol Heart Circ Physiol 290:
H1763-1769, 2006.
Hansch A, Rzanny R, Heyne JP, Leder U, Reichenbach JR, and Kaiser WA. Noninvasive
measurements of cardiac high-energy phosphate metabolites in dilated cardiomyopathy by using 31P
spectroscopic chemical shift imaging. Eur Radiol 15: 319-323, 2005.
Hardy CJ, Weiss RG, Bottomley PA, and Gerstenblith G. Altered myocardial high-energy
phosphate metabolites in patients with dilated cardiomyopathy. Am Heart J 122: 795-801, 1991.
He L, Kim T, Long Q, Liu J, Wang P, Zhou Y, Ding Y, Prasain J, Wood PA, and Yang Q. Carnitine
palmitoyltransferase-1b deficiency aggravates pressure overload-induced cardiac hypertrophy caused
by lipotoxicity. Circulation 126: 1705-1716, 2012.
Heidenreich PA, Albert NM, Allen LA, Bluemke DA, Butler J, Fonarow GC, Ikonomidis JS,
Khavjou O, Konstam MA, Maddox TM, Nichol G, Pham M, Pina IL, and Trogdon JG. Forecasting
the impact of heart failure in the United States: a policy statement from the American Heart
Association. Circ Heart Fail 6: 606-619, 2013.
How OJ, Larsen TS, Hafstad AD, Khalid A, Myhre ESP, Murray AJ, Boardman NT, Cole M,
Clarke K, Severson DL, and Aasum E. Rosiglitazone treatment improves cardiac efficiency in hearts
from diabetic mice. Arch Physiol Biochem 113: 211-220, 2007.
Kannel WB, and McGee DL. Diabetes and cardiovascular risk factors: the Framingham study.
Circulation 59: 8-13, 1979.
Kolwicz SC, Jr., Olson DP, Marney LC, Garcia-Menendez L, Synovec RE, and Tian R. Cardiacspecific deletion of acetyl CoA carboxylase 2 prevents metabolic remodeling during pressure-overload
hypertrophy. Circ Res 111: 728-738, 2012.
Koves TR, Ussher JR, Noland RC, Slentz D, Mosedale M, Ilkayeva O, Bain J, Stevens R, Dyck
JR, Newgard CB, Lopaschuk GD, and Muoio DM. Mitochondrial overload and incomplete fatty acid
oxidation contribute to skeletal muscle insulin resistance. Cell Metab 7: 45-56, 2008.
Labinskyy V, Bellomo M, Chandler MP, Young ME, Lionetti V, Qanud K, Bigazzi F, Sampietro
T, Stanley WC, and Recchia FA. Chronic activation of peroxisome proliferator-activated receptoralpha with fenofibrate prevents alterations in cardiac metabolic phenotype without changing the onset
of decompensation in pacing-induced heart failure. J Pharmacol Exp Ther 321: 165-171, 2007.
Larsen TS, and Aasum E. Metabolic (in)flexibility of the diabetic heart. Cardiovasc Drugs Ther 22:
91-95, 2008.
Lemieux H, Semsroth S, Antretter H, Hofer D, and Gnaiger E. Mitochondrial respiratory control
and early defects of oxidative phosphorylation in the failing human heart. Int J Biochem Cell Biol 43:
1729-1738, 2011.
Li YH, Reddy AK, Taffet GE, Michael LH, Entman ML, and Hartley CJ. Doppler evaluation of
peripheral vascular adaptations to transverse aortic banding in mice. Ultrasound Med Biol 29: 12811289, 2003.
Liao R, Jain M, Cui L, D’Agostino J, Aiello F, Luptak I, Ngoy S, Mortensen RM, and Tian R.
Cardiac-specific overexpression of GLUT1 prevents the development of heart failure attributable to
pressure overload in mice. Circulation 106: 2125-2131, 2002.
Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, and Stanley WC. Myocardial fatty acid
metabolism in health and disease. Physiol Rev 90: 207-258, 2010.
Luptak I, Balschi JA, Xing Y, Leone TC, Kelly DP, and Tian R. Decreased contractile and metabolic
reserve in peroxisome proliferator–activated receptor-α–null hearts can be rescued by increasing
glucose transport and utilization. Circulation 112: 2339-2346, 2005.
Maslov MY, Chacko VP, Stuber M, Moens AL, Kass DA, Champion HC, and Weiss RG. Altered
high-energy phosphate metabolism predicts contractile dysfunction and subsequent ventricular
remodeling in pressure-overload hypertrophy mice. Am J Physiol Heart Circ Physiol 292: H387-391,
2007.
Mildaziene V, Baniene R, Nauciene Z, Marcinkeviciute A, Morkuniene R, Borutaite V,
Kholodenko B, and Brown GC. Ca2+ stimulates both the respiratory and phosphorylation subsystems
in rat heart mitochondria. Biochem J 320 ( Pt 1): 329-334, 1996.
Motaal AG, Coolen BF, Abdurrachim D, Castro RM, Prompers JJ, Florack LM, Nicolay K, and
Strijkers GJ. Accelerated high-frame-rate mouse heart cine-MRI using compressed sensing
reconstruction. NMR Biomed doi 10.1002/nbm.2883, 2012.
Neubauer S. The failing heart--an engine out of fuel. N Engl J Med 356: 1140-1151, 2007.
Neubauer S, Horn M, Cramer M, Harre K, Newell JB, Peters W, Pabst T, Ertl G, Hahn D,
Ingwall JS, and Kochsiek K. Myocardial phosphocreatine-to-ATP ratio is a predictor of mortality in
patients with dilated cardiomyopathy. Circulation 96: 2190-2196, 1997.
Neubauer S, Krahe T, Schindler R, Horn M, Hillenbrand H, Entzeroth C, Mader H, Kromer EP,
Riegger GA, Lackner K, and et al. 31P magnetic resonance spectroscopy in dilated cardiomyopathy
and coronary artery disease. Altered cardiac high-energy phosphate metabolism in heart failure.
Circulation 86: 1810-1818, 1992.
Oakes ND, Thalen P, Aasum E, Edgley A, Larsen T, Furler SM, Ljung B, and Severson D.
Cardiac metabolism in mice: tracer method developments and in vivo application revealing profound
metabolic inflexibility in diabetes. Am J Physiol Endocrinol Metab 290: E870-881, 2006.
Pereira RO, Wende AR, Olsen C, Soto J, Rawlings T, Zhu Y, Anderson SM, and Abel ED.
Inducible overexpression of GLUT1 prevents mitochondrial dysfunction and attenuates structural
remodeling in pressure overload but does not prevent left ventricular dysfunction. J Am Heart Assoc 2:
e000301, 2013.
Cardiac metabolic adaptations in heart failure
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
Pereira RO, Wende AR, Olsen C, Soto J, Rawlings T, Zhu Y, Riehle C, and Abel ED. GLUT1
deficiency in cardiomyocytes does not accelerate the transition from compensated hypertrophy to
heart failure. J Mol Cell Cardiol 2014.
Randle PJ, Garland PB, Hales CN, and Newsholme EA. The glucose fatty-acid cycle. Its role in
insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1: 785-789, 1963.
Reader AJ, Ally S, Bakatselos F, Manavaki R, Walledge RJ, Jeavons AP, Julyan PJ, Sha Z,
Hastings DL, and Zweit J. One-pass list-mode EM algorithm for high-resolution 3-D PET image
reconstruction into large arrays. Nuclear Science, IEEE Transactions on 49: 693-699, 2002.
Rosca MG, Vazquez EJ, Kerner J, Parland W, Chandler MP, Stanley W, Sabbah HN, and
Hoppel CL. Cardiac mitochondria in heart failure: decrease in respirasomes and oxidative
phosphorylation. Cardiovasc Res 80: 30-39, 2008.
Schafers KP, Reader AJ, Kriens M, Knoess C, Schober O, and Schafers M. Performance
evaluation of the 32-module quadHIDAC small-animal PET scanner. J Nucl Med 46: 996-1004, 2005.
Scheuermann-Freestone M, Madsen PL, Manners D, Blamire AM, Buckingham RE, Styles P,
Radda GK, Neubauer S, and Clarke K. Abnormal cardiac and skeletal muscle energy metabolism in
patients with type 2 diabetes. Circulation 107: 3040-3046, 2003.
Sharov VG, Goussev A, Lesch M, Goldstein S, and Sabbah HN. Abnormal mitochondrial function
in myocardium of dogs with chronic heart failure. J Mol Cell Cardiol 30: 1757-1762, 1998.
Sharov VG, Todor AV, Silverman N, Goldstein S, and Sabbah HN. Abnormal mitochondrial
respiration in failed human myocardium. J Mol Cell Cardiol 32: 2361-2367, 2000.
Shipp JC, Opie LH, and Challoner D. Fatty acid and glucose metabolism in the perfused heart.
Nature 189: 1018-1019, 1961.
Sidell RJ, Cole MA, Draper NJ, Desrois M, Buckingham RE, and Clarke K. Thiazolidinedione
treatment normalizes insulin resistance and ischemic injury in the zucker Fatty rat heart. Diabetes 51:
1110-1117, 2002.
Stefan D, Cesare FD, Andrasescu A, Popa E, Lazariev A, Vescovo E, Strbak O, Williams S,
Starcuk Z, Cabanas M, van Ormondt D, and Graveron-Demilly D. Quantitation of magnetic
resonance spectroscopy signals: the jMRUI software package. Meas Sci Technol 20: 104035, 2009.
Stegger L, Heijman E, Schafers KP, Nicolay K, Schafers MA, and Strijkers GJ. Quantification of
left ventricular volumes and ejection fraction in mice using PET, compared with MRI. J Nucl Med 50:
132-138, 2009.
Steinbusch LK, Luiken JJ, Vlasblom R, Chabowski A, Hoebers NT, Coumans WA, Vroegrijk IO,
Voshol PJ, Ouwens DM, Glatz JF, and Diamant M. Absence of fatty acid transporter CD36 protects
against Western-type diet-related cardiac dysfunction following pressure overload in mice. Am J
Physiol Endocrinol Metab 301: E618-627, 2011.
Taegtmeyer H, Golfman L, Sharma S, Razeghi P, and van Arsdall M. Linking gene expression to
function: metabolic flexibility in the normal and diseased heart. Ann N Y Acad Sci 1015: 202-213,
2004.
van Bilsen M, van Nieuwenhoven FA, and van der Vusse GJ. Metabolic remodelling of the failing
heart: beneficial or detrimental? Cardiovasc Res 81: 420-428, 2009.
van Nierop BJ, van Assen HC, van Deel ED, Niesen LB, Duncker DJ, Strijkers GJ, and Nicolay
K. Phenotyping of left and right ventricular function in mouse models of compensated hypertrophy and
heart failure with cardiac MRI. PLoS ONE 8: e55424, 2013.
Weiss RG, Bottomley PA, Hardy CJ, and Gerstenblith G. Regional myocardial metabolism of highenergy phosphates during isometric exercise in patients with coronary artery disease. N Engl J Med
323: 1593-1600, 1990.
Zhong M, Alonso CE, Taegtmeyer H, and Kundu BK. Quantitative PET imaging detects early
metabolic remodeling in a mouse model of pressure-overload left ventricular hypertrophy in vivo. J
Nucl Med 54: 609-615, 2013.
Zhou YT, Grayburn P, Karim A, Shimabukuro M, Higa M, Baetens D, Orci L, and Unger RH.
Lipotoxic heart disease in obese rats: implications for human obesity. Proc Natl Acad Sci U S A 97:
1784-1789, 2000.
167
Chapter 10
Summarizing discussion and future perspective
Chapter 10
The role of altered fatty acid metabolism in
cardiomyopathy and heart failure
An in vivo magnetic resonance imaging and spectroscopy approach
Alterations in cardiac substrate metabolism have been implicated in the development of diabetic
and obesity-related cardiomyopathy. In the healthy heart, fatty acid oxidation (FAO) provides
~60-90% of energy, while the diabetic heart almost exclusively relies on FAO for energy
production. The increased preference towards FAO in the diabetic heart has been shown to be
associated with reduced function, suggesting that high FAO is detrimental for the heart.
However, recent data show that maintaining FAO during the progression of heart failure may be
beneficial, and that increased glucose oxidation further exacerbates the heart failure status.
Data in this field, however, largely originate from ex vivo perfused heart studies. In the
perfused heart setup, the heart is supplied with a constant concentration of substrates, while in
vivo, the types and the concentrations of substrates may change during the progression of heart
failure. The ex vivo setup is therefore not representative of the in vivo condition, and moreover,
does not allow to study the time course of changes in cardiac energy metabolism in relation to
the decline in cardiac function. The use of in vivo techniques is therefore crucial to study the
time dependence of metabolic adaptations in cardiomyopathy and heart failure.
The aim of this thesis was to elucidate the role of FAO in the development of cardiomyopathy
and heart failure in mice, using state-of-the-art in vivo magnetic resonance imaging (MRI) and
spectroscopy (MRS) techniques. In Chapter 2, we extensively reviewed the current progress in
in vivo MRI, 1H MRS, and
31
P MRS techniques for cardiac applications in mice. In Chapter 3, we
studied the current literature on the effect of alterations in fatty acid (FA) metabolism on
cardiac function. We reviewed data on the metabolic shift during the development of ischemic
and non-ischemic heart failure, and how the modulation of FA metabolism at different levels
along the FA metabolic pathway affects cardiac function.
Implementation of in vivo cardiac MRI, 1H MRS, and
31
P MRS
MRI is the gold standard method to assess cardiac function. However, mouse cardiac imaging is
very challenging because of the small size of the mouse heart, as well as the high mouse heart
and breathing rates. The time resolution of routine mouse cardiac MRI is usually not sufficient to
visualize the two diastolic filling phases necessary for the assessment of diastolic function.
Diastolic function measurements are however very important, as diabetic or obesity-related
cardiomyopathy is usually preceded by sub-clinical diastolic dysfunction at the early stage.
Therefore, we developed high temporal resolution cardiac MRI to measure diastolic function in
mice, as described in Chapter 4. Reduced cardiac function in diabetic db/db mice has mainly
been observed in ex vivo setups, while it often remains undetected in vivo. In a previous study,
in vivo cardiac function of db/db mice was normal even at 18 weeks of age, and cardiomyopathy
was only revealed when the heart was under dobutamine stress (6). Using high temporal
resolution cardiac MRI, we detected subtle impairments in diastolic function in diabetic db/db
mice.
Increased FA availability in diabetes leads to increased myocardial FA uptake, which is not
matched by a sufficient increase in FAO. An imbalance between FA uptake and oxidation may
170
Summarizing discussion and future perspective
result in myocardial lipid accumulation, which could lead to lipotoxicity-induced cardiac
dysfunction. In Chapter 5, we implemented a localized 1H MRS technique using point resolved
spectroscopy (PRESS), to quantify myocardial triglyceride (TG) content in vivo in a 1 x 2 x 2
mm3 voxel in the septum of the heart (Fig. 10.1A). We reported that diabetic db/db mice had
indeed higher myocardial TG content compared with control mice, which confirmed ex vivo
biochemical findings from a previous study (1). We also showed that treatment using the antidiabetic drug pioglitazone normalized myocardial TG levels in the diabetic db/db mice.
Pioglitazone has been shown to improve diabetes by increasing FA uptake and oxidation in
adipose tissue and by redirecting lipids from ectopic sites into adipose tissue (26), which could
explain the reduction in ectopic lipid accumulation in the myocardium.
Figure 10.1. Increased FAO has been associated with (A) excessive myocardial lipid accumulation,
when the increased FAO does not match the increase in FA supply, and (B) lower cardiac energetics,
which may both contribute to (C) the development of cardiomyopathy. In vivo 1H MRS, 31P MRS, and
cine MRI techniques have been implemented to investigate these mechanisms.
Increased FAO has also been associated with reduced cardiac energy efficiency, i.e. reduced ATP
production per oxygen molecule consumed. On top of the intrinsic oxygen-inefficiency of FAO
(i.e. compared to glucose oxidation, palmitate oxidation produces ~10% less ATP per oxygen
molecule consumed), FA may induce mitochondrial uncoupling, in which respiration is less
coupled with ATP synthesis and energy from substrate oxidation is dissipated as heat. Cardiac
energetics can be assessed using
31
P MRS, to determine the phosphocreatine-to-ATP (PCr/ATP)
ratio as a measure of cardiac energy status (Fig. 10.1B). Reduced myocardial PCr/ATP ratio has
been observed in diabetic patients. In patients with dilated cardiomyopathy, a lower myocardial
PCr/ATP ratio has been shown to be a strong predictor of mortality. In Chapter 6, we describe
the implementation of 3D image selected in vivo spectroscopy (ISIS) to determine the
171
Chapter 10
myocardial PCr/ATP ratio in the mouse heart. We validated our technique in a mouse model of
pressure overload heart failure induced by transverse aortic constriction (TAC), and showed that
the myocardial PCr/ATP ratio was indeed lower in mice after 7 weeks of TAC compared with
controls.
Applications in mouse models of altered FAO
Using the developed high temporal resolution cardiac MRI, 1H MRS, and
31
P MRS techniques, we
then investigated the effect of altered FAO on cardiac function and metabolism in different
mouse models. In Chapter 7, we studied a mouse model of diet-induced obesity, promoted by
20 weeks of high fat diet (HFD) feeding. We observed that HFD-fed mice exhibited hypertrophy
and cardiac diastolic dysfunction. HFD feeding promoted mitochondrial adaptations for increased
utilization of FA, which however could not match the increased FA supply. Increased levels of
myocardial TG, and potentially toxic lipid intermediates i.e. diacylglycerols and long-chain
acylcarnitines, were observed in HFD-fed mice, while in vivo cardiac PCr/ATP ratio and intrinsic
mitochondrial function were not impaired. Our data suggest that the diastolic dysfunction in
HFD-fed mice was associated with lipotoxicity-induced myocardial oxidative stress, fibrosis, and
disturbed calcium homeostasis, rather than impaired cardiac energetics.
In Chapter 8, we investigated cardiac function and metabolism in a novel mouse model of
altered FAO, the Cpt1bE3A/E3A knockin mouse. Cpt1b is an enzyme which is essential for the
transport of FA into the mitochondria, and of which the activity is inhibited by malonyl-CoA. We
showed that Cpt1bE3A/E3A has a lower sensitivity to malonyl-CoA compared with Cpt1bWT/WT,
resulting in an increased FAO flux in hearts of Cpt1bE3A/E3A knockin mice, as observed using
perfused heart setups. Increased FAO flux was paralleled by decreased cardiac energy status,
increased LV mass, and reduced cardiac function in vivo. In contrast to the HFD-fed mice,
increased FAO in the Cpt1bE3A/E3A knockin mice was not associated with increased FA supply.
Consequently, myocardial TG levels were not affected. However, there was a trend for increased
long-chain acylcarnitine levels, which indicates that the increased FAO flux was not paralleled by
concomitant increases in the downstream pathways, such as the oxidative phosphorylation.
Experiments on isolated heart mitochondria of Cpt1bE3A/E3A mice showed suppressed oxidative
capacity for both glucose- and FA-derived substrates, which may explain the compromised in
vivo cardiac energetics in these mice.
The previous two chapters showed that increased FAO can be associated with both lipotoxicity
and impaired cardiac energetics, and that either mechanism may independently contribute to
cardiac dysfunction (Fig. 10.1A-C). These data support studies in diabetic mice showing that
increased reliance on FAO may be detrimental for the heart. Studies in wild-type (non-diabetic)
mice showed that the development of heart failure is generally accompanied by a shift away
from FAO towards glucose oxidation (2, 4, 5, 16, 23, 27). However, less is known about how
the diabetic heart adapts during heart failure development. The diabetic heart is proposed to be
metabolically inflexible to switch to glucose oxidation. Thus, if the switch towards glucose
oxidation is a necessary compensatory rescue mechanism, the progression to heart failure will
probably be accelerated in the diabetic heart.
172
Summarizing discussion and future perspective
In Chapter 9, we followed cardiac metabolic and functional adaptations in non-diabetic and
diabetic mice at different time points during the progression of heart failure. Mice were
subjected to TAC-induced pressure overload, and measured before, and at 1, 5, and 12 weeks
post TAC. In addition to MRI and MRS measurements, we used positron emission tomography
(PET) with
18
F-fluorodeoxyglucose (18F-FDG) administration to measure myocardial glucose
uptake in vivo. In non-diabetic mice, we showed that TAC induced progressive LV hypertrophy
and a decline in cardiac function. In the short term, this was correlated with increased
myocardial FDG uptake, while cardiac energy status was still maintained. In the long term (i.e.
12 weeks post TAC), myocardial FDG uptake tended to reduce, which was accompanied with a
decrease in cardiac energy status. The impaired cardiac energetics may be a chronic effect of
TAC, which occurs at the later stage, when total substrate oxidation might possibly be
suppressed (summarized in Fig. 10.2).
Figure 10.2. Time-dependent changes in LV mass, ejection fraction, myocardial
lipid content, myocardial FDG uptake, and myocardial PCr/ATP ratio during the
development of pressure-overload heart failure in non-diabetic mice. Pressureoverload induces increased LV mass and reduced ejection fraction. Myocardial lipid
content is not altered. In the short term, changes in cardiac morphology and
function are associated with increased myocardial FDG uptake, while myocardial
PCr/ATP ratio is maintained. At a later stage, myocardial FDG uptake is reduced,
together with a decrease in PCr/ATP ratio.
Compared with non-diabetic mice at baseline, diabetic mice had 51% lower FDG uptake, 42%
higher myocardial lipid content, 40% lower PCr/ATP ratio, but normal cardiac function.
Surprisingly, TAC only mildly affected cardiac function in the diabetic mice. The PCr/ATP ratio
was maintained in the diabetic mice up to 12 weeks post TAC. Similar to non-diabetic mice,
myocardial FDG uptake was also increased in the diabetic mice upon TAC compared with
baseline, suggesting that metabolic flexibility was not completely lost in the diabetic mice,
although total myocardial FDG uptake remained lower than in the non-diabetic mice. These
findings in the diabetic mice upon TAC suggest that increased FAO may not necessarily be
detrimental during the development of heart failure.
173
Chapter 10
In summary, MRI and MRS techniques proved to be powerful tools in studying cardiac function
and metabolism in mice. We showed in mouse models of altered FAO that a mismatch between
the upstream and downstream pathways of FA metabolism may lead to lipotoxicity-induced
cardiomyopathy associated with the excessive accumulation of lipid intermediates. Additionally,
increased FAO may also be associated with reduced cardiac energy status. However, during
pressure overload, maintained FAO may have a protective role, as shown by preserved cardiac
function and energetics in the diabetic mice upon TAC. The mechanisms by which FAO is
protective during pressure overload, however, warrant further investigation.
Future perspective
In this thesis, we set up non-invasive MRI, MRS, and PET techniques to assess in vivo cardiac
function and metabolism. These in vivo techniques allow longitudinal mapping of cardiac
metabolic and functional changes during the development and progression of heart failure in the
same animals, which provides a platform to investigate the cause-and-effect relationships
between derangements in cardiac metabolism and function in heart failure. Using in vivo PET,
we observed increased myocardial FDG uptake in mice upon TAC, suggesting a shift from FAO
towards glucose utilization upon TAC-induced pressure overload. The increase in myocardial
FDG uptake correlated strongly with a decline in cardiac function of non-diabetic mice as
determined by MRI. Interestingly, these changes in myocardial FDG uptake and function were
already observed as early as one week after TAC, therefore preceding the reduction in cardiac
energy status detected using
31
P MRS at 12 weeks after TAC. As one week post TAC was our
earliest time point measured, our study setup prevented us to investigate the cause-and-effect
relationship between derangements in myocardial FDG uptake and cardiac function. Therefore,
future studies should include measurements at even earlier time points (<1 week) after
pressure overload induction. Furthermore, it will be interesting to map the in vivo cardiac
metabolic and functional changes during development of heart failure from various etiologies,
such as ischemic (e.g. myocardial infarction) versus non-ischemic (e.g. pressure overload) heart
failure.
With regard to studies in ischemic heart failure, the development and implementation of in vivo
MRI and MRS methods which give insight into spatial variations in functional and metabolic
parameters are necessary (25). Unlike TAC, which gives a rather homogenous effect on the
myocardium, a myocardial ischemic insult will differentially affect the infarcted and remote
regions. While cine MRI assesses global cardiac function, methods such as MR tagging can be
useful to study regional deformation in myocardium (7, 13). Furthermore, spatially encoded
MRS techniques, such as 2D-chemical shift imaging (2D-CSI) can be implemented to assess
spatial variations in e.g. myocardial lipid content and cardiac energy status (10).
Current strategies in heart failure treatment aim at inhibition of FAO, however, with varying
degrees of success (9, 17). Heart failure involves complex mechanisms, in which the associated
changes in cardiac metabolism may likely depend both on the cause and the stage of heart
failure (18). For example, whereas inhibition of FAO during reperfusion after a myocardial
ischemic insult is beneficial, inhibition of FAO at the later stage of (non-ischemic) heart failure is
probably not optimal, as mitochondrial oxidative capacity is known to be reduced and every
contribution to energy production is crucial. Furthermore, it has been shown that a depletion of
174
Summarizing discussion and future perspective
free fatty acids (FFA) acutely decreases cardiac work and efficiency in heart failure (24),
questioning the approach of FAO inhibition as an effective intervention in heart failure. This is
also supported by the finding that maintained/high FAO may actually be beneficial to prevent
deterioration of cardiac function and energetics in the development of heart failure (15, 16, 21),
as we also observed in our diabetic mice upon pressure overload. These observations support
the importance of FAO, which is also the main energy providing pathway in the healthy heart.
Therefore, a comprehensive mapping of cardiac metabolic changes during the development of
heart failure, in various etiologies of HF, may aid in finding new targets for heart failure
treatment. Furthermore, such detailed characterization would open up possibilities to tailor
treatment to different causes and stages of heart failure.
The observation on the metabolic shift towards glucose during heart failure in this thesis is
based on myocardial glucose uptake, which was measured using PET with administration of the
glucose analogue
18
F-FDG. Changes in myocardial FA uptake were however not measured. The
use of FA PET tracers in future studies will be invaluable to gain insight into myocardial FA
utilization. Several
18
F-FDG,
18
18
F-based FA PET tracers have been developed (20); unfortunately, unlike
F-FA tracers are not widely and commercially available, thus limiting their
application. Other than
11
18
F-based PET tracers,
11
C-based PET tracers, such as
C-palmitate, can also be used. However, the application of
by the short half-life of
11
11
11
C-glucose and
C-based tracers is often limited
C (20.4 minutes), which thus requires the availability of a cyclotron on
site. As an alternative to PET, emerging
13
13
C MRS techniques using hyperpolarized
13
C-labeled
13
substrates, such as [1- C]-pyruvate and [1- C]-butyrate, have been demonstrated to be
powerful in studying cardiac substrate metabolism in vivo (3, 8, 14, 22). Chemical shift in the
13
C MR spectra allows identification of different metabolites, providing the opportunity to
measure various fluxes in the metabolic pathway.
Myocardial substrate oxidation is essential for energy production. It is widely accepted that the
failing heart is ‘energy starved’, and that its cardiac energy status, i.e. the PCr/ATP ratio, is
reduced (12, 19). Using
31
P MRS, we showed that although an increase in myocardial glucose
uptake did not directly lead to a reduced PCr/ATP ratio in the early stage of pressure overload,
the PCr/ATP ratio tended to decrease at a later stage of pressure overload when myocardial
glucose uptake also decreased. However, the use of the PCr/ATP ratio may obscure changes in
the absolute concentrations of PCr and/or ATP. It was shown previously that in fact both PCr
and ATP concentrations were decreased in the failing heart, but that the decrease in PCr
concentration
was
greater
than
the
decrease
in
the
ATP
concentration
(11).
Thus,
measurements of time-dependent changes in the absolute concentrations of PCr and ATP would
provide important insight into changes in cardiac energetics. Therefore, further development of
the
31
P-MRS technique to quantify absolute PCr and ATP concentrations in vivo should be a
priority in future work.
In conclusion, further advancement and application of non-invasive in vivo measurement
techniques, such as MRI, MRS, and PET, would allow a more complete characterization of
cardiac functional and metabolic changes during the development of heart failure. In turn, this
will assist in elucidating the etiology of heart failure and in finding new treatment strategies.
175
Chapter 10
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
176
Aasum E, Hafstad AD, Severson DL, and Larsen TS. Age-dependent changes in metabolism,
contractile function, and ischemic sensitivity in hearts from db/db mice. Diabetes 52: 434-441, 2003.
Allard MF, Schonekess BO, Henning SL, English DR, and Lopaschuk GD. Contribution of
oxidative metabolism and glycolysis to ATP production in hypertrophied hearts. Am J Physiol 267:
H742-750, 1994.
Ball DR, Rowlands B, Dodd MS, Le Page L, Ball V, Carr CA, Clarke K, and Tyler DJ.
Hyperpolarized butyrate: a metabolic probe of short chain fatty acid metabolism in the heart. Magn
Reson Med 71: 1663-1669, 2014.
Christe ME, and Rodgers RL. Altered glucose and fatty acid oxidation in hearts of the spontaneously
hypertensive rat. J Mol Cell Cardiol 26: 1371-1375, 1994.
Christian B, El Alaoui-Talibi Z, Moravec M, and Moravec J. Palmitate oxidation by the
mitochondria from volume-overloaded rat hearts. Mol Cell Biochem 180: 117-128, 1998.
Daniels A, van Bilsen M, Janssen BJA, Brouns AE, Cleutjens JPM, Roemen THM, Schaart G,
van der Velden J, van der Vusse GJ, and van Nieuwenhoven FA. Impaired cardiac functional
reserve in type 2 diabetic db/db mice is associated with metabolic, but not structural, remodelling.
Acta Physiol 200: 11-22, 2010.
Del-Canto I, Lopez-Lereu MP, Monmeneu JV, Croisille P, Clarysse P, Chorro FJ, Bodi V, and
Moratal D. Characterization of normal regional myocardial function by MRI cardiac tagging. J Magn
Reson Imaging 2013.
Dodd MS, Ball V, Bray R, Ashrafian H, Watkins H, Clarke K, and Tyler DJ. In vivo mouse cardiac
hyperpolarized magnetic resonance spectroscopy. J Cardiovasc Magn Reson 15: 19, 2013.
Fillmore N, and Lopaschuk GD. Targeting mitochondrial oxidative metabolism as an approach to
treat heart failure. Biochim Biophys Acta 1833: 857-865, 2013.
Flogel U, Jacoby C, Godecke A, and Schrader J. In vivo 2D mapping of impaired murine cardiac
energetics in NO-induced heart failure. Magn Reson Med 57: 50-58, 2007.
Gupta A, Chacko VP, and Weiss RG. Abnormal energetics and ATP depletion in pressure-overload
mouse hearts: in vivo high-energy phosphate concentration measures by noninvasive magnetic
resonance. Am J Physiol Heart Circ Physiol 297: H59-64, 2009.
Ingwall JS. Energy metabolism in heart failure and remodelling. Cardiovasc Res 81: 412-419, 2009.
Jeung MY, Germain P, Croisille P, El ghannudi S, Roy C, and Gangi A. Myocardial tagging with
MR imaging: overview of normal and pathologic findings. Radiographics 32: 1381-1398, 2012.
Josan S, Park JM, Hurd R, Yen YF, Pfefferbaum A, Spielman D, and Mayer D. In vivo
investigation of cardiac metabolism in the rat using MRS of hyperpolarized [1-13C] and [2-13C]pyruvate.
NMR Biomed 26: 1680-1687, 2013.
Kolwicz SC, Jr., Olson DP, Marney LC, Garcia-Menendez L, Synovec RE, and Tian R. Cardiacspecific deletion of acetyl CoA carboxylase 2 prevents metabolic remodeling during pressure-overload
hypertrophy. Circ Res 111: 728-738, 2012.
Labinskyy V, Bellomo M, Chandler MP, Young ME, Lionetti V, Qanud K, Bigazzi F, Sampietro
T, Stanley WC, and Recchia FA. Chronic activation of peroxisome proliferator-activated receptoralpha with fenofibrate prevents alterations in cardiac metabolic phenotype without changing the onset
of decompensation in pacing-induced heart failure. J Pharmacol Exp Ther 321: 165-171, 2007.
Lionetti V, Stanley WC, and Recchia FA. Modulating fatty acid oxidation in heart failure. Cardiovasc
Res 90: 202-209, 2011.
Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, and Stanley WC. Myocardial fatty acid
metabolism in health and disease. Physiol Rev 90: 207-258, 2010.
Neubauer S. The failing heart--an engine out of fuel. N Engl J Med 356: 1140-1151, 2007.
Pandey MK, Bansal A, and DeGrado TR. Fluorine-18 labeled thia fatty acids for PET imaging of fatty
acid oxidation in heart and cancer. Heart Metab 51: 15-19, 2011.
Pereira RO, Wende AR, Olsen C, Soto J, Rawlings T, Zhu Y, Riehle C, and Abel ED. GLUT1
deficiency in cardiomyocytes does not accelerate the transition from compensated hypertrophy to
heart failure. J Mol Cell Cardiol 2014.
Rider OJ, and Tyler DJ. Clinical implications of cardiac hyperpolarized magnetic resonance imaging. J
Cardiovasc Magn Reson 15: 93, 2013.
Taegtmeyer H, and Overturf ML. Effects of moderate hypertension on cardiac function and
metabolism in the rabbit. Hypertension 11: 416-426, 1988.
Tuunanen H, Engblom E, Naum A, Nagren K, Hesse B, Airaksinen KE, Nuutila P, Iozzo P,
Ukkonen H, Opie LH, and Knuuti J. Free fatty acid depletion acutely decreases cardiac work and
efficiency in cardiomyopathic heart failure. Circulation 114: 2130-2137, 2006.
Weiss RG, Bottomley PA, Hardy CJ, and Gerstenblith G. Regional myocardial metabolism of highenergy phosphates during isometric exercise in patients with coronary artery disease. N Engl J Med
323: 1593-1600, 1990.
Yang KJ, Noh JR, Kim YH, Gang GT, Hwang JH, Yang SJ, Yeom YI, and Lee CH. Differential
modulatory effects of rosiglitazone and pioglitazone on white adipose tissue in db/db mice. Life Sci 87:
405-410, 2010.
Zhong M, Alonso CE, Taegtmeyer H, and Kundu BK. Quantitative PET imaging detects early
metabolic remodeling in a mouse model of pressure-overload left ventricular hypertrophy in vivo. J
Nucl Med 54: 609-615, 2013.
Acknowledgements
Acknowledgements
“If you want to go fast, go alone. If you want to go far, go together.”
-African proverb
I am very fortunate to have had the opportunity to do my PhD in the Biomedical NMR group of
Prof. Dr. Klaas Nicolay. Dear Klaas, it was really a privilege to be part of your group. I
immensely enjoyed the stimulating work environment and the nice atmosphere in which
everyone works hand-in-hand. Thank you for the invaluable support, discussion, and advice for
my personal and professional growth.
My PhD would not have been the same without the support of my co-promotor Dr. Jeanine
Prompers. Dear Jeanine, no words can describe my gratitude. Thank you for trusting me to be
your PhD student. I admire your attention to details, your critical and logical thinking, your
writing skills, and your well-organized working style. Thank you for always being available for
discussion and help behind the scanner. I enjoyed our trips to Padua, Salt Lake City, and the
countless NVDO trips. But foremost, I am thankful for your support throughout my PhD that
extended beyond the scientific content. People say that doing a PhD is like raising a baby, but
you made it possible for me to raise two during this time. Your support really made combining
research and motherhood not only doable, but also enjoyable, and something that I will always
treasure. Oh, and thank you for your support for my third baby, angklung! I am indeed very
lucky to have you in this special period of my life!
I remember attending a lecture on ‘The Fundamental of Biomedical Engineering’ during my
undergraduate study in Indonesia. My professor (Prof. Dr. Soegijardjo Soegijoko, to whom I am
thankful for his encouragement for me to pursue a graduate study) could not stress enough that
Biomedical Engineering is a multidisciplinary field. Indeed, during my PhD I experienced it firsthand that Biomedical Engineering is not a field for a one-man show. This thesis is a collection of
five year work worth of fruitful collaborations, which I owe to the people whom I had a great
privilege to work with.
Dear Ot, we have been working together since the beginning of my PhD. Thank you so much for
the chat and your support during ups and downs. I had a good time during our trips to Padua
and Oxford. It was really nice working with you on the development of the
31
P MRS method in
Chapter 6. It was a long way from ‘grass’ to the spectra that we have today! Thank you for the
work in Chapter 5 in this thesis. The 1H MRS method has proven successful! I am impressed by
your writing skills and your way of doing science. Your drive, work, and achievements during
your PhD really inspired me. Congratulations on winning the well-deserved Gorter Prize 2014!
Dear Miranda, your coming in the last stage of my PhD was a blessing. It was nice to have you
to brainstorm about experiment plans, to actually do it together, and to be excited about the
results. Your curiousity, enthusiasm, and optimism were very motivating. I really enjoyed our
work together: the PET pilot, the experiments in Mϋnster and Amsterdam, the interesting long
discussion, and –of course– surviving those 65++ hour experiments in 4 days (which we
repeated for 3 more times!) with some secret singing and dancing in the wee hours. Our work
resulted in Chapter 9 and led to a very special trip to Tromso in the arctic circle, during the
special time when the sun never sets. We are indeed the ‘DA-MN’ team! And oh, that review in
178
Acknowledgements
Chapter 3! I am glad that we decided to do it! I learned a lot from it. Thank you also for your
contribution to Chapter 7 and 8. I wish you the very best in your Veni research and your career!
Dear Bart, my officemate for 5+ years. It was really nice to start our PhD together, to go
through the PhD roller-coaster, and to share not only frustration but also hope. I really enjoyed
having you in the office –yes, including those writing sessions in the weekend :). Your presence
and initiative always enlightened our group. The writers club was very useful and stimulating!
Thanks for your support, help, and great company in and out of the lab (moving house –
including carrying the super heavy piano, the first angklung workshop, NVDO trips), and for
many interesting chats on science, work, and life. Thank you for the respiration measurements
in Chapter 7! I wish you the very best for your PhD defense and career!
Dear Sharon, it was nice sharing the office with you after the move to HTC. I enjoyed your
company very much: the science and non-science chats, the ice cream, sports, and shopping
time, the biking during the Ardennen weekend, and the yearly NVDO trips. You were always
around to help; I am really happy that we could help each other when performing biochemical
assays. All the best with finishing up your PhD!
Dear Jolita, we worked together on the high-fat diet study. I learned a lot from your efficient
way of working in the lab. Your expertise in biochemical experiments and your thorough
understanding in the field contributed to a very nice article in Chapter 7.
Dear Gustav, you were always available for all MRI-related questions and help at the scanner. I
am happy that I had a chance to work with you, particularly on Chapter 4 with Bram. Dear
Bram, I had a good time working with you and supervising a student project together. Dear
Abdallah, I applied your reconstruction techniques not only to Chapter 4, but also to my other
experimental work. Thank you for your availability for reconstruction-related questions (and for
‘No problem, Desi, at least you have the k-space!’ ;)), and for the Egyptian dinner and karaoke.
Dear Bastiaan, thank you for your contribution to the
31
P work in Chapter 6. It was also nice
arranging the Ardennen weekend together! Dear Inge, your master project served as a basis for
the
31
P work. Dear Tom G, I enjoyed building the
31
P coil with you, thank you for teaching me
1
everything. Your master thesis work on H MRS contributed to Chapter 5.
The work in Chapter 8 was based on the collaboration with Dr. Sander Houten. Dear Sander,
thank you for the opportunity to work with you on your Cpt1b mouse model. Thank you for your
insight and discussion, your willingness to be on my PhD committee, and your invaluable
feedback on the thesis. I would also like to thank you for your contribution to Chapter 5 and 6.
Dear Michel and Riekelt, thank you for the breeding, the vast amount of (biochemical) work on
the characterization of the mice, and the discussion. I would also like to thank Dr. Coert
Zuurbier and Rianne Nederlof for the perfused heart measurements. It was nice to visit your lab
and to see your delicate work. Dear Coert, thank you for teaching me how to fish in the
Norwegian waters around Tromso, and for the drink and nice discussion.
Regarding the work in Chapter 9, I am very grateful for the opportunity to collaborate with the
groups of Prof. Dr. Cornelius Faber and Prof. Dr. Michael Schäfers in Mϋnster, Germany. Dear
179
Acknowledgements
Prof. Dr. Cornelius Faber, thank you for your willingness to become one of my committee
members and your feedback on the thesis. Dear Verena and Philipp, I really appreciate your
hard work on our project –never out-of-trouble, long MR measurements. I enjoyed our
discussion and time together behind the scanner; it was never boring! Dear Verena, your
enthusiasm and dedication to science are really admirable! Dear Sven and Michael K, thank you
for the PET measurements, surgery, and invaluable discussion. Dear Roman, Sarah, Nina,
Florian, Ingrid, I am thankful for your (technical) support for the experiments. Dear Saeedeh,
my stay in Mϋnster was never lonely because of your company. Thank you for inviting me to
stay with you, for the chat and the delicious Iranian dishes!
Dear Katia and Suzanne, thank you for teaching me how to work with PET and how to prepare
radiotracers during the pilot PET experiments at HTC. I learned a lot and enjoyed it very much.
Dear Prof. Dr. Jan Glatz, I really appreciate your contribution to the review in Chapter 3! I also
thank you for your willingness to be on my PhD committee. Dear Dr. Joost Luiken, I am grateful
for your contribution for the review in Chapter 3. Thank you for the intramyocellular lipid
determination and your invaluable feedback to Chapter 7.
I am thankful to Prof. Dr. Jolanda van der Velden and Dr. Etto Eringa for the try-out on
myocardial stiffness measurements.
I would also like to thank Prof. Dr. Peter Hilbers, Prof. Dr. Jan Smit, and Prof. Frans van de
Vosse, for their willingness to serve on my PhD committee.
Dear Leonie N, David, and Jo, thank you for your all-around support in the animal lab. Leonie, I
am really happy that our (tricky) surgeries worked out really well! Dear Larry, your technical
support with the coil, setup preparation, and scanner was invaluable, thank you!
Dear Sin Yuin, it was always nice to share thoughts and ‘curhat’ with you about work, finishing
PhD, and life. Thanks for your support (I remember you and Igor soothing a hungry and sleepy
baby outside, while I was doing my experiment in the lab), for dinners (and bringing me one
during one of my evening experiments), and for being the most loyal angklung supporter!
Dear Valentina, I enjoyed your company in and outside the lab. Your risotto is the best! Thanks
for your help with cutting tissues and histology. Richard, thank you for helping me getting
started with the 9.4 T. Dear Wolter, your enthusiasm is contagious, and thank you for the time
we spent with ultrasound. Dear Katrien, you were always encouraging and motivating. Dear
Floortje, the N-emmers were very lively with you around! Thank you for all the help with the
administration. Dear Rik, Igor, Stefanie, it was always nice to stop by your office and to have a
nice chat. Dear Tom S, thank you for letting me drag you into my coil work. Dear Tessa and
Leonie P, I learned some histological techniques from you. Dear Pedro, thank you for the chat
and dinner. Dear Martijn, it was great chasing ISMRM Benelux sponsors with you. I would also
like to thank Jules, Luc, Marloes, Siem, Steffie, Holger, Tiemen, Esther, Laura, Rene, and former
group members Ewelina, Nicole, Maarten, Glenda, Geralda, Erik, Sander, Hedwig, Prashant,
Wouter, for the nice work environment.
180
Acknowledgements
During my PhD, I had opportunities to work with a few students. Esther, your
contributed to the development of our
31
31
P-CSI work
P MRS method. It was admirable how much you could
accomplish in such a short period! Tessa, Sophie, and Angela, thank you for your work on the
method for absolute quantification of
31
P MRS. Hopefully the method will be up-and-running in
the near future. Rene, your literature review on PET cardiac metabolic studies is very useful for
our reference. Laura, thank you for your contribution in analyzing diastolic data.
Living far away from my home country, I am blessed to be surrounded by good friends and a
supportive community where I could always turn to. Dear Agnese, Maria, Angela, Anitha, and
Andrey, I started the chapter in The Netherlands with you around, and 8 years later, I am really
happy that we are still friends! For my families in Belgium and The Netherlands: Wa Ann and
Wa Francis, Tante Yvette, and Wa Enung, thank you for your support and encouragement!
Angklung Eindhoven made my life in Eindhoven more colourful. We started with nothing and
reached our dream one by one. Dear Teh Ida, thank you to jump right in at the idea of building
the group. From you, I learned to set the bar high and not to settle for less. I still remember
when we first put our steps at Frits Philips, what an unforgettable night! Dear Arnaud, Denny,
Burhan, Aji, thank you for always being available for the whole group. Dear Dody, thank you for
putting us into priority despite your busy schedule. You opened up opportunities and new
experiences for us (we certainly never dreamed to take part in an event for Princess Beatrix and
have a few seconds of fame). Dear Ibu Ine, you played a significant role during the
establishment of our group and I learned a lot from you. Pak Joz, thank you for your support for
us. And for the board members whom I had been lucky to work with: Dina W (thanks for the
deep conversation and laughter!), Pak Joz, Amel, Rizki, Burhan, Aji, Nabil, Ferry, Nisa, Brian,
Fitria, Ayuta, Dhanya, Mukhlis, Paskal, Linda, Tiaan. And for all members, without you all no
symphony was possible.
Dear Elena, every week I had (and still have) time with you, to stop the clock and to enjoy my
own time, to immerse in my own music-making. Thank you for your trust and for giving me
opportunities to exploit my full potential, and for encouraging me to do nothing but the best.
I also wish to thank to the Indonesian community in Eindhoven who always made me feel at
home in Eindhoven: Musihoven and ‘Kumpul Bocah’. They kept me sane and provided my family
with social support which, being far away from my extended family, was a luxury. Special
thanks are extended to the family of Intan–Zulfan, for taking care of Mentari regularly and
being her second family, also to the family of Dina–Ricky, Teh Rita–Mas Umar, Mbak Ina-Pak
Haji Rosyid, and Ibu Rosita.
For the occasion of my defense, I would like to thank Dina V, Kakak Bella, Teh Iin, Teh Rita, and
Mbak Nur for taking care of the reception and dinner. Dear Heather (and T!NT board), thank
you for making it possible to have the studentenkapel for the evening! Dear Tiaan, thank you
for your help in putting my ideas onto the cover of this thesis, and for putting up with my
countless and never-ending requests.
Dear Pak Nanang Harijanto, during my undergraduate study your office was always open for me.
Your encouragement motivated me to do research and pursue a graduate study. Thank you!
181
Acknowledgements
Going and living abroad had been my childhood dream. I remember looking at a picture of my
parents with magical autumn trees and ground covered with falling leaves. The brown-yellowish
colour of those leaves caught the imagination of the 9-year-old who was admiring it with
wonder. It lit her curiosity of how autumn smelt, how falling leaves on the ground felt on her
feet as she walked, and how the air felt on her face as she breathed. At that time, she promised
herself that she would, at one point of her life, go and greet autumn herself.
It was my parents’ faith in me that kept me going to keep my promise. I know that it was in
their prayers for me to be able to go conquer the world. So, I really owe it to them that I am
now living my childhood dream. Their love brought me the world, and I know that nothing could
pay it back. I do hope that they can capture a glimpse of my gratitude to them in this thesis, for
this journey would never happen without their blessing. –Ibu & Papay, this thesis is for you.
Dede, my beloved sister, even though we live in different hemispheres, you are never a text
message away. Thank you for your prayers, love, and support. We grew up sharing and learning
a lot of things together. I am happy that we still can do it although in a different way. I am very
proud of you, and wish you the very best with Edo, Faisal, and Alisha!
I would also like to thank my parents-in-law, Bapak dan Ibu di Malang, and my brothers- and
sister-in-law: Mbak Teti, Mas Iin, Mas Yoyon, Edo, Wa Maman–Wa Betty, Wa Mimin–Wa Iya,
Mang Arry-Bi Ratih, and my big family in Indonesia. Knowing that you are always around
warmed our stay here in The Netherlands. Sita and Iie, thank you for the beautiful friendship!
Mas Zalfany, our little family thrives because of your love, patience, and understanding. You
made it possible for me to do a PhD and nurture a family. I could not have asked for a better
husband whom I would like to spend the rest of my life with. Dear our Mentari, our sunshine, I
hope this thesis and this period in our family will once become a beautiful memory that you can
always treasure.
I received so much support during my PhD from so many people, that I might have
inadvertently missed someone in this acknowledgement. For those, please accept my sincere
apology and thanks.
Eindhoven, November 2014
Desiree
MAH
11/2014
Alhamdulillahi rabbil alamin
All praise and thanks be to Allah, the Lord of Existence
182
List of publications
List of publications
Journal publications
1. D. Abdurrachim, A.J. Bakermans, R.P.M. Moonen, A.G. Motaal, J.J. Prompers, G.J.
Strijkers, K. Vandoorne, K. Nicolay. Small animal MR imaging and spectroscopy of
cardiovascular disorder. Prog Nucl Magn Reson Spectrosc. Invited Review.
2. D. Abdurrachim, J.J.F.P. Luiken, K. Nicolay, J.F.C. Glatz, J.J. Prompers, M. Nabben.
Good and bad consequences of altered fatty acid metabolism in heart failure: Evidence
from mouse models. Submitted.
3. A.J. Bakermans*, D. Abdurrachim*, B.J. van Nierop, I. van der Kroon, G.J. Strijkers,
S.M. Houten, K. Nicolay, J.J. Prompers. In vivo
31
P MR spectroscopy of the mouse heart
using respiratory-gated and cardiac-triggered 3D ISIS. Submitted.
4. P. Bovenkamp, T. Brix, F. Lindemann, R. Holtmeier, D. Abdurrachim, M.T. Kuhlmann,
G.J. Strijkers, J. Stypmann, K. Hinrichs, V. Hoerr. Velocity mapping of the aortic flow at
9.4 T in healthy mice and mice with induced heart failure using time-resolved threedimensional phase contrast MRI (4D PC MRI). MAGMA 2014; DOI: 10.1007/s10334014-0466-z.
5. D. Abdurrachim*, J. Ciapaite*, B. Wessels, M. Nabben, J.J.F.P. Luiken, K. Nicolay,
J.J. Prompers. Cardiac diastolic dysfunction in high-fat diet fed mice is associated with
lipotoxicity without impairment of cardiac energetics in vivo. Biochim Biophys Acta
2014; 1841:1525-1537.
6. A.J. Bakermans, D. Abdurrachim, T.R. Geraedts, S.M. Houten, K. Nicolay,
J.J. Prompers. In vivo proton T1 relaxation times of mouse myocardial metabolites at
9.4 T. Magn Reson Med 2014; DOI: 10.1002/mrm.25340.
7. B. Coolen*, D. Abdurrachim*, A. Motaal, K. Nicolay, J.J. Prompers, G.J. Strijkers.
High frame-rate retrospectively triggered Cine MRI for assessment of murine diastolic
function. Magn Reson Med 2013; 69(3):648-56.
8. A. Motaal, B. Coolen, D. Abdurrachim, R.M. Castro, J.J. Prompers, L.M. Florack,
K. Nicolay, G.J. Strjikers. Accelerated high-frame-rate mouse heart cine-MRI using
compressed sensing reconstruction. NMR Biomed 2012; 26(4):451-457.
Conference abstracts (first author only)
1. M. Nabben*, D. Abdurrachim*, V. Hoerr, M.T. Kuhlmann, P. Bovenkamp, M.
Schaefers, K. Nicolay, C. Faber, S. Hermann, J.J. Prompers. Metabolic adaptations in
diabetes protect the heart from pressure-overload induced failure: a murine in vivo PET,
MRI, and MRS approach. Dutch Association for Diabetes Research (NVDO) meeting,
Oosterbeek, the Netherlands, November 27-28, 2014. (Oral)
2. D. Abdurrachim*, M. Nabben*, V. Hoerr, M.T. Kuhlmann, P. Bovenkamp, M.
Schaefers, K. Nicolay, C. Faber, S. Hermann, J.J. Prompers. Cardiac metabolic
adaptations of the failing mouse heart: an in vivo PET, MRI, and MRS approach. 12th
Annual Scientific Sessions of Society of Heart and Vascular metabolism (SHVM),
Tromso, Norway, Jun 24-27, 2012. Early Investigator Commendation award. (Oral)
184
List of publications
3. D. Abdurrachim, J. Ciapaite, M. van Weeghel, B. Wessels, K. Nicolay, S.M. Houten, J.J.
Prompers. In vivo cardiac 1H MRS,
31
P MRS, and MRI in mouse models of increased fatty
acid oxidation with and without myocardial lipid accumulation. 21st Annual Meeting and
Exhibition of International Society of Magnetic Resonance in Medicine (ISMRM), Salt
Lake City, USA, April 20-26, 2013. Magna Cum Laude merit award. (Oral)
4. D. Abdurrachim, J. Ciapaite, B. Wessels, K. Nicolay, J. J. Prompers. Long-term high-fat
feeding induces cardiac diastolic dysfunction without affecting cardiac energy status in
mice. Dutch Association for Diabetes Research (NVDO) meeting, Oosterbeek, the
Netherlands, November 29-30, 2012. (Oral)
5. D. Abdurrachim, M. van Weeghel, K. Nicolay, S.M. Houten, J.J. Prompers. In vivo
magnetic resonance study of cardiac lipid content, energetics, and function in a mouse
model expressing malonyl-CoA insensitive Cpt1b. 10th Annual Scientific Sessions of
Society of Heart and Vascular metabolism (SHVM), Oxford, United Kingdom, Jun 24-27,
2012. (Poster)
6. D. Abdurrachim, J. Ciapaite, J. Jeneson, K. Nicolay, J.J. Prompers. The effects of highfat diet feeding on cardiac function, energetics, and lipid content studied in mice using
MRI,
31
P- and 1H-MRS. 4th Scientific Meeting of ISMRM Benelux Chapter, Leuven,
Belgium, January 16, 2012. (Poster)
7. D. Abdurrachim, K. Nicolay, J.J. Prompers. Pioglitazone treatment reduces myocardial
triglyceride accumulation and prevents diastolic dysfunction. 71st Scientific Sessions of
American Diabetes Association, San Diego, CA, June 24-28, 2011. (Audio poster)
8. D. Abdurrachim, K. Nicolay, J.J. Prompers. In vivo assessment of the effects of
pioglitazone on myocardial triglyceride content and cardiac function in diabetic mice
using 1H MRS and MRI. 19th Annual Meeting and Exhibition of International Society of
Magnetic Resonance in Medicine (ISMRM), Montreal, Canada, May 7-13, 2011. (Poster).
9. D. Abdurrachim, K. Nicolay, J.J. Prompers. In vivo assessment of the effects of
pioglitazone on myocardial triglyceride content and cardiac function in diabetic mice
using 1H MRS and MRI. 3rd Scientific Meeting of ISMRM Benelux Chapter, Bovendonk,
the Netherlands, January 19, 2011. (Oral)
10. D. Abdurrachim, K. Nicolay, J.J. Prompers. Pioglitazone treatment reduces myocardial
triglyceride accumulation and prevents diastolic dysfunction. Dutch Association for
Diabetes Research (NVDO) meeting, Oosterbeek, the Netherlands, December 2-3, 2010.
(Oral)
11. D. Abdurrachim, K. Nicolay, J.J. Prompers. Myocardial lipid accumulation and diabetic
cardiomyopathy status studied in vivo in a diabetic mouse model using 1H MRS and MRI.
9th Biennial Scientific Meeting of the International Society for the Study of Fatty Acids
and Lipids, Maastricht, the Netherlands, May 29-June 2, 2010. (Oral)
12. D. Abdurrachim, K. Nicolay, J.J. Prompers. Cardiac function and metabolism in a
mouse model of early-stage diabetic cardiomyopathy assessed in vivo using MRI and 1H
MRS. 2nd Scientific Meeting of ISMRM Benelux Chapter, Utrecht, the Netherlands,
January 17, 2010. (Poster)
* both authors contributed equally
185
About the author
Desiree Abdurrachim was born on October 1, 1983, in Bandung,
Indonesia. After finishing her high school in 2001 at SMU Negeri 3
Bandung, she proceeded with her undergraduate study at the
Bandung Institute of Technology (ITB), Indonesia. She chose a major
in
Electrical
Engineering
with
a
specialization
in
Biomedical
Engineering. In the fourth year of her study she was granted a
scholarship to participate in the Young Scientist Exchange Program at
the Tokyo Institute of Technology, Japan. She obtained an award for
the best student at ITB (Ganesha Award) in 2004, and completed her
undergraduate study in 2006 cum laude. Fully funded by the Shell
Centenary Scholarship Fund, she continued with a master’s degree in Biomedical Engineering at
Eindhoven University of Technology, The Netherlands. She performed a summer internship at
the neuroimaging group of the Howard Florey Institute in Melbourne, Australia, and completed
her master’s study in 2008 with great appreciation. Her master thesis was on fluorine magnetic
resonance imaging (MRI) for quantitative molecular imaging, which was performed in the
Biomedical NMR group of Prof. Dr. Klaas Nicolay in collaboration with Philips Research
Eindhoven.
In 2009 she started her PhD project under supervision of Dr. Jeanine J. Prompers and Prof. Dr.
Klaas Nicolay in the Biomedical NMR group at Eindhoven University of Technology. During her
PhD study, she was part of the organizing committee for the annual meeting of the Benelux
Chapter of the International Society for Magnetic Resonance in Medicine (ISMRM) in 2012. She
received a travel grant from the Boehringer Ingelheim Fonds to perform experiments at the
European Institute for Molecular Imaging (EIMI) and the Department of Clinical Radiology at the
University Hospital of Mϋnster, Germany. The results of her PhD project are described in this
thesis. The research described in Chapters 7 and 8 was awarded with a ‘Magna Cum Laude Merit
Award’ at the 21st Annual Meeting of the ISMRM in 2013, and the work presented in Chapter 9
with an ‘Early Investigator Commendation’ at the 12th Annual Meeting of the Society for Heart
and Vascular Metabolism (SHVM) in 2014.
187