CARDIAC MYOFILAMENT CALCIUM SENSITIVITY IN HEALTH AND DISEASE
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Kenneth D. Varian
*****
The Ohio State University 2008
Dissertation Committee: Approved by Dr. Paul M.L. Janssen, Ph.D. Advisor
Dr. Jack Rall, Ph.D. ______Dr. George Billman, Ph.D. Advisor Integrated Biomedical Sciences Dr. William Carson III, M.D. Graduate Program
Dr. David Feldman, M.D., Ph.D.
ABSTRACT
Heart failure is the leading cause of death in much of the western world.
Progressive deterioration of cardiac contractility and/or relaxation, resulting in cardiac output below metabolic demands is the central theme in heart failure. Although treatments are improving, we still do not have a full understanding of the fundamental pathophysiology that impairs contractility and relaxation. The force frequency relationship (FFR) is ubiquitously altered in failing myocardium regardless of the
etiology. The FFR, which is normally positive (greater contractile force at higher
contraction frequency) in healthy myocardium, is blunted, flat, or negative in
dysfunctional myocardium. While the FFR has been studied extensively, the role
myofilament properties play in this effect is unknown. To address deficiencies in our
understanding of cardiac contraction, we worked from the following hypothesis:
Alterations in myofilament calcium sensitivity significantly contribute to blunted
force frequency response, contractile dysfunction, and impairment of myocardial
relaxation of failing myocardium.
First, we set out to develop a method which would enable quantitative analysis of shifts in myofilament calcium sensitivity due to changes in frequency in isolated, yet intact muscles at physiologic temperature. We found that we could induce slowly forming contractures by superfusing the intact isolated rat trabeculae with a solution of
ii high potassium and lower sodium. The contractures allowed for development of a
myofilament calcium sensitivity curve sensitive to interventions known to shift myofilament calcium sensitivity (beta adrenergic simulation and pH). We then aimed at determining if myofilament calcium sensitivity shifted with changes in frequency. We hypothesized that an increase in frequency would lead to a decreased in myofilament
calcium sensitivity which was needed for frequency dependent acceleration of relaxation.
We found that myofilament calcium sensitivity did decrease with frequency in intact
rabbit trabeculae. This effect correlated with an increase in the phosphorylation status at
Troponin I and Myosin Light Chain-2, as well as relaxation acceleration. Using
staurosporine (non-specific kinase inhibitor), we found inhibition of the myofilament
calcium sensitivity shift between 1 and 4 Hz. With the exception of blunting this shift in
myofilament calcium sensitivity, staurosporine did little else to the force frequency relationship, suggesting that most of the effect was not due to phosphorylation of key proteins.
To examine how inter-beat duration independent of posttranslational modifications occurs, we performed an analysis of twitch contractions in rabbit trabeculae stimulated randomly at 5 different cycle lengths; this was done in order to isolate the effect of cycle time on contractility and relaxation. We found that the primary cycle length correlated positively with force, while the secondary and tertiary cycle lengths correlated negatively with force.
Finally, we set out to determine how frequency dependent modulations of force, relaxation, and myofilament calcium sensitivity differed in a model of right ventricular hypertrophy. We used pulmonary artery banded rabbits and examined twitch
iii contractions, intracellular calcium, and myofilament calcium sensitivity (potassium contractures) at 1 through 4 Hz. We found that the shift in myofilament calcium sensitivity observed in control animals was nearly abolished in animals with right ventricular hypertrophy. This blunting of myofilament desensitization was accompanied by elevated diastolic force compared to controls, possibly reflecting some diastolic dysfunction in the hypertrophied tissue. In conclusion, myofilament calcium desensitization in response to elevated heart rate is a regulated, kinase specific effect that may be necessary to prevent insufficient diastolic filling at times of high heart rates.
iv
Dedicated to my grandparents: Ken and Virginia Varian, and George and Myrtle
Davidson
v
ACKNOWLEDGMENTS
In completion of my dissertation, there are several groups of people I wish to mention. First I thank my advisor Paul Janssen for his careful guidance and support throughout the project. Paul has been an excellent motivator and mentor for me over the past 4 years. From him I have learned autonomy in research, independent thinking, and better developed skills in experimental design. My co-workers in the Janssen lab have also been supportive. Carlos Torres has given me very helpful discussions on research and acted as a mentor for me outside the lab. Michelle Monasky, Nitisha Hiranandani,
Ying Xu, and Ko Bupha Intr have shared projects and ideas with me extending the breadth of my experiences in the lab. Anil Birdi, Ben Canan, Erin Shaffer, and
Annemarie Hoffman have helped me immensely in preparing experiments. I am in debt to Anusak Kijtawornrat from the department of Veterinary Biosciences for his dedicated help with the pulmonary artery banding procedure.
There are several faculty members outside the Janssen lab I would like to thank.
Doctors Mark Ziolo and Jonathan Davis have both held extensive discussions with me about my research many of which prompted me to perform experiments that helped shape the project. My dissertation advisory committee members Doctors William Carson III,
George Billman, David Feldman, and Jack Rall have all given me helpful guidance and pushed me to be more productive. The Medical Scientist program director, Allen Yates
vi has also been instrumental in my success here at OSU. He acted as a research advisor and mentor when I was an undergrad at Mount Union College and provided me with sound advice and support during my years as an M.D./Ph.D. student.
Finally, I would like to thank my family for their support through the years. My wife, Melinda, has been very supportive of my career and will be instrumental in my future success. My father, mother and brother have been encouraging to me my whole life and I certainly would not be here to today without their support.
vii
VITA
February 11th, 1980………………….Born, Johns Hopkins Hospital, Baltimore Maryland
1998-2002………………………………….Mount Union College BS Biology/Chemistry
2002-2004……Medical Scientist Fellow, The Ohio State University College of Medicine
2004-2006…...Graduate Research Associate, Department of Physiology and Cell Biology
2006-2008………………...Graduate Fellow, Department of Physiology and Cell Biology
PUBLICATIONS
1. Varian KD, Raman S, Janssen PML. Measurement of Myofilament Calcium Sensitivity at Physiological Temperature in Intact Cardiac Trabeculae. Am J Physiol Heart Circ Physiol 2006 May;290(5):H2092-7
2. Hiranandani N*, Varian KD*, Monasky MM, Janssen PML. Frequency-Dependant Contractile Response of Isolated Cardiac Trabeculae Under Hypo-,Normo-, and Hyper- thermic Conditions. J. Appl. Physiol 2006 May;100(5):1727-32. * these authors contributed equally.
3. Varian, KD and Janssen, PML. Frequency Dependent Acceleration of Relaxation Involves Decreased Myofilament Calcium Sensitivity. Am J Physiol Heart Circ Physiol 2007 May;292(5):H2212-9
4. Monasky MM, Varian KD, Janssen PML “Gender comparison of contractile performance and beta-adrenergic response in isolated rat cardiac trabeculae.” J Comp Physiol 2008 Mar;178(3):307-13
5. Monasky MM, Varian KD, Janssen PML “Dissociation of force decline from calcium decline by preload in isolated rabbit myocardium.” Pflugers Arch 2008 May;456(2):267- 76
viii
FIELDS OF STUDY
1. Integrated biomedical sciences 2. Cardiac contractile physiology
ix
TABLE OF CONTENTS
Page
ABSTRACT...... ii
DEDICATION...... v
ACKNOWLEDEMENTS...... vi
VITA...... viii
LIST OF TABLES...... xiv
LIST OF FIGURES...... xv
LIST OF ABBREVIATIONS...... xvii
CHAPTERS
1. Introduction……………………………………………………………………...1
1.1 General introduction…………………………………………………..1
1.2 Excitation contraction coupling……………………………..………...4 1.2.1 The cardiac action potential………………………..………..4 1.2.2 The myocyte calcium transient……………………...…...... 5 1.2.3 The cardiac myofilaments………………………………..….7
1.3 Myofilament calcium sensitivity……………………………..………..9
1.4 Regulation of cardiac output and contractility…………………...…..11 1.4.1 Sympathetic modulation...... 12 1.4.2 The force frequency relationship...... 13
1.5 Cardiac ventricular hypertrophy...... 14 1.5.1 Diastolic dysfunction and ventricular hypertrophy...... 15
1.6 Introduction to the methods and techniques...... 17 1.6.1 Measurement of intracellular calcium...... 18 x
1.7 Overall aims, objectives, and rationales…………………...…...…....19
1.7.1 Develop a technique to measure MCS in intact trabeculae at body temperature...... …..19 1.7.2 Assess MCS modulations due to changes in frequency in rabbit trabeculae...... 20 1.7.3 Elucidate the mechanism of frequency dependent myofilament desensitization...... 21 1.7.4 Isolate and characterize the effect of cycle length on force and relaxation...... 21 1.7.5 Determine how frequency dependent shifts in MCS were altered in RVH...... 21
2. Measurement of Myofilament Calcium Sensitivity in Intact Rat Trabeculae at Body Temperature...... 24
2.1 Introduction...... 24
2.2 Methods...... 26 2.2.1 Muscle preparation, solutions, and experimental set-up…...26 2.2.2 Potassium contractures...... 27 2.2.3 Intracellular calcium measurements...... 28 2.2.4 Data analysis and statistics...... 29
2.3 Results...... 29
2.4 Discussion...... 33 2.4.1 Limitations of the Study...... 36
3. Frequency Dependent Acceleration of Relaxation Involves Decreased Myofilament Calcium Sensitivity...... 44
3.1 Introduction...... 44
3.2 Methods...... 45
3.3 Results...... 48
3.4 Discussion...... 52
4. Staurosporine Inhibits Frequency Dependent Modulations of Myofilament Function...... ……...... 66
4.1 Introduction...... ….66
xi 4.2 Materials and Methods...... 68 4.2.1 Measurement of intracellular calcium and force frequency.69 4.2.2 Measurement of steady state myofilament activation…...... 70 4.2.3 Statistics...... 70
4.3 Results...... 71
4.4 Discussion...... 73
5. Contribution of Cycle Length History to Myocardial Contractility in Isolated Rabbit Myocardium...... …….81 5.1 Introduction……………...... ….81
5.2 Methods………………………...... 83 5.2.1 Ethical information...... 83 5.2.2 Tissue isolation...... 83 5.2.3 Variable cycle length simulation protocol...... 83 5.2.4 Data analysis...... 84
5.3 Results...... 85
5.4 Discussion...... 88
6. Frequency Dependent Myofilament Desensitization is Impaired in Rabbit Right Ventricular Hypertrophy...... 97
6.1 Introduction...... 97
6.2 Materials and methods...... 99 6.2.1 Animal ethics...... 99 6.2.2 Pulmonary artery banding (PAB)...... 99 6.2.3 Measurement of twitch contractions, calcium transients, and force-pCa...... 100 6.2.4 Western blots...... 101 6.2.5 Statistics...... 102
6.3 Results...... 102
6.4 Discussion...... 104
7. Discussion
7.1 Principle findings...... 116
7.2 Potassium induced contractures...... 117 xii 7.2.1 Mechanism of potassium induced contractures...... 117 7.2.2 Intracellular pH and potassium induced contractures...... 118
7.3 Force-calcium relationship in cardiac trabeculae, twitches vs. steady State...... 119
7.4 Myofilament Calcium Sensitivity, Frequency, and Diastole...... 121 7.4.1 Myofilament Calcium Sensitivity as a Modulator or Relaxation Rate...... 122 7.4.2 Myofilament Calcium Sensitivity as a Modulator of Diastolic Tension...... 125
7.5 Future Directions and Final Reflections...... 126
LIST OF REFERENCES...... 133
xiii
LIST OF TABLES
Tables Page
1. Data on frequency dependent myofilament desensitization with individual kinase inhibitors...... 131
xiv
LIST OF FIGURES
Figure Page
1. Diagrammatic representation of excitation contraction coupling…..……………22
2. Diagrammatic representation of myofibril and sarcomere structures……………23
3. Representative potassium contracture……………………………………………38
4. Pre and post contracture twitches parameters……………………………………39
5. Intracellular calcium and force during a contracture…………………………….40
6. Representative force-calcium relationship…………………………………….…41
7. Representative twitches and calcium transients control and isoproterenol……...42
8. Phase plane force-calcium versus steady state and average sensitivity shift…….43
9. Example rabbit calcium transients, force tracings and phase planes………...... 58
10. Average twitch force and relaxation times at each frequency………...... 59
11. Average systolic and diastolic calcium for all 4 frequencies………...... …..60
12. Example potassium contracture and curve fit of force-calcium data……...... ….61
13. Example and average shifts in calcium sensitivity with frequency…...... 62
14. Relationship between EC50 and RT50………………………………...... ……63
15. Relationship between intracellular pH (BCECF) and frequency……...... 64
16. Pro-Q Diamond phosphoprotein state of myofilaments 1 Hz, 4 Hz, and ISO...…65
17. Example twitches, calcium transients, and phase planes with and without staurosporine…...... 77 xv
18. Average twitch and calcium transient data with and without staurosporine….…78
19. Example potassium contracture and steady state relationship with and without staurosporine...... ……...... …79
20. Average EC50 and Fmax of myofilament calcium sensitivity curves with and without staurosporine...... 80
21. Chart of trabecular twitches during the variable cycle length protocol...... …92
22. Example and average correlations and slopes of cycle lengths versus Fdev...... 93
23. Example and average correlations and slopes of cycle lengths versus RT90...... 94
24. Average difference between low and high cycle lengths and percent contribution to twitch force...... 95
25. 3 dimensional plot of primary CL and secondary CL versus Fdev...... …96
26. Cross section of PAB and sham hearts, and western blot of TnI and GAPDH...110
27. Force frequency normalized to 1 Hz 4 weeks post-op in PAB trabeculae...... 111
28. Example twitches and calcium transients in PAB and sham trabeculae...... 112
29. Average twitch and calcium transient parameters in PAB and sham trabeculae.113
30. Average change in diastolic tension in PAB and sham trabeculae...... 114
31. Steady state force-calcium relationships PAB and sham...... 115
32. Phase planes and steady state force calcium in rabbit...... 129
33. Rat-rabbit comparison of twitches and calcium transients...... 130
xvi
LIST OF ABBREVIATIONS
ATP Adenosine Triphosphate
BDM 2,3 Butadione Monoxime
CAMKII Calcium-Calmodulin Kinase 2 cAMP Cyclic Adenosine Monophosphate
HF Heart Failure
CL Cycle Length
CRCP Contraction relaxation coupling point
EC50 Effective Concentration to reach 50% of maximum force
ECC Excitation Contraction Coupling
FDAR Frequency Dependent Acceleration of Relaxation
Fdev Developed Force
FFR Force Frequency Relationship
Fmax Maximum Force production
GAPDH glyceraldehyde 3-phosphate dehydrogenase
L-Type Long type
MCS Myofilament calcium sensitivity
MLC-2 Myosin light chain 2
xvii NCX Sodium calcium exchanger
PAGE Polyacrylamide gel electrophoresis
PAB Pulmonary artery banding
PKA Protein Kinase A
PLB Phospholamban
RT50 Relaxation Time to 50% of max force
RyR Ryanodine receptor
RVH Right Ventricular Hypertrophy
SDS sodium do-decyl sulphate
SERCA Sarco-endoplasmic reticulum calcium ATPase
SR Sarcoplasmic Reticulum
TnC Troponin C
TnI Troponin I
TnT Troponin T
xviii
CHAPTER 1
INTRODUCTION
1.1 General introduction
The circulatory system provides the body’s tissues with a continuous supply of
oxygen, nutrients, and waste removal. Blood, the carrier of these essentials, is pumped
through the body and lungs by the heart. The heart is a muscular organ that contracts and
relaxes rhythmically producing the force to move blood at high pressure into the great
arteries for delivery to the target tissues. The heart’s contractile function has been
extensively studied for the last 100 years such that we now understand the basic
mechanisms by which the heart contracts and relaxes. Nevertheless, therapies for heart
disease are currently insufficient and further investigation is warranted to extend our
understanding and better treat patients with heart disease.
Heart failure (HF) is a syndrome characterized by cardiac output that cannot meet
the body’s metabolic needs 100% of the time. HF presently is the number one cause of
mortality in developed countries (Dayer & Cowie, 2004). Currently, more than 5 million
people living in the US suffer from heart failure and over 500 thousand new cases are diagnosed each year. Though etiologies vary, two of the most common types of heart failure are ischemic cardiomyopathy and pressure overload (hypertrophic) cardiomyopathy (Cowie et al., 1997). The prevalence of coronary artery disease and hypertension found in the general population are the two main factors that contribute to 1 the disease. Until recently, HF had been described as a disease mainly of cardiac
contraction in which the heart’s ejection fraction is compromised leading to a decrease cardiac output. It is now apparent that diastolic dysfunction (insufficient ventricular filling) is a common finding in patients with HF and can directly lead to low cardiac output (Desai & Fang, 2008). It is estimated that at least half of HF patients display
diastolic dysfunction, some of which have normal ejection fractions (Kass et al., 2004;
Satpathy et al., 2006). It is clear that the heart’s ability to relax and fill with blood is
critical to normal cardiac function and that this process appears to be dysfunctional in many patients with HF.
The obstacles we face in understanding diastolic dysfunction lie in our lack of understanding of how cardiac relaxation is regulated in healthy myocardium. Relaxation
has been viewed as a passive return to a resting state. It is now clear that cardiac
relaxation is a highly regulated, energy dependent process that follows a different
pathway than contraction. The two main determinants of cardiac muscle force decline
are removal of cytosolic calcium into intracellular stores or the extracellular space, and
myofilament properties of calcium dissociation and force producing cross bridge
detachment (Bers, 2002a). Despite our extensive knowledge on how the individual
mechanisms work, we do not fully understand the role each process plays in myocardial
force decline under physiological conditions. Until we have a workable model of
relaxation, which includes the effects of both the cytosolic calcium removal and
myofilament properties under physiologically relevant conditions in normal myocardium,
we will struggle to find the right approach to treat diastolic dysfunction.
2 Cardiac output (liters/min blood pumped) can be increased several fold during times of increased metabolic demand. The main mechanism the body uses to increase cardiac output is through modulation of heart rate. While an increase in rate itself will increase the output of the heart, there is an additional heart rate dependent increase in contractility (force-frequency relationship - FFR) and rate of relaxation (frequency dependent acceleration of relaxation - FDAR). The increase in contractility helps to increase cardiac output by increasing stroke volume, while the enhanced relaxation helps to complete the filling phase of each cardiac cycle under greater time restraints. This mechanism, critical to maintaining normal cardiac function, is still not fully understood.
The heart’s reaction to changes in frequency is severely disabled in failing hearts such that a blunted or negative force frequency is often observed. In fact, an altered FFR is considered by many experts to be a signature of heart failure (Rossman et al., 2004). The normal modulation of relaxation with heart rate may also be altered and contribute to diastolic dysfunction observed in certain cases of HF. Many aspects of these basic properties of cardiac muscle have remained elusive. An all inclusive model with the mechanisms by which the heart responds to changes in rate would be a significant step toward a more complete understanding of healthy cardiac contractile physiology and how it is altered in HF.
The gaps in our knowledge regarding heart rate dependent alterations in contractility and relaxation have not been left unattended by accident. Methodological restraints have set significant road blocks to elucidating physiological mechanisms. For example, the vast majority of contractile physiology data on the FFR have been gathered using ex-vivo heart preparations at room temperature in small rodents such as the mouse
3 or rat. Nonphysiological temperature prevents accurate extrapolation to the in vivo situation due to non linear kinetic dependence on temperature for the many processes involved (de Tombe & Stienen, 2007). Small rodents, such as the mouse and rat, have calcium handling characteristics and contractile protein isoforms that differ significantly from human (Maier et al., 2000). Therefore, we set out to assess how changes in heart rate alter myofilament calcium sensitivity (MCS) under physiologic conditions in a larger
mammal with human-like calcium handling and myofilament protein isoforms. We
hypothesized that MCS decreased with increasing frequency and contributed to the
enhanced diastolic function observed at higher heart rates. In addition, we hypothesized
that pressure overload (the major etiology that results in diastolic dysfunction) would
alter the effect of heart rate on the myofilaments.
1.2 Excitation contraction coupling
Excitation contraction coupling is the process by which electrical action potentials
on the myocyte cell membrane translate to transient cytosolic calcium flux and
consequently contraction. An overview of this process is necessary in support of
subsequent chapters that provide no such review.
1.2.1 The cardiac action potential
The heart’s ability to contract and relax rhythmically to eject blood into the
arteries resides in the ability of the cardiac myocytes to shorten and re-lengthen. Each
cell has electrical excitability translated into mechanical force. All myocytes are connected at their long ends to other myocytes via electrically coupled gap junctions
4 allowing an action potential to travel from cell to cell. The action potential created when the ventricular cell depolarizes starts with an initial upstroke (depolarization) when voltage-gated sodium channels open. Calcium channels (long duration (L) type) open next allowing calcium to enter the myocyte creating the plateau phase of the action potential. The highest concentration of L-type calcium channels is located in the transverse tubules and near the intracellular calcium stores of the sarcoplasmic reticulum
(SR). It is here that the calcium entering through the L-type calcium channel interacts with the ryanodine receptor (RyR) of the SR causing the sudden release of calcium into the cytoplasm. Voltage-gated potassium channels open to initiate repolarization and closure of the calcium channels (for review see Bers, Nature 2002).
The plateau phase of the action potential serves two purposes. First, as mentioned above, it represents the time in which the L-type calcium channels are open allowing for calcium entry for contraction initiation. Second, it lengthens the refractory period to an extent that prevents a possible state of muscular tetanus which is observed in skeletal muscle that has been stimulated with a train of action potentials. This is necessary to ensure that the heart is able to relax (at least partially) after a contraction before a second contraction is initiated. The length of the plateau phase is dependent largely on the species in question. Smaller mammals, such as mice and rats, have much shorter plateau phases than larger mammals required for their faster heart rates (Bassani et al., 2004).
1.2.2 The myocyte calcium transient
The calcium transient is the sharp rise and fall of intracellular calcium that occurs with each cardiac beat to initiate contraction and relaxation. The entry of calcium into
5 the cytoplasm is induced by the action potential; its removal is controlled by pumps such
as the sarco-endoplasmic reticulum calcium ATPase (SERCA) and the sodium calcium
exchanger (NCX) in the sarcolemma. Calcium enters the cytoplasm from extracellular
sources (most L-type calcium current) and intracellular stores (SR). To maintain calcium
homeostasis, the amount entering from the extracellular space and the SR must be
removed from the cell or pumped back into the SR respectively. The SR’s main calcium
loading protein in the heart is SERCA2a. It is a primary active transporter using one
ATP to remove 2 calcium ions from the cytosol. SERCA is regulated primarily by a neighboring SR protein phospholamban (PLB), which when bound to SERCA, lowers its affinity for calcium and decreases its pumping rate. The sarcolemma’s main transporter is the NCX. This transmembrane protein uses the gradient of 3 sodium ions entering the cell to drive one calcium ion out of the cell (Bers, 2002b). The quantitative contribution of each source of calcium to the calcium transient is species dependent. In the mouse, up to 95% of total cytosolic calcium comes from the SR. This percent can be as low as 70% in larger mammals (rabbits, dogs, humans) and even lower in human heart failure (50%)
(Maier et al., 2000). Each source of calcium entry and mechanism of removal has a very different rate determining process and thus, animals with different percent contributions have calcium transients with vastly different time courses. In general, the more the SR contributes to the calcium transient, the faster it rises and falls. This can partially account
for the sharper rises and more rapid declines in systolic pressure observed in smaller
mammals. Modulation of contractility is largely accomplished by changes in peak
systolic calcium, which vary the level of activation of the myofilaments.
6 1.2.3 Cardiac myofilaments
The main target for calcium entering the sarcoplasm is the myofilament proteins.
They occupy the majority of the space within each myocyte and are arranged in parallel long myofibrils that stretch the length of each cell (Figure 2). Each myofibril is comprised of a repeating pattern of proteins with each unit being one sarcomere.
Microscopically, the sarcomeres appear to have two types of protein filaments named the thick and thin filaments. The thick filaments are composed mostly of myosin molecules.
The myosin molecules are composed of 6 subunits: 2 heavy chains, and 4 light chains.
Two myosin “heads” per myosin molecule bind to the thin filament to form a cross bridge. The light chains include both the essential light chain and the regulatory light chain (also known as myosin light chain 2 or MLC-2). MLC-2 is the only one of the myosin subunit that can be phosphorylated by a kinase. Phosphorylation of MLC-2 mainly through myosin light chain kinase is thought to increase myofilament calcium sensitivity (Morano & Ruegg, 1986; Wang et al., 2006). The thin filaments are comprised of 3 major proteins. First, polymerized actin monomers provide the backbone of the thin filament and the binding site for the myosin heavy chain. Second, tropomyosin is an actin binding protein that binds to 7 actins and provides a partial block of the myosin binding site on the actin when the myofilaments are in the resting (closed) state. Third, troponin, the third major myofilament protein, is the calcium sensor in cardiac muscle. Troponin contains 3 subunits: Troponin C (TnC), which binds and releases calcium entering the sarcoplasm; Troponin T (TnT), which binds tropomyosin, and Troponin I (TnI), which binds actin when calcium is not bound to TnC and releases actin when calcium is bound. This movement acts as a switch which helps expose the
7 myosin binding site on the actin. TnI and TnT can both be phosphorylated by various kinases active in the cardiac myocyte. The effect of TnT phosphorylation on contractility is not well understood. However, the effect of TnI phosphorylation has been studied extensively. Key phosphorylation sites at serines 23 and 24 have been implicated in protein kinase A (PKA) dependent desensitization of the myofilaments to calcium
(Kranias & Solaro, 1982). Other sites such as serines 43 and 45 have been implicated in affecting the maximum force generating capacity of the myofilaments (Montgomery et al., 2002) and threonine 144 is thought to possibly play a role in length dependent cardiac activation (Tachampa et al., 2007).
In the resting state of the myofilaments, calcium is not bound to TnC and myosin and actin have few, if any, force producing cross bridges. ATP is bound to the myosin heavy chain providing it with future energy for force production. The myosin ATPase hydrolyzes the ATP to an ADP plus a free organic phosphate, both of which are still bound to myosin. The ATP hydrolysis energizes the myosin and equips it for the force production. When calcium binds to TnC, the conformational change that subsequently occurs exposes the myosin binding site on the actin. This permits myosin to bind to actin and begin the so-called “power stroke”, or the force generating action of the cross bridge, thereby releasing the phosphate (keeping the ADP) in the process. With the myosin still bound to actin, a new ATP replaces the ADP on the myosin allowing for the actin-myosin interaction to break. With calcium still bound to TnC, the cycle will continue. If calcium has left TnC, the myofilaments return to their resting state.
8
1.3 Myofilament calcium sensitivity
Myofilament calcium sensitivity (MCS) is defined as the relationship between the
concentration of intracellular free calcium and the force produced by myofilament
activation when the two are at equilibrium. Much like the hemoglobin dissociation
curve, the MCS curve has a sigmoidal shape indicating a cooperative effect of force
producing cross bridges on the production of more force. Three important parameters
can be derived from the MCS curve. The maximum force, or Fmax, is the force produced
by the myofilaments that cannot be increased by adding more calcium. The EC50 is the effective concentration of calcium resulting in force production to 50% of maximum force. Shifts in MCS are defined as changes in the EC50. A myofilament sensitization
would be measured as a decrease in the EC50 where less calcium is needed to reach 50%
of the maximum force. MCS can also be displayed on a log scale as the pCa50 (pCa50 = -
LOG (EC50). The final parameter, the Hill coefficient (nHILL) defines the level of
cooperativity, or positive feedback, that force production has on additional force
production.
Shifts in the force-calcium relationship can directly influence cardiac contraction
and relaxation. Important modulators of MCS are intracellular pH, re-dox state of the
myocyte, and myofilament protein phosphorylation (Solaro et al., 1989; Kentish et al.,
2001). In general, increases in MCS can result in greater force production and slower
relaxation. If diastolic calcium levels are sufficient, increased MCS can result in
increased diastolic force. Decreases in MCS can cause decreased force production and faster relaxation. The most widely studied modulation of myofilament calcium
9 sensitivity is the phosphorylation of TnI at serines 23 and 24 by adrenergic response kinase PKA. TnI phosphorylation at these sites can produce an increase in the EC50 but no change in the Fmax thereby decreasing force production at non-saturating calcium concentrations. This protein modification has been implicated in adrenergic dependent acceleration of relaxation (Zhang et al., 1995; Yasuda et al., 2007).
The measurement of MCS requires that intracellular calcium concentration be held constant for long enough that it comes to an equilibrium with force. This cannot be done in a stimulated cardiac twitch because calcium enters and leaves the cytoplasm too quickly for it to equilibrate with force. Therefore, measurement of MCS has classically been performed in so called “skinned fibers.” Skinned fibers are muscle preparations
(either skeletal or cardiac) that have been treated with a detergent to dissolve away the sarcolemma and membrane bound intracellular organelles. This method gives the investigator direct access to the myofilaments and allows for controlled myofilament activation with buffered calcium solutions. By exposing the myofilaments to various calcium concentrations and measuring force, one can construct a MCS curve. This method has several key disadvantages that prompted us to develop a new technique to measure MCS. First, after the preparations are demembranated, they can no longer respond to normal physiological stimuli such as neurohumoral modulators (which require receptors) and electrical simulations (which require membrane action potentials). This limitation prevents direct assessment of MCS in a normal contracting cardiac muscle.
Skinned fiber experiments are also classically done at temperatures lower (15-25 ºC) than physiologic to extend the preparation’s life span.
10 An alternative method to measure MCS besides skinned fibers was developed by
Gao et al in which high frequency stimulations were used in the presence of ryanodine
and thapsigargan (to prevent SR cycling) at room temperature (Gao et al., 1994). This
induced a constant tetanic contracture while measuring intracellular calcium with
florescent indicators. This technique has the advantage of measuring MCS in an intact
cardiac preparation, but SR poisoning agents and subphysiologic temperatures were
required to induce the contractures. We felt this technique was inappropriate for
measuring frequency dependent shifts in MCS since frequency itself must be changed to
induce the contractures. Addressing this problem is the primary topic of Chapter 2 and of
our publication on the potassium contracture protocol (Varian et al., 2006).
1.4 Regulation of cardiac output and contraction
Cardiac Output is the number of liters of blood per minute the heart ejects and is
the product of heart rate (in beats per minute) and stroke volume (liters). As metabolic
demands change, cardiac output must also change to meet the new demands of blood
flow. Cardiac output that is insufficient to meet the body’s metabolic demand is the
central theme in heart failure and thus, a complete understanding of its regulation is
necessary to develop better therapies. Stroke volume can be increased by enhancing
contractility (myocardial force production at certain muscle length) or by increasing end diastolic volume (increasing force via the Frank-Starling mechanism). Heart rate is mainly modulated by the autonomic nervous system and directly increases cardiac output and also enhances contractility (the force frequency relationship). These systems are
11 complex, yet their background is important for future chapters and thus, two of them will
be discussed in further detail.
1.4.1 Sympathetic modulation of contraction
The two branches of the autonomic nervous system differentially affect cardiac
function. The sympathetic neurotransmitters bind to adrenergic receptors on the
myocardium to enhance contractility and heart rate. Epinephrine released from the
adrenal medulla, and to a lesser extent norepinephrine from sympathetic neurons, can modulate contractility and heart rate by binding to beta (type 1) receptors (Rockman et al., 2002). Beta-1 receptors activate intracellular G-proteins that in turn activate adenylyl cyclase. Adenylyl cyclase, using ATP as a substrate, produces cyclic adenosine monophosphate (cAMP) which acts as a second messenger moving into the sarcoplasm. cAMP can activate PKA which can phosphorylate a number of key proteins in the cardiac myocytes. First, PKA can phosphorylate PLB relieving its inhibitory action on
SERCA2a thereby resulting in a greater calcium pumping velocity into the SR. This enhancement of calcium reuptake can accelerate relaxation (by removing calcium from the sarcoplasm causing TnC calcium release) and increase SR calcium load. Calcium transient amplitude can be enhanced by increased SR calcium load directly effecting peak force production. Thus, the phosphorylation of phospholamban can produce both enhanced contractility (inotropic effect) and enhanced relaxation (lusitropic effect). PKA can also phosphorylate TnI which has been shown to decrease TnC’s affinity for calcium and accelerate TnC-calcium dissociation and myofilament relaxation (Yasuda et al.,
2007). Taken together, the phosphorylation of TnI and phospholamban enhances
12 contractility by increasing peak systolic calcium and accelerated relaxation by enhanced calcium reuptake and myofilament calcium desensitization.
1.4.2 The force frequency relationship
An increase in heart rate can induce some of the same physiologic changes as a sympathetic neurotransmitter. The greater number of contractions per second results in a greater calcium transient amplitude and accelerated calcium reuptake. It has been debated for some time exactly how the calcium transient is enhanced upon an increase in contraction frequency. One proposed theory ascertains that calcium dependent kinases, such as CaMKII (abundant in cardiac myocytes), become activated when frequency is increased due to the greater amount of calcium entering through L-type channels in the sarcolemma. CaMKII then phosphorylates key calcium handling and myofilament proteins resulting in the observed inotropic and lusitropic effects (DeSantiago et al.,
2002; Picht et al., 2007). This attractive hypothesis has been tested by several investigators in many different models with widely varying results. Some studies have shown that CaMKII phosphorylates PLB resulting in enhanced SR calcium reuptake.
These studies utilized the CaMKII inhibitor KN93 to block the effect of the kinase.
Blockage of the kinase resulted in reduced frequency dependent acceleration of the calcium transient decline. However, some groups find that CaMKII plays little or no role in enhancing force production and relaxation (Valverde et al., 2005). A more recent study has examined the time course of PLB phosphorylation and its association with the changes in contraction and relaxation. This study found that while PLB is
13 phosphorylated (to a limited extent) by CaMKII, this event is temporally dissociated from the physiological changes (Huke & Bers, 2007).
Frequency dependent modulations of the calcium transient have been the focus of nearly all studies initiated to elucidate the mechanism of the FFR. Frequency dependent modulations of MCS and other myofilament properties have been largely ignored and the few studies published show conflicting results. Some have shown frequency dependent myofilament calcium sensitization. Increases in MCS with frequency would contribute directly to the positive FFR but would (in theory) slow relaxation (Gao et al., 1998).
Others have shown that frequency has no effect on MCS and attribute all of the FFR and
FDAR to changes in calcium handling (Kassiri et al., 2000). Finally, some investigators have alluded to possible frequency dependent myofilament desensitization antagonizing the rise in force but contributing to FDAR (Tong et al., 2004). All studies on frequency dependent modulations of MCS have one aspect in common; none of them actually measured MCS in intact contracting cardiac preparations at physiologic temperature.
1.5 Cardiac ventricular hypertrophy
Cardiac hypertrophy is defined as the enlargement of the cardiac myocytes attributable to an increase in the number of myofibrils. Hypertrophy can be divided into two general pathways known as physiologic and pathologic hypertrophy. Physiologic hypertrophy occurs in trained athletes where the response to chronic exercise results in a positive effect on cardiac function. The pathologic version, often precipitated by hypertension or valvular stenosis, can ultimately result in HF. The normal progression of pathologic ventricular hypertrophy begins with an enlargement of the heart which is
14 initially compensatory. If the insulting stress continues, the myocardium may begin to show signs of dysfunction, a process known as decompensation (Pokharel et al., 2003).
It is at this point where normal physiological responses to heart rate begin to change. The
FFR, normally positive, will begin to deteriorate. The decompensating hypertrophy stage
ends when either the stress is removed from the heart and hypertrophy regresses, or more
commonly when the patient begins to feel symptoms of heart failure (ushering in the HF
phase) (Gradman & Alfayoumi, 2006). By the time patients are in HF the FFR is usually
negative (although sometimes flat). Interestingly, derangement of the FFR is a common
finding in all failing myocardium of any etiology. It does not matter if the initial assault
was hypertension, neurohumoral deregulation, ischemia, inherited cardiomyopathy, an infections agent, or a combination of more than one; all will result in a blunted or negative FFR (Rossman et al., 2004). It is therefore imperative that we understand how the FFR is regulated in healthy myocardium so we can definitively determine what has been altered to result in the FFR observed in patients with HF.
1.5.1 Diastolic dysfunction and ventricular hypertrophy
As the heart progresses from compensatory hypertrophy through decompensation and ultimately to HF, diastolic and sometimes systolic abnormalities begin to develop.
Diastolic dysfunction is observed in most patients with pressure overload hypertrophy and is considered by some experts to be the “hallmark of hypertensive heart disease”
(Gradman & Alfayoumi, 2006). Patients tend to be older females with hypertension.
Diastolic dysfunction can be defined using several different parameters but clinically, the
15 most common parameter is elevated end diastolic pressure (Kass et al., 2004). It is most easily assessed through Doppler echocardiography.
There are several candidate mechanisms for diastolic dysfunction. First, hypertrophy often results in myocardial fibrosis, or altered collagen deposition. Total collagen levels in the heart may correlate with the level of myocardial stiffness resulting in slower filling and the need for higher filling pressures to drive blood into the ventricles
(Kass et al., 2004). Other studies have found that only certain types of collagen or collagen post-translational modifications result in increased stiffness (Badenhorst et al.,
2003). Either way, it remains unclear what role the extracellular matrix plays in the elevation of end diastolic pressures. Myocyte cytoskeletal abnormalities, especially that of titin, can affect myocyte length tension relationship (Wu et al., 2000). Titin can undergo isoform shifts and/or changes in phosphorylation status possibly resulting in decreased compliance of myocytes and thus the whole heart (Cazorla et al., 2000; Radke et al., 2007). Slowed myocyte relaxation can also influence filling pressures mostly by delaying the early filling phase of diastole. Rate of relaxation can be even more of an issue when heart rate is high and filling times are reduced (Gelpi et al., 1991). Thus, blunting of FDAR may result in diastolic dysfunction at high, but not low heart rates.
This possibility is controversial as it remains unclear whether a slight delay in relaxation time would elevate end-diastolic filling pressures. Finally, active diastolic cross bridge cycling resulting in increased diastolic tension may also contribute to diastolic dysfunction. Increased MCS or increased diastolic calcium can increase active force generation when the heart is supposed to be filling. In Chapter 6 this issue is addressed with the finding that frequency dependent myofilament desensitization is impaired in a
16 model of right ventricular hypertrophy and this finding is associated with elevated
diastolic tension in trabeculae from hypertrophied hearts compared to controls.
1.6 Introduction to the methods and techniques
Experiments performed in this manuscript were all done using fresh isolated
cardiac tissue outside the bodies (ex vivo) of either rats or rabbits. Measuring cardiac function ex vivo can be done in three different models. The most physiologically relevant technique is whole heart experiments exemplified by those done on the Langendorrf apparatus (Farkas et al., 2008). Pressure development and decline can be measured from the right or left ventricle as it contracts at its own intrinsic (sinus) rate. The advantage to this method are its relevance to the in vivo situation (whole, intact heart). However, one cannot assess myofilament function because muscle length changes with every beat in the intact heart which then (due to the Frank-Startling mechanism) changes the force-calcium relationship. Uncontrolled muscle length prevents the assessment of shifts in MCS.
Also, intracellular calcium is technically very difficult to measure (Hayashi et al., 2008) and certainly not nearly as accurate as other methods. On the other end of the spectrum, cell shortening measurements can be made in isolated myocytes. Advantages to this model include more accurate measurement of intracellular calcium than whole heart, controlled environment (no influence of other cells and connective tissue), and the ability to modify the proteome. However, physiologic myofilament function is difficult to
assess. Because they have no surrounding attachments to the extracellular matrix, their
resting length is shorter (1.8 µm) than even a contracted myocyte in the whole heart
(Delbridge & Roos, 1997). The principal model we used in this project was intact right
17 ventricular trabeculae. Trabeculae are linear cardiac muscle strips that can be found
lining the inside of the right and left ventricular free wall. They provide a unique
representation of intact myocardium that contains all cells (myocytes, fibroblasts,
endothelial cells, etc.) of the normal heart. The myocytes in trabeculae all point and
contract in the same direction allowing for the measurement of tension as an indicator of contraction. This allows the length of the muscle to be set and kept constant to assess myofilament function. Other advantages include accurate calibratable measurement of calcium transients and electrical stimulation (not sinus rhythm) allowing for measurement of contractions at any frequency. Since no two trabeculae are the same, inter-muscle variability can be a disadvantage. In addition, not every animal has suitable trabeculae for experimentation.
1.6.1 Measurement of intracellular calcium
All but one of the subsequent studies described in this document utilized a technique to measure intracellular calcium in which the potassium salt version of a calcium indicator is iontophoretically loaded into trabecular myocytes as originally described by Backx and Ter Keurs (Backx & Ter Keurs, 1993). Therefore, a thorough introduction of this technique is required to appreciate a majority of this document. The iontophoretic loading technique was first developed to overcome limitations of using the acetomethoxyester (AM) form of various calcium indicators. Since organic calcium indicators generally need to be negatively charged to bind calcium, they are unable to cross membranes without some sort of a conduit. A solution to this problem is to hide negative charges of carboxylate groups with a methyl ester group which is subsequently
18 cleaved by non-specific esterases after the dye crosses the membrane. This effectively
traps the molecule inside the cell since the negatively charged carboxylate groups are reformed. Intracellular organelle compartmentalization, mostly in the mitochondria and
SR, is the major disadvantage of the AM form of calcium indicators. Calcium measurements subsequently made using this technique arise from a mixture of the cytosol and any other compartment that contains the dye. The goal of the loading via iontophoresis is to load the cytoplasm exclusively using a small glass micropipette.
Using this method, a single cell is punctured with the tip of the micropipette and the salt
form of the dye is electrically driven into the cell. Gap junctions that connect myocytes
and make their cytoplasm continuous allow for uniform spreading of the indictor
throughout the trabeculae. Fluorescent signals can be calibrated to calcium concentration
using this technique.
1.7 Overall aims, objectives, and rationales
Below is a list of aims that were set forth at the beginning of my research and
modified throughout the project as results were obtained and hypotheses modified. Each
aim is explained along with its rationale.
1.7.1 Develop a technique to measure MCS in intact trabeculae at body temperature
The primary obstacle to assessing physiological modulations of MCS was a lack
of established techniques available to measure MCS at body temperature, in intact living
trabeculae, and multiple times in the same muscle. The two standard techniques
available for assessing MCS were the skinned fiber technique and the high frequency
19 stimulation technique. Skinned fibers are demembranated, and thus could not be
stimulated to contract, nor could we assess any physiologic modulations of MCS. The
high frequency stimulation technique was limited to room temperature and required the
use of two SR poisoning drugs, ryanodine and thapsigargin. The use of potassium has
been used in the past to induce contractures in smooth muscle (Mundey et al., 1998) and skeletal muscle (Yamada et al., 2007) and in cardiac muscle (Allen et al., 1983;
Holubarsch, 1983). The theory behind the approach is that slow membrane depolarization caused by the high potassium, low sodium solution would induce a slow rise in intracellular calcium allowing for a pseudo-steady state contracture to develop. To
develop and validate this technique, we set out to measure MCS using potassium induced
contractions under two major conditions: normal and under beta adrenergic stimulation.
It has been established for the last 25 years that beta adrenergic stimulation induces TnI
phosphorylation and desensitization of the myofilaments. We needed to show that our
technique was sensitive to this known shift in MCS.
1.7.2 Assess MCS modulations due to changes in frequency in rabbit trabeculae
A clear use for the technique developed in aim-1 was not obvious until after we
switched from the use of rat to rabbit as the primary model in the lab. The first calcium
transients measured in rabbit trabeculae (in our lab) showed an astonishing difference
from those measured in rat trabeculae. Diastolic calcium increased significantly with an
increase in frequency, a phenomenon that most likely would induce a rise in diastolic
force and/or slowed relaxation without an accompanying decrease in MCS. This
prompted us to measure MCS in rabbit trabeculae at frequencies within the physiologic
20 heart rates of the rabbit to quantify this apparent shift and determine a possible molecular
mechanism for frequency dependent myofilament desensitization.
1.7.3 Elucidate the mechanism of frequency dependent myofilament desensitization
After finding an association between TnI and MLC-2 phosphorylation status and
the shift in MCS with frequency, we set out to inhibit this effect with organic kinase
inhibitors. This was done to prove a causative relationship between the phosphorylations and the shift in myofilament function as well as to show the effect of inhibiting the shift in MCS on trabeculae contractility and relaxation.
1.7.4 Isolate and characterize the effect of cycle length on force and relaxation
Next, we set out to isolate the effect of changes in cycle time between twitches on contractility independent of any post-translation modifications of calcium handling or myofilament proteins. This was done to show that most of the FFR was due to intrinsic responses of the cardiac muscle to cycle length and not due to net phosphorylations or other longer lasting, slower acting modifications.
1.7.5 Determine how frequency dependent desensitization is altered in hypertrophy
Patients with ventricular hypertrophy often suffer from diastolic dysfunction. We viewed our findings of frequency dependent decrease in MCS as a mechanism for FDAR or prevention of diastolic calcium induced diastolic dysfunction. Therefore, we set out to determine if frequency dependent myofilament desensitization is altered in right ventricular hypertrophy and how this affects diastolic function.
21
Figure 1. A diagrammatic representation of a myocyte sarcolemma and sarcoplasm showing the transverse tubule, components of the myocyte action potential, and components of the calcium handling system. The figure shows how the action potential, calcium transient, and myocyte contraction are related in time (taken from Bers, 2002 Nature)
22
Figure 2. Diagrammatic representation of myofibril and sarcomere structures. The bottom of the figure shows the longitudinal view of actin, myosin and regions when they overlap.
23
CHAPTER 2
MEASUREMENT OF MYOFILAMENT CALCIUM SENSITIVITY IN INTACT RAT TRABECULAE AT BODY TEMPERATURE
2.1 Introduction
Cardiac contraction-relaxation coupling is linked to both the rise and fall of the intracellular calcium concentration and intrinsic properties of the myofilaments (Backx et al., 1995). Although changes in calcium handling alone or myofilament properties alone can each have effects on the dynamic cardiac twitch, there is bilateral feedback between calcium concentration and myofilament properties. Up until now, these two determinants of contractile function have only been assessed independently or under nonphysiologic conditions.
In a dynamic cardiac twitch, the myofilaments and intracellular calcium are virtually never in a equilibrium (Backx et al., 1995); the only time during the cardiac cycle at which a pseudo-equilibrium exists is right before stimulation, when active force development is absent and the intracellular calcium concentration is nearly stable at diastolic values (50-100 nM). Upon stimulation of a cardiac myocyte, the intracellular calcium rapidly rises, while force production is much slower. When peak intracellular calcium is reached, force development is still rising, and usually only at approximately
50% of peak force development. Well before time of peak force production, the intracellular calcium concentration declines sharply, to nearly resting values, while
24 during most part of the relaxation phase, intracellular calcium has already reached and
remains at resting levels (Janssen et al., 2002). This implies that although the intracellular calcium and myofilament response are linked, changes in calcium and force production are de-coupled in the time domain. MCS reflects the contractile response of the myofilaments to a given calcium concentration at equilibrium. Thus, this sensitivity plays a major role in determining the contractile response to the calcium transient. In
order to elucidate how force production depends on intracellular calcium under
physiologically relevant conditions, we set out to develop a technique where MCS could
be assessed at body temperature in intact cardiac muscle.
Assessment of MCS has almost exclusively been done in skinned
(demembranized) fibers where myofilaments are exposed to a series of buffered calcium
solutions (Fabiato & Fabiato, 1978). Our goal was to assess MCS in intact muscle under
near physiologic conditions; this would have the advantage that factors which are known
to influence MCS (such as pH, temperature, and protein phosphorylation status (Fabiato
& Fabiato, 1978; Kranias & Solaro, 1982; Stephenson & Williams, 1985)) are close to their in vivo settings. Although MCS has been assessed in intact muscle before, these
experiments have thus far been limited to room temperature.
We developed a protocol and technique based on iontophoresis coupled with
potassium-contractures to construct the steady state relationship, while still allowing for
assessment of the dynamic force-calcium relationship measurements multiple times under
different conditions all in the same trabeculae preparation (to eliminate inter-muscle
variation). The results show that we can repeatedly assess the dynamic as well as steady
state force-calcium relationships under physiological preload, frequency, and temperature
25 in intact myocardium, and that the obtained force-calcium relationships shift in the
expected directions under β-adrenergic stimulation, alkalosis, and acidosis.
2.2 Materials and Methods
2.2.1 Muscle preparation, solutions, and experimental set-up
The investigation conforms to the Guide for the Care and Use of Laboratory
Animals published by the US National Institutes of Health (NIH Publication No. 85-23,
revised 1996). LBN-F1 rats (male, 175-200 g) were anesthetized with an intraperitoneal
injection of pentobarbital sodium. After bilateral thoracotomy and intracardiac
heparinization, the hearts were rapidly excised and placed in Krebs-Henseleit (K-H)
buffer containing (in mM) 137 NaCl, 5 KCl, 1.2 NaH2PO4, 1.2 MgSO4, 20 NaHCO3, 0.25
CaCl2, and 10 Glucose. 20 mM 2,3,-butanedione monoxime (BDM) was added to the
dissection buffer to prevent cutting injury. Exposure to BDM for a short time has been
shown to be reversible (Mulieri et al., 1989; Janssen & Hunter, 1995). Hearts were
cannulated via the ascending aorta and retrogradely perfused with K-H + BDM buffer in
equilibrium with 95% O2 / 5% CO2, resulting in a constant pH of 7.4. After blood was
washed out, the right ventricle was opened and thin, uniform, non branched trabeculae
were removed leaving a block of right ventricular free wall tissue on one end and a piece
of tricuspid valve on the other (ter Keurs et al., 1980). The dimensions of the trabeculae
(n=8) were measured with a reticule on the inverted fluorescence microscope (100x) and
were 0.17±0.03 mm wide, 0.08±0.01 mm thick and 2.0±0.2 mm long. The cross sectional area was calculated assuming an ellipsoid shape. Muscle dimensions have been
shown to be a determinant of contractile performance (due to diffusion distance) and
26 thus, only trabeculae with a smallest dimension of less than 150 µm were selected and
included (Raman et al., 2006).
Using the dissection microscope, muscles were mounted between a platinum-
iridium basket-shaped extension of a force transducer (KG7, Scientific Instruments
GmbH, Heidelberg, Germany) and a hook (valve end) connected to a micromanipulator.
Muscles were superfused with the same buffer at 37.5 °C as above (with the exception
that BDM was omitted) and stimulated at 4 Hz using 3 ms stimuli of approximately
150% threshold (typically 3-5 V.). Extracellular calcium was raised to 1.5 mmol/L and
muscles were allowed to stabilize for at least 30 minutes before the experimental protocol
was initiated. Muscles were stretched to a length where a small increase in length
resulted in nearly equal increases in resting tension and active developed tension. This length (optimal length) is slightly below the length where active force development is maximal, and was selected to be comparable to the maximally-attained length in vivo at the end of diastole (around 2.2 µm sarcomere length) (Rodriguez et al., 1992).
2.2.2 Potassium contractures
At room temperature, cardiac muscle can be tetanized by a rapidly pacing the muscle (at 15 Hz) (Perez et al., 1999). However, at a physiological temperature, even at pacing rates as fast as 20 Hz, healthy, well-perfused muscles almost fully relax after each beat. Therefore, we resorted to potassium contractures in order to obtain levels of steady state force development, which are needed to construct the steady state force–calcium relationship. After recording normal twitches at 4 Hz, the superfusion solution was switched to a modified K-H solution containing (in mM) 142 KCl, 0 NaCl, and 3 CaCl2.
27 The rest of the contents were identical to that of the normal K-H buffer. The high potassium solution was applied for 30 seconds then washed out.
2.2.3 Intracellular calcium measurements
Trabeculae were iontophorically injected with bis-fura 2 (Texas Fluorescence
Labs) as described previously (Backx & Ter Keurs, 1993; (Janssen et al., 2002). Bis-fura
2 was chosen due to its higher signal to calcium buffering ratio (allowing for a loading of the dye to 5-10 times background without impacting buffering) and its slightly higher Kd
(370 nM in vitro) than Fura-2, allowing for still accurate diastolic values, but a better resolution at higher intracellular calcium. Iontophoretic loading of the dye was performed at room temperature (at body temperature the rate of dye leak is higher). We loaded the bis-fura-2 until the photomultiplier output at baseline 380 nm excitation was between 6 and 10 (average 8.7±1.0) times over background. After loading and spreading of the dye at room temperature (1 Hz), the system was returned to 37°C. The muscle was stimulated to contract at 4 Hz while force and fluorescent emission measurements
(excitation 340 and 380 nm) were recorded. The superfusion solution was then switched to the one with 142 mM K+. Excitation wavelength was repeatedly switched back and
forth between 340 and 380 nm so that the ratio (340/380) could be taken at given points
along the contracture. After the contracture relaxed and twitches at 4 Hz resumed, 1
µmol/L isoproterenol was administered to the Krebs-Henseleit solution and the protocol
was repeated. In a separate set of experiments, the pH of the K-H and high potassium
solutions were set at 7.7 and 7.1 to mimic alkalosis and acidosis respectively. Potassium contractures were performed at pH 7.7, 7.1, and 7.4, and force was plotted against ratio
28 (340/380) of fluorescence. Data was stored on a computer for off line analysis using
custom written software.
2.2.4 Data analysis and statistics
All force values were normalized to cross sectional area and comparisons were
done within each muscle to eliminate inter-muscle variations. Paired and unpaired t-tests
were performed with a two-tailed p with a value < 0.05 being considered significant.
2.3 Results
Defining the relationship between the dynamic cardiac twitch force-calcium
relationship and the steady state force-calcium relationship required exploring an
approach to reversibly “tetanize” cardiac muscle while measuring calcium under
physiologic conditions. Exposing cardiac muscle to high potassium solutions causes a
slow depolarization of myocyte membranes (Holubarsch, 1983) due to the resting
potential’s dependence on the potassium gradient. This most likely results in the opening
of voltage dependent calcium channels. We found that brief exposure (30 seconds) to a
modified Krebs-Henseleit solution with 142 mM K+ elicited a slow forming contracture
in the muscle resembling a twitch but occurring around 200-400 times slower (data not
shown). Unlike tetani at room temperature (Gao et al., 1996; Kogler et al., 2001), the level of contracture could not be regulated by varying extracellular calcium. However, the contracture occurred very slow, and compared to the extreme rapid changes in intracellular calcium in a twitching muscle, the rise in intracellular calcium in potassium contractures was >1000-1200 times slower (data not shown). From the observations we
29 concluded that during the potassium contracture, we may consider the obtained values for force and intracellular calcium to be in steady-state. Figure 3 shows a representative tracing of such a potassium contracture. The muscle is twitching at 4 Hz while the solution flowing through the bath is switched to that with high potassium. Note that the twitches begin to rise in amplitude before stimulation is stopped. This is most likely because the potassium solution contains a higher calcium concentration (3 mmol/L) in order to reach a near maximal myofilament activation during the contracture. Figure 4A and 4B shows that both twitch amplitudes and relaxation time (RT50) before and after the contracture were not significantly different indicating the reversibility and absence of significant rundown. Other contractile and kinetic parameters (such as time-to-peak and diastolic tension) were likewise unaltered after a potassium contracture. We have repeated this potassium contracture several times in the same muscle, as it is completely reversible. Twitch forces and kinetics return to pre-contracture values within minutes after the contracture. We found that dye leakage limited the lifespan of the experiments, not contractile performance. We observed bis-fura-2 leakage at a rate of 15±3% (n=10) per 10 minutes after switching to 37°C. Our protocol lasted around 20 minutes and thus at the end of the experiment we retained 71% ± 4% of the baseline 380 nm fluorescence.
Due to the very low background-to-bis-fura-2 fluorescence ratio (1 to ~6-8), ratios can still be accurately obtained.
Figure 5A shows the output of the photomultiplier during a potassium contracture, with the excitation filter being alternated between 340 and 380 nm roughly every half second. As expected, calcium enters the sarcoplasm slowly indicated by the slow decrease in fluorescence emission at 380 nm excitation and an increase at 340 nm. In
30 figure 5B raw recordings of force and analyzed calcium fluorescence ratios are depicted.
It can be seen that calcium and force rise simultaneously. When ratios were plotted against force up to the peak of the potassium contracture, a sigmoidal curve is observed typical of a MCS curve. An example of such a curve is shown in Figure 6. In pilot experiments, using the pH indicator BCECF, we observed no changes in intracellular pH until peak contracture was reached. We did not observe open-loop behavior similar to the dynamic calcium transient vs. force relationship. Instead, either there was no hysteresis at all (downstroke fell right on top of upstroke) or there was a very small amount of open loop behavior, but in the opposite direction as the twitch (calcium still high but force falling). The latter behavior coincided with a drop in intracellular pH. In addition, when K+-contractures were applied for a short time (15 seconds), the up and down strokes of both force and calcium were identical, underlining the fact that during potassium contractures force and calcium are in equilibrium.
With the ability to repeat the protocol and compare the steady state force-calcium plots to the twitch force-calcium, we used a set of preparations to show that the observed force-calcium curve truly represents myofilament responsiveness. We used the β- adrenergic agonist isoproterenol to evoke an expected change in MCS (reduction in sensitivity due to PKA mediated phosphorylation of TnI). Figure 7 shows the simultaneous measurement of intracellular calcium (as ratio 340/380) and force during a cardiac twitch with and without a maximal isoproterenol (1 µM) response. Figure 8A shows the resulting phase-plane plot when intracellular calcium is plotted against force in a single cardiac trabeculae with and without isoproterenol. In this example, we also show
the steady-state force-calcium relationship which can be directly compared to the phase
31 plane plot. We averaged the contracture-derived MCS curves (n=8) to show the reproducible rightward shift in the curve upon isoproterenol administration (Figure 8B).
The average 340/380 ratio at 50% of maximal developed force was 0.65 and shifted to a higher calcium value of ratio 0.85 upon administration of isoproterenol. Comparing the two steady state curves a clear rightward shift is observed upon administration of isoproterenol probably resulting from myofilament protein phosphorylation such as TnI
(Kranias & Solaro, 1982).
Assessment of dynamic and steady-state calcium relationships now allows us to better analyze the phase-plane loops of force versus calcium. We can now relate the active twitch curves to the steady state values and determine, for instance, at what isochrone the dynamic force-calcium value equals that obtained in the steady state relationship. We termed this point the Contraction-Relaxation Coupling Point (CRCP) in units of milliseconds from initiation of stimulation. Before reaching this point, the cytosolic calcium concentration exceeds that predicted by steady state behavior, while thereafter the opposite holds; force is higher than predicted from the steady state relationship. As expected, the CRCP after isoproterenol administration shifted to lower values, from 43±2 ms to 29±2 ms (n=8, P<0.0001).
To further validate this novel assessment of MCS, we demonstrated that the technique is sensitive to other forms of physiologic modi