University of Groningen Molecular adaptations in human

University of Groningen

Molecular adaptations in human atrial fibrillationBrundel, Bianca Johanna Josephina Maria

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2000

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Brundel, B. J. J. M. (2000). Molecular adaptations in human atrial fibrillation: mechanisms of proteinremodeling. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 03-04-2022

https://research.rug.nl/en/publications/3a13bf15-9321-428d-9521-b037345a9d99

Molecular Adaptations in Human Atrial Fibrillation:

Mechanisms of Protein Remodeling

ISBN 90-367-1248-3

Nugi code 743

© Copyright 2000 Bianca J.J.M. Brundel

All rights are reserved. This publication is protected by copyright. No part of it

may be reproduced, stored in a retrieval system, or transmitted, in any form or by

any means — electronic, mechanical, photocopy, recording, or otherwise — with-

out the prior written permission of the author.

Lay-out: Peter van der Sijde, Groningen NL

Druk: Ponsen en Looijen bv, Wageningen NL

The studies described in this thesis were supported by grants 94.014 and 96.051

from the Netherlands Heart Foundation.

Publication of this thesis was financial supported by :

The Netherlands Heart Foundation

Groningen University Institute of Drug Exploration (GUIDE)

Faculteit Medische Wetenschappen RUG

Bristol-Myers Squibb BV

AstraZeneca BV

ASTA Medica BV

Novartis Pharma BV

Dr. Saal van Zwanenberg stichting

Cover:

Top of the picture shows the view from the ‘mountain’ of Kardinge

(Groningen). Bottom of the picture shows electron microscopic detail of a

human atrial myocyte (magnification x 4500).

Backside of the cover shows a picture of Captain Hook (Myrthe), Wendy (Jona)

and Peter Pan (Joachim).

Rijksuniversiteit Groningen

Molecular Adaptations in Human Atrial

Fibrillation:

Mechanisms of Protein Remodeling

Proefschrift

ter verkrijging van het doctoraat in

de Medische Wetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. D.F.J. Bosscher,

in het openbaar te verdedigen op

woensdag 29 november 2000

om 16.00 uur

door

Bianca Johanna Josephina Maria Brundel

geboren op 7 februari 1971

te Lichtenvoorde

Promotores:Prof. Dr. H.J.G.M. Crijns

Prof. Dr. W.H. van Gilst

Referenten:Dr. I.C. van Gelder

Dr. R.H. Henning

Beoordelingscommissie:Prof. Dr. M. Borgers

Prof. Dr. A.A.M. Wilde

Prof. Dr. H.W.G.M. Boddeke

Table of contents

Chapter 1 Introduction 9

Part I Gene expression of proteins influencing calcium

homeostasis

Chapter 2 Gene expression of proteins influencing calcium homeostasis

in patients with persistent and paroxysmal atrial fibrillation

Cardiovascular Research 1999 (42) 443-454 21

Chapter 3 Alterations in gene expression of proteins involved in

calcium handling in patients with atrial fibrillation

Journal of Cardiovascular Electrophysiology

1999 (10) 552-560 39

Part II Gene expression of ion-channels

Chapter 4 Alterations in potassium channel gene expression in atria of

patients with persistent and paroxysmal atrial fibrillation

Differential regulation of protein and mRNA levels for

K+-channels

Journal of the American College of Cardiology, provisionally

accepted 55

Chapter 5 Ion channel remodeling is related to intra operative atrial

refractory periods in patients with paroxysmal and persistent

atrial fibrillation

Circulation, in press 69

Part III Gene expression of neurohormones

Chapter 6 Gene expression of the natriuretic peptide system in atrial tissue

of patients with paroxysmal and persistent atrial fibrillation

Journal Cardiovascular Electrophysiology 1999, (10) 827-835 85

Chapter 7 Endothelin-1 mRNA is upregulated in patients with persistent

atrial fibrillation with underlying valve disease

Submitted 99

Part IV Calpain activation, a new adaptive mechanism in AF

Chapter 8 Activation of proteolysis by calpain during paroxysmal and

persistent atrial fibrillation

Submitted 113

Chapter 9 Calpain activity is related to ion-channel, structural and electrical

remodeling in human paroxysmal and persistent atrial fibrillation

Submitted 121

Chapter 10 General discussion 137

Summary 153

Samenvatting 155

Dankwoord 157

9

Introduction

Introduction

Atrial Fibrillation, clinical aspects

Atrial fibrillation (AF) is currently the most common sustained clinical arrhythmia

and is responsible for a substantial proportion of hospital costs incurred in the treatment of

cardiac rhythm disorders.1 AF becomes increasingly common with age, having an incidence

averaging <0.5% in patients <40 years of age and reaching >5% in patients >65.2 Thus AF

is likely to become increasingly important with the ageing of the population. The arrhythmia

is defined by a very rapid atrial rate (generally >400/min in humans) along with irregular

atrial activation and lack of repetitive pattern of co-ordinated atrial activity. AF is associated

with a variety of complications, including thromboemboli resulting from coagulation in

the relatively static atrial blood pool, a loss of the fine adjustment of ventricular rate to the

body’s precise metabolic needs, potential impairment of cardiac function and subjective

symptoms like palpitations, dizziness, breathlessness and chest pain.

AF can occur in paroxysms of a duration shorter than 24 hours (but longer lasting

paroxysms are not unusual) with intermittent sinus rhythm. Paroxysmal AF either converts

spontaneously or is terminated with an intravenously administered antiarrhythmic drug.3,4

In contrast, during persistent AF, the arrhythmia is continuously present until the moment

of investigation, i.e. at least two consecutive electrocardiograms of AF more than 24 hours

apart and without intercurrent sinus rhythm. Persistent AF does not convert spontaneously.3,4

AF has the tendency to become more persistent over time. This is illustrated by the

fact that about 30% of patients with paroxysmal AF eventually will develop persistent

AF.5 Also pharmacological and electrical cardioversion and maintenance of sinus rhythm

thereafter become more difficult the longer the arrhythmia exists.6 The cumulative

percentage of patients who maintained sinus rhythm after serial cardioversion treatment

was not more than 42% after 1 year and 27% after 4 years.6 This relates to progression of

the underlying disease and possibly also to electrical remodeling of the atria.7

Mechanism: multiple-wavelet reentry

In 1959 Moe & Abildskov8 showed that AF could be produced by experimental

paradigms of both multiple circuit reentry and rapid activity and they suggested that either

type of mechanism might cause clinical AF. Moe put forward the ‘multiple wavelet

hypothesis’ of AF in 1962.9 This concept described the propagation of reentry waves as

involving multiple independent wavelets circulating around functionally refractory tissue.

The maintenance of AF then depends on the probability that electrical activity can be

sustained by a sufficient number of active wavelets at any time. Experimental support for

Moe’s ideas was obtained subsequently with the use of computerized mapping of AF

maintained in the presence of acetylcholine in dog hearts.10 It was demonstrated that during

10

Chapter 1

Figure 1

Schematic representation of inward and outward ionic currents involved in the ventricular action potential, resting

membrane potential and cytoplasmic Ca2+ transients.

The numbers 0 through 4 indicate the different phases of the action potential: 0, upstroke; 1, fast early repolarization

phase; 2, plateau phase; 3, repolarization phase; 4, resting membrane potential. Adapted from The Sicilian Gambit.47

INa

, sodium current encoded by the gene SNC5A; ICaT

, calcium current encoded by T-type Calcium channel; ICaL

,

calcium current L-type Calcium channel; ICl(Ca)

, calcium dependent transient outward current encoded by the

gene Kv4.2; INaCa

, sodium calcium exchanger; ITo

, transient outward current encoded by Kv4.3; IK1

, inward rectifier

K+ current encoded by Kir2.1; IKs

, slow delayed rectifier K+ current encoded by minK and KvLQT; IKr

, rapid

delayed rectifier K+ current encoded by HERG, IKACh

, acetylcholine dependent potassium current encoded by

Kir3.1 and Kir3.4; INa-K

, sodium potassium pump.

11

Introduction

AF, multiple independent wavelets activate the atria in a random reentrant way. Individual

wavelets could brake-up, fuse or collide with each other and wavelets would extinct when

they reached the border of the atria or met refractory tissue. From time to time a varying

number of wavelets was present in the atria and the duration of each individual wavelet

lasted only several hundreds of a millisecond. Further it was shown that the number of

wavelets that fit into the atria determined the perpetuation of AF. Below a critical number

of wavelets (between 3 and 6), there was a considerable chance for the wavelets to die out

all at the same time. When more than 6 independent wavelets were present, the arrhythmia

would not convert spontaneously anymore. The number of wavelets that fit into the atria

depends on the atrial refractory period, conduction velocity and atrial mass.8,9 During the

last decade, mapping studies in humans with AF have further confirmed the multiple wavelet

theory.11

Electrophysiology of the atria

The processes that signal the heart to contract (excitation-contraction coupling) begin

when an action potential depolarizes the plasma membrane surrounding the myocardial

cell. This electrical signal is generated by the passage of ions through ion channels in the

plasma membrane that changes the electrical potential of the interior of the cardiac cell

relative to the extracellular space. These ion fluxes include two major inward currents that

depolarize the heart. The phase 0 depolarization is initiated by a rapid inflow of sodium

ions through the voltage-gated Na-channels and later by calcium ions via the L-type calcium

channel (ICaL

) (Figure 1).12,13 A large transient outward current has been held responsible

for the phase 1 repolarization. This current is composed of a 4-aminopyridine-sensitive

component (ITo1

) and a Ca2+ activated and verapamil blocked Cl- current (ITo2

or ICl(Ca)

).14-16

During the plateau phase (phase 2), there is a delicate balance of inward and outward

currents. Inward currents can be carried through the Na+ channel and the Ca2+ channels.

During reverse-mode, the Na+/Ca2+ exchanger will transport three Na+ ions outside and

exchanged for one Ca2+ resulting in a net movement of charge across the plasma membrane

from inside to outside. The influx of calcium through the L-type calcium channels is

responsible for the plateau phase during repolarization and initiates calcium release from

the sarcoplasmic reticulum that binds to the contractile filaments of the myocardial cell,

causing contraction (see calcium homeostasis in normal cardiac cells).17 Outward currents

during the plateau are carried by a number of K+ channels (IKs

, IKr

, IKACh

) and the Na+-K+

pump. For the initiation of phase 3 the contributions of rapidly and slowly activating

delayed rectifier K+ currents are of key importance. Here the outward potassium current

IK1

is activated and the action potential returns to its transmembrane resting potential,

which remains at that level during phase 4 of the action potential until the cell becomes re-

activated.

12

Chapter 1

The duration of the repolarization phase is different between atrial and ventricular

myocardial cells. This phase is shorter in the atrium compared to ventricle.18 An other

difference is the distribution and magnitude of the ionic currents responsible for the rest-

ing membrane potential. The inward rectifier current, which is responsible for mainte-

nance of the resting membrane potential, IKs

is smaller in atrial cells compared with ven-

tricular cells.19,20 Secondly, the acetylcholine-dependent potassium current IKACh

is very

important in maintaining and hyperpolarizing the resting membrane potential in atrial

cells, while less active in ventricular cells.21 It is this current which is responsible for

hyperpolarization and shortening of the atrial action potential during parasympathic stimu-

lation. Vagal stimulation is indeed one of the oldest models to induce sustained atrial

fibrillation by applying vagal stimulation. The resulted increased parasympathetic tone

will shorten the atrial refractory period through opening of these acetylcholine-dependent

potassium channels. Furthermore, inhomogeneous distribution of vagal nerve endings will

increase the spatial dispersion in refractoriness.22 In this way, atrial fibrillation will persist

after induction with premature stimuli or atrial burst pacing as long as the parasympathetic

system is stimulated, either by vagal nerve stimulation or by acetylcholine application.

Figure 2

Schematic representation of the calcium triggered calcium release in the myocardial cell. Small amount of

calcium enters the cell via the Na+/Ca2+ exchanger, but mainly through voltage gated L-type calcium channels,

which open in response to the action potential. This causes a much larger amount of calcium to be released by

the ryanodine receptors (RyR) in the sarcoplasmic reticulum (SR). Calcium will bind to the troponin C

resulting in contraction. The calcium pump of the SR (SR Ca2+ ATPase) has a high affinity for calcium and

reduces cytosolic Ca2+ concentrations to levels low enough to dissociate this cation from its binding sites on

troponin C. The SR Ca2+ ATPase is inhibited by phospholamban (PLB). When PLB is phosphorylated this

inhibition is reversed.

13

Introduction

Calcium homeostasis in normal cardiac cells

When an action potential depolarizes the plasma membrane surrounding the

myocardial cell, the processes that signal the heart to contract begins. Calcium transients

underlying this excitation-contraction coupling in cardiac cells result mainly from calcium

release from the sarcoplasmic reticulum triggered by calcium entry during the action

potential. This process is called calcium-induced calcium release (Figure 2). Under normal

circumstances, calcium entry into cardiac myocytes is carried primarily via IcaL, whereas

additional fractions can enter via on reverse-mode Na+/Ca2+ exchange (one calcium ion is

transported inside the cell against three sodium ions outside) and ICaT

. All three pathways

are capable of triggering sarcoplasmic reticulum (SR) calcium release and contraction,

but the relative contribution and efficiency is largest for ICaL

.23,24 Upon depolarization,

influx of calcium through discrete clusters of L-type calcium channels in the plasma

membrane triggers the opening of ryanodine receptors in the sarcoplasmic reticulum

membrane, resulting in the major release of calcium into the cytoplasm which in turn

triggers myofibril contraction by binding of troponin C. Following contraction, the SR

calcium ATPase enzyme in the network of sarcoplasmic reticulum surrounding the

myofibrils, rapidly pumps the calcium back into the SR lumen, causing the myofibrils to

relax. The SR calcium ATPase is regulated by phospholamban, a small protein and when

phosphorylated it will activate SR calcium ATPase and the calcium content of the SR will

increase.25

Modern concepts of excitation-contraction coupling rely on a ‘local-control theory’ a

close association between L-type calcium channels and ryanodine receptors and

subsequently received experimental support documenting local functional interaction

between these channels.26 Changes in the close association between the L-type calcium

channels and ryanodine receptors result in a reduction of calcium release from the

sarcoplasmic reticulum and reduced contractility of the cardiac myocyte.26

The membrane machinery allows each myocyte to function as an autonomous

contractile unit. To produce a heart beat, the contractile capabilities of myocytes that

make up the heart have to work in a highly synchronous fashion. This requires both an

orderly spread of the wave of electrical activation and effective transmission of contractile

force from one cell to the next, throughout the heart.

Electrical remodeling of the atria

Over the past several years, AF-induced remodeling and its underlying mechanisms

have been studied in substantial detail. Wijffels and coworkers published the first study in

1995 which part of the underlying electrophysiological changes explaining the progressive

nature of AF was demonstrated.7 In healthy goats they showed that repetitive induction of

AF increased the duration of successive episodes of AF, until AF finally did not convert

14

Chapter 1

spontaneously any more. They discovered that the increased tendency of the atria to fibrillate

was paralleled by a progressive shortening of the atrial effective refractory period (AERP)

and loss of the physiological rate adaptation of the refractory period which they termed

atrial electrical remodeling.7 After cardioversion of atrial fibrillation that had been present

for two to four weeks, this so-called atrial electrical remodeling appeared to be completely

reversible within one week after restoration of sinus rhythm.

In 1995 Morillo et al.27 also showed that rapid atrial pacing (400 bpm) strongly

promotes the ability to maintain AF in dogs, with changes quite similar to those observed

by Wijffels et al.7 Later observations by Wijffels et al. suggested that acute volume loading,

opening of the ATP-dependent potassium channels, neurohumoral activation or an increase

in ANP were not responsible for the altered electrophysiological characteristics in this

experimental model. This supports the idea that AF-induced remodeling is primarily due

to the rapid atrial activation rates caused by AF.28 Consequentially, many investigators

have used rapid atrial pacing in experimental models to study the electrophysiological

changes caused by sustained atrial tachycardia, hoping to gain insights into the atrial

electrophysiological changes caused by AF in man.

Experimental and clinical studies have shown that sustained AF decreases the atrial

effective refractory period (AERP).7,29-32 AERP changes occur over a period of days to

weeks7,27,33,34, but AF can decrease AERP over a time interval as short as several minutes.32

Although the AERP reduction caused by AF favours arrhythmia maintenance, it cannot be

the only factor involved because AF-induced AERP alterations become maximal well before

AF-promoting effects stabilize.7,34 One of the AF-promoting effects is tachycardia induced

atrial conduction slowing.27,33,34 It has a slower time course than AERP changes, probably

due to structural changes and could account for at least a part of the continued development

of AF promotion after AERP changes have stabilized.

Tieleman and coworkers found that the AERP shortened with loss of the normal rate

adaptation in response to 24 hours of rapid atrial pacing in goats.35 Here the tendency of

the atria to fibrillate increased. With resumption of sinus rhythm after cessation of pacing,

the refractory period normalized over a period of slightly more than 24 hours. This electrical

remodeling could be modulated using several pharmacological agents. First, the calcium

channel blocker verapamil reduced atrial electrical remodeling; suggesting that tachycardia-

induced calcium overload might trigger the shortening of the refractory period.35 By contrast,

digoxin delayed the recovery from electrical remodeling of the atria.36 This could be due

to the effect of digoxin on calcium handling, preventing effective wash-out of calcium

after cessation of pacing.

Contractile dysfunction of the atrium

Besides the electrical remodeling observed in experimental studies, clinical studies

15

Introduction

demonstrated atrial contractile dysfunction after AF.37,38 Leistad and co-workers investigated

the contractile dysfunction in an experimental model for AF.39 They demonstrated that the

atrial contractile dysfunction after acute atrial fibrillation is reduced by the calcium channel

antagonist verapamil, which suggests that transsarcolemmal calcium influx contributed to

this dysfunction. The calcium agonist BAY K8644 increased postfibrillation atrial contractile

dysfunction. A remarkable finding was that atrial contractility increased in the first seconds

after atrial fibrillation before a longer period of reduced atrial contractility ensued. On the

basis of these observations they hypothesised that cytosolic calcium overload due to rapid

depolarization during the preceding fibrillation might be responsible for the atrial contractile

dysfunction. This hypothesis was also suggested by Shapiro et al. in 1988.40It should be

noted that in the pig model of Leistad and coworkers39 atrial contractile dysfunction was

observed after only five minutes of AF, indicating that contractile dysfunction, like electrical

remodeling in the goat7, starts early.39

Structural remodeling

Until now, research focused on electrical and contractile remodeling. AF is also

associated with structural changes.27,31,41 Ausma and coworkers described and quantified

the structural remodeling in atrial myocardium due to sustained atrial fibrillation in the

goat.41 They maintained atrial fibrillation in normal goats for a prolonged period of time.

After 9 to 23 weeks of sustained atrial fibrillation, several areas of the right and left atria

were examined by light and electron microscopy. They found that a substantial proportion

of the atrial myocytes (up to 92%) revealed marked changes in their cellular substructures,

such as loss of myofibrils, accumulation of glycogen, changes in mitochondrial shape and

size, fragmentation of sarcoplasmic reticulum and dispersion of nuclear chromatin. They

pointed out that these changes are the same as observed in chronically hibernating

myocardium. This is a condition that occurs in patients as a result of low flow ischemia

caused by stenosis of one or more coronary arteries.41 The clinical condition is defined as

the ability of the myocardium to adapt to chronic ischemia by down-regulating its contractile

function, thereby maintaining cell viability for a prolonged period of time. Furthermore, it

was found that these hibernating myocytes resembled a form of dedifferentiation as a

result of chronic atrial fibrillation.42 Some of the atrial cardiomyocytes acquired a

dedifferentiated phenotype, as deduced from the re-expression of α-smooth muscle actin,

the disappearance of cardiotin and the staining patterns of titin, which resembled those of

embryonic cardiomyocytes.

Considerations for the thesis

Thus, experimentally it has been shown that contractile dysfunction starts accutely

after AF onset and was reduced by verapamil but increased by BAY K8644.39 These results

16

Chapter 1

strongly indicate that changes in the calcium homeostasis triggered by tachycardia induced

intracellular calcium overload 43-45, play a pivotal role in the induction of atrial contractile

dysfunction.

Shortening of the atrial effective refractory period was an other important factor contributing

to the maintenance of AF 7,46. This shortening could also be mediated by verapamil and

gave a reduction of the tachycardia induced electrical remodeling of the atria.35 This finding

also suggests that electrical remodeling could be due to changes in the calcium homeostasis.

Aim of the thesis

The main goal was to study the molecular remodeling in human atrial fibrillation. We

focussed on gene expression of proteins which influence the calcium homeostasis and

action potential duration in human AF. The impact of modulating systems like the natriuretic

peptide system and the endothelin system were also studied.

For the purpose of the thesis, right and left atrial appendages were collected during

four years from different groups of patients undergoing cardiac surgery. The patients clinical

characteristics (underlying heart disease, type of AF, electrocardiograms, medication and

exercise tolerance) were assessed. The result was a unique human atrial appendage collection

of around 150 individual patients. Because of this amount of different atrial appendages it

was possible to match AF patients with control patients in sinus rhythm for age, sex,

underlying heart disease, left ventricular function and as far as possible for medication

use. In this way, dissecting the effect of AF was optimized.

During the progression of the study a discrepancy between alterations in mRNA and

protein expression of several ion-channels in patients with paroxysmal AF was observed.

No alterations in mRNA expression of ion-channels compared to important reductions on

the protein level were found in human paroxysmal AF. This new finding prompted us to

explore an adaptive mechanism unknown to occur in AF. We hypothesized reduction in

protein channels due to calcium activated neutral protease calpain. To test this the calpain

activity, protein expression and localization were determined in tissue of patients with

atrial fibrillation. Finally the role of calpain activity in ion-channel protein, structural and

electrophysiological remodeling was studied.

17

Introduction

References

1. Waktare JEP, Camm AJ. Acute treatment of atrial fibrillation: why and when to maintain sinus rhythm. J

Am Coll Cardiol 1998; 81:3C-15.

2. The National Heart Lung and Blood Institute Working Group on Atrial Fibrillation. Atrial fibrillation:

current understandings and research imperatives. J Am Coll Cardiol 1993; 22:1830-1834.

3. Gallagher MM, Camm AJ. Classification of atrial fibrillation. Pacing Clin Electrophysiol 1997; 20:1603-

1605.

4. Levy S, Breithardt G, Campbell RW, et al. Atrial fibrillation: current knowledge and recommendations for

management. The Working Group on Arrhythmias of the European Society of Cardiology. Eur Heart J

1998; 19:1294-1320.

5. Godtfredsen J. Etiology, course and prognosis. A follow-up study of 1212 cases. Copenhagen: University

of Copenhagen. Thesis 1975

6. Van Gelder IC, Crijns HJGM, Tieleman RG, et al. Value and limitation of electrical cardioversion in

patients with chronic atrial fibrillation - importance of arrhythmia risk factors and oral anticoagulation.

Arch Intern Med 1996; 156:2585-2592.

7. Wijffels MC, Kirchhof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation. A study in awake

chronically instrumented goats. Circulation 1995; 92:1954-1968.

8. Moe GK, Abildskov JA. Experimental and laboratory reports. Atrial fibrillation as a self-sustained arrhythmia

independent of focal discharge. Am Heart J 1959; 58:59-70.

9. Moe GK. On the multiple wavelet hypothesis of atrial fibrillation. Arch Int Pharmacodyn Ther 1962;

140:183-188.

10. Allessie MA, Lammers WJEP, Bonke FIM, et al. Experimental evaluation of Moe’s multiple wavelet

hypothesis of atrial fibrillation, in Zipes DP, Jalife J (eds). Cardiac Electrophysiology and Arrhythmias

1985; New York, Grune & Stratton:265-275.

11. Konings KT, Kirchhof CJ, Smeets J, et al. High-density mapping of electrically induced atrial fibrillation

in humans. Circulation 1994; 89:1665-1680.

12. Gettes LS, Reuter H. Slow recovery from inactivation of inward currents in mammalian myocardial fibres.

J Physiol (Lond) 1974; 240:703-724.

13. Weidman S. The effect of the cardiac membrane potential on the rapid availability of the sodium-carrying

system. J Physiol 1955; 127:213-224.

14. Coraboeuf E, Carmeliet E. Existence of two transient outward currents in sheep cardiac Purkinje fibers.

Pflugers Arch 1982; 1982:352-359.

15. Tseng GN, Hoffman BF. Two components of transient outward current in canine ventricular myocytes.

Circ Res 1989; 68:424-437.

16. Zygmunt AC, Gibbons WR. Calcium-activated chloride current in rabbit ventricular myocytes. Circ Res

1991; 68:424-437.

17. Fabiato A. Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol

1983; 245:C1-14.

18. Whalley DW, Wendt DJ, Grant A. Basic concepts in cellular cardiac electrophysiology: Part I: Ion channels,

membrane currents, and the action potential. Pacing Clin Electrophysiol 1995; 18:1556-1574.

19. Giles WR, Imaizumi Y. Comparison of potassium currents in rabbit atrial and ventricular cells. J Physiol

(Lond) 1988; 405:123-145.

20. Heidbuchel H, Vereecke J, Carmeliet. Three different potassium channels in human atrium. Contribution

to the basal potassium conductance. Circ Res 1990; 66:1277-1286.

21. Yang ZK, Boyett MR, Janvier NC, et al. Regional differences in the negative inotropic effect of acetylcholine

within the canine ventricle. J Physiol 1996; 492:789-806.

22. Liu L, Nattel S. Differing sympathetic and vagal effects on atrial fibrillation in dogs: role of refractoriness

heterogeneity. Am J Physiol 1997; 273:H805-H816

23. Sipido KR, Carmeliet E, Van De Werf F. T-type Ca2+ current as trigger for Ca2+ release from the sarcoplasmic

reticulum in guinea-pig ventricular myocytes. J Physiol 1998; 508:439-451.

24. Sipido KR, Maes M, Van de Werf F. Low efficiency of Ca2+ entry through the Na+-Ca2+ exchanger as

trigger for Ca2+ release from the sarcoplasmic reticulum: a comparison between L-type Ca2+ current and

reverse mode Na+-Ca2+ exchange. Circ Res 1997; 80:1034-1044.

25. Sasaki T, Inui M, Kimura Y, et al. Molecular mechanism of regulation of Ca2+ pump ATPase by

phospholamban in cardiac sarcoplasmic reticulum. Effects of synthetic phospholamban peptides on Ca2+

pump ATPase. J Biol Chem 1992; 267:1674-1679.

18

Chapter 1

26. Gómez AM, Valdivia HH, Cheng H, et al. Defective excitation-contraction coupling in experimental

cardiac hypertrophy and heart failure. Science 1997; 276:800-805.

27. Morillo CA, Klein GJ, Jones D, et al. Chronic rapid atrial pacing. Structural, functional, and

electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation 1995; 91:1588-

1595.

28. Wijffels MC, Kirchhof CJ, Dorland R, et al. Electrical remodeling due to atrial firbrillation in chronically

instrumented conscious goats: role of neurohumoral changes, ischemia, atrial stretch, and high rate of

electrical activation. Circulation 1997; 96:3710-3720.

29. Franz MR, Karasik PL, Li C, et al. Electrical remodeling of the human atrium: similar effects in patients

with chronic atrial fibrillation and atrial flutter. J Am Coll Cardiol 1997; 30:1785-1792.

30. Yu WC, Chen SA, Lee SH, et al. Tachycardia-induced change of atrial refractory period in humans. Rate

dependency and effects of antiarrhythmic drugs. Circulation 1998; 97:2331-2337.

31. Goette A, Honeycutt C, Langberg JJ. Electrical remodeling in atrial fibrillation. Time course and mecha-

nisms. Circulation 1996; 94:2968-2974.

32. Daoud EG, Knight BP, Weiss R, et al. Effect of verapamil and procainamide on atrial fibrillation-induced

electrical remodeling in humans. Circulation 1997; 96:1542-1550.

33. Elvan A, Wylie K, Zipes DP. Pacing-induced chronic atrial fibrillation impaires sinus node function in

dogs: electrophysiological remodeling. Circulation 1996; 94:2953-2960.

34. Gaspo R, Bosch RF, Talajic M, et al. Functional mechanisms underlying tachycardia-induced sustained

atrial fibrillation in a chronic dog model. Circulation 1997; 96:4027-4035.

35. Tieleman RG, De Langen CDJ, Van Gelder IC, et al. Verapamil reduces tachycardia-induced electrical

remodeling of the atria. Circulation 1997; 95:1945-1953.

36. Tieleman RG, Blaauw Y, Van Gelder IC, et al. Digoxin delays the recovery from electrical remodeling of

the atria in the goat. Circulation 1999; 100:1836-1842.

37. Manning WJ, Silverman DI, Katz SE, et al. Impaired left atrial mechanical function after cardioversion:

relation to the duration of atrial fibrillation. J Am Coll Cardiol 1994; 23:1535-1540.

38. Daoud EG, Marcovitz P, Knight B, et al. Short-term effect of atrial fibrillation on atrial contractile func-

tion in humans. Circulation 1999; 99:3024-3027.

39. Leistad E, Aksnes G, Verburg E, et al. Atrial contractile dysfunction after short-term atrial fibrillation is

reduced by verapamil but increased by BAY K8644. Circulation 1996; 93:1747-1754.

40. Shapiro EP, Effron MB, Lima, et al. Transient atrial dysfunction after conversion of chronic atrial fibrilla-

tion to sinus rhythm. Am J Cardiol 1988; 62:1202-1207.

41. Ausma J, Wijffels M, Thone F, et al. Structural changes of atrial myocardium due to sustained atrial

fibrillation in the goat. Circulation 1997; 96:3157-3163.

42. Ausma J, Wijffels M, Van Eys G., et al. Dedifferentiation of atrial cardiomyocytes as a result of chronic

atrial fibrillation. Am J Pathol 1997; 151:985-997.

43. Lee HC, Clusin WT. Cytosolic calcium staircase in cultured myocardial cells. Circ Res 1987; 61:934-

939.

44. De Pauw M, Borgers M, Heyndrickx GR. Ultrastuctural calcium distribution in cardiac myocytes after

48h of rapid pacing in dogs. Circulation 1996; 94:I-604

45. Schouten VJA, Morad M. Regulation of Ca2+ current in frog ventricular myocytes by the holding poten-

tial, cAMP and frequency. Pflugers Arch 1989; 415:1-11.

46. Yue L, Feng J, Gaspo R, et al. Ionic remodeling underlying action potential changes in a canine model of

atrial fibrillation. Circ Res 1997; 81:512-525.

47. The Sicilian Gambit. a new approach to the classification of antiarrhythmic drugs based on their actions on

arrhythmogenic mechanisms. Task force of the Working Group on Arrhythmias of the European Society of

Cardiology. Circulation 1991; 84:1831-1851.

Part I

Gene expression of proteins influencing

calcium homeostasis

21

Gene expression of proteins influencing calcium homeostasis in patients

Chapter 2

Gene Expression of Proteins Influencing Calcium

Homeostasis in Patients with Persistent and Paroxysmal

Atrial Fibrillation

Bianca J. J. M. Brundela,b, Isabelle C. Van Geldera, Robert H. Henningb,

Anton E. Tuinenburga, Leo E. Deelmanb, Robert G. Tielemana,

Jan G. Grandjeanc, Wiek H. van Gilstb, Harry J. G. M. Crijnsa

From the departments of Cardiology (a), Clinical Pharmacology (b),

and Thoracic Surgery (c), Thoraxcenter, University Hospital Groningen,

Groningen, The Netherlands.

Cardiovascular Research 42 (1999) 443-454

Abstract

Objective: Persistent atrial fibrillation (AF) results in an impairment of atrial function.

In order to elucidate the mechanism behind this phenomenon, we investigated the gene

expression of proteins influencing the calcium handling. Methods: Right atrial appendages

were obtained from 8 patients with paroxysmal AF, 10 with persistent AF (> 8 months)

and 18 matched controls in sinus rhythm. All controls underwent coronary artery bypass

grafting whereas most AF patients underwent Cox’s MAZE surgery (n=12). All patients

had a normal left ventricular function. Total RNA was isolated and reversely transcribed

into cDNA. In a semi-quantitative polymerase chain reaction the cDNA of interest and of

glyceraldehyde-3-phosphate dehydrogenase were coamplified and separated by ethidium

bromide stained gel-electrophoresis. Slot blot analysis was performed to study protein

expression. Results: L-type calcium channel α1 and sarcoplasmic reticulum Ca2+-ATPase

mRNA (-57%, p=0.01 and -28%, p=0.04, respectively) and protein contents (-43%, p=0.02

and –28%, p=0.04, respectively) were reduced in patients with persistent AF compared to

the controls. mRNA contents of phospholamban, ryanodine receptor type 2 and sodium/

calcium exchanger were comparable. No changes were observed in patients with paroxysmal

AF. Conclusions: Alterations in gene expression of proteins involved in the calcium

homeostasis occur only in patients with long-term persistent AF. In the absence of underlying

heart disease, the changes are rather secundary than primary to AF.

22

Chapter 2

Introduction

Atrial fibrillation (AF) is the most common cardiac arrhythmia affecting millions of

people worldwide and its incidence increases with age [1]. Clinical observations showed

that immediately after restoration of sinus rhythm, atrial contractile function is severely

impaired or even absent [2, 3]. The contractile dysfuntion is reversible after restoration of

sinus rhythm, its time course being related to the previous duration of AF [2]. Restoration

does not seem to be complete in all patients presumably related to the extent of damage

occuring during AF.

In an experimental pig model atrial contractile dysfunction was observed after cessation

of pacing induced AF of a duration of only 1 to 30 minutes, indicating that contractile

remodeling, like electrical remodeling in the goat [4], is an early process [5]. There are

strong indications that abnormalities in the calcium handling, in response to tachycardia

induced intracellular calcium overload [6_8], play a pivotal role in the induction of atrial

contractile dysfunction. [5, 9_13]. The proteins and ion channels involved in the adaptation

processes during AF have not been clarified yet. Most likely, identification of the signaling

pathways and their target genes may lead to new therapeutic options for the treatment of

AF. Therefore, we investigated alterations in mRNA and protein expression of proteins

involved in the calcium handling of right atrial appendages (RAA) of patients with

paroxysmal and persistent AF undergoing cardiac surgery. To overcome the problem

whether changes were caused by AF itself, or by the concomitant underlying heart disease,

we selected AF patients with a normal left ventricular function and matched them for age,

sex and left ventricular function with patients in sinus rhythm who underwent coronary

artery bypass surgery.

Methods

Patients

The day before surgery, one investigator (AET) assessed the clinical characteristics

of the patients. The patients’ history and previous electrocardiograms were used to establish

type and duration of AF. In addition, the patients were asked for medication use and

exercise tolerance (according New York Heart Association classification). Echocardio-

graphy data were obtained within 3 months before surgery. RAAs were obtained from 8

patients with paroxysmal AF and from 10 with persistent AF without valvular heart disease

and a normal left ventricular function. The AF patients were matched for age, sex, left

ventricular function, and as far as possible for medication with 18 clinically stable patients

in sinus rhythm undergoing coronary artery bypass surgery. The Institutional Review Board

approved the study, and all patients gave written informed consent. Immediately after

excision, the RAAs were snap-frozen in liquid nitrogen and stored at -85 °C.

23


RNA isolation and cDNA synthesis

Total RNA was isolated from RAAs using the method of acid guanidinium thiocyanate/

-phenol/chloroform extraction followed by a RNeasy kit for RNA minipreps from tissues

(Qiagen). The amount of RNA was evaluated by absorption at 260 nm, using a GeneQuant

II (Pharmacia Biotechnology, The Netherlands). The ratio of absorption (260-280 nm) of

all preparations was between 1.8 and 2.0. First strand cDNA was synthesized by incubation

of 1 µg of total RNA, reverse transcription 10x buffer and 200 ng of random hexamers

with 200 units of Moloney Murine Leukemia Virus Reverse Transcriptase, 1mM of each

dNTP and 1 unit of RNase inhibitor (Promega, The Netherlands) in 20 µl. The synthesis

reaction lasted 10 minutes at 20 °C, 20 minutes at 42 °C, 5 minutes at 99 °C and 5 minutes

at 4 °C, respectively. All the products were checked on contaminating DNA (data not

shown).

Semi quantitative PCR analyses

Since a linear relationship between the amount of input template and amplification

product exists within the exponential range of amplification, a semi-quantitative polymerase

chain reaction (PCR) was developed [14]. The cDNA of interest and the cDNA of the

ubiquitously expressed housekeeping gene glyceraldehyde-3-phosphate dehydrogenase

(GAPDH) were coamplified in a single PCR reaction. Primers were designed for

Sarcoplasmic Reticulum Calcium ATPase (SR Ca2+-ATPase), Phospholamban (PLB), L-

type Calcium Channel α1 subunit (L-type Ca2+), Sodium/ Calcium exchanger (Na+/Ca 2+

exchanger), Ryanodine receptor type 2 (RyR2) and the housekeeping enzyme GAPDH

(Table 1). Eurogentec (Belgium) synthesized the oligonucleotides. For the semi-quantitative

PCR co-amplification of 1 µl of cDNA mixture, 0.5 unit of Taq polymerase (Eurogentec,

Belgium) was added to 17.5 nM dNTPs, 10x PCR buffer provided with Taq polymerase,

2.5 mM MgCl2, 40 pmol of sense and antisense primer for the gene of interest, 40 pmol of

sense and antisense GAPDH primer and water to bring the final volume to 50 µl. All

reaction mixtures were overlaid with 50 µl of mineral oil (Sigma, The Netherlands). After

3 min denaturation at 94 °C, n cycles (Table 1) of amplification were performed, each for

1 min at 94°C, 1 min at 56 °C and 1 min at 72 °C, using the thermocycler Perkin Elmer

480 (The Netherlands). After the last cycle the 72 °C elongation step was extended to 5

min. The PCR products were separated on 1-2.5% agarose gels by gel-electrophoresis and

stained with ethidium bromide. The densities of the PCR products were quantified by

densitometry (Aldus PhotoStyler 2.0, Grafic Workshop and ImageQuant Version 3.3).

During the PCR for L-type Ca2+ the three isoforms of this channel were amplified.

The primers were designed to amplify the IVS1 upto IVS5 region of the α1 subunit. The

PCR fragment contained the IVS3B adult isoform, the IVS3B deleted D1 isoform and the

IVS3A fetal isoform, which could be identified as products of 472 bp, 442 bp and 355 bp,

24

Chapter 2

respectively, after digestion with DdeI (Promega, The Netherlands).

The RyR2 PCR fragment was designed to contain the alternative splicing site, a 24

bp insert between residues 11145 and 11146 with a restriction site for BamH1. The RyR2

isoform with the 24 bp insert could be identified as products of 310 bp and 326 bp after

digestion with BamH1 (Promega, The Netherlands).

Determination of the absolute alterations of mRNA

To validate the semi quantitative PCRs the changes observed in ratios for L-type Ca2+

and SR Ca2+ ATPase were determined. Increasing amounts of gene of interest were added

to a fixed amount of GAPDH. Therefore, known amounts of GAPDH input template (range

10-2 to 10-4 ng, 410 bp) were added to a PCR sample mixture of 0.5 unit of Taq polymerase

(Eurogentec, Belgium),17.5 nM dNTPs, 10x PCR buffer, 2.5 mM MgCl2, 40 pmol of

sense and antisense GAPDH primer, 0.7 µl of cDNA mixture and water to bring the final

volume to 50 µl, and amplified for 25 cycles. Thereafter, a fixed amount of GAPDH input

template was used in a PCR with known amounts of the SR Ca2+-ATPase input template

(range 10-1.5 ng to 10-3.5 ng, 657 bp) or L-type Ca2+ input template (range 10-1.5 ng to 10-3.5

ng, 563 and 530 bp). The ratios SR Ca 2+ ATPase input or L-type Ca2+ input versus GAPDH

input were calculated.

Table 1. Sequence for the primers

Sequence cycles annealing

temp (0C)

Glyceraldehyde-3-phosphate dehydrogenase:

F 5'-CCC ATC ACC ATC TTC CAG GAG CG-3', - 56

R 5'-GGC AGG GAT GAT GTT CTG GAG AGC C-3'

Phospholamban:

F 5'-ATG GAG AAA GTC CAA TAC CTC ACT CGC-3', 25 56

R 5'-TCA GAG AAG CAT CAC GAT GAT ACA GAT CAG-3'

Sarcoplasmic reticulum calcium ATPase:

F 5’-TGT TCA TTC TGG ACA GAG TGG AAG G-3’ 25 56

R 5’-TTA ATA AAG TTG GCA GAG TCC TCA AGG-3’

L-type calcium channel :

F 5’-CTG GAC AAG AAC CAG CGA CAG TGC G-3’, 32 56

R 5’-ATC ACG ATC AGG AGG GCC ACA TAG GG-3’

Sodium-calcium exchanger:

F 5’- CTA CCA AGT CCT AAG TCA GCA GC3’, 27 56

R 5’-GAT CCG AGG CAA GCA AGT GTA GA-3’

Ryanodine receptor type 2:

F 5’-AAG GCA TCG GGC TGT CAA TCT-3’ 28 56

25


Protein Preparation and Slot-Blot Analysis

Frozen RAAs of 5 patients in sinus rhythm, 5 patients with paroxysmal and 5 with

persistent AF were homogenized in RIPA buffer (1% NP40, 0.5% sodium deoxycholate,

0.1% sodium dodecyl sulfate (SDS), 10 mM β mercapto-ethanol, 10mg/ml PMSF, 5µl/ml

aprotinin, 100 mM sodium orthovanadate, 5µl/ml benzamidine, 5µl/ml pepstatine A, 5 µl/

ml leupeptine in 1x PBS by use of an ultraturrax (Polytron, The Netherlands) with 10

seconds intervals. The homogenate was centrifuged at 14.000 rpm for 20 minutes at 4°C.

After centrifugation the supernatant was carefully removed and used for protein

concentration measurement. This was done according to the Bradford method (Sigma,

The Netherlands) with bovine albumin used as a standard. Samples of 10 µg protein were

denaturated by heating to 95°C before spotting on a TBS (10 mM Tris-HCl pH 8.0, 150

mM NaCl ) wetted nitrocellulose membrane (Bio-Rad, The Netherlands) by use of a slot

blot apparatus (Bio-Rad, The Netherlands). The membrane was washed twice with 200 µl

TBS buffer and the transfer was checked by staining the nitrocellulose membrane with

Ponceau S solution (Sigma, The Netherlands). Blocking was performed for 20 minutes in

blocking buffer (5% nonfat milk, TBS and 0.1% Tween 20). After three times washing for

5 minutes in TBS with 0.1% Tween 20 the membranes were incubated for 90 minutes with

primary antibodies: SR Ca 2+ ATPase 1:2500, PLB 1: 200 (gifts from dr. F. Wuytack

University Leuven, Belgium), GAPDH 1:5000 (Affinity Reagents, USA) or L-type Ca2+

anti α1-subunit (Alomone Labs, Israel). Immunodetection of the primary antibody was

performed, after three times washing for 5 minutes with TBS and 0.1% Tween 20, with

peroxidase conjugated secondary antibody anti-rabbit and anti-mouse IgG (Santa Cruz

Biotechnology, The Netherlands) for 60 minutes. The primary and secondary antibodies

were diluted in blocking buffer. The blot was washed two times for 5 minutes in TBS and

0.1% Tween 20 and one time in TBS also for 5 minutes. Consequently, the blot was

incubated with the ECL-detection reagent (Amersham, The Netherlands) for 1 minute,

and exposed to a X-OMAT x-ray film (Kodak, The Netherlands) for 15 to 90 seconds. The

band densities were evaluated by densitometric scanning using a Snap Scan 600 (Agfa,

The Netherlands). To test the linearity of the immunodetection system distinct amounts of

protein were analyzed. There was a linear relation between protein amounts spotted on the

membrane and the immunoreactive signals of L-type Ca2+, SR Ca 2+-ATPase, PLB and

GAPDH (data not shown).

Definitions

Persistent AF: continuous presence of AF until the moment of cardiac surgery, i.e. at

least two consecutive electrocardiograms of AF more than 24 hours apart, without

intercurrent SR. Persistent AF has a non-spontaneously converting character [15, 16].

Previously, this type of AF was classified as chronic AF.

26

Chapter 2

Paroxysmal AF: AF typically occurring in episodes of a duration shorther than 24

hours (but longer lasting paroxysms are not unusual) with intermittent sinus rhythm.

Paroxysmal AF is either converting spontaneously or is terminated with an intravenously

administered antiarrhythmic drug [15, 16]. It is non-controlled whether paroxysmal AF is

present at the moment of cardiac surgery.

Statistical Analysis

All PCRs and SDS-PAGEs were performed in duplo series. The mean values of the

ratios were used for statistical analysis. To compare the baseline characteristics between

groups for normally distributed variables, mean values and standard deviations are reported.

In case of skewed distribution of variables, the median values and ranges are given. Baseline

comparison between groups for normally distributed variables was performed by one-way

ANOVA for skewed distributed variables by the Wilcoxon two-sample test. The Chi-square

test with continuity correction or Fisher’s exact test was performed for group comparison

for categorical variables when appropriate. To determine which variables influenced mRNA

levels of proteins, univariate regression analyses were performed. Only variables with a p

value < 0.15 were selected for multiple regression analysis. To determine differences in

mRNA levels of these proteins between the four groups a Tukey correction for multiple

comparison was performed.

For determination of correlations the Spearman correlation test was used. The Mann-

Whitney U-test was performed for group to group comparisons of small numbers.

All p-values are two-sided, a p-value <0.05 was considered statistically significant.

SAS version 6.12 (Cary, NC) was used for all statistical evaluations.

Results

Patients

Included were 8 patients with paroxysmal and 10 patients with persistent AF. These

two groups were compared with two groups of control patients in sinus rhythm, who were

matched for sex, age and left ventricular function (Table 2). Six of the 8 patients with

paroxysmal AF suffered from intractable AF and were scheduled for Cox’s MAZE surgery.

The median duration of sinus rhythm before surgery was 1.5 days. The median frequency

of paroxysms was once a day (median duration of each paroxysm was 3 hours). Three

patients with paroxysmal AF had AF at the moment of surgery and harvesting of the RAA.

Control RAAs were obtained from clinically stable patients in sinus rhythm who were

scheduled for coronary artery bypass surgery. Although the AF groups and their controls

in sinus rhythm differed with respect to the underlying heart disease, all had a normal left

ventricular function and were in the functional class I or II for exercise tolerance. Also,

echocardiographic atrial and left ventricular dimensions were similar among groups.

27


Alterations in mRNA Levels in Paroxysmal and Persistent AF

Changes in mRNA levels of the gene of interest were determined by comparison of

gene-of-interest/GAPDH ratios between patients with persistent AF or with paroxysmal

AF, and their matched controls in sinus rhythm. The densities of the amplified GAPDH of

the 4 groups of patients were the same for all the genes investigated (data not shown).

Only patients with persistent AF showed a significant reduction of the cDNA ratios of L-

type Ca2+/GAPDH (-57%) and SR Ca2+-ATPase/GAPDH (-28%) (Figure 1A and 1B). The

cDNA ratios of Phospholamban/GAPDH , Na+/Ca2+ exchanger/GAPDH and RyR2/GAPDH

were unchanged compared to the controls in sinus rhythm (Figure 1C, 1D and 1E).

Table 2. Baseline characteristics of patients with paroxysmal AF, persistent AF and patients in sinus

rhythm at the moment of surgery.

PAF SR (PAF) CAF SR (CAF)

N 8 8 10 10

Male/ female (n) 6/2 6/2 6/4 6/4

Age 51 ±7 56 ± 11 63 ± 11 65 ± 17

Previous duration of AF (median, range (months) - - - 18 (8-64) -

Duration SR before surgery (median, range (days) 1.5 (0-30) - - -

Underlying heart disease (n)

Coronary artery disease 2* 8 4* 10

Hypertension 1 1 3 2

Lone AF 6* 0 5* 0

Surgical procedure

Coronary Artery Bypass Grafting 2* 8 4* 10

MAZE 6* 0 6* 0

New York Heart Association for exercise tolerance

Class I 7 5 6 5

Class II 1 3 4 5

Left atrial diameter (long axis, mm) 43 ± 7 41 ± 3 45 ± 7 44 ± 5

Left atrial diameter (apical, mm) 60 ± 6 64 ± 3 63 ± 4 64 ± 6

Right atrial diameter (apical, mm) 54 ± 9 54 ± 4 62 ± 7 57 ± 4

Left ventricular end-diastolic diameter (mm) 48 ± 4 49 ± 8 53 ± 3 53 ± 6

Left ventricular end-systolic diameter (mm) 35 ± 4 35 ± 7 33 ± 6 35 ± 4

Beta blockers 1* 5 3 6

Calcium antagonists 0 3 3 3

Digitalis 0 1 5 3

ACE inhibitors 0 1 4 2

* p-value < 0.05 compared to the control group

Values are presented as mean value ± SD. ACE indicates Angiotensin Converting Enzyme; CAF, chronic

persistent atrial fibrillation; PAF, paroxysmal atrial fibrillation; SR (CAF), matched controls in sinus rhythm of

patients with persistent AF; SR (PAF), matched controls in sinus rhythm of patients with paroxysmal AF.

28

Chapter 2

No changes were observed in patients suffering from paroxysmal AF. Table 3 shows that

the cDNA ratio L-type Ca2+/GAPDH in patients with persistent AF was neither influenced

Figure 1.

Individual cDNA ratios for L-type Ca2+/GAPDH (1A), SR Ca2+ ATPase/GAPDH (1B), PLB/GAPDH (1C), Na+/

Ca2+ exchanger/GAPDH (1D) and RyR2/GAPDH (1E). Mean values ar given for the 4 different groups ± SEM.

SR-CAF is the sinus rhythm control group for patients with chronic persistent AF, SR-PAF is the sinus rhythm

control group for patients with paroxysmal AF. All data are represented in density units/density units. Significance

was determined using the Tukey correction for multiple regression analysis for each mean of the four groups.

29


by the underlying heart disease nor by any (calcium handling influencing) drug. However,

patients in the sinus rhythm control group for paroxysmal AF who were treated with a beta

blocker (n=5) had an increased cDNA ratio for L-type Ca2+/GAPDH compared to those

who were not treated with the drug (n=3).

Validation of the Absolute mRNA Contents

To assess the amounts of L-type Ca2+ and SR Ca2+-ATPase in the cDNA mixtures in

different groups, a semi-quantitative PCR was developed with increasing amounts of input

cDNA of interest to a standard amount of GAPDH cDNA (0.018 ng).The ratios were

determined and plotted in a logarithmic way. This resulted in a straight line demonstrating

the validity of the method and enabling estimations of differences of L-type Ca2+ and SR

Ca2+ ATPase in the cDNA mixture (Figure 2A and 2B).

Figure 2.

Plot showing a significant correlation between the increasing doses of L-type Ca2+ (A) and SR Ca2+ ATPase (B)

input template and the ratios of L-type Ca2+/GAPDH and SR Ca2+ ATPase/GAPDH respectively. Correlation was

determined by the Spearman correlation test.

30

Chapter 2

Alterations in Proteins Levels in Paroxysmal and Persistent AF

Changes in protein levels were studied for the genes which showed alterations in

mRNA ratios (L-type Ca2+ and SR Ca2+-ATPase) and for phospholamban. Sufficient RAA

tissue to carry out protein isolations were available from 5 patients with paroxysmal AF, 5

with persistent AF and 5 patients in sinus rhythm. The protein levels of gene-of-interest/

GAPDH ratios were determined. Figure 3A shows the specificity of the antibodies used,

figure 3B the results of the slot-blot analysis. The protein ratio of L-type Ca2+/GAPDH

and SR Ca2+-ATPase/GAPDH were significantly reduced in patients with persistent AF

compared to the controls (-43%, p=0.02 and –28%, p=0.04, respectively, Figures 4A and

B). The PLB/GAPDH ratio showed a slight upregulation in patients with persistent AF

(Figure 4C). The GAPDH levels were comparable between the different groups (mean

values not shown). No alterations were found in the protein levels of patients with

paroxysmal AF (Figure 4). A positive correlation between the mRNA ratios and the protein

ratios of L-type Ca2+ and SR Ca2+-ATPase for patients with persistent AF and sinus rhythm

could be demonstrated (Figure 5).

Table 3. Effects of underlying heart disease and medication on the mRNA levels of l-type Ca2+ channel.

SR-CAF CAF SR-PAF PAF

CABG yes 2.4 (1.2–3.1, n=10) 1.6 (1.5–1.9, n=4) 2.6 (1.2–3.7, n=8) 2.6 (2.6-2.6, n=2)

no - 1.8 (0.7-2.4, n=6) - 2.2 (1.9-2.6,n=6)

MAZE yes - 1.8 (0.7-2.4, n=6) - 2.6 (2.6-2.6, n=6)

no 2.4 (1.2-3.1, n=10) 1.8 (1.5-2.2, n=4) 2.6 (1.2-3.7, n=8) 2.2(1.9-2.6, n=2)

Beta blockers yes 2.5 (1.9-3.1, n=6) 1.7 (1.5-1.9, n=3) 3.1 (2.5-3.7, n=5)* 2.6 (n=1)

no 2.2 (1.2-2.6, n=4) 1.8 (0.7-2.4, n=7) 1.8 (1.2-2.4, n=3)* 2.3 (1.9-2.6, n=7)

Calcium entry yes 2.7 (2.5-2.9, n=3) 1.8 (1.5-1.9, n=3) 2.4 (1.7-3.2, n=3) -

Blockers no 2.3 (1.2-3.1, n=7) 1.8 (0.7-2.4, n=7) 2.7 (1.2-3.7, n=5) 2.3 (1.9-2.6, n=8)

Digitalis yes 2.4 (1.9-2.7, n=3) 1.8 (1.5-2.2, n=5) 2.5 (n=1) -no 2.4 (1.2-3.1, n=7) 1.7 (0.7-2.4, n=5) 2.6 (1.2-3.7, n=7) 2.3 (1.9-2.6, n=8)

ACE-inhibitors yes 2.0 (1.9-2.1, n=2) 1.5 (0.7-2.2, n=4) 3.2 (n=1) -

no 2.5 (1.2-3.1,n=8) 1.9 (1.5-2.4, n=6) 2.5 (1.2-3.7, n=7) 2.3 (1.9-2.6, n=8)

* p value is 0.01 between SR-PAF with and without beta blocker (Mann-Whitney U-test)

The values are expressed as mean (range, number of patients). CAF, chronic persistent atrial fibrillation;

PAF, paroxysmal atrial fibrillation; SR (CAF), matched controls in sinus rhythm of patients with persistent

AF; SR (PAF), matched controls in sinus rhythm of patients with paroxysmal AF.

31


Figure 3.

A.Western blot analysis showing the specificity of the antibodies on human RAA. Anti-L-type calcium channel

α1 subunit (200 kD, lane 1), anti-SR Ca2+ ATPase (105 kD,lane 2), anti-phospholamban (6.5 kD, lane 3), GAPDH

(36 kD,lane 4) and the Marker (lane 5, Bio Rad, The Netherlands).

B. Slot-blot analysis of L-type Ca2+, SR Ca2+ ATPase, PLB and GAPDH of 5 control patients, 5 patients with

chronic persistent AF (CAF) and 5 patients with paroxysmal AF (PAF).

1 2 3 4 M

200 kD

116 kD97 kD

66 kD

45 kD

31 kD

22 kD14 kD

6.5 kD

A

GAPDH

PLB

SR Ca2+ ATPase

Control

CAF

PAF

Control

Control

CAF

CAF

PAF

PAF

PAF

Control

CAF

L-type Ca2+B

A

B

32

Chapter 2

Isoforms of L-type Calcium Channel

To investigate whether dedifferentiation occurred as an adaptation process in patients

with AF, the mRNA expression level of the fetal isoform compared to the adult isoforms

of L-type Ca2+ was determined. The amplified PCR fragment of L-type Ca2+ contained the

IVS3A (fetal form), the IVS3B (adult form) and the IVS3B deleted D1 form after digestion

with DdeI (Figure 6). No differences in percentages of the adult and fetal isoforms of L-

type Ca2+ were found in relation to the total quantity of the amplified L-type Ca2+ in the

different groups (Figure 7A). The ratio of the IVS3B form was significantly reduced in

patients with persistent AF (p=0.01, Figure 7B). No significant changes were seen in the

IVS3B deleted D1 form and the IVS3A form. Any significant alteration in the L-type Ca2+

ratios was observed in patients with paroxysmal AF (Figure 7B, 7C and 7D).

Figure 4.

Protein ratios of L-type Ca2+/GAPDH (A), SR Ca2+ ATPase/GAPDH (B) and PLB/GAPDH (C) of the individual

patients. SR is the sinus rhythm control group. All data are presented as density units/density units. Values are

mean ± SEM. Significant differences are indicated (Mann-Whitney U-test).

33


Figure 5.

Relationship between the mRNA and protein expression of L-type Ca2+(A), SR Ca2+ ATPase (B) and PLB (C).

(•) represents chronic persistent AF patients, (ο) represents SR patients. Correlation was determined by the

Spearman correlation test.

Discussion

Main Findings

We examined five different genes which play an important role in the calcium

homeostasis of the myocardial cell. By examining the mRNA and protein expression in

patients with paroxysmal AF and persistent AF, and for age, sex and cardiac function

matched controls in sinus rhythm we observed two important features. First, the mRNA

and protein contents of both SR Ca2+-ATPase and L-type Ca2+ were significantly reduced

in patients with persistent AF. Secondly, significant changes occurred only in patients with

34

Chapter 2

Figure 7.

Figure A shows the percentage IVS3B, IVS3B deletion and IVS3A in relation to the total L-type Ca2+ protein

mRNA expression. No differences in percentage were found between the three isoforms. The cDNA ratios of the

IVS3B/GAPDH was significantly lower in patients with persistent AF (B). No alterations were observed between

the groups of cDNA ratios of IVS3B deletion/GAPDH (C) and the fetal form IVS3A/GAPDH (D). SR-CAF are

the matched control patients in sinus rhythm for chronic persistent AF and SR-PAF for paroxysmal AF. All data

are presented as density units/density units. Values are mean ± SEM. Significant differences are indicated (Mann-

Whitney U-test).

Figure 6.

Typical example of an agarose gel. Here, the L-type Ca2+ isoforms (IVS3B, 472 bp; IVS3B deletion, 442 and

IVS3A, 355 bp) of 4 patients with chronic persistent AF and their matched controls in sinus rhythm and patients

with paroxysmal AF and their controls are shown.

IVS3BIVS3B del.

IVS3A

GAPDH

CAF CAF CAF CAF CAF PAF PAF

SR SR SR SR SR M SR SR

35


persistent AF but not in those with paroxysmal AF.

Alterations in Gene Expression of Ion Channels and Proteins Involved in the Calcium

Handling

The alterations observed in the present study of a reduction of the mRNA and protein

contents for the L-type Ca2+ without changes in L-type Ca2+ isoforms in patients with

persistent AF, fit in with the described changes in the above mentioned experimental models.

Clearly, a lower mRNA and protein expression of L-type Ca2+ may underlie reduced L-

type Ca2+ current densities [10, 11, 17]. Whereas in the experiments of Yue et al. significant

reduction of L-type Ca2+ current densities was observed within 6 weeks of rapid continuous

atrial pacing, we, however, observed significant alterations only in patients with persistent

longstanding (> 8 months) AF.

Data on alterations in the sarcoplasmic reticulum proteins influencing the calcium

handling in RAA tissue of patients with AF are lacking. The observed reduction of the SR

Ca2+-ATPase mRNA and protein expression in atrial tissue of patients with persistent AF

is comparable to data on left ventricular tissue of patients with severe heart failure due to

dilated and ischemic cardiomyopathy [18_23]. In our population both mRNA and protein

expression were reduced. In contrast, a discrepancy between adaptation of the SR Ca2+-

ATPase mRNA, protein and activity levels have been reported in ventricular tissue of

patients suffering from severe heart failure [19, 20, 23_25]. The mRNA and protein

expression of phospholamban, the regulatory protein of SR Ca2+-ATPase, was not

significantly different between the AF groups and their controls. A reduced gene expression

of SR Ca2+-ATPase and a not significantly changed expression of phospholamban yields a

reduced ratio of SR Ca2+-ATPase to phospholamban in patients with persistent AF. If we

assume that the stoichiometry of phospholamban to SR Ca2+-ATPase determines the level

of SR Ca2+-ATPase inhibition, this finding may indicate that in the basal low-phosphorylated

state, depression of SR calcium uptake is even more pronounced than would be expected

from the lower SR Ca2+-ATPase mRNA level in patients with persistent AF. This

interpretation would be consistent with functional abnormalities observed in the failing

human ventricular myocardium [22].

For the ryanodine receptor the amounts of a RyR2 region with or without the 24 bp

insert were examined. Alternative transcripts with or without this insertion might provide

a means for altering the binding affinity of this putative site for calcium [26]. No differences

were found in the expression of RyR2 with or without insert between the groups. This

suggests that persistent AF did not induce differences in the calcium binding affinity and

changes in fundamental nature of the RyR2. Furthermore, no changes in mRNA and protein

expression of RyR2 were observed in patients with AF.

36

Chapter 2

No Dedifferentiation of the L-type Calcium Channel During AF

Each IVS3 isoform is encoded by a separate but adjacent exon within a single genomic

clone and the various isoforms are generated by a developmentally regulated, mutually

exclusive exon splicing of the primary transcript [27, 28]. No reversion of the adult isoform

of L-type Ca2+ to its fetal isoform was observed in patients with AF. In contrast, however,

Gidh-Jain et al. demonstrated that in patients with left ventricular hypertrophy a significant

increase of the mRNA contents of the fetal isoform and reversion of fetal/ adult isoform

ratio to the fetal phenotype was observed in ventricular tissue [29]. In atrial goat tissue,

Ausma et al. showed that during pacing induced AF proteins which are present in embryonic/

fetal myocardial cells, e.g. α-smooth muscle actin, were reexpressed [30]. A reversion to

fetal isoforms of certain ion channels and proteins might be hypothesized to occur in

situations comparable to the embryonic situation, e.g. during higher heart rates as is the

case during AF. Therefore, a reexpression of fetal proteins might have occurred during AF.

No Changes in Gene Expression in Patients with Severe Paroxysmal AF

We observed significant alterations in expression of the investigated genes only in

patients with longstanding (> 8 months) persistent AF. No significant changes were observed

in those patients suffering from paroxysmal AF. Importantly, the included paroxysmal AF

patients had severe AF with daily episodes of AF. Moreover, 3 patients had AF at the

moment of surgery, i.e. at the moment of harvesting of the right atrial appendage. This

may suggest that episodes of sinus rhythm in between episodes of AF protect the myocardial

cell from alterations in gene expression.

Limitations of the Study

Drugs and differences in underlying diseases may influence gene expression of proteins

and ion channels influencing the calcium handling. In this study, to minimize the influence

of particular clinical parameters on gene expression, we included only patients with a

normal left ventricular function and, when possible, drugs were discontinued before surgery.

The present study does not clarify when the adaptive mechanisms of the atrial

myocardial cell, i.e. alterations in gene expression, start. Although our data suggest that

this is a late process, no patients with persistent AF with a duration between 1 day and 8

months were included .

No matched controlled analysis could be performed for determination of protein levels.

However, no significant changes in mRNA levels between the control groups were observed.

Therefore, in our opinion, a comparison between persistent AF, paroxysmal AF and sinus

rhythm patients seems to be justified.

37


Acknowledgments

Dr. Van Gelder was supported by Grant 94.014 of the Netherlands Heart Foundation, The

Hague, The Netherlands. The study was supported by Grant 96.051 of the Netherlands

Heart Foundation, The Hague, The Netherlands. We are indebted to Pieter J. De Kam for

statistical analysis.

References

1 Kannel WB, Abbott RD, Savage DD, McNamara PM. Epidemiologic features of chronic atrial fibrilla-

tion: the Framingham study. N Engl J Med 1982;306:1018-1022.

2 Manning WJ, Silverman DI, Katz SE, et al. Impaired left atrial mechanical function after cardioversion:

relation to the duration of atrial fibrillation. J Am Coll Cardiol 1994;23:1535-1540.

3 Van Gelder IC, Crijns HJ, Blanksma PK, et al. Time course of hemodynamic changes and improvement

of exercise tolerance after cardioversion of chronic atrial fibrillation unassociated with cardiac valve

disease. Am J Cardiol 1993;72:560-566.

4 Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study

in awake chronically instrumented goats. Circulation 1995;92:1954-1968.

5 Leistad E, Aksnes G, Verburg E, Christensen G. Atrial contractile dysfunction after short-term atrial

fibrillation is reduced by verapamil but increased by BAY K8644. Circulation 1996;93:1747-1754.

6 Lee HC, Clusin WT. Cytosolic calcium staircase in cultured myocardial cells. Circ Res 1987;61:934-

939.

7 De Pauw M, Borgers M, Heyndrickx GR. Ultrastuctural calcium distribution in cardiac myocytes after

48h of rapid pacing in dogs. Circulation 1996;94:I-604 (Abstract)

8 Schouten VJA, Morad M. Regulation of Ca2+ current in frog ventricular myocytes by the holding

potential, cAMP and frequency. Pflugers Arch 1989;415:1-11.

9 Ausma J, Wijffels M, Thone F, Wouters L, Allessie M, Borgers M. Structural changes of atrial myocar-

dium due to sustained atrial fibrillation in the goat. Circulation 1997;96:3157-3163.

10 Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S. Ionic remodeling underlying action potential changes

in a canine model of atrial fibrillation. Circ Res 1997;81:512-525.

11 Van Wagoner DR, Lamorgese M, Kirian P, Cheng Y, Efimov IR, Mazgalev TN. Calcium current density

is reduced in atrial myocytes isolated from patients in chronic atrial fibrillation. Circulation 1997;96:I-

180 (Abstract)

12 Daoud EG, Knight BP, Weiss R, et al. Effect of verapamil and procainamide on atrial fibrillation-in-

duced electrical remodeling in humans. Circulation 1997;96:1542-1550.

13 Sun H, Gaspo R, Leblanc N, Nattel S. Cellular mechanism of atrial contractile dysfuntion caused by

sustained atrial tachycardia. Circulation 1998;98:719-727.

14 Brundel BJ, Van Gelder IC, Henning RH, Tuinenburg AE, Van Gilst WH, Crijns HJ. Downregulation of

the mRNA expression of the acethylcholine-activated potassium channel and L-type calcium channel in

patients with chronic atrial fibrillation. Pacing Clin Electrophysiol 1997;20:II:1050 (Abstract)

15 Gallagher MM, Camm AJ. Classification of atrial fibrillation. Pacing Clin Electrophysiol 1997;20:1603-

1605.

16 Levy S, Breithardt G, Campbell RW, et al. Atrial fibrillation: current knowledge and recommendations

for management. The Working Group on Arrhythmias of the European Society of Cardiology. Eur

Heart J 1998;19:1294-1320.

17 Yue L, Wang Z, Gaspo R, Nattel S. The molecular mechanism of ionic remodeling of repolarization in a

dog model of atrial fibrillation. Circulation 1998;96:I-180 (Abstract)

18 Mercadier JJ, Lompre AM, Duc P, et al. Altered sarcoplasmic reticulum Ca2+-ATPase gene expression in

the human ventricle during end-stage heart failure. J Clin Invest 1990;85:305-309.

19 Hasenfuss G, Reinecke H, Studer R, et al. Relation between myocardial function and expression of

sarcoplasmic reticulum Ca2+-ATPase in failing and nonfailing human myocardium. Circ Res 1994;75:434-

442.

20 Linck B, Boknik P, Eschenhagen T, et al. Messenger RNA expression and immunological quantification

of phospholamban and SR-Ca2+-ATPase in failing and nonfailing human hearts. Cardiovasc Res

1996;31:625-632.

21 Studer R, Reinecke H, Bilger J, et al. Gene expression of the cardiac Na+-Ca2+ exchanger in end-stage

human heart failure. Circ Res 1994;75:443-453.

38

Chapter 2

22 Meyer M, Schillinger W, Pieske B, et al. Alterations of sarcoplasmic reticulum proteins in failing human

dilated cardiomyopathy. Circulation 1995;92:778-784.

23 Flesch M, Schwinger RH, Schnabel P, et al. Sarcoplasmic reticulum Ca2+ATPase and phospholamban

mRNA and protein levels in end-stage heart failure due to ischemic or dilated cardiomyopathy. J Mol

Med 1996;74:321-332.

24 Movsesian MA, Karimi M, Green K, Jones LR. Ca2+-transporting ATPase, phospholamban, and

calsequestrin levels in nonfailing and failing human myocardium. Circulation 1994;90:653-657.

25 Schwinger RH, Böhm M, Schmidt U, et al. Unchanged protein levels of SERCA II and phospholamban

but reduced Ca2+ uptake and Ca2+-ATPase activity of cardiac sarcoplasmic reticulum from dilated cardi-

omyopathy patients compared with patients with nonfailing hearts. Circulation 1995;92:3220-3228.

26 Tunwell RE, Wickenden C, Bertrand BM, et al. The human cardiac muscle ryanodine receptor-calcium

release channel: identification, primary structure and topological analysis. Biochem J 1996;318:477-

487.

27 Diebold RJ, Koch WJ, Ellinor PT, et al. Mutually exclusive exon splicing of the cardiac calcium channel

alpha 1 subunit gene generates developmentally regulated isoforms in the rat heart. Proc Natl Acad Sci

USA 1992;89:1497-1501.

28 Feron O, Octave JN, Christen MO, Godfraind T. Quantification of two splicing events in the L-type

calcium channel alpha-1 subunit of intestinal smooth muscle and other tissues. Eur.J.Biochem.

1994;222:195-202.

29 Gidh JM, Huang B, Jain P, Battula V, el Sherif N. Reemergence of the fetal pattern of L-type calcium

channel gene expression in non infarcted myocardium during left ventricular remodeling. Biochem

Biophys Res Commun 1995;216:892-897.

30 Ausma J, Wijffels M, Van Eys G., et al. Dedifferentiation of atrial cardiomyocytes as a result of chronic

atrial fibrillation. Am J Pathol 1997;151:985-997.

39

Alterations in gene expression of proteins

Chapter 3

Alterations in Gene Expression of Proteins Involved in the

Calcium Handling in Patients with Atrial Fibrillation

Isabelle C. Van Gelder, MD; Bianca J.J.M. Brundel, MSc; Robert H. Henning,

MD,#; Anton E. Tuinenburg, MD; Robert G. Tieleman, MD; Leo Deelman,

MSc#; Jan G. Grandjean, MD*Pieter Jan De Kam, MSc; Wiek H. Van Gilst,

PhD#; Harry J.G.M. Crijns, MD

From the Departments of Cardiology, Clinical Pharmacology (#) and Thoracic Surgery

(*), Thoraxcenter, University Hospital Groningen, Groningen, The Netherlands.

Journal of Cardiovascular Electrophysiology 10 (1999) 552 - 560

Abstract

Introduction: Atrial fibrillation (AF) leads to a loss of atrial contraction within hours

to days. During persistence of AF cellular dedifferentiation and hypertrophy occur,

eventually resulting in degenerative changes and cell death. Abnormalities in the calcium

handling in response to tachycardia induced intracellular calcium overload play a pivotal

role in these processes. Methods and Results: The purpose was to investigate the mRNA

expression of proteins and ion channels influencing the calcium handling in patients with

persistent AF. Right atrial appendages were obtained from 18 matched controls in sinus

rhythm (group 1) and 18 patients with persistent AF undergoing elective cardiac surgery.

Previous duration of AF was ≤ 6 months in 9 (group 2) and > 6 months in 9 patients (group

3). In a single semi-quantitative polymerase chain reaction the mRNA of interest and of

glyceraldehyde-3-phosphate dehydrogenase were coamplified and separated by gel-

electrophoresis. L-type calcium channel α1 subunit mRNA content was inversely related

to the duration of AF: -26% in group 2 compared to group 1 (p=0.2), and -49% in group 3

compared to group 1 (p=0.01). Inhibitory guanine nucleotide-binding protein iα2 mRNA

content was reduced in group 3 compared to group 1 (-30%, p=0.01). Sarcoplasmic

reticulum calcium ATPase, phospholamban and sodium-calcium exchanger mRNA contents

were not affected by AF. Conclusions: AF-induced alterations in mRNA contents of proteins

and ion channels

40

Chapter 3

involved in the calcium handling seem to occur in relation to the previous duration of AF.

In the present patient population these changes were significantly only if AF had lasted >

6 months.

Introduction

Atrial fibrillation (AF) is responsible for patient discomfort, thromboembolic

complications and heart failure.1-3 AF has the tendency to become permanent over time,4

which is illustrated by the fact that cardioversion to and maintenance of sinus rhythm

become increasingly difficult the longer the arrhythmia exists.5-7

AF leads to atrial contractile dysfunction, i.e. loss of atrial contraction, within hours

to days in both humans and experimental models.8-12 This process appears to be reversible,

its time course being related to the previous duration of AF.10-11,13,14 During persistence of

AF cellular dedifferentiation (resembling hibernation) and cellular hypertrophy may

occur,15,16 eventually resulting in degenerative changes like fibrosis and cell death.17-19 There

are strong indications that abnormalities in the calcium handling, in response to tachycardia

induced intracellular calcium overload,20-22 play a pivotal role in these adaptive

processes.10,12,15,23,24 The proteins and ion channels involved in the adaptation of the atria

during AF have not been clarified yet. Most likely, identification of the signaling pathways

and their target genes may lead to new therapeutic options for the treatment of AF. We

hypothesized that changes in mRNA expression of proteins and ion channels involved in

the calcium handling in the atria during AF would be comparable to those observed in the

ventricles during pacing-induced heart failure, and would be more pronounced the longer

AF had lasted. Therefore, it was our purpose to investigate the alterations in mRNA contents

of proteins and ion channels involved in the calcium handling in right atrial appendages of

patients undergoing cardiac surgery. A second aim was to determine whether these changes

were more pronounced in patients with AF of a duration > 6 months compared to those

with AF of a duration ≤ 6 months.

Methods

Patients

The day before surgery clinical characteristics of the patient were assessed by one

investigator (AET). Presence, type and duration of AF were assessed by patient’s complaints

and previous electrocardiograms. In addition, the patient was asked for medication use

and exercise tolerance (according to the New York Heart Association classification).

Echocardiography was performed within 3 months of date of surgery. Right atrial

appendages were obtained from 18 patients with persistent AF and from 18 controls in

sinus rhythm who were matched for age and sex, and as far as possible for underlying

disease and cardiac function. All patients were scheduled for elective cardiac surgery. The

41


study was approved by the Institutional Review Board, and written informed consent was

given by all patients. Immediately after excision the right atrial appendages were frozen in

liquid nitrogen and stored at -850C.

RNA Isolation and cDNA Synthesis

Total RNA was isolated from right atrial appendages using the method of acid

guanidinium thiocyanate/phenol/chloroform extraction followed by a RNeasy kit for RNA

minipreps for tissues (Qiagen, Hilden, Germany). The amount of RNA was evaluated by

absorption at 260 nm, using a GeneQuant II (Pharmacia LKB Biotechnology, The

Netherlands). The ratio of absorption (260-280 nm) of all preparations was between 1.8

and 2.0. First strand cDNA was synthesized by incubation of 1 µg of total RNA, reverse

transcription 10x buffer (Promega, The Netherlands), 200 ng of random hexamers (Promega,

The Netherlands) with 200 units of Moloney Murine Leukemia Virus Reverse Transcriptase

(Promega, The Netherlands), 1mM of each dinucleotidetriphosphate (dNTP) and 1 unit of

RNAse inhibitor (Promega, The Netherlands). The total volume was 20 µl. The synthesis

reaction lasted 10 minutes at 200C, 20 minutes at 420C, 5 minutes at 990C and 5 minutes at

40C, respectively. To assure that the amplification products did not arise from contaminating

DNA, a reverse transcriptase was performed without the addition of the enzyme. After

amplification of this product, no PCR products were detectable on an agarose gel implying

the absence of contaminating DNA in the RNA preparations.

Semi-Quantitative Polymerase Chain Reaction Analyses

Since a linear relationship between the amount of input template and amount of

amplification product exists within the exponential range of amplification, a semi-

quantitative polymerase chain reaction (PCR) was developed.25,26 To validate the results

obtained with this semi-quantitative PCR a competititive PCR method was performed. In

our experiments, the cDNA of interest and the cDNA of the ubiquitously expressed

housekeeping enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were

coamplified in a single PCR. Primers were designed for L type calcium channel α1 subunit

(L-type Ca2+), Sarcoplasmic Reticulum Calcium-ATPase 2A (SR Ca2+-ATPase),

phospholamban, sodium-calcium exchanger (Na+-Ca2+ exchanger), inhibitory guanine

nucleotide-binding protein iα2 (Giα

2), and the housekeeping gene GAPDH. The

oligonucleotides were synthesized by Eurogentec (Belgium). The sequence for each primer

is given in Table 1.

For the semi-quantitative PCR co-amplification of SR Ca2+-ATPase, phospholamban,

Na+-Ca2+ exchanger and Giα2 with GAPDH, 0.5 unit of Taq polymerase (Eurogentec,

Belgium) was added to 17.5 nM of each dNTP, 10x PCR buffer provided with Taq

polymerase, 2.5 mM MgCl2, 40 pmol of sense and antisense primer for the gene of interest,

42

Chapter 3

40 pmol of sense and antisense GAPDH primer, 1 µl of cDNA mixture and water to bring

the final volume to 50 µl. All reaction mixtures were overlaid with 50 µl of mineral oil

(Sigma). After 3 min denaturation at 940C, n cycles of amplification (Table 1) were

performed, each for 1 min at 940C , 1 min at the annealing temperature (Table 1) and 1 min

at 720C, using the thermocyclar Pharmacia LKB Gene. After the last cycle, the 720C

elongation step was extended to 5 min. The conditions for the semi-quantitative PCR

analysis of L-type Ca2+ with GAPDH were the same as above except that the MgCl2

concentration was reduced (2.0 mM). The PCR products were separated on a 1-2% agarose

gel by gel-electrophoresis and stained with ethidium bromide. The density of the PCR

products were quantified by densitometry (Aldus PhotoStyler 2.0, Graphic Workshop and

ImageQuant Version 3.3).

During the PCR for L-type Ca2+ the three isoforms of this channel were amplified.

The primers were designed to amplify the IVS1 up to IVS5 region of the α1 subunit. The

PCR fragment contained the IVS3B (adult form), the IVS3B deleted D1, and IVS3A (fetal

TABLE 1.Sequence for Primers

Protein sequence cycles annealing temp

(0C)


5'-CCC ATC ACC ATC TTC CAG GAG CG-3' - 56

5'-GGC AGG GAT GAT GTT CTG GAG AGC C-3'

L type calcium channel α1 subunit :

5’-CTG GAC AAG AAC CAG CGA CAG TGC G-3’ 32 56

5’-ATC ACG ATC AGG AGG GCC ACA TAG GG-3’

Sarcoplasmic Reticulum Calcium-ATPase 2A:

5'-TGT TCA TTC TGG ACA GAG TGG AAG G-3' 25 56

5'-TTA ATA AAG TTG GCA GAG TCC TCA AGG-3'

Phospholamban:

5'-ATG GAG AAA GTC CAA TAC CTC ACT CGC-3' 25 56

5'-TCA GAG AAG CAT CAC GAT GAT ACA GAT CAG-3'

sodium-calcium exchanger:

5’- CTA CCA AGT CCT AAG TCA GCA GC3’ 27 56

5’-GAT CCG AGG CAA GCA AGT GTA GA-3’

Inhibitory guanine nucleotide-binding protein α2:

5’-AGG GAA GAG CAC CAT CGT CAA GCA G-3’ 25 56

5’-AGC ACC AAG TCA TAG GCG CTC AAG G-3’

43


form) isoforms. These products could be identified as products of 475 bp, 442 bp and 355

bp, respectively, after digestion with DdeI (Promega, The Netherlands).

Determination of the absolute alterations of mRNA

To validate the semi-quantitative PCR the changes observed in the ratio for L-type

Ca2+ were determined. Increasing amounts of gene of interest were added to a fixed amount

of GAPDH. Therefore, known amounts of GAPDH input template (range 10-2 to 10-4 ng,

410 bp) were added to a PCR sample mixture of 0.5 unit of Taq polymerase (Eurogentec,

Belgium),17.5 nM dNTPs, 10x PCR buffer, 2.0 mM MgCl2, 40 pmol of sense and antisense

GAPDH primer, 0,7 µl of cDNA mixture and water to bring the final volume to 50 µl, and

amplified for 25 cycles. Thereafter, a fixed amount of GAPDH input template was used in

a PCR with known amounts of L-type Ca2+ input template (range 10-2.5 ng to 10-3.5 ng, 563

and 530 bp). The ratio L-type Ca2+ input versus GAPDH input was calculated.


A two-sided probability level < 0.05 was considered to indicate statistical significance.

Mean values and standard deviations or standard errors are reported for normally distributed

variables. In case of skewed distributed variables, the median values and ranges are given.

Categorical variables are presented by frequencies. For the comparison of clinical

characteristics between the persistent AF groups and the sinus rhythm group a student’s t

test was performed if variables were normally distributed. The chi-square test with continuity

correction or Fisher’s exact test were performed for categorical variables. The Wilcoxon

two-sample test was performed if variables were not normally distributed. To determine

parameters related to the mRNA levels of the investigated genes multivariate regression

analysis was performed. Only covariates with p values ≤ 0.15 in the univariate analysis

were entered in this model. The backward selection procedure was employed using the

clinical relevant first order interactions to derive a model with statistically significant

predictors. To determine differences in mRNA levels between the three groups (sinus

rhythm, AF ≤ 6 months and AF > 6 months), Tukey correction for multiple comparisons

was performed. Correlation between the increasing amounts of L-type Ca2+ input template

and the ratio L-type Ca2+/GAPDH was determined by the Spearman correlation test. The

analysis was performed by SAS statistical software (SAS, version 6.12, Cary, NC).

Results

Patients

Included were 18 patients with persistent AF and 18 patients in sinus rhythm at the

moment of surgery (Table 2). Age, sex distribution, exercise tolerance and underlying

heart disease were comparable between groups. However, more patients in sinus rhythm

44

Chapter 3

underwent coronary artery bypass grafting. Only two patients had lone AF and were

scheduled for Cox’s maze III surgery.27 Patients with severe left ventricular dysfunction

were excluded from this study.

Changes in mRNA transcription of different proteins and ion channels

Changes in transcription of the gene of interest were determined by comparison of

the gene of interest/ GAPDH ratio between patients with AF and patients with sinus rhythm

at the moment of surgery. The densities of amplified GAPDH did not differ between the

groups in the respective PCR reactions for the different genes (not shown).

The cDNA ratio of L-type Ca2+/ GAPDH was significantly lower in patients with

persistent AF, predominantly caused by the patients with AF of a duration > 6 months.

These patients had a significantly lower cDNA ratio L-type Ca2+/ GAPDH compared to

those in sinus rhythm (-49%, p<0.01). In contrast, patients with persistent AF ≤ 6 months

showed no significant reduction (-26%, p=0.20, Figure 1A). Patients with AF > 6 months

had a significantly lower content of the IVS3B form (475 bp), the IVS3B deleted D1 form

(442 bp) and the sum of the three isoforms. No significant changes were observed in the

fetal isoform IVS3A (335bp) of L-type Ca2+ (Figure 1B). Neither functional class for

exercise tolerance (Figure 1C), nor the use of calcium handling influencing drugs (Table

3), nor any other baseline characteristic, were related to the ratio cDNA L-type Ca2+/

GAPDH.

The cDNA ratios of SR Ca2+-ATPase/ GAPDH, phospholamban/ GAPDH, and Na+-

Ca2+ exchanger/ GAPDH were neither altered by the presence of persistent AF (Figure 2,

Panel A, B and C), nor by any other parameter listed in Table 2. The cDNA ratio of Giα2/

GAPDH was significantly lower in patients with persistent AF of a duration > 6 months (-

30%, p=0.01, Figure 2D). No other clinical parameters as presented in Table 2 influenced

the latter ratio.

Validation of the absolute mRNA contents

To assess the amount of L-type Ca2+ in the cDNA mixtures in different groups, a

semi-quantitative PCR was developed with increasing amounts of input cDNA of interest

to a standard amount of GAPDH cDNA (0.018 pg).The ratios were determined and plotted

in a logarithmic way. This resulted in a straight line demonstrating the validity of the

method and enabling estimations of differences of L-type Ca2+ in the cDNA mixture (Figure

3). The mean ratio cDNA for patients in sinus rhythm was 2.3, of patients with AF < 6

months duration 1.7 and of patients with AF > 6 months 1.2. (Figure 1A), which means an

amount of L-type Ca2+ of 2.1 pg, 0.85 pg and 0.41 pg, respectively (Figure 3).

45


Table 2. Baseline Characteristics of Patients with Persistent Atrial Fibrillation and Their Matched

Controls in Sinus Rhythm (at Surgery)

persistent atrial sinus rhythm

fibrillation

all≤ 6 months > 6 months p value

Duration of atrial fibrillation

(median months, range) 3 (0.1-6) 24 (12-240) -

Age (mean “ SD) 65 ± 10 60 ± 9 69 ± 9 63 ± 11 n.s.

Male/ female (n) 12 / 6 2 / 7 5 / 4 14 / 4 n.s.

Underlying heart disease (n)*

Coronary artery disease 4 0 4 10 0.10

previous myocardial infarction 4 2 2 5 n.s.

Mitral valve disease 4 4 0 3 n.s.

Mitral stenosis 0 0 0

Mitral regurgitation 4 4 0 2

Aortic valve disease 6 3 3 7 n.s.

Aortic stenosis 5 2 3 4

Aortic regurgitation 4 1 3 3

Hypertension 3 0 3 2 n.s.

“Lone” atrial fibrillation 2 1 1 - n.s.

New York Heart Association Class for Exercise tolerance (n) 0.23

Class I 5 2 3 5

Class II 3 1 2 8

Class III 10 6 4 5

Drugs at surgery (n)

Digitalis 14 7 7 1 0.01

Calcium entry blockers 5 1 4 7 n.s.

Beta-blockers 4 1 3 8 n.s.

Angiotensin converting enzyme inhibitors 8 5 3 6 n.s.

Acenocoumarol 8 4 4 0 0.01

Left atrial long axis view (mean±SD, mm) 51±11 50±14 51±5 47±7 n.s.

Left ventricular end diastolic diameter

(mean±SD, mm 53±10 53±9 53±11 49±6 n.s.

Left ventricular end systolic diameter

(mean±SD, mm) 36±14 37±3 35±10 35±3 n.s.

n.s.= not statistically significant. p value indicates comparison between all AF and sinus rhythm patients.

* > 1 underlying disease per patient scored.

Discussion

This study demonstrates selective changes in mRNA contents of proteins and ion

channels involved in the calcium handling in patients suffering from persistent AF. L-type

Ca2+ and Giα2 mRNA contents were reduced in patients with longstanding persistent AF,

whereas SR Ca2+-ATPase, phospholamban and Na+-Ca2+ mRNA contents were not affected

by AF. No significant alterations in the mRNA expression of proteins and ion channels

were observed in patients with AF of a duration ≤ 6 months.

46

Chapter 3

AF AF AF AF

SR SR SR SR

IVS3B 475 bp

IVS3B 442 bp

IVS3A 335 bp

GAPDH

Figure 1. A, The cDNA ratio L-type Ca2+ (CaLtype)/ GAPDH was significantly reduced (-49%, p=0.01) in

patients with AF of a duration > 6 months compared to the controls in sinus rhythm (SR). Mean values ± SEM.

B, Typical example of an agarose gel showing the results of a semi-quantitative PCR of L-type Ca2+ and GAPDH

of 4 patients with persistent AF at the moment of surgery and 4 controls in sinus rhythm. In the AF group there

was a lower expression the IVS3B form (475 bp), the IVS3B deleted D1 form (442 bp) and the sum of the three

isoforms of the L-type Ca2+. C, No influence of the functional class for exercise tolerance (NYHA) was demon-

strated on the ratio cDNA ratio L-type Ca2+/ GAPDH (mean ± SEM). All data are represented in density units/

density units.

A

B

C

47


Table 3.Influence of different drugs influencing the calcium handling on the cDNA ratio L-type Ca2+/

GAPDH (median value (range))

persistent atrial fibrillation sinus rhythm

≤ 6 months > 6 months

Digitalis yes (n=7) 2.2 (0.5-3.1) yes (n=7) 1.4 (0.2-3.2) yes (n=1) 4.1

no (n=2) 0.7 (0.6-0.8) no (n=2) 0.6 (0.3-0.9) no (n=17) 2.3 (0.5-3.3)

Calcium entry blockers yes (n=1) 3.1 yes (n=4) 1.2 (0.3-2.8) yes (n=7) 2.7 (2.4-3.1)

no (n=8) 1.6 (0.5-2.9) no (n=4) 1.3 (0.2-3.2) no (n=11) 2.2 (0.5-4.1)

Beta-blockers yes (n=1) 0.8 yes (n=3) 0.5 (0.2-0.9) yes (n=8) 2.7 (0.5-4.1)

no (n=8) 1.8 (0.5-3.1) no (n=6) 1.6 (0.3-3.2) no (n=10) 2.2 (1.3-3.3)

ACE inibitors yes (n=5) 1.8 (0.8-2.9) yes (n=3) 1.5 (0.9-2.8) yes (n=6) 2.2 (1.3-3.1)

no (n=4) 1.7 (0.5-3.1) no (n=6) 1.1 (0.2-3.2) no (n=12) 2.5 (0.5-4.1)

Figure 2. No differences in cDNA ratios SR Ca2+-ATPase/ GAPDH (panel A), phospholamban (PLB)/GAPDH

(Panel B), and Na+-Ca2+ exchanger/ GAPDH (Panel C) between patients with and without AF were observed.

The cDNA ratio Giα2/ GAPDH was reduced in patients with AF of a duration > 6 months compared to the

controls in sinus rhythm (- 30%, p=0.01, Panel D). All data are represented in density units/ density units and

mean ± SEM.

48

Chapter 3

Figure 3. Plot showing a significant correlation between the increasing amounts of L-type Ca2+ input template

and the ratio L-type Ca2+/GAPDH. The ratios are expressed in a logarithmic way. Each value represents the mean

of five measurements.

Changes in Gene Expression during AF

Several studies suggested that abnormalities in the calcium handling play a pivotal

role in the adaptive processes which occur during persistent AF. Experiments of pacing

induced AF,10,28,29 or rapid atrial pacing30 revealed that calcium lowering by verapamil

reduced both electrical and contractile dysfunction, while calcium loading or the Ca2+

agonist Bay 8644 had an opposite effect.10,28 Furthermore, depletion of contractile material

was demonstrated in atria of goats with AF induced for 9 to 23 weeks.15 Finally, calcium

lowering drugs, when administered in humans with persistent AF, reduced the recurrence

rate of AF after restoration of sinus rhythm.31 Only a few data are available on alterations

in gene expression or in ionic currents occurring in atrial tissue subjected to rapid atrial

rates due to AF or atrial pacing. Yue et al. demonstrated that 6 weeks of rapid atrial pacing

in dogs led to significant changes in ionic currents.23 The induction of atrial electrical

remodeling,32, i.e. a reduced atrial refractoriness together with loss of the physiological

rate adaptation to heart rate, was accompanied by a time-dependent decrease of L-type

Ca2+ current,23 transient outward current,23 and sodium current densities.33 Preliminary data

49


revealed that downregulation of the mRNA expression of the these ion channels was

responsible for the reduction of latter currents.34 The lower L-type Ca2+ current density has

been confirmed in humans with persistent AF24 and dilated atria.35,36 Protein expression of

the potassium channel Kv1.5, together with transient outward current and a sustained

outward potassium current densities were reduced in atrial myocardium of patients with

persistent AF.37 In contrast to the data of Yue et al.,23 the present study revealed significant

reductions only if AF had lasted > 6 months. This discrepancy between our data of patients

selected for cardiac surgery and data of the experimental canine model may be related to

(a) the difference in species (human versus dog); and (b) the presence of co-morbidity and

drug treatment in our patients versus the “clean” canine model.

Differences and Similarities between Changes in Atrial Gene Expression during AF

and Ventricular Gene Expression during Heart Failure.

Rapid ventricular pacing induces ventricular failure, and is frequently used as a model

to investigate alterations in gene expression during congestive heart failure in humans.38

Both AF and rapid atrial pacing induce atrial contractile dysfunction.10-14 We hypothesized

that during AF alterations in gene expression in atrial myocardium would be similar to

those demonstrated in ventricular myocardium during (pacing-induced) heart failure.39-44

However, in contrast to what was expected, the present study revealed important

differences between fibrillating atrial and failing ventricular myocardium. In the present

study, the expected reduction of mRNA contents of SR Ca2+-ATPase42,45-50 and

phospholamban,47-49 and upregulation of Na+-Ca2+ exchanger mRNA content46,51,52 could

not be demonstrated. Furthermore, in contrast to the increased level of Giα2 in failing

ventricular myocardium,40,48 a reduced mRNA content of Giα2 was observed in the right

atrial appendage. However, in agreement with most studies investigating the expression

or density of the L-type Ca2+ at the ventricular level, the present study showed a reduced

mRNA content of L-type Ca2+.41-45,52,53 Thus, the adaptation processes in the atria during

AF, as far as transcription is concerned, seem to be different from those occurring at the

ventricular level during heart failure. Furthermore, in contrast to the observed diffusealterations in the failing ventricular myocardium, changes in mRNA expression of proteins

and ion channels involved in the calcium handling occurred selectively in atrial myocardium

during AF. The mechanism behind these differences in genetic reprogramming, which

have to be confirmed in an animal model investigating gene expression in both the atrium

and ventricle, are unknown. However, (1) differences in distribution, number and type of

ion channels and proteins between atrial and ventricular myocardium,54 and (2) differences

in neurohumoral activation during AF and (pacing-induced) heart failure may have

influenced the latter.

50

Chapter 3


The present study has several important limitations. First, only mRNA expression

was determined which neither reflects the protein expression nor its functional activity.

Furthermore, not a quantitative but a semi-quantitative PCR was used. Finally, drugs,

differences in hemodynamic function and underlying diseases may influence gene

expression of proteins and ion channels in humans. To minimize the influence of the above

mentioned clinical parameters on gene expression, no patients with severe left ventricular

dysfunction were included, the patients were matched, and we used multivariate analysis.

Using this statistical analysis, parameters were identified which independently influenced

alterations in the mRNA expression of the investigated proteins and ion channels. This

analysis revealed that in the present population the mRNA expressions were neither affected

by the underlying disease, nor by left atrial or left ventricular dimensions, nor by the

degree of heart failure, nor by any (calcium handling affecting) drug.

References

1. Kannel WB, Abbott RD, Savage DD, et al: Epidemiologic features of chronic atrial fibrillation. N Engl J

Med 1982;306:1018-1022.

2. Önundarson PT, Thorgeirsson G, Jonmundsson E, et al: Chronic atrial fibrillation-epidemiologic features

and 14 years follow-up: a case control study. Eur Heart J 1987;8:521-527.

3. Krahn AD, Manfreda J, Tate RB, et al: The natural history of atrial fibrillation: incidence, risk factors, and

prognosis in the Manitoba follow-up study. Am J Med 1995;98:476-484.

4. Godtfredsen J: Atrial fibrillation. Etiology, course and prognosis. A follow-up study of 1212 cases. Copen-

hagen: University of Copenhagen, 1975. Thesis

5. Bjerkelund C, Orning OM: An evaluation of DC shock treatment of atrial arrhythmias. Acta Med Scand

1968;184:481-491.

6. Van Gelder IC, Crijns HJGM, Tieleman RG, et al: Value and limitation of electrical cardioversion in


Arch Int Med 1996;156:2585-2592.

7. Sopher SM, Camm AJ: Atrial fibrillation - maintenance of sinus rhythm versus rate control. Am J Cardiol

1996;77:24A-38A.

8. Graettinger JS, Carleton RA, Muenster JJ: Circulatory consequences of changes in cardiac rhythm pro-

duced in patients by transthoracic direct-current shock. J Clin Invest 1964;43:2290-2302.

9. Braunwald E: Symposium on cardiac arrhythmias: introduction with comments on the hemodynamic sig-

nificance of atrial systole. Am J Med 1964;37:665-669.

10. Leistad E, Aksnes G, Verburg E, et al: Atrial contractile dysfunction after short-term atrial fibrillation is

reduced by verapamil but increased by BAY K8644. Circulation 1996;93:1747-1754.

11. Manning WJ, Silverman DI, Katz SE, et al: Impaired left atrial mechanical function after cardioversion:

relation to duration of atrial fibrillation. J Am Coll Cardiol 1994;23:1535-1540.

12. Sun H, Gaspo R, Leblanc N, et al: Cellular mechanism of atrial contractile dysfunction caused by sus-

tained atrial tachycardia. Circulation 1998;98:719-727.

13. Shapiro EP, Effron MB, Lima S, et al: Transient atrial dysfunction after conversion of chronic atrial fibril-

lation to sinus rhythm. Am J Cardiol 1988;62:1202-1207.

14. Gallagher MM, Obel OA, Camm AJ: Tachycardia-induced atrial myopathy: an important mechanism in

the pathology of atrial fibrillation? J Cardiovasc Electrophysiol 1997;1065-1074.

15. Ausma J, Wijffels M, ThonJ F, et al: Structural changes of atrial myocardium due to sustained atrial

fibrillation in the goat. Circulation 1997;96:3157-3163.

16. Morillo CA, Klein GJ, Jones DL, et al: Chronic rapid atrial pacing: structural, functional and

electrophysiologic characteristics of a new model of sustained atrial fibrillation. Circulation 1995;91:1588-

1595.

51


17. Bailey GW, Braniff BA, Hancock W, et al: Relation of left atrial pathology to atrial fibrillation in mitral

valvular disease. Ann Int Med 1968;69:13-20.

18. Davies MJ, Pomerance A: Pathology of atrial fibrillation in man. Br Heart J 1972;34:520-525.

19. Mary-Rabine L, Pham TD, Hordof A, et al: The relationship of human atrial cellular electrophysiology to

clinical function and ultrastructure. Circ Res 1983;52:188-199.

20. Lee H-C, Clusin WT: Cytosolic calcium staircase in cultured myocardial cells. Circ Res 1987;61:934-939.

21. Schouten VJA, Morad M: Regulation of Ca2+ current in frog ventricular myocytes by the holding poten-

tial, c-AMP and frequency. Pflugers Arch 1989;415:1-11.

22. De Pauw M, Borgers M, Heyndrickx GR: Ultrastructural calcium distribution in cardiac myocytes after

48h of rapid pacing in dogs. Circulation 1996;94:I-604. (Abstract).

23. Yue L, Feng J, Gaspo R, et al: Ionic remodeling underlying action potential changes in a canine model of

atrial fibrillation. Circulation 1997;81:512-525.

24. Van Wagoner DR, Lamorgese M, Kirian P, et al: Calcium current density is reduced in atrial myocytes

isolated from patients in chronic atrial fibrillation. Circulation 1997;96:I-180. (Abstract).

25. Chelly J, Montarras D, Pinset C, et al: Quantitative estimation of minor mRNAs by cDNA polymerase

chain reaction - application to dystrophin mRNA in cultured myogenic and brain cells. Eur J Biochem

1990;187: 691-698.

26. Noonan KE, Beck C, Holzmayer TA, et al: Quantitative analysis of MDR1 (multi drug resistance) gene

expression in human tumors by polymerase chain reaction. Proc Natl Acad Sci USA 1990;87:7160-7164.

27. Cox JL, Jaquiss RD, Schuessler RB, et al: Modification of the maze procedure for atrial flutter and atrial

fibrillation. II. Surgical technique of the maze III procedure. J Thorac Cardiovasc Surg 1995;110:485-495.

28. Goette A, Honeycutt C, Langberg JJ: Electrical remodeling in atrial fibrillation. Time course and mecha-

nisms. Circulation 1996;94:2968-74.

29. Daoud EG, Knight BP, Weiss R, et al: Effect of verapamil and procainamide on atrial fibrillation - induced

electrical remodeling in humans. Circulation 1997;96:1542-1550.

30. Tieleman RG, De Langen CDJ, Van Gelder IC, et al: Verapamil reduces tachycardia induced electrical

remodeling of the atria. Circulation 1997;95:1945-53.

31. Tieleman RG, Van Gelder IC, Crijns HJGM, et al: Early recurrences of atrial fibrillation after electrical

cardioversion: a result of fibrillation-induced electrical remodeling of the atria? J Am Coll Cardiol

1998;31:167-173.

32. Wijffels MCEF, Kirchof CJHJ, Dorland R, et al: Atrial fibrillation begets atrial fibrillation. A study in

awake chronically instrumented goats. Circulation 1995;92:1954-1968.

33. Gaspo R, Bosch RF, Bou-Abboud E, et al: Tachycardia-induced changes in Na+ current in a chronic dog

model of atrial fibrillation. Circ Res 1997;81:1045-1052.

34. Yue L, Wang Z, Gaspo R, et al: The molecular mechanism of ionic remodeling of repolarization in a dog

model of atrial fibrillation. Circulation 1997;96:I-180. (Abstract).

35. Le Grand B, Hatem S, Deroubaix E, et al: Depressed transient outward and calcium currents in dilated

human atria. Cardiovasc Res 1994;28:548-556.

36. Ouadid H, Albat B, Nargeot J: Calcium currents in diseased human cardiac cells. J Cardiovasc Pharmacol

1995;25:282-91.

37. Van Wagoner DR, Pond AL, McCarthy PM, et al: Outward K+ current densities and Kv1.5 expression are

reduced in chronic human atrial fibrillation. Circ Res 1997;80:772-781.

38. Shinbane JS, Wood MA, Jensen N, et al: Tachycardia-induced cardiomyopathy: a review of animal mod-

els and clinical studies. J Am Coll Cardiol 1997;29:709-715.

39. Spinale FG, Tanaka R, Crawford FA, et al: Changes in myocardial blood flow during development of and

recovery from tachycardia-induced cardiomyopathy. Circulation 1992;85:717-729.

40. Vatner DE, Sato N, Galper JB, et al: Physiological and biochemical evidence for coordinate increases in

muscarinic receptors and Gi during pacing-induced heart failure. Circulation 1996;94:102-107.

41. Colston JT, Kumar P, Chambers JP, et al: Altered sarcolemmal calcium channel density and calcium-pump

ATPase activity in tachycardia heart failure. Cell Calcium 1994;16:349-356.

42. Cory CR, Shen H, O’Brien PJ: Compensatory asymmetry in down-regulation and inhibition of the myo-

cardial Ca2+ cycle in congestive heart failure produced in dogs by idiopathic dilated cardiomyopathy and

rapid ventricular pacing. J Mol Cell Cardiol 1994;26:173-184.

43. Hasenfuss G: Alterations of calcium-regulatory proteins in heart failure. Cardiovasc Res 1998;37:279-

289.

44. Mukherjee R, Hewett KW, Walker JD, et al: Changes in L-type calcium channel abundance and function

during transition to pacing-induced congestive heart failure. Cardiovasc Res 1998;37:432-444.

45. Takahashi T, Allen PD, Lacro RV, et al: Expression of dihydropyridine receptor (Ca2+ channel) and

calsequestrin genes in the myocardium of patients with end-stage heart failure. J Clin Invest 1992;90:927-

52

Chapter 3

935.

46. Studer R, Reinecke H, Bilger J, et al: Gene expression of the Na+-Ca2+ exchanger in end-stage human heart

failure. Circ Res 1994;75:443-453.

47. Schwinger RHG, Böhm M, Schmidt U, et al: Unchanged protein levels of SERCA II and phospholamban

but reduced Ca2+ uptake and Ca2+-ATPase activity of sarcoplasmic reticulum from dilated cardiomyopathy

patients compared with patients with nonfailing hearts. Circulation 1995;92:3220-3228.

48. Flesch M, Schwinger RHG, Schnabel P, et al: Sarcoplasmic reticulum calcium ATPase and phospholamban

mRNA and protein levels in end-stage heart failure due to ischemic or dilated cardiomyopathy. J Mol Med

1996;74:321-332.

49. Linck B, Boknik P, Eschenhagen T, et al: Messenger RNA expression and immunological quantification of

phospholamban and SR-Ca2+-ATPase in failing and nonfailing human hearts. Cardiovasc Res 1996;31:625-

632.

50. Movsesian MA, Schwinger RHG: Calcium sequestration by the sarcoplasmic reticulum in heart failure.

Cardiovasc Res 1998;37:352-259.

51. Flesch M, Schwinger RHG, Schiffer F, et al: Evidence for functional relevance of and enhanced expres-

sion of the Na+-Ca2+ exchanger in failing human myocardium. Circulation 1996;94:992-1002.

52. De Tombe PP: Altered contractile function in heart failure. Cardiovasc Res 1998;37:367-380.

53. Dixon IMC, Lee SL, Dhalla NS: Nitrendipine binding in congestive heart failure due to myocardial infarc-

tion. Circ Res 1990;66:782-788.

54. Schwinger RHG, Böhm M, Pieske B, et al: Different $-adrenoceptor-effector coupling in human ventricu-

lar and atrial myocardium. Eur J Clin Invest 1991;21:443-451.

55. Romanin C, Karlsson JO, Schindler H: Activity of L type Ca2+ channels is sensitive to cytoplasmic

calcium. Pflugers Arch 1992;421:516-518.

Part II

Gene expression of ion-channels

55

Alterations in potassium channel gene expression in atria

Chapter 4

Alterations in Potassium Channel Gene Expression in

Atria of Patients with Persistent and Paroxysmal Atrial

Fibrillation Differential regulation of protein and mRNA

levels for K+-channels

Bianca J. J. M. Brundel, MSc*,†; Isabelle C. Van Gelder, MD*; Robert H.

Henning, MD†; Anton E. Tuinenburg, MD*; Mirian Wietses†; Jan G. Grandjean,

MD‡; Arthur A. M. Wilde, MD§, Wiek H. Van Gilst, PhD†; Harry J. G. M.

Crijns, MD*.

Departments of Cardiology*, Clinical Pharmacology†, and Thoracic Surgery,

Thoraxcenter University Hospital Groningen and Department of Cardiology

University of Amsterdam and Utrecht§, The Netherlands.

Journal of the American College of Cardiology: Provisionally Accepted

AbstractObjectives: Our propose was to determine whether patients with persistent and

paroxysmal atrial fibrillation (AF) show alterations in potassium channel expression.

Background: Persistent AF is associated with a sustained shortening of the atrial action

potential duration and atrial refractory period. Underlying molecular changes have not

been studied in humans. We investigated whether a changed gene expression of specific

potassium channels is associated with these changes in patients with persistent and

paroxysmal AF.

Methods: Right atrial appendages were obtained from 8 patients with paroxysmal AF, 10

with persistent AF and 18 matched controls in sinus rhythm. All controls underwent CABG

whereas most AF patients underwent Cox’s MAZE surgery (n=12). All patients had normal

left ventricular function. mRNA levels were measured by semi-quantitative polymerase

chain reaction and protein content by Western blotting.

Results: mRNA levels of transient outward channel (Kv4.3), acetylcholine dependent

potassium channel (Kir3.4) and ATP dependent potassium channel (Kir6.2) were reduced

in patients with persistent AF (-35%, -47%, -36%, respectively, p<0.05), whereas only

Kv4.3 mRNA level was reduced in patients with paroxysmal AF (-29%, p=0.03). No

56

Chapter 4

changes were found for Kv1.5 and HERG mRNA levels in both groups. Protein levels of

Kv4.3, Kv1.5 and Kir3.1 were reduced both in patients with persistent (-39%, -84%, -

47%, respectively, p<0.05) and paroxysmal AF (-57%, -64%, -40%, respectively, p<0.05).

Conclusions: Persistent AF is accompanied by reductions in mRNA and protein levels of

several potassium channels. In patients with paroxysmal AF these reductions were observed

predominantly at the protein level and not at the mRNA level suggesting a post-

transcriptional regulation.

Abbreviations

AF atrial fibrillation

SR sinus rhythm

GAPDH glyceraldehyde-3-phosphate dehydrogenase

HERG gene encoding rapid component of the delayed recitifier IKr

Kir3.1 gene encoding part of the IKACH

, together with IKir3.4

Kir3.4 gene encoding part of the IKACh

, together with IKir3.1

Kir6.2 gene encoding part of the IKATP

Kv1.5 gene encoding ultra rapid component of the delayed rectifier IKur

Kv4.3 gene underlying calcium independent transient outward current ITo1

Introduction

Atrial fibrillation (AF) is a common cardiac arrhythmia affecting millions of people

worldwide (1). AF has the tendency to become more persistent and increasingly difficult

to treat over time. During recent years, experimental studies showed that shortening of the

atrial effective refractory period was one important factor contributing to the maintenance

of AF (2,3). This shortening has been confirmed in patients suffering from AF and atrial

flutter (4,5). Experimental and human data revealed that AF or tachycardia induced

shortening of atrial effective refractory period and action potential duration were associated

with a reduction of ICaL

, ITo1

and INa

currents due to reduced mRNA expression of these

channels (6-9). Previously we have demonstrated that mRNA and protein expression of

the L-type calcium channel in both patients with persistent AF with more severe underlying

heart disease (10) and the present patient population (11) were significantly reduced. No

alterations, however, were observed in either patients with paroxysmal or short term

persistent AF.

Theoretically, action potential duration and atrial effective refractory period can be

shortened by 1) an increase in K+ channel gene expression and activity, or 2) a decrease in

L-type Ca2+ channel (L-type Ca2+) gene expression and activity, or 3) a combination of

both. The present study was undertaken to evaluate the impact of persistent and paroxysmal

AF on gene expression of potassium channels in human right atrial appendages. Therefore,

57


the mRNA and protein expression of Kv4.3 (gene underlying the calcium independent

transient outward current ITo1

) (12), HERG (gene encoding the rapid component of the

delayed rectifier) (13), Kv1.5 (gene encoding the ultra rapid delayed rectifier, IKur

) (14,15),

Kir3.1/Kir3.4 (heterotetrameric complex of these two genes forms the acetylcholine

dependent K+ current, IKACh

) (16) and Kir6.2 (gene encoding the inward rectifier K+ current,

forming IKATP

with the sulfonylurea receptor) (17) were examined in patients with persistent

and paroxysmal AF undergoing cardiac surgery. Patients with lone AF or patients with AF

scheduled for coronary artery bypass grafting were matched with patients in sinus rhythm

without history of AF and undergoing coronary artery bypass grafting.

Materials and Methods

Patient selection and atrial tissue collecting


of the patient. Patient’s history and previouw electrocardiograms were used to establish

type and duration of AF. In addition, medication use and exercise tolerance (according to

the NYHA classification) was determined. Echocardiography data were obtained within 3

months prior to surgery. Right atrial appendages were obtained from 10 patients with

persistent AF and from 8 patients with paroxysmal AF. All patients were euthyroid. The

AF patients were matched for age, sex and degree of heart failure with 18 clinically stable

patients in sinus rhythm undergoing CABG. The Institutional Review Board approved the

study and all patients gave written informed consent. Immediately after excision, the right

atrial appendages were snap-frozen in liquid nitrogen and stored at -85 °C.


Total RNA was isolated and processed as described previously (11). Briefly, first

strand cDNA was synthesized by incubation of 1 µg of total RNA in reverse transcription

10x buffer, 200 ng of random hexamers with 200 units of Moloney Murine Leukemia

Virus Reverse Transcriptase, 1mM of each dNTP and 1 unit of RNase inhibitor (Promega,

The Netherlands) in 20 µl. Synthesis reaction was performed for 10 minutes at 20 °C, 20

minutes at 42 °C, 5 minutes at 99 °C and 5 minutes at 4 °C. All the products were checked

for contaminating DNA.


We described and validated these methods before (11). In short, the cDNA of interest

and the cDNA of the ubiquitously expressed housekeeping gene glyceraldehyde-3-

phosphate dehydrogenase (GAPDH) were coamplified in a single PCR. Primers

(Eurogentec, Belgium) were designed for Kv4.3, HERG, Kv1.5, Kir3.4, Kir6.2 and the

housekeeping gene GAPDH (Table 1).

58

Chapter 4

The PCR products were separated on agarose gel by electrophoresis and stained with

ethidium bromide. The density of the PCR products was quantified by densitometry.

Linearity for the PCR was established by making a correlation between the number of

cycles and the density of gene of interest and GAPDH (data not shown).

Protein Preparation and Western Blotting

Frozen right atrial appendages of 5 patients in sinus rhythm, 5 patients with paroxysmal

AF and 5 patients with persistent AF were homogenized in RIPA buffer as previously

described (11). Protein concentration was determined according to the Bradford method

(Sigma, The Netherlands) with bovine albumin as a standard. Protein expression was

determined by Western-blot analysis and expressed as the ratio to levels of GAPDH.

Therefore, denatured protein (10 µg) was separated by SDS-PAGE, transferred to

nitrocellulose membranes (Stratagene, The Netherlands) and incubated with primary

antibodies against GAPDH (Affinity Reagents, USA), anti Kir3.1, anti Kv4.3 and anti

Kv1.5 (Alomone Labs, Israel). Anti-mouse IgG (Santa Cruz Biotechnology, The

Netherlands) was used as secondary antibody. Signals were detected by the ECL-detection

method (Amersham, The Netherlands) and quantified by densitometry. The specificity of

the band was tested by pre-incubation of the antibody with the antigen. The band densities

were evaluated by densitometric scanning using a Snap Scan 600 (Agfa, The Netherlands).

There was a linear relation between protein amounts on the membrane and immunoreactive

signals of Kir3.1, Kv4.3, Kv1.5 and GAPDH (data not shown).

Table 1. The sequence for the primers.

protein sequence

cycles annealing

temp (°C)

GAPDH: F 5'-CCC ATC ACC ATC TTC CAG GAG CG-3', var. var.


To1: F 5’-CAG CAA GTT CAC AAG CAT CC-3’ 31 52

Kv4.3 R 5’-AGC TGG CAG GTT AGA ATT GG-3’

Kr: F 5’-GTC AAT GCC AAC GAG GAG GT-3’, 27 58

HERG R 5’-CTG GTG GAA GCG GAT GAA CT-3’

Kur: F 5’-AAC GAG TCC CAG CGC CAG GT-3’ 32 64

Kv1.5 R 5’-AGG CGG ATG ACT CGG AGG AT-3’

KACh F 5’-CAC CCT GGT GGA CCT CAA GTG GCG C-3’ 30 56

Kir3.4 R 5’-AGC TCC GGG CTT GGC AGG TCA TGC-3’

KATP F 5’-CAC GCT GGT GGA CCT CAA GT-3’ 29 59

Kir6.2 R 5’-CAC GAT GAG GCT CAG GAT GG-3’

59


Definitions

Persistent AF: continuous presence of AF until the moment of cardiac surgery, i.e. at least

two consecutive electrocardiograms of AF more than 1 week apart, without intercurrent

sinus rhythm. Persistent AF has a non-spontaneously converting character. Previously,

this type of AF was classified as chronic AF (18).

Paroxysmal AF: AF typically occurring in episodes with a duration shorther than 24

hours (but longer lasting paroxysms are not unusual) with intermittent sinus rhythm.

Paroxysmal AF is either converting spontaneously or terminated with intravenously

administered antiarrhythmic drug. It is non-controlled whether paroxysmal AF is present

at the moment of cardiac surgery (18).


All PCR and SDS-PAGE procedures were performed in duplicate series and mean

values were used for statistical analysis. For determination of correlation’s the Spearman

correlation test was used. One-way ANOVA was used for all group to group comparisons.

All p-values are two-sided, a p-value <0.05 was considered statistically significant. SAS

version 6.12 (Cary, NC) was used for all statistical evaluations.

Results

Patients

Included were 10 patients with persistent and 8 patients with paroxysmal AF. These

two groups were compared with two groups of controls in sinus rhythm, which were matched

for sex, age and left ventricular function (Table 2). Six of the 8 patients with paroxysmal

AF suffered from intractable paroxysmal AF without any underlying heart disease and

were scheduled for Cox’s MAZE surgery. The median duration of sinus rhythm before

surgery was 1.5 days. The median frequency of paroxysms was once a day with a median

duration of 3 hours. Three patients with paroxysmal AF were in AF at the moment of

surgery and harvesting of the right atrial appendage. Control right atrial appendages were

obtained from clinically stable patients in sinus rhythm who were scheduled for coronary

artery bypass surgery. Although the AF groups and their controls in sinus rhythm differed

with respect to the underlying heart disease, all had a normal left ventricular function and

were in the functional class I or II for exercise tolerance. Also, atrial and left ventricular

dimensions were similar among groups (data not shown).

Alterations in mRNA Levels in Persistent and Paroxysmal AF


gene-of-interest/GAPDH ratios between patients with persistent AF or paroxysmal AF

60

Chapter 4

and their matched controls in sinus rhythm. No differences in GAPDH densities were

found between the groups (data not shown).

Patients with persistent AF showed significant reductions of mRNA contents for Kv4.3

(-35%, p=0.02), Kir3.4 (-47%, p=0.0003) and Kir6.2 (-36%, p=0.03) (Table 3). Patients

with paroxysmal AF showed only reduction of the Kv4.3 mRNA level (-29%, p=0.03,

Table 3). No differences in mRNA contents for Kv1.5 and HERG were found between

patients with persistent and paroxysmal AF compared to patients in sinus rhythm (Table

3). Although the group samples are small, the mRNA levels of Kv4.3, Kv1.5 and Kir3.4 in

both patients with persistent and paroxysmal AF seemed not to be influenced by any drug

(data not shown).

Table 2. Baseline characteristics of patients with paroxysmal AF, persistent AF and matched control patients

in sinus rhythm of both groups at the moment of surgery.

PAF SR (PAF) CAF SR (CAF)

8 8 10 10

M/F (n) 6/2 6/2 6/4 6/4

Age 51 ±7 56 ± 11 63 ±11 65 ± 17

Duration of AF: median, range (months) 18 (8 - 64)

Duration SR before surgery: median, range (days) 1.5 (0-30)


coronary artery disease 2* 8 3* 10

hypertension 1 1 3 2

lone AF 6* 0 5* 0

Surgical procedure

CABG 2* 8 4* 10

MAZE 6* 0 6* 0

New York Heart Association for exercise tolerance

Class I 7 5 6 5

Class II 1 3 4 5

Medication

Beta blockers 1* 5 3 6

Calcium antagonist 0 3 3 3

Digitalis 0 1 5 3

ACE inhibitor 0 1 4 2

* p-value < 0.05 compared to the control group

Values are presented as mean value ± SD or number of patients. ACE, inhibitor indicates Angiotensin

Converting Enzyme; AF, atrial fibrillation; CABG, Coronary Artery Bypass Grafting; CAF, chronic persistent

atrial fibrillation; M/F, male/female; PAF, paroxysmal atrial fibrillation; SR (CAF), matched controls in

sinus rhythm of patients with persistent AF; SR (PAF), matched controls in sinus rhythm of patients with

paroxysmal AF

61


Figure 1. The top of each panel shows a typical Western blot analysis of 10 µg of protein homogenates of 3

patients in sinus rhythm (SR), 3 patients with chronic persistent AF (CAF) and 3 patients with paroxysmal AF

(PAF). The immunoblot swere done for anti-Kv4.3 (A), anti-Kv1.5 (B), and anti-Kir3.1 (C) with GAPDH (37

kD) as an internal control.

All data are presented as density units/ density units. Values are mean ± SEM.

Alterations in Proteins Levels in Persistent and Paroxysmal AF

From the total patient group there were 5 patients with persistent AF, 5 with paroxysmal

AF and 5 patients in sinus rhythm with enough right atrial appendage tissue to isolate

proteins. Changes in protein expression were studied for Kv4.3, Kv1.5 and Kir3.1 in rela-

tion to protein levels of GAPDH. The protein expression of Kv1.5/GAPDH and Kir3.1/

62

Chapter 4

Table 3. Comparison of mRNA and protein expression for patients with persistent and paroxysmal AF with

their matched controls in sinus rhythm.

mRNA expression Correlation

mRNA and Protein

SR1 CAF SR2 PAF r

Kv4.3 1.12 ± 0.1 0.73 ± 0.09* 1.08 ± 0.11 0.77 ± 0.1* 0.75**

Kv1.5 1.35 ± 0.12 1.28 ± 0.11 1.37 ±0.14 1.56 ± 0.14 0.31

HERG 0.43 ± 0.04 0.42 ± 0.03 0.52 ± 0.04 0.62 ± 0.03 n.a.

Kir3.4/Kir3.1 1.9 ± 0.1 1.03 ± 0.1** 1.92 ± 0.11 1.59 ± 0.13 0.74**

Kir6.2 1.04 ± 0.07 0.67 ± 0.07* 1.1 ± 0.07 0.96 ± 0.07 n.a.

*, p<0.05, **, p<0.005

Figure 2. Correlation between the mean protein expression of Kv4.3 ( , r=0.67, p>0.05), Kv1.5 ( , r=0.97,

p=0.02) and Kir3.1 ( , r=0.56, p>0.05) in patients with paroxysmal AF and the duration of sinus rhythm before

surgery.

GAPDH was markedly reduced in patients with persistent AF compared to patients in

sinus rhythm (-84%, p=0.001 and -47%, p=0.002, respectively) and also in patients with

paroxysmal AF (-64%, p=0.005 and -40%, p=0.007, respectively, Figure 1B and C). Simi-

lar results were obtained for Kv4.3/GAPDH protein content, i.e. both a reduction in pa-

tients with persistent AF (-39%, p=0.04) and paroxysmal AF (-57%, p=0.001, Figure 1A).

A positive correlation between mRNA levels and protein levels of Kv4.3 and Kir3.1 for

patients with paroxysmal AF, persistent AF and sinus rhythm could be demonstrated, but

not for Kv1.5 (Table 3). Although the group samples are small the protein ratio of Kv4.3,

Kv1.5 and Kir3.1 seemed not to be influenced by any drug (data not shown).

63


Importantly, in patients with paroxysmal AF the mean protein expression of Kv1.5

appeared to be related to the duration of sinus rhythm after the last episode of AF. Patients

in AF at the moment of surgery showed the lowest protein expression, comparable to

patients with persistent AF. Patients in sinus rhythm at the moment of surgery showed the

highest protein expression (Figure 2).

Discussion

Both experimental (2,3,19,20) and human (4,5,21,22) AF is accompanied by shortening

of the action potential duration and effective refractory period. This shortening can be

mediated by either an increase in K+ channel gene products and/or activity, or a decrease

in L-type Ca2+ channel gene products and/or activity. Previously, we demonstrated a reduced

mRNA and protein expression of the L-type calcium channel in patients with longstanding

AF, but not in paroxysmal AF (10,11). The present study shows that in patients with long-

standing persistent AF, the mRNA and protein expression of almost all investigated

potassium channel genes were reduced. In paroxysmal AF patients, reduction in mRNA

levels was confined to Kv4.3, whereas the investigated protein levels (Kv4.3, Kv1.5 and

Kir3.1) were all importantly decreased. Finally, there was a significant positive correlation

between the duration of sinus rhythm after the last episode of paroxysmal AF and content

of protein expression of Kv1.5, suggesting a protective effect of high protein contents or

normalization of protein content after a longer duration of sinus rhythm.

Differences in mRNA and protein expression

We determined mRNA and protein levels of genes encoding a number of potassium

channels. Unfortunately, no antibodies against all the potassium channels have yet been

generated. Therefore, we could only study protein expression of Kv4.3, Kv1.5 and Kir

3.1. Nevertheless, this study reports profound changes in protein expression in both

persistent and paroxysmal AF. In contrast, reduction of mRNA contents seems almost an

exclusive feature for persistent AF. The observed reduction in Kv4.3 mRNA expression

(gene encoding the calcium independent transient outward current) in patients with both

persistent and paroxysmal AF is in accordance with experimental and human studies (3,8).

In dogs subjected to rapid atrial pacing (400 bpm) the transient outward current was reduced

by 70% after 6 weeks (3) with a concomitant reduction of mRNA and protein expression

(6).

In the heart, Kir3.1 and Kir3.4 gene products appear to be responsible for the

acetylcholine-activated K+ current (16), representing an important atrial inwardly rectifying

current. Activation of this channel, e.g. by vagal stimulation, shortens the action potential

duration and refractory period. The Kir3.4 gene was used for mRNA expression

determination and a reduction in Kir3.4 mRNA expression was found in patients with

64

Chapter 4

persistent AF. For Western blotting Kir3.1 was analyzed. A reduction in protein level was

observed in patients with both persistent and paroxysmal AF, which may occur to protect

the cell against further shortening of the action potential duration during AF. The

downregulation observed in our study is, however, in contrast to findings by others on the

electrophysiological level. In a comparable group of patients with persistent AF, an increase

in inward rectifying currents (IK1

and IKACh

) was measured in isolated myocardial cells

(23). This apparent inconsistency between protein level and current density can only be

explained by assuming a change in single channel properties in patients with persistent

AF, such as an increase of mean open-time, an increase in channel conductance or a change

in voltage dependency.

The reduction of Kir6.2 mRNA levels in patients with persistent AF may be related to

depletion of ATP by an increase in metabolic demand during AF. This depletion of ATP

could promote opening of Kir6.2 leading to enhanced repolarization (24) and subsequently

increased expression of this channel (25). When activation of Kir6.2 continues the myocyte

may eventually respond by reducing the gene expression of this channel. There is still

uncertainty whether atrial ischemia indeed plays a role in triggering electrical remodeling

by AF. First, in a canine model White et al. demonstrated that induced AF immediately

caused increase in coronary atrial perfusion and oxygen consumption of atrial myocardium,

but without induction of ischemia (26). On the other hand a progressive increase in metabolic

demand during persistent AF may lead to repeated episodes of atrial ischemia, contributing

to activation of the ATP dependent potassium channel. The latter is suggested by results of

Ausma et al. who demonstrated similarities between cellular structural changes induced

by AF and those seen in hibernating myocardium (27).

The observed reduction in protein expression of Kv1.5 in patients with persistent and

paroxysmal AF could be due to post-transcriptional changes, since at the mRNA level no

changes were found between the groups. The reduction in protein expression is in agreement

with the previous data in patients with persistent AF of Van Wagoner et al. (8). However,

in a canine model of the group of Nattel no changes could be found in the current density

of IKur

(3). It should be pointed out that the molecular species underlying canine I

Kur, probably

Kv3.1, is likely different than that underlying human atrial IKur

, Kv1.5 (28).

No changes in mRNA expression were found for the HERG gene, the gene encoding

the rapid component of the delayed rectifier. This is in accordance with data of Yue and

coworkers (3) and suggests that the HERG gene is less involved in repolarization at the

atrial level during AF.

Finally, we observed a positive correlation between the duration of sinus rhythm

before surgery and the protein levels of Kv1.5 in patients with paroxysmal AF; patients in

AF at the moment of surgery had lower protein levels compared with patients in sinus

rhythm. This finding may suggest that alterations in protein expression and possibly also

65


structural changes occur early (most paroxysms lasted shorter than 24 hours) and could be

reversible.

Underlying mechanisms

The observed reduction in gene expression of three potassium currents clearly can

not explain the observed shortening of effective refractory period and action potential

duration. One may hypothesize that reduction in potassium channels gene expression serves

as an adaptation mechanism to prolong the initially reduced atrial effective refractory

period and action potential duration.

The observed discrepancy between alterations in mRNA and protein expression in

patients with paroxysmal AF may suggest the influence of a different compensatory

mechanism. We hypothesize that reduction in protein channels occurs due to calcium

overload (20,29) and structural changes, including atrophy (27,30), in atrial tissue during

AF by an increased expression of proteolytic enzymes (31). An increased expression of

the proteolytic system is observed in heart tissue during atrophy, calcium overload and

stunning (32-36). Increased protein degradation in muscle atrophy and calcium overload

seemed predominantly induced by activation of a non-lysosomal ATP dependent proteolytic

process. Medina et al. showed that the ubiquitin-proteasome dependent pathway, a highly

conserved pathway consisting of ubiquitin, ubiquitin-conjugating enzymes, deubiquitinases

and proteasome, is activated in atrophying muscles of the heart during starvation (31).

Another common cytosolic proteinase regulating pathway in eukaryotes is the calcium-

dependent pathway, which consists of a diverse group of calcium-dependent cysteine

proteinases (calpains in vertebrate tissues) (37). The increase in cytosolic calcium (29,38)

during AF, could be an important activator of this calcium-dependent pathway by promoting

activation of neutral proteases such as calpains which, once accomplished, leads to

proteolysis of numerous cytoskeletal, membrane-associated and regulatory proteins (32-

35) leading to degeneration of the myocardial cell.


Drugs and differences in underlying diseases may influence gene expression of ion

channels. In this study, to minimize the influence of particular clinical parameters on gene

expression, we included only patients with a normal left ventricular function and, when

possible, drugs were discontinued before surgery.

Because of the limited amount of tissue available, no matched controlled analysis could

be performed for determination of protein levels. However, no significant changes in mRNA

levels between the various control groups of patients in sinus rhythm were observed.

Therefore, in our opinion, a comparison between persistent AF, paroxysmal AF and sinus

rhythm patients seems to be justified.

66

Chapter 4

The paroxysmal AF patients, included in this study, represent patients who were

difficult to treat and underwent predominantly MAZE surgery. Furthermore, it should be

noted that in all groups the number of patients was small. Therefore, the present data can

not be extrapolated uncritically to all (paroxysmal) AF patients.

Acknowledgments

Dr. Van Gelder was supported by Grant 94.014 of the Netherlands Heart Foundation, The

Hague, The Netherlands. The study was supported by Grant 96.051 of The Netherlands

Heart Foundation, The Hague, The Netherlands.

References

1. Kannel WB, Abbott RD, Savage DD, McNamara PM. Epidemiologic features of chronic atrial fibrillation:

the Framingham study. N Engl J Med 1982;306:1018-22.

2. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in


3. Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S. Ionic remodeling underlying action potential changes in

a canine model of atrial fibrillation. Circ Res 1997;81:512-25.

4. Franz MR, Karasik PL, Li C, Moubarak J, Chavez M. Electrical remodeling of the human atrium: similar

effects in patients with chronic atrial fibrillation and atrial flutter. J Am Coll Cardiol 1997;30:1785-92.

5. Pandozi C, Bianconi L, Villani M, et al. Electrophysiological characteristics of the human atria after

cardioversion of persistent atrial fibrillation. Circulation 1998;98:2860-5.

6. Yue L, Melnyk P, Gaspo, Wang Z, Nattel S. Molecular mehanisms underlying ionic remodeling in a dog


7. Gaspo R, Bosch RF, Bou-Abboud E, Nattel S. Tachycardia-induced changes in Na+ current in a chronic

dog model of atrial fibrillation. Circ Res 1997;81:1045-52.

8. Van Wagoner DR, Pond AL, McCarthy PM, Trimmer JS, Nerbonne JM. Outward K+ current densities and

Kv1.5 expression are reduced in chronic human atrial fibrillation. Circ Res 1997;80:1-10.

9. Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM. Atrial L-Type Ca2+

Currents and Human Atrial Fibrillation. Circ Res 1999;85:428-36.

10. Van Gelder IC, Brundel BJJM, Henning RH, et al. Alterations in gene expression of proteins involved in

the calcium handling in patients with atrial fibrillation. J Cardiovasc Electrophysiol 1999;10:552-60.

11. Brundel BJJM, Van Gelder IC, Henning RH, et al. Gene expression of proteins influencing the calcium

homeostasis in patients with persistent and paroxysmal atrial fibrillation. Cardiovasc Res 1999;42:443-54.

12. Dixon JE, Shi W, Wang HS, et al. Role of the Kv4.3 K+ channel in ventricular muscle. A molecular corre-

late for the transient outward current [published erratum appeared in Circ Res 1997 Jan;80(1):147]. Circ

Res 1996;79:659-68.

13. Curran ME, Splawski I, Timothy KW, Vincent GM, Green ED, Keating MT. A molecular basis for cardiac

arrhythmia: HERG mutations cause long QT syndrome. Cell 1995;80:795-803.

14. Wang Z, Fermini B, Nattel S. Sustained depolarization-induced outward current in human atrial myocytes:

evidence for a novel delayed rectifier K+ current similar to Kv1.5 cloned channel currents. Circ Res

1993;73:1061-76.

15. Fedida D, Wible B, Wang Z, et al. Identity of a novel delayed rectifier current from human heart with a

cloned K+ channel current. Circ Res 1993;73:210-6.

16. Krapivinsky G, Gordon EA, Wickman K, Velimirovic B, Krapivinsky L, Clapham DE. The G-protein-

gated atrial K+ channel IKACh

is a heteromultimer of two inwardly rectifying K(+)-channel proteins. Nature

1995;374:135-41.

17. Inagaki N, Gonoi T, Clement IV J, et al. Reconstitution of IKATP

: An inward rectifier subunit plus the sulfo-

nylurea receptor. Science 1995;270:1166-70.

18. Gallagher MM, Camm AJ. Classification of atrial fibrillation. Pacing Clin Electrophysiol 1997;20:1603-5.


remodeling of the atria. Circulation 1997;95:1945-53.

67



nisms. Circulation 1996;94:2968-74.


electrical remodeling in humans. Circulation 1997;96:1542-50.


dependency and effects of antiarrhytmic drugs. Circulation 1998;97:2331-7.

23. Bosch RF, Zeng X., Grammer JB, Popovic K, ewis C, ühlkamp V. Ionic mechanisms of electrical remodel-

ing in human atrial fibrillation. Cardiovasc Res 1999;44:121-31.

24. Shaw RM, Rudy Y. Electrophysiologic effects of acute myocardial ischemia. A theoretical study of altered

cell excitability and action potential duration. Cardiovasc Res 1997;35:256-72.

25. Akao M, Otani H, Horie M, et al. Myocardial ischemia induces differential regulation of KATP

channel gene

expression in rat hearts. J Clin Invest 1997;100:3053-9.

26. White C, Holida M, Marcus M. Effects of acute atrial fibrillation on the vasodilator reserve of the canine

atrium. Cardiovasc Res 1986;20:683-9.

27. Ausma J, Wijffels M, Thone F, Wouters L, Allessie M, Borgers M. Structural changes of atrial myocardium

due to sustained atrial fibrillation in the goat. Circulation 1997;96:3157-63.

28. Yue L, Feng J, Wang Z, Nattel S. Adrenergic control of the ultrarapid delayed rectifier current in canine

atrial myocytes. Journal of Physiology 1999;516:385-98.

29. Sun H, Leblanc N, Nattel S. Effects of atrial tachycardia on intracellular Ca2+ and cellular contractility.

Circulation 1999;100:I-200

30. Morillo CA, Klein GJ, Jones D, Guiraudom CM. Chronic rapid atrial pacing. Structural, functional, and

electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation 1995;91:1588-

95.

31. Medina R, Wing SS, Goldberg AL. Increase in levels of polyubiquitin and proteasome mRNA in skeletal

muscle during starvation and denervation atrophy. Biochem J 1995;307:631-7.

32. Bartus RT, Elliott PJ, Hayward NJ, et al. Calpain as a novel target for treating acute neurodegenerative

disorders. Neurol Res 1995;17:249-58.

33. Atsma DE, Bastiaanse EM, Jerzewski A, Van Der Valk LJ, Van Der Laarse A. Role of calcium-activated

neutral protease (calpain) in cell death in cultured neonatal rat cardiomyocytes during metabolic inhibition.

Circ Res 1995;76:1071-8.

34. Gorza L, Menabo R, Di Lisa F, Vitadello M. Troponin T cross-linking in human apoptotic cardiomyocytes.

Am J Pathol 1997;150:2087-97.

35. Gorza L, Menabo R, Vitadello M, Bergamini CM, Di Lisa F. Cardiomyocyte troponin T immunoreactivity

is modified by cross-linking resulting from intracellular calcium overload. Circulation 1996;93:1896-904.

36. Gao WD, Atar D, Liu Y, Perez NG, Murphy AM, Marban E. Role of troponin I proteolysis in the pathogen-

esis of stunned myocardium. Circ Res 1997;80:393-9.

37. Mykles DL. Intracellular proteinases of invertebrates: calcium-dependent and proteasome/ubiquitin-de-

pendent systems. International Review of cytology 1998;184:157-289.

38. Ausma J, Dispersyn GD, Duimel H, et al. Changes in ultrastructural calcium distribution in goat atria

during atrial fibrillation. J Mol Cell Cardiology 2000;32:355-64.

68

69

Ion channel remodeling is related to intra operative atrial refractory periods in patients

Chapter 5

Ion Channel Remodeling is Related to Intra- Operative

Atrial Effective Refractory Periods in Patients with

Paroxysmal and Persistent Atrial Fibrillation

Bianca J. J. M. Brundel, MSc1,2; Isabelle C. Van Gelder, MD1;

Robert H. Henning, MD2; Robert G. Tieleman, MD1; Anton E. Tuinenburg,

MD1; Mirian Wietses2; Jan G. Grandjean, MD3, Wiek H. Van Gilst, PhD2;

Harry J. G. M. Crijns, MD1.

Departments of Cardiology1, Clinical Pharmacology2 and Thoracic Surgery3,

Thoraxcenter University Hospital Groningen, The Netherlands.

Circulation: in press

Abstract

Background Sustained shortening of the atrial effective refractory period (AERP),

probably due to reduction in L-type calcium current, is a major factor in the initiation and

maintenance of AF. We investigated underlying molecular changes by studying the rela-

tion between gene expression of the L-type calcium channel and potassium channels and

the AERP in patients with AF. Methods and Results mRNA and protein expression were

determined in left and right atrial appendages of 13 patients with paroxysmal AF, 16 with

persistent AF and 13 controls in sinus rhythm, by RT-PCR and slot-blot, respectively. The

mRNA content of almost all investigated ion-channel genes was reduced in persistent but

not in paroxysmal AF. Protein levels for L-type Ca2+ channel and five potassium channels

(Kv4.3, Kv1.5, HERG, minK and Kir3.1) were significantly reduced in both persistent

and paroxysmal AF. Furthermore, AERPs were determined intra-operatively with pro-

grammed electrical stimulation at 5 basic cycle lengths (BCLs) (between 250 and 600

ms). Patients with persistent and paroxysmal AF displayed significant shorter AERPs.

Protein levels of all ion-channels investigated correlated positively with the AERP (at

BCL: 600, 500, 400 and 300 ms) and with the rate adaptation of AERP. Patients with

reduced ion-channel protein expression revealed shorter AERP duration and poorer rate

adaptation.

70

Chapter 5

Conclusions AF is predominantly accompanied by decreased protein contents of the L-

type Ca2+ channel and several potassium channels. Reductions in L-type Ca2+ channel

correlated with AERP and rate adaptation, and represent a probable explanation for the

electrophysiological changes during AF.

Introduction

Atrial fibrillation (AF) is a common arrhythmia affecting millions of people world-

wide.1 AF has the tendency to become more persistent and increasingly difficult to treat

over time. During recent years, experimental and human studies showed that rapid short-

ening of the atrial effective refractory period (AERP) is an important factor contributing

to the maintenance of AF.2-6 Rapid shortening of the AERP in AF involves functional

changes in ion channels. Animal experimental data revealed that the L-type Ca2+ channel

plays a main role in shortening of AERP and action potential duration (APD).7,8 These

observations are supported by blocking of AERP shortening with the L-type Ca2+ antago-

nist, verapamil, in other experimental studies.9,10 In addition, human data on AF have dem-

onstrated reductions in ICaL

11,12 and gene expression of L-type Ca2+ channel.13

However, shortening of AERP could also be explained by an increase in (repolariz-

ing) K+ channel activity. Indeed, one study found increased IKACh

and IK1

in isolated human

atrial cells of patients with persistent AF.11 In contrast, other studies support a decrease in

K+ channels in AF. In human atrial myocytes reductions in ITo

and IKsus

and a reduced gene

expression of Kv1.5, Kv4.3, Kir3.1, Kir3.4 and Kir6.213-15 were found.

Until now, the relationship between changes in AERP and ion channel gene expres-

sion has not been investigated in human tissue of patients with AF. The aim of the present

study was to investigate the regulation of L-type Ca2+ channel and K+ channels and its

relation to AERP in patients with persistent and paroxysmal AF. We included both patients

with lone AF and with mitral valve disease (MVD), since the occurrence of MVD seems

to prolong the AERP.16,17


Patients and atrial tissue collecting

Prior to surgery, one investigator assessed the clinical characteristics of patients

(Table 1). The patient’s arrhythmia history was classified according to Gallagher.18 The

persistent and paroxysmal AF group contained patients with lone AF or AF with underly-

ing MVD. All patients underwent MAZE surgery, were euthyroid and had normal left

ventricular function. Coumarin therapy was interrupted 3 days before surgery and class I

and III anti-arrhythmic drugs were discontinued for at least 5 half-times. During surgery

the AERPs were determined with use of temporary epicardial pacing leads. AERPs were

measured intra-operatively at 5 different basic cycle lengths (BCL; 600, 500, 400, 300 and

71


250 ms) at the right atrial appendage (RAA) and left atrial appendage (LAA) using pro-

grammed electrical stimulation.

LAAs and RAAs were obtained except for the control patients undergoing CABG

from whom only the RAA was gathered. After excision, the RAAs and LAAs were imme-

diately snap-frozen in liquid nitrogen and stored at -85 °C. The study was approved by the

Institutional Review Board and written informed consent was given by all patients.

Table 1. Baseline characteristics of patients with lone paroxysmal AF, lone persistent AF and control patients

in sinus rhythm

Lone AF AF with MVD

SR (CABG) PAF CAF SR(MVD) PAF CAF

N 9 7 7 4 6 9

Age 61±8 48±7 50±7 60±9 47±9 56±10

Previous duration of AF

(median, range (months)) - - 13.6 (0.1-56) - - 8(0.4-32)

Duration SR before surgery

(median, range (days)) - 2 (0.5-12) - - 75(10-210) -


and /surgical procedure

Coronary artery disease/CABG 9 0 0 0 0 0

Lone AF / MAZE 0 7 7 0 0 0

MVD /MV replacement/repair 0 0 0 4 6 9

New York Heart Association for

exercise tolerance

Class I 9 6 4 1 0 0

Class II 0 1 3 3 3 4

Class III 0 0 0 0 3 5

Echocardiography

Left atrial diameter (parasternal) 36± 5 39± 5 45± 8 47±5 45±9 51±10

Left ventricular end-diastolic diam

eter (mm) 37± 7 50± 4 49± 8 60±7 54±8 54±5

Left ventricular end-systolic diam

eter (mm) 29± 8 37± 4 29± 13 38±6 38±6 38±7

Medication (n)

Ace-inhibitors 0 0 0 4 5 7

Digitalis 0 1 4 0 0 2

Verapamil 2 2 3 2 1 0

Beta-blocker 4 2 2 2 1 2

Values are presented as mean value ± SD or number of patients. ACE inhibitor indicates Angiotensin Con-

verting Enzyme; AF, atrial fibrillation; CABG, Coronary Artery Bypass Grafting; CAF, chronic persistent

atrial fibrillation; PAF, paroxysmal atrial fibrillation; SR, control patients in sinus rhythm.

72

Chapter 5


Total RNA was isolated and processed as described previously.13 Briefly, first strand

cDNA was synthesized by incubation of 1 µg of total RNA in reverse transcription 10x

buffer, 200 ng of random hexamers with 200 units of Moloney Murine Leukemia Virus

Reverse Transcriptase, 1mM of each dNTP and 1 unit of RNase inhibitor (Promega, The

Netherlands) in 20 µl. Synthesis reaction was performed for 10 minutes at 20 °C, 20

minutes at 42 °C, 5 minutes at 99 °C and 5 minutes at 4 °C. All the products were checked

for contaminating DNA.


We described and validated these methods before.13 In short, the cDNA of interest

and the cDNA of the ubiquitously expressed housekeeping gene glyceraldehyde-3-phos


protein sequence cycles ann. temp(°C)

GAPDH: F 5'-CCC ATC ACC ATC TTC CAG GAG CG-3', var. var.

R 5'-GGC AGG GAT GAT GTT CTG GAG AGC C-3'.

Na-channel: F 5'-ATG CAG CTG TGG ACT CCA GG-3' 27 56

SCN5A R 5'-CAG GCG GAT GAC TCG GAA GA-3'

L-type Ca2+ F 5'-CTG GAC AAG AAC CAG CGA CAG TGC G-3' 30 56

channel: R 5'-ATC ACG ATC AGG AGG GCC ACA TAG GG-3'

To1: F 5’-CAG CAA GTT CAC AAG CAT CC-3’ 31 52

Kv4.3 R 5’-AGC TGG CAG GTT AGA ATT GG-3’

Ks: F 5'-AGC AGA AGC AGA GGC AGA AG-3' 28 58

KvLQT1 R 5'-GAC GGA GAT GAA CAG TGA GG-3'

Kr: F 5’-GTC AAT GCC AAC GAG GAG GT-3’, 27 58

HERG R 5’-CTG GTG GAA GCG GAT GAA CT-3’

Kur: F 5’-AAC GAG TCC CAG CGC CAG GT-3’ 32 64

Kv1.5 R 5’-AGG CGG ATG ACT CGG AGG AT-3’

KACh F 5’-CAC CCT GGT GGA CCT CAA GTG GCG C-3’ 30 56

Kir3.4 R 5’-AGC TCC GGG CTT GGC AGG TCA TGC-3’

KATP F 5’-CAC GCT GGT GGA CCT CAA GT-3’ 29 59

Kir6.2 R 5’-CAC GAT GAG GCT CAG GAT GG-3’

GAPDH, glyceraldehyde-3-phosphate dehydrogenase; SCN5A, gene encoding Na-channel; l-type Ca2+ chan-

nel, voltage gated L type calcium channel α1 subunit; Kv4.3, gene probably encoding the calcium indepen-

dent transient outward current Ito1

; KvLQT1, gene encoding the slow delayed rectifier current together with

minK; HERG, gene encoding the rapid component of the delayed rectifier; Kv1.5, gene encoding the ultra-

rapid component of the delayed rectifier IKur

; Kir3.4, gene encoding part of the heterotetrameric complex of

this gene together with Kir3.1 which forms the acetylcholine dependent potassium current, IKACh

; Kir6.2,

gene encoding the inward rectifier K+ current, forming IKATP

with sulfonylurea receptor 2.

73


phate dehydrogenase (GAPDH) were co-amplified in a single PCR. Primers (Eurogentec,

Belgium) were designed for SCNA1, L-type calcium channel, Kv4.3, HERG, Kv1.5, Kir3.4

and Kir6.2 and the housekeeping gene GAPDH (Table 2).


ethidium bromide. The density of the PCR products was quantified by densitometry. Lin-

earity of the PCR was established by a good correlation between the number of cycles and

the density of gene of interest and GAPDH (data not shown).

Protein Preparation and Slot Blotting

Frozen atrial appendages of patients in sinus rhythm, patients with paroxysmal AF

and patients with persistent AF were homogenized in RadioImmunoPrecipitationAssay

(RIPA) buffer as described before.13 The homogenate was centrifuged at 14.000 rpm for

20 minutes at 4°C. The supernatant was used for protein concentration measurement ac-

cording to the Bradford method (Bio-Rad, The Netherlands) with bovine albumin used as

a standard. Samples of 10 µg heat denatured protein were spotted on nitrocellulose mem-

branes (Stratagene, The Netherlands) and checked by staining with Ponceau S solution

(Sigma, The Netherlands). Blocking was performed for 20 minutes in blocking buffer (5%

nonfat milk, TBS and 0.1% Tween 20). After washing three times in TBS with 0.1%

Tween 20 the membranes were incubated with primary antibody against GAPDH (Affin-

ity Reagents, USA), L-type calcium channel α1 subunit, Kv4.3, HERG, minK, Kir3.1

and Kv1.5 (all Alomone Labs, Israel). Immunodetection of the primary antibody was per-

formed with peroxidase conjugated secondary antibody anti-mouse IgG (Santa Cruz Bio-

technology, The Netherlands). The blot was incubated with the ECL-detection reagent

(Amersham, The Netherlands) for 1 minute, and exposed to an X-OMAT x-ray film (Kodak,

The Netherlands) for 15 to 90 seconds. The band densities were evaluated by densitomet-

ric scanning using a Snap Scan 600 (Agfa, The Netherlands). The amount of protein cho-

sen was in the linear immunoreactive signal area and the specificity of the antibody was

checked by SDS-PAGE and pre-incubation with the control peptide antigen.

Rate adaptation coefficient

To quantify the change in AERP at the different BCLs, we calculated the rate adapta-

tion coefficient for individual patients as the slope of the linear regression after logarith-

mic transformation of BCL. Three patients were excluded, because AERP was obtained at

less than 4 BCLs.


All PCR and slot-blotting procedures were performed in duplo series and mean val-

ues were used for statistical analysis. Comparison between groups for normally distrib

74

Chapter 5

uted variables was performed by one-way ANOVA and for skewed variables by Wilkoxon

two-sample test. For determination of correlations the Spearman correlation test was used.

The Mann-Whitney U-test was performed for group to group comparisons of small num-

bers. All p-values are two-sided, a p-value <0.05 was considered statistically significant.

SPSS version 8.0 was used for all statistical evaluations.

Results

mRNA Remodeling


gene-of-interest/GAPDH ratios between patients with persistent AF, paroxysmal AF and

their controls in sinus rhythm (Table 3). No differences in GAPDH amount between the

groups were found for all the genes investigated (data not shown). Persistent, lone AF was

associated with reductions in mRNA amount of Kv4.3, L-type Ca2+ channel and Kir3.4.

The mRNA amounts of HERG and KvLQT1 showed an additional reduction in persistent

AF with MVD. In general the reduction in mRNA expression was less pronounced in

paroxysmal AF.

Table 3. Percentage ion-channel remodeling: comparison of lone AF versus AF with MVD

mRNA Protein

lone +MVD lone +MVD lone +MVD lone +MVD

PAF PAF CAF CAF PAF PAF CAF CAF

Na-channel +35 ±6 -* - - na na

Kv4.3 -20 ±4 -20 ±5 -19 ±5 -25 ±3 -46 ±10-75 ±6* -33 ±8 -64 ±4*

L-type calcium channel -13 ±5 - -27 ±4 -20 ±4 -48 ±8 -65 ±5 -55 ±6 -57 ±6

HERG - - - -22 ±4 -28 ±10 -38 ±5 -25 ±8 -42 ±6

Kv1.5 - - - - -46 ±10 -69 ±5 -47 ±8 -61 ±7

KvLQT1/minK - +82 ±8* - +56 ±6* -35 ±5 -44 ±3 -50 ±6 -39 ±7

Kir3.4/Kir3.1 - - -27 ±5 -34 ±5 -41 ±7 -67 ±5* -42 ±6 -62 ±6*

Kir6.2 +28 ±4 -* - -42 ±5* na na

Only significant changes (p<0.05) are given for the mRNA or protein content of interest/GAPDH. CAF means

patients with chronic, persistent AF, PAF means paroxysmal AF, MVD means mitral valve disease; - means

not significant; na means not available. *, means significant differences between lone PAF and PAF with

MVD or lone CAF and CAF with MVD.

75


Protein Remodeling

Proteins were isolated from RAA and LAA and used for immunological detection of

L-type Ca2+ channel, Kv4.3, HERG, Kv1.5, minK and Kir3.1. Changes in protein expres-

sion were studied in relation to protein levels of GAPDH and to total amount of protein

spotted on the membrane. Because the GAPDH density and total protein amount density

showed a highly significant positive correlation (r=0.92, p<0.001), we used the protein of

interest/GAPDH ratio for further investigation. The protein expression of L-type Ca2+ chan-

nel, Kv4.3, Kv1.5, HERG, Kir3.1 and minK was reduced in both patients with both persis-

tent and paroxysmal AF (Figure 1 and 2, Table 3). Furthermore, ion-channel protein levels

did not correlate with mRNA levels, duration of persistent AF or with the duration of sinus

rhythm before surgery (data not shown).

Significant differences in protein remodeling between the lone AF group and the AF

group with MVD were observed. Reductions in protein expression of Kv4.3, minK and

Kir3.1 were more pronounced in patients with underlying MVD (Table 3).

Figure 1. Slot blot analysis of 10

µg of protein homogenates of 6

patients in sinus rhythm (SR), 6

patients with paroxysmal AF

(PAF) and 6 patients with

chronic, persistent AF (CAF).

The immunoblots were done for

(A) anti GAPDH, (B) anti L-type

calcium channel α1 subunit, (C)

anti Kv4.3, (D) anti-Kv1.5, (E)

anti HERG, (F) anti minK and

(G) anti-Kir3.1.

SRPAFCAF

SR

SR

PAF

PAF

CAF

CAF

A

B

C

SR

SR

SR

PAF

PAF

PAF

CAF

CAF

CAF

F

E

D

CAFPAFSRG

76

Chapter 5

Atrial Effective Refractory Period and Protein Remodeling

The AERP at 5 different basic cycle lengths (BCLs: 600, 500, 400, 300 and 250 ms)

was determined in the RAA and LAA of patients during surgery. Patients with persistent

and paroxysmal AF had significantly shorter AERPs than patients in sinus rhythm (Table

4). The relation between AERP and the amount of ion-channel protein was investigated,

because protein amounts are anticipated to represent the amount of functional ion-channel

closer than mRNA levels. A significant positive correlation was found at BCLs of 600,

500, 400 and 300 ms for all the proteins investigated in patients with AF (Figure 3, Table

5). Patients with reduced ion-channel protein expression exhibited the shortest AERP.

Furthermore, no significant correlation was found between the GAPDH amount and AERP

(data not shown).

Relation Rate Adaptation and Protein Remodeling

The rate adaptation coefficient was determined for every RAA and LAA. The rate

adaptation coefficient was significantly reduced by 32% in persistent AF compared to

sinus rhythm (mean persistent AF: 104 ± 53; paroxysmal AF: 133 ± 62 and sinus rhythm:

Figure 2. Protein ratios for (A) L-type calcium channel, (B) Kv4.3, (C) Kv1.5, (D) HERG, (E) minK and (F)

Kir3.1 for patients in sinus rhythm (SR), with paroxysmal AF (PAF) and with chronic, persistent AF (CAF). *,

p<0.01, **, p<0.001

SR PAF CAF0

1

2

SR PAF CAF0,0

0,5

1,0

SR PAF CAF0,0

0,5

1,0

1,5

SR PAF CAF0,0

0,5

1,0

1,5

SR PAF CAF0,0

0,5

1,0

SR PAF CAF

Ion-channel protein expression

0,0

0,5

1,0

1,5

* ** ** **

****

****

Ion-

chan

nel p

rote

in e

xpre

ssio

n

A B

C D

E F

77


153 ± 32), indicating a poorer adaptation to higher heart rates in patients with AF. Signifi-

cant positive correlations were observed between ion-channel protein expression and the

adaptation coefficient (Figure 4). AF patients with reduced ion-channel protein expression

demonstrated poorer rate adaptation.

Furthermore, significant differences were observed between lone paroxysmal AF and

patients with paroxysmal AF and MVD. Lone paroxysmal AF demonstrated a poorer rate

adaptation compared to paroxysmal AF with MVD (109 ± 40 and 164 ± 76, p=0.04,

respectively).

Discussion

Both experimental and human AF is accompanied by electrical remodeling2,4-6,10,19

and ion-channel remodeling.7,8,11,12-15,20 This is the first study which demonstrates in human

paroxysmal and persistent AF (1) a positive correlation between the ion-channel protein

remodeling and the AERPs, irrespective of the underlying heart disease, (2) a correlation

between ion-channel protein remodeling and changes in rate adaptation and (3) dis-

Figure 3. Correlation between

the ion-channel protein expres-

sion of (A) L-type calcium chan-

nel, (B) Kv4.3, (C) Kv1.5, (D)

HERG, (E) minK and (F) Kir3.1

and the AERP measured at BCL

of 500 ms in RAA and LAA.

( ) represents control patients

in sinus rhythm undergoing

CABG, ( ) patients with lone

paroxsymal AF, ( ) patients

with lone persistent AF, ( ) pa-

tients in sinus rhythm with un-

derlying MVD, ( ) patients

with paroxysmal AF and MVD,

( ) patients with persistent AF

and MVD.

0 1 2 3

AERP(msBCL500ms)

150200250300350

0 1 2150200250300350

r=0.77, p<0.001

r=0.65, p<0.001

L-type Ca2+ channel

Kv1.5

0,0 0,5 1,0 1,5r=0.43, p<0.001

Kv4.3

HERG0 1 2r=0.52, p<0.001

Kir3.10 1 2 3minK0,0 0,5 1,0 1,5150200250300350

r=0.55, p<0.001r=0.54, p<0.001

AE

RP

(m

s, B

CL

500 m

s)

A B

C D

E F

78

Chapter 5

Table 4. AERP measured at the different BCLs.

AERP (ms)

lone AF AF with MVD

BCL (ms) SR (CABG) PAF CAF SR (MVD) PAF CAF

600 291±53 222±15* 208±39* 271±19 274+25 226±42

500 277±42 224±24* 207±29* 268±22 259±35 228±34

400 252±34 216±24* 203±25* 256±20 243±32 217±33*

300 224±16 202±20* 189±24* 226±21 217±33 200±40

250 184±5 185±19 172±17 180±5 187±28 170±11

Data expressed as mean ± SD

*, means p<0.05

Table 5. Relation AERP and protein remodeling for the different basic cycle lengths (BCL).

BCL L-type Ca2+ Kv4.3 Kv1.5 HERG Kir3.1 minK

channel

r p-value r p-value r p-value r p-value r p-value r p-value

600 0.67 <0.001 0.32 0.001 0.57 <0.001 0.34 0.008 0.39 0.003 0.4 0.002

500 0.77 <0.001 0.43 <0.001 0.65 <0.001 0.53 <0.001 0.55 <0.001 0.54 <0.001

400 0.68 <0.001 0.35 0.004 0.61 <0.001 0.47 <0.001 0.45 <0.001 0.48 <0.001

300 0.53 <0.001 0.29 0.04 0.47 0.009 0.32 0.009 0.34 0.006 0.33 0.008

250 0.47 0.001 ns 0.42 0.004 0.3 0.03 ns 0.310.02

ns, not significant; r, regression coefficient

crepancies between mRNA and protein remodeling. These data suggest ion-channel protein

remodeling represent an important adaptation mechanism during AF, that may contribute

to intractability of AF and inactivity of antiarrhythmic drugs instituted for the prevention

of AF.

Relation ion channel remodeling and AERP

The observed ion-channel protein remodeling in this study is associated with the

occurrence of AF. Patients with paroxysmal and persistent AF showed marked reductions

in ion-channel protein expression of both L-type Ca2+ channel and several K+ channels.

Furthermore, low ion-channel protein levels were associated with short AERP and poor

79


rate adaptation. This indicates that electrical remodeling2 and structural remodeling21 are

paralleled by ion-channel protein remodeling as part of the adaptation mechanisms during

AF. Furthermore, patients with paroxysmal AF showed a reduction in ion-channel protein

expression comparable to persistent AF in the absence of mRNA reductions, suggesting

that paroxysms of AF are able to induce changes in ion-channel protein expression via

activation of a proteolytic system. Indeed, we have observed activation of the calpain

system in human paroxysmal and persistent AF (Brundel et al., submitted).

As stated above, AF is accompanied by shortening of the AERP and action potential

duration (APD). It has been suggested that the short-term decrease of APD and its reduced

rate adaptation is mainly due to a ± 70 % reduction of the L-type calcium current in

animal experimental studies and human AF.7,8,11,12 This assumption is further supported by

the observation that administration of the L-type Ca2+ channel agonist Bay K 8644 largely

restored the plateau phase of the action potential in remodeled cells.22 If the main role for

Kir3.10 1 2 3

Kv1.50 1 2

0100200300

Kv4.30,0 0,5 1,0 1,5L-type Ca2+ channel 0 1 2 3

adaptationcoefficient

0

100

200

300

r=0.50, p<0.001 r=0.34, p=0.004

r=0.34, p=0.007

minK0,0 0,5 1,0 1,5

0100200300

HERG0 1 2

r=0.37, p=0.003

r=0.36, p=0.003 r=0.28, p=0.02

adapta

tion c

oeff

icie

nt

Figure 4. Correlation between

the ion-channel protein expres-

sion and the rate adaptation

coefficient for (A) L-type cal-

cium channel, (B) Kv4.3, (C)

Kv1.5, (D) HERG, (E) minK

and (F) Kir3.1. ( ) represents

control patients in sinus rhythm

undergoing CABG, ( ) pa-

tients with lone paroxsymal AF,

( ) patients with lone persist-

ent AF, ( ) patients in sinus

rhythm with underlying MVD,

( ) patients with paroxysmal

AF and MVD, ( ) patients

with persistent AF and MVD.

80

Chapter 5

L-type Ca2+ channels in APD is correct, the observed reduction in protein expression of L-

type Ca2+ channel in this study explains the present AERP shortening and decrease in its

adaptation to rate.

The other possibility that may mediate AERP shortening is an increase in (repolariz-

ing) K+ channel gene products and/or activity. However, we observed a reduction of K+

channel gene expression. Similar results were obtained in animal experimental studies

showing reductions in ITo

and Kv4.3 mRNA amount without reductions in delayed inward

rectifier K+ current and Kir2.1 expression.7 The group of Van Wagoner et al. and our group

examined the adaptation in gene expression of several potassium channels in patients with

AF.13-15 The current of ITo

and the protein expression of Kv1.5 were reduced rather than

elevated during persistent AF.15 Our previous study, in a different patient group, showed

reductions in gene expression of Kv4.3, Kv1.5, Kir3.1 and Kir6.2.14 Only one study in

isolated RAA cells of patients with persistent AF showed that shortening of the human

action potential by AF was related to a 70% reduction in ICaL

and ITo

and a 30% increase in

IK1

and IKACh

.11 The downregulation of potassium channel protein amounts observed in our

study are in contrast with the few reports on the electrophysiological level. This possible

inconsistency between decrease in protein level and increase in current density may be

explained by a change in single channel properties in patients with persistent AF, such as

an increase of mean open-time, an increase in channel conductance or a change in voltage

dependency. Thus, a reduced expression of L-type Ca2+ channels probably plays a main

role in AERP shortening. Secondary to this process, the myocardial cell may further adapt

to high rate by reducing the expression of potassium channels to counteract the shortening

of the AERP.

We did not find differences in ion-channel protein expression between AF patients

with lone AF and AF with underlying MVD. Nevertheless, AERP was prolonged in MVD,

as previously reported in experimental studies.6,16,17 Also an association between AF with

MVD and severe cellular degeneration was observed.23 The results indicate that other

factors beside AF are probably involved in the regulation of the duration of the effective

refractory period. One of most likely candidates would be morphological changes, as AF

is promoted by structural changes induced during experimental heart failure, which cause

important local conduction abnormalities that could play an additional role in the vulner-

ability of AF.24,25

Post-transcriptional regulation?

The observed discrepancy between alterations in mRNA and protein expression in

patients with paroxysmal AF suggests the activation of proteolysis. Recently, we found

that activation of the calpain system in human persistent and paroxysmal AF, in the ab-

sence of activation of the proteasome pathway (Brundel et al., submitted). As calpain are

81


activated by calcium overload in the myocard cell26,27, calpain activation would serve to

protect the cells to additional damage by down-regulation of multiple ion-channels. How-

ever, this would be at the cost of proteolysis of several cytoskeletal, membrane-associated

and regulatory proteins26,28-32 Whether interference with the calpain system represents a

valuable therapeutic strategy in AF remains to be investigated.

In conclusion, the observed correlation between ion-channel protein amounts and

AERP strongly suggest that ion-channel protein remodeling, beside the electrical remod-

eling and structural remodeling33 may play an important role in the vulnerability of AF

after restoration of sinus rhythm.


The patients with lone AF included in this study represent patients who were difficult

to treat and underwent finally MAZE surgery. Therefore, the present data cannot be ex-

trapolated uncritically to all AF patients. Furthermore, it should be noted that in all groups

the number of patients was small.

Acknowledgments

Dr. Van Gelder was supported by Grant 94.014 of the Netherlands Heart Foundation,

The Hague, The Netherlands. The study was supported by Grant 96.051 of The Nether-

lands Heart Foundation, The Hague, The Netherlands.

References

1. Kannel WB, Abbott RD, Savage DD, et al. Epidemiologic features of chronic atrial fibrillation: the

Framingham study. N Engl J Med 1982; 306:1018-1022.




cardioversion of persistent atrial fibrillation. Circulation 1998; 98:2860-2865.




dependency and effects of antiarrhytmic drugs. Circulation 1998; 97:2331-2337.

6. Tieleman RG, Van Gelder IC, Tuinenburg AE, et al. Intra- and post-operative atrial refractory periods in

relation to atrial arrhythmia history and the presence of mitral regurgitation. Circulation 1999; 100:I-361

7. Yue L, Melnyk P, Gaspo, et al. Molecular mehanisms underlying ionic remodeling in a dog model of atrial

fibrillation. Circ Res 1999; 84:776-784.







11. Bosch RF, Zeng X., Grammer JB, et al. Ionic mechanisms of electrical remodeling in human atrial fibril-

lation. Cardiovasc Res 1999; 44:121-131.

82

Chapter 5

12. Van Wagoner DR, Pond AL, Lamorgese M, et al. Atrial L-Type Ca2+ Currents and Human Atrial Fibrilla-

tion. Circ Res 1999; 85:428-436.


homeostasis in patients with persistent and paroxysmal atrial fibrillation. Cardiovasc Res 1999; 42:443-

454.

14. Brundel BJJM, Van Gelder IC, Henning RH, et al. Changes in mRNA and protein content of ion channels

in patients with paroxysmal and persistent atrial fibrillation. PACE 1999; 22:II-753

15. Van Wagoner DR, Pond AL, McCarthy PM, et al. Outward K+ current densities and Kv1.5 expression are

reduced in chronic human atrial fibrillation. Circ Res 1997; 80:1-10.

16. Boyden PA, Tilley LP, Albala A, et al. Mechanisms for atrial arrhythmias associated with cardiomyopa-

thy: a study of feline hearts with primary myocardial disease. Circulation 1984; 69:1036-1047.

17. Boyden PA, Tilley LP, Pham T, et al. Effects of left atrial enlargement on atrial transmembrane potentials

and structure in dogs with mitral valve fibrosis. Am J Cardiol 1982; 49:1896-1908.

18. Gallagher MM, Camm AJ. Classification of atrial fibrillation. Pacing Clin Electrophysiol 1997; 20:1603-

1605.



20. Gaspo R, Bosch RF, Bou-Abboud E, et al. Tachycardia-induced changes in Na+ current in a chronic dog

model of atrial fibrillation. Circ Res 1997; 81:1045-1052.





23. Thiedemann KU, Ferrans VJ. Left atrial ultrastructure in mitral valvular disease. Am J Pathol 1977;

89:575-604.

24. Li D, Melnyk P, Feng J, et al. Effects of experimental heart failure on atrial cellular and ionic electrophysi-

ology. Circulation 2000; 101:2631-2638.

25. Li D, Fareh S, Leung TK, et al. Promotion of atrial fibrillation by heart failure in dogs; atrial remodeling

of a different sort. Circulation 1999; 100:87-95.


Circulation 1999; 100:I-200


during atrial fibrillation. J Mol Cell Cardiology 2000; 32:355-364.

28. Bartus RT, Elliott PJ, Hayward NJ, et al. Calpain as a novel target for treating acute neurodegenerative

disorders. Neurol Res 1995; 17:249-258.

29. Atsma DE, Bastiaanse EM, Jerzewski A, et al. Role of calcium-activated neutral protease (calpain) in cell

death in cultured neonatal rat cardiomyocytes during metabolic inhibition. Circ Res 1995; 76:1071-1078.

30. Gorza L, Menabo R, Di Lisa F, et al. Troponin T cross-linking in human apoptotic cardiomyocytes. Am

J Pathol 1997; 150:2087-2097.

31. Gorza L, Menabo R, Vitadello M, et al. Cardiomyocyte troponin T immunoreactivity is modified by cross-

linking resulting from intracellular calcium overload. Circulation 1996; 93:1896-1904.

32. Gao WD, Atar D, Liu Y, et al. Role of troponin I proteolysis in the pathogenesis of stunned myocardium.

Circ Res 1997; 80:393-399.

33. Ausma J, Lenders MH, Duimel H, et al. Time course of structural changes in atria after atrial fibrillation

in goats: existence of hibernating myocardium. in press 2000

Part III

Gene expression of neurohormones

85

Gene expression of the natriuretic peptide system in atrial tissue

Chapter 6

Gene Expression of the Natriuretic Peptide system in

Atrial Tissue of Patients with Paroxysmal and

Persistent Atrial Fibrillation

Anton E. Tuinenburg, Bianca J.J.M. Brundel, Isabelle C. Van Gelder,

Robert H. Henning*, Maarten P. Van Den Berg, Cécile Driessen*,

Jan G. Grandjean†, Wiek H. Van Gilst*, Harry J.G.M. Crijns

From the Departments of Cardiology, Clinical Pharmacology*, and Thoracic Surgery†,

Thoraxcenter, University Hospital Groningen, Groningen, The Netherlands

J Cardiovasc Electrophysiol 1999;10:827-835

Abstract

Introduction: Circulating cardiac natriuretic peptides play an important role in

maintaining volume homeostasis, especially during conditions affecting hemodynamics.

During atrial fibrillation (AF), plasma atrial natriuretic peptide (ANP) becomes elevated.

It was the aim of the study to gather information about gene expression of the natriuretic

peptide system on the level of the atrium in patients with AF.

Methods and Results: Right atrial appendages of 36 patients with either paroxysmal or

persistent AF were compared with 36 case matched controls in sinus rhythm for mRNA

expression of pro- atrial natriuretic peptide (pro-ANP), pro-brain natriuretic peptide (pro-

BNP), and their natriuretic peptide receptor type-A (NPR-A). We investigated patients

without (n=36) and with (n=36) valvular disease. Persistent AF was associated with higher

mRNA expression of pro-BNP (+66%, p=0.04, in patients without valvular disease, and

+69%, p<0.01, in patients with valvular disease) and lower mRNA expression of NPR-A

(-58%, p=0.02, in patients without valvular disease, and -62%, p<0.01, in patients with

valvular disease). The mRNA content of pro-ANP was only increased in patients with

valvular disease (+12%, p=0.03). No changes were observed in patients with paroxysmal

AF. Conclusion: This study demonstrates that persistent AF, but not paroxysmal AF, induces

alterations in gene expression on the level of the atrium of pro-BNP and NPR-A. Although

AF is generally associated with an increase of plasma ANP, a change in mRNA content of

pro-ANP is only observed in the presence of concomitant valvular disease and is of minor

magnitude.

86

Chapter 6

Introduction

The pathophysiological mechanisms contributing to the perpetuation of atrial

fibrillation (AF), e.g. electrical remodeling and (ultra)cellular changes, are slowly being

unravelled.1-5 Abnormal calcium handling by the atrial myocyte,6-10 in response to

tachycardia induced intracellular calcium overload,11-13 is of pivotal importance in this

respect. However, the signaling pathways involved in these processes remain to be clarified.

The cardiac natriuretic peptide system plays an important role in

maintaining volume homeostasis via renal and cardiovascular actions, especially in

conditions that affect hemodynamics.14,15 In response to cardiac volume or pressure overload,

either caused by valvular disease, left ventricular dysfunction, hypertension, or AF, resultant

atrial stretch induces cardiac natriuretic peptide production. Data in AF, however, pertain

mainly to circulating atrial natriuretic peptide (ANP).16-19 Studies about brain natriuretic

peptide (BNP) in relation to AF are sparse and observational,20,21 merely stating that AF

(without specification) is one of the factors explaining plasma BNP level. ANP and BNP

modulate cardiac calcium handling indirectly via the autonomic nervous system,22 and

directly through cyclic guanosinemonophosphate (cGMP) mediated pathways (via the

cardiac natriuretic receptor), leading to reduced intracellular concentrations of cyclic

adenosinemonophosphate (cAMP), in turn lowering cAMP-dependent ion channel

activation.23 Among others, the L-type calcium channel becomes inactivated, whilst the

acetylcholine dependent potassium channel (KACh) becomes potentiated. Furthermore, ANP

stimulates calcium efflux from isolated atrial myocytes,24 underscoring the involvement

of natriuretic peptides in cellular calcium handling.

We hypothesized that AF would enhance gene expression of pro-atrial natriuretic

peptide (pro-ANP) and pro-brain natriuretic peptide (pro-BNP), and attenuate gene

expression of the cardiac natriuretic peptide receptor type-A (NPR-A) on the level of the

atrium. Therefore, it was our aim to investigate alterations in mRNA (messenger ribonucleic

acid) expression of pro-ANP, pro-BNP, and their NPR-A in the right atrial appendage

(RAA) of patients with paroxysmal and persistent AF undergoing cardiac surgery. To single

out the effect of AF, case matched control patients in SR were used. Since hemodynamic

overload of the heart per se is known to cause changes in natriuretic peptides,25-27 patients

with and without valvular disease were investigated.

Methods

Study patients

Patients were enrolled by using the in-house waiting list for cardiac surgery. The day

before elective cardiac surgery, the clinical characteristics of each patient (including

medication use and exercise tolerance according to the New York Heart Association

classification) were assessed by one investigator (AET). Patients with renal dysfunction

87


(serum creatinine > 150 µmol/L) were excluded from the study. Presence, type, and duration

of AF were assessed based on the patient’s history and previous electrocardiograms.

Echocardiographic data were obtained while patients were on the waiting list, but within

3 months prior to cardiac surgery. RAA was obtained from 36 patients with (paroxysmal

or persistent) AF and from 36 controls in sinus rhythm who were optimally case matched

for age, sex, and left ventricular function. During cardiac surgery, RAA was removed,

immediately snap-frozen in liquid nitrogen, and stored at -850C. The RAA’s were analyzed

separately in two series: one series consisting of patients without valvular disease, and

another series with patients with mitral or aortic valve disease of hemodynamic significance

to such extent that valvular surgery was deemed indicated. The investigation conforms to

the principles outlined in the Declaration of Helsinki, was approved by the Institutional

Review Board, and written informed consent was given by all patients.


Total RNA (ribonucleic acid) was isolated from RAA’s using the method of acid

guanidinium thiocyanate/phenol/chloroform extraction followed by a RNeasy kit for RNA

minipreps for tissues (Qiagen). The amount of RNA was evaluated by absorption at 260

nm, using a GeneQuant II (Pharmacia LKB Biotechnology, The Netherlands). The ratio

of absorption (260-280 nm) of all preparations was between 1.8 and 2.0. First strand

cDNA (copy-deoxyribonucleic acid) was synthesized by incubation of 1 µg of total RNA,

reverse transcription 10x buffer and 200 ng of random hexamers with 200 units of Moloney

Murine Leukemia Virus Reverse Transcriptase, 1mM of each dNTP

(dinucleotidetriphosphate) and 1 unit of ribonuclease (RNAse) inhibitor (Promega, The

Netherlands) in 20 µl. The synthesis reaction lasted 10 minutes at 20°C, 20 minutes at

42°C, 5 minutes at 99°C and 5 minutes at 4°C, respectively. All the products were checked

on contaminating DNA (data not shown).

Semiquantitative polymerase chain reaction analyses

We decribed these methods before.28,29 Validation of the present semi-quantitative

polymerase chain reaction (PCR) was performed by determination of the absolute alterations

of mRNA.28,29 In short, the cDNA of interest and the cDNA of the ubiquitously expressed

housekeeping enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were

coamplified in a single PCR. Primers were designed for pro-ANP, pro-BNP, the NPR-A

receptor, and the housekeeping gene GAPDH (Table 1). The oligonucleotides were

synthesized by Eurogentec (Belgium).

For the semi-quantitative PCR co-amplification 1 µl of cDNA mixture, 0.5 unit of

Taq polymerase (Eurogentec, Belgium) was added to 17.5 nM of dNTP’s, 10x PCR buffer

provided with Taq polymerase, 2.5 mM MgCl2

, 40 pmol of sense and antisense primer for

88

Chapter 6

the gene of interest, 40 pmol of sense and antisense GAPDH primer and water to bring the

final volume to 50 µl. All reaction mixtures were overlaid with 50 µl of mineral oil (Sigma,

The Netherlands). After 3 min denaturation at 94°C, n cycles of amplification (Table 1)

were performed, each for 1 min at 94°C , 1 min at annealing temperature (Table 1), 1 min

at 72°C, using the thermocycler Perkin Elmer 480 (The Netherlands). After the last cycle,

the 72°C elongation step was extended to 5 min. The PCR products were separated on a

1.5% agarose gel by gel-electrophoresis and stained with ethidium bromide. The densities

of the PCR products were quantified by densitometry (Aldus PhotoStyler 2.0, Graphic

Workshop and ImageQuant Version 3.3). Linearity for the PCR reactions was established

by making a correlation between the number of cycles and the ratio of the densities of

gene of interest / GAPDH (data not shown).

Table 1. Primer sequences

Sequence Cycles Annealing

Temperature


F 5'-CCC ATC ACC ATC TTC CAG GAG CG-3' 26 -


Pro-ANP:

F 5'-CCA TGT ACA ATG CCG TGT CC-3' 26 56 °CR 5'-GCT CCA ATC CTG TCC ATC CT-3'

pro-BNP:

F 5'-GTT ACA GGA GCA GCG CAA CC-3' 26 56 °CR 5'-AGG CCA CTG GAG GAG CTG AT-3'

NPR-A receptor:

F 5'-CTT GCT CGG CAT TCT GAT TG-3' 26 56 °CR 5'-CAC GCA GTT GGA TGA CTT GA-3

ANP= atrial natriuretic peptide, BNP= brain natriuretic peptide,

NPR-A= natriuretic receptor type-A

Definitions

Persistent AF: continuous presence of AF until the moment of cardiac surgery, i.e.

at least two consecutive electrocardiograms with AF more than 24 hours apart, without

intercurrent SR. Persistent AF does not spontaneously convert to SR, but is considered

cardiovertible.30 Previously, this type of AF was classified as chronic AF.

Paroxysmal AF: AF typically occurring in episodes of shorter duration than 24 hours

(though paroxysms may occasionally last longer) with intermittent sinus rhythm.

Paroxysmal AF either converts spontaneously or can be terminated with an intravenously

administered antiarrhythmic

89


drug.30 Due to its spontaneous character, paroxysmal AF might be present at the moment

of cardiac surgery. The intensity of the arrhythmia was scored using a recently proposed

classification.31

Statistical analysis

All PCRs were performed twice. Mean values of the ratios are presented. Unless

stated otherwise, mean values and standard deviations are reported. For the comparison

between groups, a Student’s t-test was used for normally distributed variables and a

Wilcoxon-Mann-Whitney test for non-normally distributed variables. In case of categorical

variables, a chi-square test with continuity correction or a Fisher’s exact test was used,

when appropriate. For determination of correlation, the Spearman correlation test was

used. A two-sided probability level < 0.05 was considered to indicate statistical significance.

The analysis was performed by SAS statistical software (SAS, version 6.12, Cary, NC).

Results

Patients without valvular disease

Eight patients with paroxysmal AF and 10 patients with persistent AF were included.

These two groups were compared with their case matched control groups in sinus rhythm

(Table 2). The distribution of underlying heart disease and type of surgery in the AF groups

differed from the control groups; most AF patients (6 and 6 patients, respectively) underwent

atrial arrhythmia surgery (Cox’s MAZE III procedure32) for intractable, symptomatic AF

whilst all control patients underwent coronary bypass surgery. Despite these unavoidable

differences, patient groups were otherwise comparable for left ventricular function, atrial

dimensions, and functional class for exercise tolerance according to the New York Heart

Association classification (NYHA class I and II).

In the persistent AF group, the median duration of AF before surgery was 16 months,

with a range of 8 months to 64 months. In the paroxysmal AF group, the median duration

of sinus rhythm before surgery was 1.5 days, and the median frequency of paroxysms was

once a day with a median duration

of 3 hours. These patients can be categorized as the most severe type of

paroxysmal AF (type IIIc; symptomatic, > 1 attack / 3 months under treatment). Of note,

3 patients with paroxysmal AF were in AF at the moment of harvesting of the RAA during

surgery.

Changes in transcription of the genes of interest were determined by comparison of

gene of interest / GAPDH cDNA ratios between the paroxysmal AF group and the control

group (SR1), and between the persistent AF group and the control group (SR2) (Figure 1).

Patients with persistent AF showed a significant change of the cDNA ratios of pro-BNP /

GAPDH (increase, +66%, p=0.04, Figure 1B) and NPR-A / GAPDH (decrease, -58%,

90

Chapter 6

p=0.02, Figure 1C). No significant correlation was found between the duration of persistent

AF and the cDNA ratios of pro-BNP / GAPDH or NPR-A / GAPDH. In contrast to the

patients with persistent AF, no changes were observed in the paroxysmal AF group.

Patients with valvular disease

By coincidence, no patients with paroxysmal AF were available. Eighteen patients

with persistent AF were included, and compared with their case matched control patients

Table 2. Characteristics of the 36 patients without valvular disease

Chacteristic PaAF SR1 PeAF SR2

Patient number 8 8 10 10

Male / female (n) 6/2 6/2 6/4 6/4

Age (years) 51±7 56±11 63±11 65±8

Cardiac surgery

Coronary bypass grafting (n) 2 † 8 4 10

Cox’s MAZE III procedure (n) 6 † 0 6 0

Underlying heart disease **

Coronary artery disease (n) 2 † 8 4 10

Hypertension (n) 1 1 3 2

Lone AF (n) 6 † - 5 -

Concomitant systemic disease

Diabetes (n) 0 1 0 3

COPD (n) 1 2 1 4

Rhythm characteristics

Duration PeAF (months) * - - 16 (8-64) -

Duration SR (days) * 1.5 (0-30) - - -

NYHA class I / II / III

For exercise tolerance (n) 7/1/0 5/3/0 6/4/0 5/5/0

For angina (n) 6/0/2 † 1/2/5 6/1/3 † 1/4/5

Echocardiographic parameters

LA long axis view (mm) 43±7 41±3 45±7 44±5

LA apical view (mm) 60±6 64±3 63±4 64±6

RA long axis view (mm) 54±9 54±4 62±7 57±4

LVEDD (mm) 48±4 49±8 53±3 53±6

LVESD (mm) 35±4 35±7 33±6 35±4

Medication

ACE inhibitor (n) 0 1 4 2

Betablocker (n) 1 † 5 3 6

Calcium entry blocker (n) 0 3 3 3

Digoxin (n) 0 1 5 3

* = values are presented as median with range. ** = per patient, more than one underlying disease might

have been present. † = p value < 0.05 compared to the control group.

ACE = angiotensin converting enzyme, ASD = atrial septal defect, COPD = chronic obstructive pulmonary

disease, LA = left atrium, LVEDD = left ventricular end diastolic diameter, LVESD = left ventricular end

systolic diameter, ND = not done, NYHA = New York Heart Association, PaAF = paroxysmal atrial fibrilla-

tion, PeAF = persistent atrial fibrillation, RA = right atrium, SR = sinus rhythm, SR1 = are sinus rhythm

control patients for PaAF, SR2 = are sinus rhythm control patients for PeAF.

91


Figure 1. cDNA ratios for pro-ANP / GAPDH (Figure 1a), pro-BNP / GAPDH (Figure 1b), and NPR-A /

GAPDH (Figure 1c) of individual patients without valvular disease. SR1 are sinus rhythm control patients for

the patients with PaAF (paroxysmal atrial fibrillation), SR2 are sinus rhythm control patients for the patients

with PeAF (persistent atrial fibrillation). All data are represented in density units / density units. • = individual

value, = mean value of group ± standard error of the mean.

Figure 2. cDNA ratios for pro-ANP / GAPDH (Figure 2a), pro-BNP / GAPDH (Figure 2b), and NPR-A /

GAPDH (Figure 2c) of individual patients with valvular disease. SR are sinus rhythm control patients, PeAF are

the patients with persistent atrial fibrillation. All data are represented in density units / density units. • = indi-

vidual value, = mean value of group ± standard error of the mean.

Figure 3. Typical agarose gel showing the cDNA levels of pro-BNP and GAPDH of patients with valvular

disease. SR are sinus rhythm control patients for the patients with PeAF (persistent atrial fibrillation).

92

Chapter 6

Characteristic PeAF SR

Patient number 18 18

Male / female (n) 11/7 14/4

Age (years) 70 ±9 66 ±11

Cardiac surgery **

Valvular surgery (n) 18 18

Coronary bypass grafting (n) 7 6

ASD closure (n) 2 0

Primary heart disease

Aortic stenosis (n) 6 8

Aortic regurgitation (n) 0 1

Aortic stenosis & regurgitation (n) 3 0

Mitral stenosis (n) 0 0

Mitral regurgitation (n) 7 3

Mitral stenosis & regurgitation (n) 1 3

Aortic and mitral valve disease (n) 1 3

Concomitant heart disease **

Coronary artery disease (n) 7 6

ASD (n) 2 0

Hypertension (n) 6

Concomitant systemic disease **

Diabetes (n) 4 2

COPD (n) 3 4


Duration PeAF (months) * 6 (0.5-240) -

NYHA class I-III

For exercise tolerance (n) 1/1/16 1/6/11

For angina (n) 13/3/2 10/4/4


LA long axis view (mm) 52±10 ‡ 44±5

LA apical view (mm) 72±9 66±7

RA long axis view (mm) 64±5 57±4

LVEDD (mm) 54±9 60±10

LVESD (mm) 38±7 43±15

Hemodynamic parameters

LV pressure max (mmHg) 167±40 178±64

LV pressure min (mmHg) 5±9 3

Wedge pressure (mmHg) 16±6 12±9

RA pressure (mmHg) 5±5 3±2

LV function I-III (n) † 13/1/3 14/0/4

Medication

ACE inhibitor (n) 11 ‡ 1

Betablocker (n) 3 4

Calcium entry blocker (n) 4 2

Digoxin (n) 14 ‡ 3

* = values are presented as median with range. ** = per patient, more than one surgical procedure might

have been performed or more than one concomitant disease might have been present. † = I indicates

normal, II indicates slightly impaired, and III indicates moderately impaired. ‡ = p value < 0.05 compared

to the control group.

LV = left ventricular, SR = are sinus rhythm control patients for PeAF. Other abbreviations are the same as

in Table 2.

Table 3. Characteristics of the 36 patients with valvular disease

93


in sinus rhythm (Table 3). Types of underlying heart disease, concomitant heart disease,

concomitant systemic disease, and surgery were equally distributed among both groups.

Patients were also comparable for NYHA functional class, left ventricular function, and

hemodynamic parameters. However, in the persistent AF group left atrial dimension (long

axis view) was larger, and angiotensin converting

enzyme inhibitors and digoxin were used more frequently. Clearly, the latter was used for

rate-control during AF. The median duration of AF before surgery was 6 months, with a

range of 0.5 months to 240 months.

Changes in transcription of the genes of interest were determined by comparison of

gene of interest / GAPDH cDNA ratios between the persistent AF group and the control

group (SR) (Figure 2). A representative agarose gel with cDNA levels of pro-BNP and

GAPDH is shown in Figure 3. Patients with persistent AF showed a significant change of

the cDNA ratios of all the genes of interest; a minor increase of pro-ANP / GAPDH (+12%,

p=0.03, Figure 2A), a substantial increase of pro-BNP / GAPDH (+69%, p<0.01, Figure

2B) and substantial decrease of NPR-A / GAPDH (-62%, p<0.01, Figure 2C). Also for the

patients with valvular disease, no significant correlation was found between the duration

of persistent AF and the cDNA ratios of pro-ANP / GAPDH, pro-BNP / GAPDH or NPR-

A / GAPDH.

Discussion

The main findings of the study are that persistent AF induces evident changes in

mRNA content of pro-BNP and NPR-A. The extent to which these changes occur are of

the same magnitude for patients with and without valvular disease. A change in mRNA

content of pro-ANP is only observed in the presence of concomitant valvular disease and

is of minor magnitude. Paroxysmal AF, although almost occurring daily in the present

patient group and therefore of clinical importance (6 of the 8 patients with paroxysmal AF

underwent Cox’s MAZE III procedure), was not associated with any change in expression

of the genes of interest.

The cardiac natriuretic peptides ANP (28-amino-acid peptide) and BNP (32-amino-

acid peptide) are produced in myocardium in the precursor forms pro-ANP and pro-BNP

that are spliced into an inactive and active peptide (N-terminal ANP / ANP and N-terminal

BNP / BNP respectively), which are released into the circulation.14 The third, more recently

discovered member of the natriuretic peptide family, C-type natriuretic peptide or CNP

(22-amino-acid peptide) is not considered to be produced in significant amounts in the

myocardium (not even in disease states),33 and was therefore not investigated in the present

study. ANP is generally considered to be primarily of atrial origin.14 BNP, on the other

hand, is mainly produced in the ventricles, and is considered to reflect left ventricular

function.34,35 ANP and BNP exercise their effects on myocardial electrophysiology and

94

Chapter 6

contractility indirectly via the autonomic nervous system, and directly via the natriuretic

receptor type-A (NPR-A), which is the principal receptor for natriuretic peptides in the

heart.22,23,36 NPR-A activates cGMP release, which in turn initiates activation of intracellular

protein kinase G.23 The pathways by which the natriuretic peptides modulate cellular

electrophysiology, calcium handling, and contractility are complex.23,24,37,38 The eventual

effects are most well described for ANP, being a decrease in intracellular calcium loading,

shortening of the action potential, an increase in conduction velocity, and a decrease in

phase 4 atrial depolarization, thus enhancing myocyte relaxation properties.22

mRNA expression of pro-ANP was unchanged in patients without valvular disease,

although the occurrence of AF is associated with a rise in plasma ANP.39,40 It should be

emphasized, however, that these observations pertain to acute AF. It is conceivable that

enhanced release of ANP from the atria, but not an altered gene expression, is responsible

for the rise in plasma ANP. A minor augmentation in mRNA content of pro-ANP was

observed only in the presence of valvular disease. In these patients, the hemodynamic

data of the AF group were comparable with the control group, but echocardiography showed

a significantly larger left atrium in the AF group, indicating higher atrial volume. This

difference was not observed in the patients without valvular disease, and might explain

the increased mRNA expression of pro-ANP in patients with valvular disease. This

possibility is supported by studies in patients with heart failure in sinus rhythm that indicate

the importance of atrial volume (stretch) for ANP release into the circulation.27 It should

also be noted that the (median) duration of AF was shorter in the presence of valvular

disease (shorter than 8 months in 10 of the 18 patients), as compared to group without

valvular disease (at least 8 months in all patients). Temporal depletion effects of plasma

ANP during the time course of longstanding AF (in a matter of months) have been

described.16,18 Pro-ANP gene expression might have been influenced the same way, thus

contributing to the differences in gene expression between patients with and without valvular

disease.

In contrast to pro-ANP, mRNA content of pro-BNP was increased irrespective of

valvular disease (+66% versus +69%, respectively). This is a surprising finding suggesting

that pro-BNP production in RAA is affected by the fibrillatory activity per se (a “load-

independent mechanism”), rather than by atrial stretch resulting from hemodynamic

overload. Analogous, during heart failure, the presence of atrial tissue-BNP is much more

pronounced as compared to ANP.26 Therefore, independent of the type of heart disease,

the atria might have the capability to vastly enhance BNP production, but not ANP

production. Another issue is that BNP seems to have a different atrial processing and

release in the circulation, as compared to ANP, especially during heart failure.26,41 The

latter might also be partly responsible for the differences in gene expression of pro-BNP

versus pro-ANP in response to AF. Only recently, a correlation between mRNA levels of

95


BNP and ANP, changing concomitantly with mean right atrial pressure, was reported for

right atrial appendage specimens of patients undergoing cardiac surgery, suggesting a

common regulation of tissue BNP and ANP.42 It should be noted, however, that patients

with and without valvular disease were mixed in this study, and that no data were given on

atrial rhythm.

Our study demonstrates that mRNA content of NPR-A is downregulated in patients

with persistent AF, either in the absence or presence of hemodynamic overload, i.e. valvular

disease. Receptor downregulation could be a result of an increase in signaling agonists, in

this case plasma ANP and BNP. mRNA expression of pro-BNP is clearly increased in the

present study, which would be in line with the observed downregulation of mRNA content

of NPR-A. Similar to the alterations of mRNA expression of pro-BNP, AF itself is likely

to be responsible for changes in mRNA content of NPR-A, because the magnitude of

change is comparable for patients with and without valvular disease (-58% versus -62%,

respectively).

In patients with frequently recurring attacks of paroxysmal AF under antiarrhythmic

treatment, to be assessed as a relevant arrhythmia burden, no changes in gene expression

of pro-ANP, pro-BNP, or NPR-A were found. It should be noted that most patients were in

sinus rhythm at the moment of the operation, with a median duration of 1.5 days (range 0-

30 days). These data suggest that intercurrent sinus rhythm between the attacks of

paroxysmal AF was enough to protect against gene alterations in the natriuretic peptide

system.

Limitations of the study

The present study has several important limitations. Firstly, atrial mRNA expression

was investigated without protein expression, binding experiments, ventricular mRNA

expression, or plasma level determinations. Such additional data might have provided

more insight in the pathophysiology. Secondly, patients with valvular disease used

“unloading” drugs, aimed at lowering of cardiac pressures and clinical stabilization. These

drugs, and differences in use of these drugs, might have interfered with the results, but the

clinical condition of most patients did not allow discontinuation (of more than 5x half-

time life) for study purposes.

Acknowledgement

Dr. Van Gelder was supported by Grant 94.014 from The Netherlands Heart

Foundation, The Hague, The Netherlands. The study was supported by Grant 96.051 from

The Netherlands Heart Foundation, The Hague, The Netherlands. We are indebted to Pieter

J. de Kam for helping us with the statistical analysis of the data.

96

Chapter 6

References

1. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA: Atrial fibrillation begets atrial fibrillation. A study in


2. Goette A, Honeycutt C, Langberg JJ: Electrical remodeling in atrial fibrillation. Time course and mechanisms.

Circulation 1996;94:2968-2974.

3. Wijffels MC, Kirchhof CJ, Dorland R, Power J, Allessie MA: Electrical remodeling due to atrial fibrillation

in chronically instrumented conscious goats: roles of neurohumoral changes, ischemia, atrial stretch, and

high rate of electrical activation. Circulation 1997;96:3710-3720.

4. Ausma J, Wijffels MC, Thone F, Wouters L, Allessie M, Borgers M: Structural changes of atrial myocardium

due to sustained atrial fibrillation in the goat. Circulation 1997;96:3157-3163.

5. Ausma J, Wijffels MC, van Eys G, Koide M, Ramaekers F, Allessie M, Borgers M: Dedifferentiation of

atrial cardiomyocytes as a result of chronic atrial fibrillation. Am J Pathol 1997;151:985-997.

6. Leistad E, Aksnes G, Verburg E, Christensen G: Atrial contractile dysfunction after short-term atrial

fibrillation is reduced by verapamil but increased by BAY K8644. Circulation 1996;93:1747-1754.

7. Van Wagoner DR, Lamorgese M, Kirian P, Cheng Y, Efimov IR, Mazgalev TN: Calcium current density is

reduced in atrial myocytes isolated from patients in chronic atrial fibrillation. Circulation

1997;96:I-180(Abstract)

8. Daoud EG, Knight BP, Weiss R, Bahu M, Paladino W, Goyal R, Man KC, Strickberger SA, Morady F:

Effect of verapamil and procainamide on atrial fibrillation-induced electrical remodeling in

humans.Circulation 1997;96:1542-1550.

9. Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S: Ionic remodeling underlying action potential changes in

a canine model of atrial fibrillation. Circ Res 1997;81:512-525.

10. Tieleman RG, De Langen CDJ, Van Gelder IC, De Kam PJ, Grandjean JG, Bel KJ, Wijffels MC, Allessie

MA, Crijns HJGM: Verapamil reduces tachycardia-induced electrical remodeling of the atria. Circulation

1997;95:1945-1953.

11. Lee HC, Clusin WT: Cytosolic calcium staircase in cultured myocardial cells. Circ Res 1987;61:934-939.

12. Schouten VJ, Morad M: Regulation of Ca2+ current in frog ventricular myocytes by the holding potential,

c-AMP and frequency. Pflugers Arch 1989;415:1-11.

13. De Pauw M, Borgers M, Heyndrickx GR: Ultrastructural calcium distribution in cardiac myocytes after

48h of rapid pacing in dogs (abstr). Circulation 1996;94:I-604

14. Levin ER, Gardner DG, Samson WK: Natriuretic peptides. N Engl J Med 1998;339:321-328.

15. Wilkins MR, Redondo J, Brown LA: The natriuretic-peptide family. Lancet 1997;349:1307-1310.

16. Seino Y, Shimai S, Ibuki C, Itoh K, Takano T, Hayakawa H: Disturbed secretion of atrial natriuretic peptide

in patients with persistent atrial standstill: endocrinologic silence. J Am Coll Cardiol 1991;18:459-463.

17. Roy D, Paillard F, Cassidy D, Bourassa MG, Gutkowska J, Genest J, Cantin M: Atrial natriuretic factor

during atrial fibrillation and supraventricular tachycardia. J Am Coll Cardiol 1987;9:509-514.

18. Van den Berg MP, Crijns HJGM, Van Veldhuisen DJ, Van Gelder IC, De Kam PJ, Lie KI: Atrial natriuretic

peptide in patients with heart failure and chronic atrial fibrillation: role of duration of atrial fibrillation. Am

Heart J 1998;135:242-244.

19. Tuinenburg AE, Van Veldhuisen DJ, Boomsma F, Van den Berg MP, De Kam PJ, Crijns HJGM: Compari-

son of plasma neurohormones in congestive heart failure patients with atrial fibrillation versus patients

with sinus rhythm. Am J Cardiol 1998;81:1207-1210.

20. Wallen T, Landahl S, Hedner T, Nakao K, Saito Y: Brain natriuretic peptide predicts mortality in the eld-

erly. Heart 1997;77:264-267.

21. Niinuma H, Nakamura M, Hiramori K: Plasma B-type natriuretic peptide measurement in a multiphasic

health screening program. Cardiology 1998;90:89-94.

22. Clemo HF, Baumgarten CM, Ellenbogen KA, Stambler BS: Atrial natriuretic peptide and cardiac electro-

physiology: autonomic and direct effects. J Cardiovasc Electrophysiol 1996;7:149-162.

23. Koller KJ, Goeddel DV: Molecular biology of the natriuretic peptides and their receptors. Circulation

1992;86:1081-1088.

24. Yoshizumi M, Houchi H, Tsuchiya K, Minakuchi K, Horike K, Kitagawa T, Katoh I, Tamaki T: Atrial

natriuretic peptide stimulates Na+-dependent Ca2+ efflux from freshly isolated adult rat cardiomyocytes.

FEBS Lett 1997;419:255-258.

25. Burnett JC, Jr., Kao PC, Hu DC, Heser DW, Heublein D, Granger JP, Opgenorth TJ, Reeder GS: Atrial

natriuretic peptide elevation in congestive heart failure in the human. Science 1986;231:1145-1147.

26. Wei CM, Heublein DM, Perrella MA, Lerman A, Rodeheffer RJ, McGregor CG, Edwards WD, Schaff HV,

Burnett JC, Jr. Natriuretic peptide system in human heart failure. Circulation 1993;88:1004-1009.

97


27. Globits S, Frank H, Pacher B, Huelsmann M, Ogris E, Pacher R: Atrial natriuretic peptide release is more

dependent on atrial filling volume than on filling pressure in chronic congestive heart failure. Am Heart J

1998;135:592-597.

28. Van Gelder IC, Brundel BJJM, Henning RH, Tuinenburg AE, Tieleman RG, Deelman LE, Grandjean JG,

De Kam PJ, van Gilst WH, Crijns HJGM: Alterations in gene expression of proteins involved in the cal-

cium handling in patients with atrial fibrillation. J Cardiovasc Electrophysiol 1999;10:552-560.

29. Brundel BJJM, Van Gelder IC, Henning RH, Tuinenburg AE, Deelman LE, Tieleman RG, Grandjean JG,

van Gilst WH, Crijns HJGM: Gene expression of proteins influencing the calcium homeostasis in patients

with persistent and paroxysmal atrial fibrillation. Cardiovasc Res 1999;42:443-454.

30. Gallagher MM, Camm AJ: Classification of atrial fibrillation. Pacing Clin Electrophysiol 1997;20:1603-1605.

31. Levy S, Breithardt G, Campbell RW, Camm AJ, Daubert J-C, Allessie M, Aliot E, Capucci A, Cosio F,

Crijns HJGM, Jordaens L, Hauer RN, Lombardi F, Luderitz B: Atrial fibrillation: current knowledge and

recommendations for management. The Working Group on Arrhythmias of the European Society of Cardi-

ology. Eur Heart J 1998;19:1294-1320.

32. Cox JL, Jaquiss RD, Schuessler RB, Boineau JP: Modification of the maze procedure for atrial flutter and

atrial fibrillation. II. Surgical technique of the maze III procedure. J Thorac Cardiovasc Surg

1995;110:485-495.

33. Takahashi T, Allen PD, Izumo S: Expression of A-, B-, and C-type natriuretic peptide genes in failing and

developing human ventricles. Correlation with expression of the Ca2+-ATPase gene. Circ Res 1992;71:9-17.

34. Richards AM, Nicholls MG, Yandle TG, Frampton C, Espiner EA, Turner JG, Buttimore RC, Lainchbury

JG, Elliott JM, Ikram H, Crozier IG, Smyth DW: Plasma N-terminal pro-brain natriuretic peptide and

adrenomedullin: new neurohormonal predictors of left ventricular function and prognosis after myocardial

infarction. Circulation 1998;97:1921-1929.

35. Nagaya N, Nishikimi T, Goto Y, Miyao Y, Kobayashi Y, Morii I, Daikoku S, Matsumoto T, Miyazaki S,

Matsuoka H, Takishita S, Kangawa K, Matsuo H, Nonogi H: Plasma brain natriuretic peptide is a bio-

chemical marker for the prediction of progressive ventricular remodeling after acute myocardial infarction.

Am Heart J 1998;135:21-28.

36. Lin X, Hanze J, Heese F, Sodmann R, Lang RE: Gene expression of natriuretic peptide receptors in myo-

cardial cells. Circ Res 1995;77:750-758.

37. Le Grand B, Deroubaix E, Couetil JP, Coraboeuf E: Effects of atrionatriuretic factor on Ca2+ current and ICa

-

independent transient outward K+ current in human atrial cells. Pflugers Arch 1992;421:486-491.

38. Rebsamen MC, Church DJ, Morabito D, Vallotton MB, Lang U: Role of cAMP and calcium influx in

endothelin-1-induced ANP release in rat cardiomyocytes. Am J Physiol 1997;273:E922-31.

39. Oliver JR, Twidale N, Lakin C, Cain M, Tonkin AM: Plasma atrial natriuretic polypeptide concentrations

during and after reversion of paroxysmal supraventricular tachycardias. Br Heart J 1988;59:458-462.

40 Christensen G, Leistad E: Atrial systolic pressure, as well as stretch, is a principal stimulus for release of

ANF. Am J Physiol 1997;272:H820-6.

41. Suzuki E, Hirata Y, Kohmoto O, Sugimoto T, Hayakawa H, Matsuoka H, Kojima M, Kangawa K, Minamino

N: Cellular mechanisms for synthesis and secretion of atrial natriuretic peptide and brain natriuretic peptide

in cultured rat atrial cells. Circ Res 1992;71:1039-1048.

42. Doyama K, Fukumoto M, Takemura G, Tanaka M, Oda T, Hasegawa K, Inada T, Ohtani S, Fujiwara T, Itoh

H, Nakao K, Sasayama S, Fujiwara H: Expression and distribution of brain natriuretic peptide in human

right atria. J Am Coll Cardiol 1998;32:1832-1838.

98

99

Endothelin-1 mRNA is upregulated in patients

Chapter 7

Endothelin 1 mRNA is Upregulated in Human Persistent

Atrial Fibrillation with Underlying Valve Disease

Brundel, Short title: Endothelin system in atrial fibrillation

Bianca J. J. M. Brundel, MSc1,2; Isabelle C. Van Gelder, MD1;

Anton E. Tuinenburg, MD1; Mirian Wietses2, Dirk J Van Veldhuisen, MD1;

Wiek H. Van Gilst, PhD2; Harry J. G. M. Crijns, MD1; Robert H. Henning, MD2

Departments of Cardiology1and Clinical Pharmacology2, Thoraxcenter University

Hospital Groningen, The Netherlands.

Submitted Journal of Cardiovascular ElectrophysiologyAbstract

Background: Activation of the endothelin system is an important compensatory

mechanism that is activated during left ventricular dysfunction. Whether this system also

plays a role at the atrial level during AF has not been thoroughly examined. The purpose

of this study was to investigate mRNA and protein expression levels of the endothelin

system in AF patients with and without concomitant underlying heart disease. Metodsand results: Right atrial appendages of 36 patients with either paroxysmal or persistent

AF were compared with 36 controls in sinus rhythm. The mRNA amounts of pro-endothelin-

1 (ET-1), endothelin receptors A (ET-A) and B (ET-B) were studied by semi-quantitative

PCR. Protein amounts of the receptors were investigated by slot-blot analysis. The

endogenous mRNA production of pro-ET-1 was induced (+ 40%, p=0.002) in patients

with persistent AF and underlying valve disease. Furthermore, the ET-A and ET-B receptor

protein amounts were significantly reduced in paroxysmal AF (-39% and –47%,

respectively) and persistent AF with (-28% and –30%, respectively) and without (-20%

and –40%, respectively) underlying valve disease. Moreover, the mRNA amounts for pro-

ET-1 and ET-A were not different in AF patients, in contrast to ET-B mRNA amounts

which were significantly reduced in persistent AF with (-30%, p<0.001) and without (-

30%, p=0.04) underlying valve disease. Conclusions: Alterations in gene expression of

the endothelin system occur in the atria during AF, especially in the presence of underlying

valve disease. These results suggest that the endothelin system might play a role in adaptive

mechanisms in these patients.

100

Chapter 7

Introduction

There exists a reciprocal relation between congestive heart failure and atrial fibrillation

(AF) (1). Congestive heart failure may lead to AF. Also, AF may promote the occurrence

of heart failure. The endothelin (ET) system plays a role in the pathophysiology of heart

failure (2). Although, it is unknown yet whether the ET system plays a role in the

pathophysiology of AF and may mediate the atrial remodeling proces. Apart from their

direct inotropic effect on the myocardium, ET-1 increases intracellular calcium

concentrations, via activation of the L-type Ca2+ channel, and cell growth (3,4). Several

studies demonstrated that intracellular calcium overload played a crucial role in the atrial

electrical and contractile remodeling (5-7). Moreover, experimental data revealed that the

L-type Ca2+ channel plays a main role in shortening of AERP and action potential duration

(APD) (8,9). There are indications that endothelin plays also a role in human AF, as

suggested by elevated ET-1 plasma concentrations found in patients with AF with

concomitant heart failure (10,11). However, the atrial expression levels of ET-1 and the

endothelin receptor A and B in human AF has not been thoroughly examined.

The present study was designed to investigate the mRNA and protein amounts of

endogenous pro ET-1 and their receptors ET-A/ET-B, in right atrial appendages of patients

with AF. Included were patients with paroxysmal and persistent AF with and without

underlying valve disease and compared to patients in sinus rhythm undergoing cardiac

surgery.


Patient selection and atrial tissue collecting


of the patient as previously described (12). Presence, type and duration of AF were assessed

by patient’s complaints and previous electrocardiograms. In addition, medication use and

exercise tolerance (according to the NYHA classification) was determined. Right atrial

appendages (RAAs) were gained from 10 patients with persistent AF without valvular

disorders and from 8 patients with paroxysmal AF. All patients were euthyroid. As controls

18 clinically stable patients in sinus rhythm undergoing CABG were used. The study was

approved by the Institutional Review Board and written informed consent was given by

all patients. Immediately after excision, the RAAs were snap-frozen in liquid nitrogen and

stored at -85 °C.


Total RNA was isolated and processed as described previously (12). Briefly, first

strand cDNA was synthesized by incubation of 1 µg of total RNA in reverse transcription

10x buffer, 200 ng of random hexamers with 200 units of Moloney Murine Leukemia

101


Virus Reverse Transcriptase, 1mM of each dNTP and 1 unit of RNase inhibitor (Promega,

The Netherlands) in 20 µl total volume. Synthesis reaction was performed for 10 minutes

at 20 °C, 20 minutes at 42 °C, 5 minutes at 99 °C and 5 minutes at 4 °C. All the products

were checked for contaminating DNA.


We described and validated these methods before (12). In short, the cDNA of interest

and the cDNA of the ubiquitously expressed housekeeping gene glyceraldehyde-3-

phosphate dehydrogenase (GAPDH) were co-amplified in a single PCR. Primers

(Eurogentec, Belgium) were designed for pro-ET-1, ET-A and ET-B receptors and the

housekeeping gene GAPDH (Table 1).


ethidium bromide. The density of the PCR products was quantified by densitometry.

Linearity of the PCR was established by a good correlation between the number of cycles

and the density of gene of interest and GAPDH (data not shown).

Protein Preparation and Slot Blotting

From a number of patients there was enough tissue to perform protein isolation and

slot-blot analysis. Two different patient groups were made. The first group consisted of 6

patients with paroxysmal AF without valve disease, 9 patients with persistent AF without

valve disease and 6 controls in sinus rhythm. The second group consisted of 7 patients


protein sequence cycles annealing

temp (°C)

GAPDH F 5'-CCC ATC ACC ATC TTC CAG GAG CG-3', var. var.

R 5'-GGC AGG GAT GAT GTT CTG GAG AGC C-3'.

Pro-ET-1 F 5’-TAC TTC TGC CAC CTG GAC AT-3’ 30 56

R 5’-CTT CCT CTC ACT AAC TGC TG-3’

ET-A F 5’-CAC GAT GAG GCT CAG GAT GG-3’ 29 56

R 5’-CTT CCT CTC ACT AAC TGC TG -3’

ET-B F 5'-CCG CAG AGA TAA TGA CGC CA-3' 29 56

GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Pro-ET-1, gene encoding the endothelin pro-peptide;

ET-A, gene encoding the endothelin receptor type A; ET-B, gene encoding the endothelin receptor type B.

102

Chapter 7

with persistent AF with valve disease and 5 control patients in sinus rhythm with valve

disease. The frozen RAAs were homogenized in RadioImmunoPrecipitationAssay (RIPA)

buffer as described before (12). The homogenate was centrifuged at 14.000 rpm for 20

minutes at 4°C. The supernatant was used for protein concentration measurement according

to the Bradford method (Bio-Rad, The Netherlands) with bovine albumin used as a standard.

Samples of 10 µg heat denatured protein were spotted on nitrocellulose membranes

(Stratagene, The Netherlands) and checked by staining with Ponceau S solution (Sigma,

The Netherlands). After blocking in blocking buffer (5% nonfat milk, TBS and 0.1% Tween

20) and washing in TBS with 0.1% Tween 20 the membranes were incubated with primary

antibody against GAPDH (Affinity Reagents, USA), ET-A and ET-B (Research Diagnostics

Inc., USA). Immunodetection of the primary antibody was performed with peroxidase

conjugated secondary antibody anti mouse (for GAPDH) or anti sheep IgG (for ET-A and

ET-B, Santa Cruz Biotechnology, The Netherlands). The blot was incubated with the ECL-

detection reagent (Amersham, The Netherlands) for 1 minute, and exposed to an X-OMAT

x-ray film (Kodak, The Netherlands) for 15 to 90 seconds. The band densities were evaluated

by densitometric scanning using a Snap Scan 600 (Agfa, The Netherlands). The amount of

protein chosen was in the linear immunoreactive signal area and the specificity of the

antibody was checked by SDS-PAGE.


All PCR and slot-blot procedures were performed in duplo series and mean values

were used for statistical analysis. Comparison between groups for normally distributed

variables was performed by one-way ANOVA. For determination of correlations the

Spearman correlation test was used. The Mann-Whitney U-test was performed for group

to group comparisons of small numbers. All p-values were two-sided, a p-value <0.05 was

considered statistically significant. SPSS version 8.0 was used for all statistical evaluations.

Results

Patients

Table 2 and 3 show the characteristics of the patients. The distribution of underlying

heart disease and type of surgery in the AF groups differed from the control groups; most

AF patients underwent atrial arrhythmia surgery (Cox maze III procedure) for intractable

symptomatic AF, whereas control patients underwent coronary bypass surgery and valvular

surgery. Other than these unavoidable differences, patient groups were comparable for left

ventricular function, atrial dimensions, and functional class for exercise tolerance according

to the NYHA classification. Of note, three patients with paroxysmal AF were in AF at the

moment of harvesting the RAA during surgery. Unfortunately no patients with underlying

valve disease and a history of paroxysmal AF could be included in this study.

103


Table 2. Characteristics of the 36 patients without valvular disease

PAF SR1 CAF SR2

Male/female (n) 6/2 6/2 6/4 6/4

Age (years) 51 ± 7 56 ± 11 63 ± 11 65 ± 8

Cardiac surgery

Coronary bypass grafting 2 8 4 10

Cox’s maze III procedure 6 0 6 0

Underlying heart disease

Coronary artery disease 3 8 4 10

Lone AF 6 - 6 -


Duration CAF (months*) - - 16 (8-64) -

Duration SR (days*) 1.5 (0-30) - - -

NYHA Class I/II/III

For exercise tolerance 7/1/0 5/3/0 6/4/0 5/5/0


LA long-axis view (mm) 43 ± 7 41 ± 3 45 ± 7 44 ± 5

LA apical view (mm) 60 ± 6 64 ± 3 63 ± 4 64 ± 6

RA long-axis view (mm) 54 ± 9 54 ± 4 62 ± 7 57 ± 4

LVEDD (mm) 48 ± 4 49 ± 8 53 ± 3 53 ± 6

LVESD (mm) 35 ± 4 35 ± 7 33 ± 6 35 ± 4

Medication

ACE inhibitor (n) 0 1 4 2

Beta blocker (n) 1 5 3 6

Calcium entry blocker (n) 0 3 3 3

Digoxin (n) 0 1 5 3

* values are presented as median (range). La = left atrium; LVEDD = left ventricular end diastolic diameter;

LVESD = left ventricular end-systolic diameter; NYHA = New York Heart Association; CAF = chronic,

persistent atrial fibrillation; PAF = paroxysmal atrial fibrillation; RA = right atrium; SR1 = sinus rhythm

control patients for PAF; SR2 = sinus rhythm control patients for CAF.

Endothelin system mRNA levels

Changes in transcription of the genes of interest were determined by comparison

gene of interest/GAPDH mRNA ratios between paroxysmal AF group and control group

(SR1), between persistent AF group and control group (SR2) and between persistent AF

group with valve disease and controls with valve disease (SR VD).

Pro-ET-1 mRNA contents were significantly increased (+40%, p=0.002) only in

patients with persistent AF and concomitant valve disease compared to control patients in

sinus rhythm with valve disease (Table 4 and Figure 1A and B).

The ET-B mRNA contents were reduced in patients with persistent AF and valve

disease (-30%, p<0.001) and without valve disease (-30%, p=0.04, Figure 2C,D, Table 4).

104

Chapter 7

Furthermore the mRNA amounts of ET-A were not different between the patient groups

(Figure 2A and B, Table 4).

Paroxysmal AF patients without valve disease showed also no differences in mRNA

amounts for pro-ET-1, ET-A and ET-B (Figure 1B and 2A,B).

Table 3. Characteristics of the 36 patients with valvular disease

CAF SR

Male/female 11/7 14/4

Age (years) 70 ± 9 66 ± 11

Cardiac surgery

Valvular surgery 18 18

Coronary bypass grafting 7 6

Underlying heart disease

Aortic stenosis 6 8

Aortic regurgitation 0 1

Aortic stenois and regurgitation 3 0

Mitral regurgitation 7 3

Mitral stenosis and regurgitation 1 3

Coronary artery disease 7 6


Duration CAF (months)* 6 (0.5 - 240)

NYHA Class I-III

For exercise tolerance 1/1/16 1/6/11


LA long axis view (mm) 52 ± 10** 44 ± 5

LA apical view (mm) 72 ± 9 66 ± 7

RA long axis view (mm) 64 ± 5 57 ± 4

LVEDD (mm) 54 ± 9 60 ± 10

LVESD (mm) 38 ± 7 43 ± 15

Medication

ACE inhibitor (n) 11** 1

Beta blocker (n) 3 4

Calcium entry blocker (n) 4 2

Digoxin (n) 14** 3

* Values are presented as median (range). ** p<0.05 compared to the control group. ASD = atrial septal

defect; LA = left atrium; LVEDD = left ventricular end diastolic diameter; LVESD = left ventricular end-

systolic diameter; NYHA = New York Heart Association; CAF = chronic, persistent atrial fibrillation; RA =

right atrium; SR = sinus rhythm control patients for CAF.

105


Figure 1.(A) Typical example of an agarose gel. Here the pro-ET-1 and GAPDH are shown of two patients with

persistent AF and valve disease (CAF VD), one patient with persistent AF (CAF), one patient with paroxysmal

AF (PAF) and controls in sinus rhythm (SR). Individual cDNA ratios for pro-ET-1/GAPDH are given for pa-

tients without valve disease (B) and with valve disease (C). SR1 and SR2 are the control groups in sinus rhythm

for persistent AF and paroxysmal AF, respectively. SR VD is the sinus rhythm control group with valve disease

for persistent AF with valve disease. Values are mean ± SEM.

Figure 2.Individual cDNA ratios are given for ET-A/GAPDH for patients without valve disease (A) and with

underlying valve disease (B). Also the individual cDNA ratios for ET-B/GAPDH of patients without valve dis-

ease (C) and with valve disease (D) are shown. SR1 and SR2 are the control groups in sinus rhythm for persistent

AF (CAF) and paroxysmal AF (PAF), respectively. SR VD is the sinus rhythm control group with valve disease

for persistent AF with valve disease (CAF VD). Values are given as mean ± SEM.

106

Chapter 7

Protein remodeling

Changes in protein expression were studied in relation to protein levels of GAPDH

and the density of total amount of protein spotted on the membrane. Because the GAPDH

density and total protein amount density showed a significant correlation, we used the

protein of interest/GAPDH ratio for further investigation.

The protein expression of ET-A and ET-B were significantly reduced in paroxysmal

(-39%, p=0.02 and –47%, p=0.02, respectively) and persistent AF without underlying

valve disease (-20%, p=0.04 and –40%, p=0.03) and persistent AF with underlying valve

disease (-28%, p=0.03 and –30%, p=0.03, respectively) (Table 4, Figure 3A and B).

Discussion

The main findings of the study are that 1) endogenous mRNA production of pro-ET-

1 is induced in patients with persistent AF with underlying valve disease, 2) ET-A and ET-

B receptor protein amounts are reduced in paroxysmal and persistent AF, irrespective to

the presence of underlying valve disease and finally 3) a discrepancy between ET-A protein

reductions in the presence of unchanged mRNA amounts is found. The results reveal that

in the atria of patients with AF, alterations in the endothelin system occur. Especially the

induction of pro-ET-1 in AF patients with valve disease may be of significant importance.

This finding can give new insights in the adaptation mehanisms during AF with respect to

the underlying valve disease reflecting the role of heart failure in AF.

Endothelin-1

Cardiac myocytes (13) as well as vascular endothelial cells (14) produce ET-1, a 21

amino acid peptide, a potent vasoconstricting peptide and exerts important cardiac effects

(2). These include positive inotropic and chronotropic effects in the heart of various species,

including humans, and growth-promoting properties. Experimental studies showed that

Table 4. mRNA and protein remodeling in patients with paroxysmal AF (PAF) and persistent AF (CAF)

without valve disease (VD) and persistent AF (CAF) with valve disease.

AF without valve disease AF with valve disease

PAF CAF CAF

mRNA protein mRNA protein mRNA protein

ET-1 ns - ns - +40% -

ET-A ns -39% ns -20% ns -28%

ET-B ns -47% -30% -40% -30% -30%

ns means not significant

107


ET-1 trigger the increase of intracellular calcium (3), likely via the L-type calcium channel

in the atrial myocyte (4). It is well known that calcium overload plays a key role in the

pathophysiology of AF (5-7).

Elevated plasma concentrations of ET-1 were observed in patients with AF in the

setting of advanced heart failure (10,11). The found increase of endogenous pro-ET-1

mRNA levels in RAA of patients with persistent AF with underlying valve disease is in

accordance with the elevated plasma levels. Moreover this finding indicates that activation

of endothelin plays a role especially in patients with AF and concomitant moderate heart

failure. At the atrial myocytes the ET-1 could trigger elevation of intracellular calcium

amounts leading to AF induced contractile dysfunction (15,16) and electrical remodeling

(6,9,17) probably via the activation of calcium overload induced protease calpain (Brundel

et al., submitted)

A reduction of ET-B mRNA and no changes in ET-A mRNA amounts were found

during persisent AF irrespective to the underlying valve disease. This result might suggest

that the elevated endothelin acts on the ET-B receptor leading to reduction of its mRNA

SR PAF CAF SR VD CAF VD0

1

2

SR PAF CAF SR VD CAF VD0

1

2

ET-A protein expressionET-B protein expression

p=0.04p=0.02

p=0.03

p=0.03p=0.02

p=0.03

A

B

SR

CAF

PAF

SR

CAF

PAF

ET-

B p

rote

in e

xpre

ssio

nE

T-A

pro

tein

exp

ress

ion

Figure 3. Protein ratios of ET-A/GAPDH (A) and ET-B/GAPDH (B) of the individual patients. The top of each

panel shows a typical slot blot analysis of 10 µg protein of control patients (SR), persistent AF (CAF) and

paroxysmal AF (PAF). Values are given as mean ± SEM.

108

Chapter 7

expression. In this case the ET-B receptor seemed to play a dominant role in the RAA

compared to ET-A. However the ET-A receptor predominates in normal myocardium and

was found to play an important role during heart failure (2). This finding suggests that the

ET-A and ET-B receptors are differentially regulated on mRNA level during AF and heart

failure. Furthermore, the ET-B mRNA reduction in patients without valve disease can not

be explained by the endogenous elevation of pro-ET-1 and plasma ET-1 should be measured

to elucidate a possible exogenous induction of ET-1, which could trigger the ET-B mRNA

reduction.

Post-transcriptional regulation?

The observed discrepancy between alterations in mRNA and protein expression mainly

in patients with paroxysmal AF suggests the activation of proteolysis. Recently, we found

that activation of the calpain system in human persistent and paroxysmal AF, in the absence

of activation of the proteasome pathway (Brundel et al., submitted). As calpain are activated

by calcium overload in the myocard cell (5), calpain activation would serve to protect the

cells to additional damage by down-regulation of proteins. However, this would be at the

cost of proteolysis of several cytoskeletal, membrane-associated and regulatory proteins

(18,19). Whether interference with the calpain system represents a valuable therapeutic

strategy in AF remains to be investigated.

In conclusion

This report describes increased levels of pro-ET-1 mRNA in patients with persistent

AF with underlying valve disease in combination with reductions in protein levels of the

ET-A and ET-B receptor. This finding indicates that alterations in gene expression of the

endothelin system occur in the atria during AF in the presence of moderate heart failure.

The endothelin system could play a significant role in adaptive mechanisms in these patients.

Moreover, the observed discrepancy between mRNA and protein levels of these receptors,

mainly in patients with paroxysmal AF, suggest the influence of proteolytic system.

109


References

1. Murgatroyd F, Camm AJ. Atrial arrhythmias. Lancet 1993;341:1317-22.

2. Ponicke K, Vogelsang M, Heinroth, Becker K, Zolk O, Bohm M, Zerkowski HR, Brodde OE. Endothelin

receptors in the failing and nonfailing human heart. Circulation 1998;97:744-51.

3. Alvarez BV, Perez NG, Ennis IL, Camilion de Hurtado M, Cingolani HE. Mechanisms underlying the

increase in force and Ca(2+) transient that follow stretch of cardiac muscle: a possible explanation of the

Anrep effect. Circ Res 1999;85:716-22.

4. He JQ, Pi Y, Walker J, Kamp TJ. Endothelin-1 and photoreleased diacylglycerol increase L-type Ca2+

current by activation of protein kinase C in rat ventricular myocytes. J Physiol (Lond) 2000;524:807-20.

5. Ausma J, Dispersyn GD, Duimel H, Thone F, Ver Donck L, Allessie M, Borgers M. Changes in ultrastruc-

tural calcium distribution in goat atria during atrial fibrillation. J Mol Cell Cardiology 2000;32:355-64.


nisms. Circulation 1996;94(11):2968-74.

7. Leistad E, Aksnes G, Verburg E, Christensen G. Atrial contractile dysfunction after short-term atrial fibril-

lation is reduced by verapamil but increased by BAY K8644. Circulation 1996;93(9):1747-54.

8. Yue L, Melnyk P, Gaspo, Wang Z, Nattel S. Molecular mehanisms underlying ionic remodeling in a dog


9. Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S. Ionic remodeling underlying action potential changes in

a canine model of atrial fibrillation. Circ Res 1997;81(4):512-25.

10. Tuinenburg AE, Van Veldhuisen DJ, Boomsma, Van Den Berg M, De Kam PJ, Crijns HJGM. Comparison

of plasma neurohormones in congestive heart failure patients with atrial fibrillation versus patients with

sinus rhythm. Am J Cardiol 1998;81:1207-10.

11. Masson S, Gorini M, Salio M, Lucci, Latini R, Maggioni AP. Clinical correlates of elevated plasma natri-

uretic peptides and Big endothelin-1 in a population of ambulatory patients with heart failure. Ital Heart J

2000;1:282-8.

12. Brundel BJJM, Van Gelder IC, Henning RH, Tuinenburg AE, Deelman LE, Tieleman RG, Grandjean JG,

Van Gilst WH, Crijns HJGM. Gene expression of proteins influencing the calcium homeostasis in patients

with persistent and paroxysmal atrial fibrillation. Cardiovasc Res 1999;42:443-54.

13. Suzuki T, Kumazaki T, Mitsui Y. Endothelin-1 is produced and secreted by neonatal rat cardiac myocytes

in vitro. Biochem Biophys Res Commun 1993;193:823-30.

14. Yanagisawa M, Kurihara H, Kimura S, Tomobe Y, Kobayashi M, Mitsui Y, Yazaki Y, Goto K, Masaki T. A

novel potent vasocontrictor peptide produced by vascular endothelial cells. Nature 1988;332:411-5.

15. Daoud EG, Marcovitz P, Knight B, Goyal R, Ching Man K, Strickberger A, Armstrong WF, Morady F.

Short-term effect of atrial fibrillation on atrial contractile function in humans. Circulation 1999;99:3024-

7.

16. Manning WJ, Silverman DI, Katz SE, Riley MF, Come PC, Doherty RM, Munson JT, Douglas PS. Im-

paired left atrial mechanical function after cardioversion: relation to the duration of atrial fibrillation. J Am

Coll Cardiol 1994;23(7):1535-40.

17. Tieleman RG, De Langen CDJ, Van Gelder IC, De Kam PJ, Grandjean JG, Bel KJ, Wijffels MC, Allessie

MA, Crijns HJGM. Verapamil reduces tachycardia-induced electrical remodeling of the atria. Circulation

1997;95(7):1945-53.

18. Gorza L, Menabo R, Vitadello M, Bergamini CM, Di Lisa F. Cardiomyocyte troponin T immunoreactivity

is modified by cross-linking resulting from intracellular calcium overload. Circulation 1996;93(10):1896-

904.

19. Gao WD, Atar D, Liu Y, Perez NG, Murphy AM, Marban E. Role of troponin I proteolysis in the pathogen-

esis of stunned myocardium. Circ Res 1997;80(3):393-9.

Part IV

Calpain activation, a new adaptive

mechanism in AF

113

Activation of proteolysis by calpain

Chapter 8

Activation of Proteolysis by Calpains in Human

Paroxysmal and Persistent Atrial Fibrillation

Brundel, Short title: Calpain activation in Atrial Fibrillation

Bianca J.J.M. Brundel, MSc1,2, Isabelle C. van Gelder, MD2, Harry J.G.M.

Crijns MD2, Wiek H. van Gilst PhD1, Robert H. Henning, MD1.

Department of Clinical Pharmacology1, Groningen University Institute for Drug

Exploration (GUIDE) and Department of Cardiology2, Thoraxcenter University Hospi-

tal, Groningen, The Netherlands.

Submitted Circulation Research

Abstract

Background: In human paroxysmal atrial fibrillation (AF), we observed reductions

of ion-channel proteins in the presence of unchanged mRNA levels. As this suggests

activation of proteolysis, we investigated two main proteolytic pathways, calpain and the

proteasome in atrial tissue of AF patients. Methods and Results: Right atrial appendages

were obtained from patients with paroxysmal (n=7) or persistent (n=10) lone AF and

compared to controls (n=10) in sinus rhythm undergoing coronary artery bypass grafting

(CABG). Proteolysis was measured using a fluorogenic substrate and expression of calpain

I and II was determined by Western-blot. Proteolytic activity was significantly increased

in paroxysmal and persistent AF, and abolished by the calpain inhibitor E-64, but unaffected

by the proteasome inhibitor lactacystin. Protein expression of calpain I was increased by

35% in persistent AF, while expression of calpain II was unchanged in AF. Tissue calpain

activity showed a significant correlation with protein levels of calpain I (r=0.61, p<0.001),

but not with calpain II levels.

Conclusions: Increased proteolytic activity during paroxysmal and persistent lone AF is

due to activation of the calpain pathway, especially of calpain I. Calpain activation may

represent an important mechanism conveying cellular changes underlying

electrophysiological, contractile and structural remodeling in AF.

114

Chapter 8

Introduction

Human atrial fibrillation (AF) is characterized by heterogeneity in electrical activation

pattern and the loss of contractile function of atrial tissue. Recent experimental research

has demonstrated AF to induce changes at the electrophysiological, protein and

morphological level, which increase the vulnerability to AF.1-3 Still, the molecular

mechanisms underlying these changes are poorly characterized. In a previous human study

we observed a reduction in protein expression of several plasma membrane ion channels

in the absence of changes in mRNA in patients with lone paroxysmal AF.4 Therefore, we

hypothesize activation of a protein degradation mechanism during AF. Different proteolytic

pathways may be involved. First, as cytosolic calcium is increased during AF2, proteolysis

may be invoked by calcium dependent neutral proteases, calpain I and calpain II, whose

activity was demonstrated in animal models of metabolic inhibition, cardiac stunning and

calcium overload.5-7 Alternatively, activation of the proteasome may underlie increased

proteolysis in cardiac cells, as shown in rat for degradation of myosin heavy chain and

connexin43 gap junctions.8,9

To examine proteolytic activation in human AF, we determined the activity of calpains

and the proteasome in atrial tissue of patients with paroxysmal and persistent lone AF and

of controls in sinus rhythm. As an increased calpain-mediated proteolysis was found in

AF, protein levels of calpain I and II were determined.


Patients and tissue collection

Right atrial appendages (RAAs) were obtained as described before10 from patients

with normal left ventricular function. Patients with paroxysmal (n=7) and persistent (n=10)

lone AF undergoing MAZE surgery were included and matched to clinically stable control

patients in sinus rhythm undergoing CABG (n=10, Table 1). The Institutional Review

Board approved of the study and patients gave written informed consent.

Protein Extraction

For analysis of proteolysis, frozen RAAs were homogenized in buffer (100 mM Tris-

HCl, 145 mM NaCl, pH=7.3) and centrifuged at 26.000 x g (30 min, 4°C). For Western-

blot analysis, parts of the same RAAs were homogenized in Radio-Immuno-Precipitation-

Assay (RIPA) buffer.10 Protein concentration was determined using the DC assay (Bio-

Rad, Netherlands) with a bovine albumin standard.

Proteolytic Assay

Suc-Leu-Leu-Val-Tyr-7-amino-4-methyl-coumarin (AMC, Sigma, Netherlands) was

used as substrate. Twenty-five µg protein extract was added to 20 µM AMC in 300 µl

115


Tris-buffered saline. AMC release was measured by fluorometry (360-nm excitation; 430-

nm emission, Spectrometer LS50B, Perkin Elmer, Netherlands) after incubation for 30

min at 25°C. Standard curves were generated using known concentrations of 7-amino-4

methyl-coumarin (Sigma, Netherlands) and 25 µg heat denatured protein. Maximal calpain

activation was assessed after reconstitution of calcium at 1 mM. E-64 (10-4 M, Roche, The

Netherlands), calpain I inhibitor (N-Acetyl-Leu-Leu-norleucinal, 10-4 M, Sigma, The

Netherlands) and calpain II inhibitor (N-acetyl-Leu-Leu-methioninal, 10-4 M Sigma, The

Netherlands) were used to assess calpain activation. Lactacystin (10-4 M, Calbiochem,

Netherlands) was used to investigate proteasome activity. Assays were conducted in

triplicate.

Western-Blot Analysis

Protein expression was determined by Western-blot and expressed as ratio to levels

of GAPDH, as described previously.10 Denatured protein (10 µg) was separated by SDS-

PAGE, transferred to nitrocellulose membranes (Stratagene, Netherlands) and incubated

with primary antibodies against GAPDH (Affinity Reagents, USA), calpain I and calpain

II (Research Diagnostics, USA). Horseradish peroxidase-conjugated anti-mouse or anti-

rabbit IgG (Santa-Cruz Biotechnology, Netherlands) was used as secondary antibody.

Signals were detected by the ECL-detection method (Amersham, Netherlands) and

quantified by densitometry.

Table 1. Baseline characteristics of patients with lone paroxysmal AF (PAF), lone persistent AF (CAF)

and control patients in sinus rhythm (SR).

Patient characteristics SR PAF CAF

N 10 7 10

Age 60±7 50±7 53±8

Previous duration AF - Median, range (months) - - 12.5 (3-56)

Duration SR before surgery - Median, range (days) - 2 (0.5-12) -

Duration last paroxysm - Median, range (hours) 9 (1-24)

Exercise tolerance

• NYHA Class I 10 6 7

• NYHA Class II 0 1 3

Medication

• Digitalis 0 1 2

• Calcium antagonists 4 2 2

• Beta-blockers 6 1* 1*

* p<0.05 compared to SR

116

Chapter 8


Results are expressed as mean ± SD. One-way ANOVA was used for multiple group

comparisons. Correlation was determined with Spearman correlation test (SPSS 8.0). p<0.05

was considered statistically significant.

Results

Proteolytic activity

Proteolytic activity was significantly increased in tissue from patients with paroxysmal

and persistent AF compared to patients in sinus rhythm (Figure 1A, Table 2).

Proteolytic activity was measured in the presence of inhibitors to assess the

involvement of different pathways (Table 2). The non-selective calpain inhibitor E-64 (10-

4 M) and calpain I inhibitor (10-4 M) significantly reduced both tissue and maximal

proteolytic activity to a similar level in all groups. Calpain II inhibitor (10-4 M) reduced

proteolysis, but only partially attenuated the increased proteolytic activity observed in AF.

In contrast, the proteasome inhibitor lactacystin (10-4 M) did not reduce proteolytic activity

in any group. Thus, increased proteolysis in AF is due to activation of calpains.

To assess maximal calpain activation, proteolytic activity was determined in the

presence of 1 mM calcium. Under these conditions, a similar increase in proteolytic activity

was found in patients with AF and controls compared with experiments in the absence of

calcium (Figure 1B, Table 2). The extra activation due to addition of calcium was abolished

by E-64, calpain I inhibitor and calpain II inhibitor, but unaffected by lactacystin (Table

2). Further, tissue calpain activity across all groups correlated significantly with the maximal

calpain activity (r=0.89, p<0.001; Figure 1C).

SR PAF CAF

Calpain activity (nM AMC/mg/30min)

01020304050607080

SR PAF CAF0

1020304050607080

0 10 20 30 40 50 600

1020304050607080

r=0.89, p<0.001

A CB

** *

*

Figure 1.

Significant increase of tissue calpain activity (A) and maximal calpain activity (B) in RAAs of patients with

sinus rhythm (SR; o), paroxysmal AF (PAF; ) and persistent AF (CAF; •). * p<0.01 compared to SR. (C)

Significant correlation for maximal and tissue calpain activity. Calpain activity was expressed as nM AMC/mg

protein/30 min.

Calp

ain

acti

vit

y (

nM

AM

C/m

g/3

0m

in)

117


Calpain I and II Protein Levels

To examine the relation between tissue calpain activity and calpain expression, protein

levels of calpain I and II were determined by Western-blotting and expressed as ratio to

the housekeeping enzyme GAPDH (Figure 2A,D). Protein levels of calpain I were

significantly increased with 35% in patients with persistent AF, compared to controls,

whereas calpain II levels were similar in all groups (Figure 2B,E). GAPDH levels did not

differ between the groups (sinus rhythm: 1139 ± 150, paroxysmal AF: 1149 ± 175, persistent

AF: 1317 ± 190, arbitrary OD units). A positive correlation was found between tissue

calpain activity and protein expression of calpain I (r=0.61, p<0.001, Figure 2C). In contrast,

no correlation was found between calpain II expression and tissue calpain activity (r=-0.2,

p=0.33, Figure 2F).

Discussion

In this study, activation of proteolysis was found in the atrial tissue of patients with

paroxysmal and persistent lone AF. Increased proteolytic activity in AF is mediated

exclusively by calpains, as the non-selective calpain inhibitor E-64 reduced proteolysis to

similar levels in all groups, even under high calcium conditions known to activate calpain

II. In addition, the selective inhibitor of the proteasome, lactacystin, did not influence

proteolysis in any group. Thus these findings represent, to the best of our knowledge, the

first report demonstrating the activation of the calpain pathway in human cardiac disease.

The relative contribution of calpain I and II is difficult to establish by using inhibitors,

due to their partial selectivity. However, calpain I inhibitor completely attenuated the

increased proteolysis in AF both under basal and high calcium conditions, whereas calpain

II inhibitor did not. In addition, Western-blot demonstrated increased protein levels of

Table 2. Proteolytic activity (nM AMC/mg protein/30min) in atrial tissue of patients with lone paroxys-

mal AF (PAF), lone persistent AF (CAF) and control patients in sinus rhythm (SR).

no calcium added + 1 mM calcium

SR PAF CAF SR PAF CAF

None 33 ± 10 53 ± 13* 58 ± 17* 41 ± 13 72 ± 17* 82 ± 16*

E-64 (10-4 M) 19 ± 4 22 ± 4 23 ± 6 20 ± 6 23 ± 8 25 ± 5

Calpain I inhibitor (10-4 M) 23 ± 3 24 ± 3 25 ± 2 25 ± 4 26 ± 3 27 ± 4

Calpain II inhibitor (10-4 M) 21 ± 5 27 ± 6 33 ± 1* 23 ± 7 34 ± 11* 37 ± 13*

Lactacystin (10-4 M) 35 ± 12 47 ± 15 69 ± 12* 32 ± 11 60 ± 16* 72 ± 16*


118

Chapter 8

calpain I in persistent AF, but unchanged levels of calpain II. Finally, calpain I protein

levels, but not calpain II levels, correlated with increased calpain activity. Taken together,

these results suggest that the increased proteolysis in AF is mainly due to activation and

up-regulation of calpain I, rather than calpain II.

Calpain activation represents a likely candidate to mediate important cellular changes

in AF. While calpains are readily activated by elevated cellular calcium levels11, calcium

overload plays a key role in the pathogenesis of AF2,12. Moreover, calpains have been

demonstrated to degrade cytoskeletal13, contractile6,7 and L-type Ca2+ channel14 proteins.

Reduction in expression of the L-type Ca2+ channel is thought to play a major role in the

electrical remodeling in AF.3 Finally, calpain activation would explain the apparently general

reduction in protein expression of multiple ion-channels in the absence of changes in

Figure 2.

Protein levels of calpain I (left panels) and calpain II (right panels) in patients with sinus rhythm (SR; o), parox-

ysmal AF (PAF; ) and persistent AF (CAF; •) expressed as ratio to GAPDH. (A,D) typical Western-blots of 10

µg protein. (B,E) group protein ratios. (C,F) Significant correlation of calpain activity and calpain I levels,

whereas absence of correlation with calpain II levels. * p<0.01 compared to SR.

Calpain I expression 0,5 1,0 1,5 2,0Calpain activity (nM

0102030405060

r = 0.61, p<0.001

Calpain II expression 0,5 1,0

0102030405060

r = -0.2, p=0.33

SR PAF CAF

Calpain expression

0,60,81,01,21,41,6

SR PAF CAF0,60,81,01,21,41,6*

Calpain I Calpain II

SR PAF CAF SR PAF CAF

Calpain I

GAPDH

Calpain II

GAPDH

A

B

C

D

E

F

Calp

ain

expre

ssio

nC

alp

ain

acti

vit

y (

nM

AM

C/m

g/3

0m

in)

119


mRNA, as observed in human paroxysmal AF.4 Thus, interference with the calpain pathway

may both represent an important tool to unravel the sequence of molecular events in AF,

and a possible future alternative for pharmacological intervention.

References







4. Brundel BJJM, Van Gelder IC, Henning RH, et al. Alterations in potassium channel gene expression in

atria of patients with persistent and paroxysmal atrial fibrillation. J Am Coll Cardiol 2000; in press






Circ Res 1997; 80:393-399.

8. Eble DE, Spragia ML, Ferguson A, et al. Sarcomeric myosin heavy chain is degraded by the proteasome.

Cell Tissue Res 1999; 296:541-548.

9. Laing JG, Tadros PN, Saffitz J, et al. Proteolysis of connexin43-containing gap junctions in normal and

heat-stressed cardiac myocytes. Cardiovasc Res 1998; 38:711-718.

10. Brundel BJJM, Van Gelder IC, Henning RH, et al.Gene expression of proteins influencing the calcium


454.

11. Suzuki K, Imajoh S, Emori Y, et al. Calcium-activated neutral protease and its endogenous inhibitor.

Activation at the cell membrane and biological function. FEBS Letters 1987; 220:271-277.



13. Papp Z, Van Der Velden J, Stienen G. Calpain-I induced alterations in the cytoskeletal structure and

impaired mechanical properties of single myocytes of rat heart. Cardiovasc Res 2000; 45:981-993.

14. Belles B, Hescheler J, Trautwein W, et al. A possible physiological role of the Ca-dependent protease

calpain and its inhibitor calpastatin on the Ca current in guinea pig myocytes. Pflugers Arch 1988; 412:554-

556.

120

121

Calpain activity is related to ion-channel, structural and electrical remodeling

Chapter 9

Calpain Activity is related to Ion-Channel, Structural and

Electrical Remodeling in human Paroxysmal and

Persistent Atrial Fibrillation

Brundel, Short title: Calpain activation during Atrial Fibrillation

Bianca J. J. M. Brundel, MSc1,2, Jannie Ausma, PhD4, Isabelle C. Van Gelder,

MD2, Harry J. G. M. Crijns, MD2, Wiek H. van Gilst, PhD1,

Robert G. Tieleman, MD2 Johan J.L. Van Der Want, PhD3;

Robert H. Henning, MD1.

Department of Clinical Pharmacology1, Groningen University Institute for Drug

Exploration (GUIDE), Department of Cardiology2, Thoraxcenter University Hospital,

Groningen, Department of Cell Biology and Electron Microscopy3, University of

Groningen and Department of Physiology4, Cardiovascular Research Institute

Maastricht, University of Maastricht, The Netherlands

Submitted CirculationAbstract

Background: Atrial fibrillation (AF) is accompanied by electrical, structural and

ion-channel protein remodeling. We tested whether calcium activated neutral protease,

calpain, is involved in AF-induced remodeling. Therefore, calpain activity and localization

in atrial tissue of patients with paroxysmal and persistent AF were determined, and correlated

with remodeling processes. Methods and Results: Right atrial appendages were obtained

from patients with paroxysmal (n=7, PAF), persistent (n=10, CAF) lone AF and controls

in sinus rhythm (n=10) and used for measuring tissue calpain activity with a fluorogenic

calpain specific substrate. Calpain I localization was studied by immunohistochemistry.

Ion-channel protein expression was determined by slot-blot analysis. Structural changes

were quantified by counting atrial myocytes showing degeneration or hibernation. Rate

adaptation was calculated from atrial effective refractory periods (AERPs) at 5 basic cycle

lengths. Tissue calpain activity was significantly increased in PAF (2-fold, p<0.001) and

CAF (3-fold, p<0.001), mainly due to calpain I activation. Calpain I was localized in the

cytosol, intercalated discs and nucleus of cardiomyocytes. Patients with AF showed

122

Chapter 9

decreased protein expression and rate adaptation, and increased structural changes compared

to controls. Tissue calpain activity correlated with ion-channel protein amounts (L-type

Ca2+ channel, r=-0.73; Kir3.1, r=-0.75; Kv1.5, r=-0.74 and minK, r=-0.79, all p<0.001),

the degree of structural changes (r=0.9, p<0.001) duration of AERP (BCL 500 ms, r=-0.6,

p<0.001) and rate adaptation of AERP (r=-0.80, p<0.001). Conclusions: Calpain activity

is induced during AF and correlates with parameters of ion-channel protein, structural

and electrical remodeling. The results strongly suggest that calpain activation represents

an important mechanism linking calcium overload to cellular adaptation mechanisms in

human AF.

Introduction

Atrial fibrillation (AF) is characterized by a heterogenic electrical activation pattern

and the loss of contractile function of atrial tissue. Clinical and animal experimental studies

indicate that atrial fibrillation has a strong tendency to promote itself.1,2 Research has been

directed at obtaining insight in the underlying mechanisms. Electrophysiological studies

have identified important factors promoting vulnerability to AF, including shortening of

atrial effective refractory period (AERP) and reduction of its rate adaptation.1,3,4 Recent

studies examining function and expression of various plasma membrane ion-channels

indicate that a reduction in the expression of the L-type Ca2+ channel plays a crucial role in

the electrical remodeling during AF.5-8 Interestingly, AF-induced electrical remodeling is

attenuated by application of calcium channel antagonists during the induction of the

arrhythmia.9,10 Therefore, the down-regulation of the L-type Ca2+ channel most likely

represents a feedback mechanism in response to the early calcium overload observed in

AF11, however, at the expense of electrophysiological changes promoting its maintenance.

Although electrical remodeling favors arrhythmia maintenance additional factors

increasing the vulnerability to AF seem to be involved, as electrical remodeling is

accomplished well in advance of the AF-promoting effect.1,12 Structural changes13 of atrial

tissue represent a likely candidate, as this could lead to conduction slowing3,12,14 and spatial

AERP heterogeneity15,16 thereby increasing the number of functional re-entry circuits in

the atrium.

The cellular mechanisms underlying the remodeling processes are not well

characterized. Two observations suggest that activation of a proteolytic pathway may

mediate the molecular changes underlying the AF-induced remodeling processes. Firstly,

we observed a discrepancy between changes in protein and mRNA expression of plasma

membrane ion channels in patients with lone paroxysmal AF.17 Whereas protein levels of

L-type Ca2+ channel, Kv1.5, Kir3.1 and minK were substantially decreased, the mRNA

levels were essentially unaffected in paroxysmal AF, suggesting activation of a protein

degradation mechanism at the post-transcriptional level. Secondly, we found increased

123


proteolysis in atrial tissue of patients with paroxysmal and persistent lone AF, which was

mainly due to activation of the calcium-activated neutral protease, calpain I (Brundel et al.

submitted).

In this study we sought to further explore the potential role of calpain activation in

the cellular adaptation mechanisms underlying AF. Therefore, we first localized the calpain

I protein in the atrial tissue by immunohistochemistry. Furthermore, we determined ion-

channel protein amounts, structural changes, AERPs and rate adaptation of AERP in atrial

tissue of AF patients and correlated these with the activity of calpain I.


Patients and electrophysiological measurements

RAAs were obtained as described before.7 In short, patients with paroxysmal (n=7)

and persistent (n=10) lone AF undergoing MAZE surgery were included and compared to

clinically stable control patients in sinus rhythm undergoing coronary bypass grafting

(CABG, n=10, Table 1). During surgery, the AERPs were determined in 8 patients with

persistent AF, 7 with paroxysmal AF and 8 control patients, with use of temporary epicardial

pacing leads. AERPs were measured at five different basic cycle lengths (BCL: 600, 500,

400, 300 and 250 ms) at the RAA using programmed electrical stimulation (Table 2). To

quantify the change in AERP at the different BCLs, we calculated the rate adaptation

Table 1. Baseline characteristics of patients with lone paroxysmal AF (PAF), lone persistent AF (CAF)

and control patients in sinus rhythm (SR).

SR PAF CAF

N 10 7 10

Age 60±7 50±7 53±8

Previous duration of AF (median, range (months) - - 12.5 (3-56)

Duration of SR before surgery (median, range (days) - 2 (0.5-12) -

Duration of last paroxysm (median, range (hours) - 9 (1-24) -

Surgical procedure

•CABG 10 0 0

•MAZE 0 7 10

Exercise tolerance

•NYHA Class I 10 6 7

•NYHA Class II 0 1 3

Medication

Digitalis 0 1 2

Calcium antagonists 4 2 2

Beta blockers 6 1* 1*


124

Chapter 9

coefficient for individual patients as the slope of the linear regression after logarithmic

transformation of BCL.

Although AF groups and their controls differed with respect to the underlying heart

disease, all had normal left ventricular function (Table 1). The Institutional Review Board

approved of the study and all patients gave written informed consent.

Protein Extraction

For analysis of calpain activity, frozen RAAs were homogenized in buffer (100 mM

Tris-HCl, 145 mM NaCl, pH=7.3) and centrifuged at 26.000 x g (30 min, 4°C). For slot-

blotting analysis parts of the same RAAs were homogenized in Radio-

ImmunoPrecipitationAssay (RIPA) buffer.7 Protein concentration was determined using

the Bio-Rad DC assay (Bio-Rad, The Netherlands) with a bovine albumin standard.

Calpain measurements

Calpain activity was measured as described previously (Brundel et al. submitted).

Immunohistochemistry was performed for six patients with paroxysmal AF, eight with

persistent AF and eight control patients in sinus rhythm. Two groups of 5 µm thick frozen

RAA sections were made. One group was only air dried before use, whereas in a second

group the sections were immediately fixed for 10 minutes in 4% paraformaldehyde (in

PBS). After three times washing in PBS for 10 minutes and 30 minutes blocking with 1%

BSA in PBS, all sections were incubated with anti-calpain I antibodies (1:100) (Research

Diagnostics, USA) overnight at room temperature. After washing the sections three times

for 10 minutes with PBS they were incubated with secondary antibody peroxidase

conjugated rabbit-anti-mouse IgG (DAKO A/S, Glostrup, Denmark) for one hour. After

three times washing with PBS for 10 minutes, peroxidase activity was detected with 3-

amino-9-ethylcarbazole (AEC, Sigma, The Netherlands). To this end, 40 mg of AEC was

dissolved in 10 ml N,N dimethylformamide (Merck, The Netherlands) and added to 190

Table 2. AERP measured at the different BCLs.

AERP (ms) correlation calpain activity

BCL (ms) SR PAF CAF r p

600 291±53 222±15* 208±39** -0.52 0.007

500 277±42 224±24* 207±29** -0.6 <0.001

400 252±34 216±24* 203±25** -0.55 0.001

300 224±16 202±20 189±24* -0.54 0.002

250 184±5 185±19 172±17 -0.33 NS

*, p<0.05; **, p<0.01; NS, not significant

125


ml of 0.05 M sodium actetate buffer (pH 4.95). Hydrogen peroxide was added to a final

concentration of 0.01% (v/v). After 10 minutes staining, the sections were rinsed with

water, counterstained with haematoxylin (Sigma, The Netherlands) and mounted with

Kaiser’s glycerol gelatin (Merck, The Netherlands).

Slot-Blot Analysis

Protein expression was determined by slot-blot analysis and expressed as ratio to

levels of GAPDH as described previously.7 Denatured protein (10 µg) was spotted on

nitrocellulose membranes (Bio-Rad, The Netherlands) and incubated with primary

antibodies against GAPDH (Affinity Reagents, USA), L-type calcium channel α1 subunit,

minK, Kir3.1 and Kv1.5 (Alomone Labs, Israel). The specificity of antibodies was

confirmed by preincubation with control peptide antigen and by SDS-PAGE. Anti-mouse

IgG (Santa-Cruz Biotechnology, The Netherlands) was used as secondary antibody. Signals

were detected by the ECL-detection method (Amersham, The Netherlands) and quantified

by densitometry. GAPDH levels did not differ between the groups (sinus rhythm: 1082 ±292, paroxysmal AF: 1190 ± 181; persistent AF: 1177 ± 222, arbitrary OD units).

Morphological evaluation

For morphological evaluation by light microscopy, a biopsy was taken from all the

RAAs and immediately fixed for at least 2 hours at 4°C in 2% glutaraldehyde (in 0.1 M

cacodylate buffer, pH 7.4). Post-fixation was performed for 2 hours in 1% osmium tetroxide

(supplemented with 1.5% K4Fe(CN)

6 in cacodylate buffer, pH 7.4) at 4°C. After dehydration

in ethanol, biopsies were embedded in Epon and semi-thin sections (1 µm) were cut and

stained with 1% toluidine blue. The degree of cellular changes was evaluated by light

microscopy in cells where the nucleus was visible in the plane of the section. To quantify

the structural changes a minimum of 300 cells from six randomly chosen regions of the

RAA were evaluated by an investigator blinded for patient groups. Two types of structural

changes were present in our biopsies, i.e. hibernation or degeneration of atrial myocytes.

Hibernating cells show areas with loss of sarcomeres and contain pale nuclei. An atrial

myocyte was defined as hibernating when >10% of the cell surface was free from

sarcomeres.18 Degenerating cells display the following characteristics: contraction band

necrosis, pyknotic nuclei with intense stained chromatin, secondary lysosomal structures

(inclusion bodies) and/or vacuoles. Myocytes were scored positive for degeneration if

>10% of the sarcomere surface contained intensely stained contraction bands.19 To verify

the structural changes on the ultrastructural level, electron microscopy was performed.

Ultrathin sections (60 nm) of the RAAs were stained with uranylacetate and lead citrate

and examined in a Philips 201 electron microscope operating at 60 kV.

126

Chapter 9

Figure 1.

Immunohistochemical localization of calpain I in myocytes. (A) at the nucleus (arrow) and (B) at the intercalated

discs (arrow) (magnification A x 400, B x 300).

Statistical Methods

Results are expressed as mean ± SEM. Parametric and non-parametric ANOVA

(Kruskal-Wallis test) were used for multiple group comparisons. Correlation was determined

using the Spearman correlation test. P<0.05 was considered statistically significant. SPSS

version 8.0 was used for all statistical evaluations.

Results

Calpain activity and localization

Calpain activity was significantly increased in tissue from patients with paroxysmal

and persistent AF compared to patients in sinus rhythm (Table 3). Both the non-selective

calpain inhibitor E64 and calpain I inhibitor significantly reduced tissue calpain activity to

a similar level in all groups.

Localization of calpain I was performed by immunohistochemical staining. Irrespective

of the fixation method used, calpain I was predominantly localized in the atrial myocytes

and to a less extend in interstitial cells. In non-fixed sections, calpain I was detected both

at the nucleus and in the cytoplasm (Figure 1A), whereas calpain I was located at the

intercalated discs and in the cytoplasm in paraformaldehyde fixed sections (Figure 1B).

A B

Table 3. Calpain activity (nM AMC/mg protein/30min) in atrial tissue of patients with lone paroxysmal

AF (PAF), lone persistent AF (CAF) and control patients in sinus rhythm (SR).

SR PAF CAF

None 33 ± 10 53 ± 13* 58 ± 17*

E-64 (10-4 M) 19 ± 4 22 ± 4 23 ± 6

Calpain I inhibitor (10-4 M) 23 ± 3 24 ± 3 25 ± 2


127


Figure 2.

Significant correlation between tissue

calpain activity and expression of vari-

ous plasma membrane ion-channels.

The top of each panel shows a typical

slot blot analysis of 10 µg protein of

control patients (SR), paroxysmal AF

(PAF) and persistent AF (CAF). The

immunoblots were done for (A) anti-

L-type calcium channel, (B) anti-

Kir3.1, (C) anti-Kv1.5 and (D) anti-

minK. ( ) indicates patients with PAF,

(•) CAF and (o) SR.

Kir3.1/GAPDH0 1 2 3

,

-100

10203040

L-type Ca2+ channel/GAPDH0 1 2 3calpain activity (nM AMC/mg)

-100

10203040

Kv1.5/GAPDH0 1 2

,

-100

10203040

minK/GAPDH0 1 2,

-100

10203040

r=-0.73, p<0.001 r=-0.75, p<0.001

r=-0.74, p<0.001 r=-0.79, p<0.001

SR PAFCAF

SR PAFCAF

PAF PAFCAFCAF

SR SR

A B

C D

Calp

ain

acti

vit

y (

nM

AM

C/m

g)

Figure 3.

(A) Atrial myocardium from a patient in sinus rhythm stained with toluidine blue showing normal structural

myocytes without myolysis and degenerative features. (B) electron microscopic detail of a normal atrial myo-

cyte without myolysis and normal nucleus. (C) Atrial myocytes from a patient show degenerative changes as

contraction band necrosis (arrows) (D) Electron microscopic detail of an atrial myoctye from a patient with PAF

showing degenerative featrues as clumping of nuclear chromatin and contraction band necrosis (arrows). (E)

Structural changes observed after persistent AF. Extensive myolysis is present perinuclear and pale nuclei were

observed. (F) Electron microscopy of a persistent myocyte showing myolysis. In this hibernating atrial myocyte

only a rim of sarcomeres is present at the border of the cell. Glycogen (g) is visual and the nucleus has a homog-

enous dispersion of chromatin. (Magnification A, C and E x 250, B x 7000, D and F x 4500)

g

A

B

C

D

E

F

128

Chapter 9

Table 4. Immunohistochemical localization of calpain I in myocytes. Intensity of the calpain I staining in

patients with paroxysmal (PAF, n=6), chronic persistent AF (CAF, n=8) and sinus rhythm (SR, n=8).

Intensity calpain I staining

SR PAF CAF

Localization -/+ + ++ +++ - + ++ +++ P - + ++ +++ P

Intercalated discs 5 3 0 0 0 2 4 0 0.04 0 0 2 6 <0.01

Nucleus 2 4 2 0 0 1 4 1 0.04 0 0 3 5 0.02

Cytoplasm 0 7 1 0 0 4 2 0 NS 0 8 0 0 NS

+/- low staining - +++ intense staining

To obtain an indication about the amount of calpain I protein in the myocytes, the intensity

of staining was scored semi-quantitatively (Table 4). The intensity of staining of the nucleus

and the intercalated discs increased significantly from sinus rhythm, via paroxysmal AF to

reach a maximum in persistent AF. In contrast, the intensity of staining in the cytoplasm

was not different between the groups.

Ion-channel Proteins

To examine the relation between tissue calpain activity and ion-channel protein

expression, protein levels of L-type calcium channel, Kir3.1, Kv1.5 and minK

weredetermined by slot-blot analysis (Figure 2). Protein levels of these ion-channels were

all significantly reduced in patients with AF as compared to controls (Table 5).

Importantly, a significant negative correlation was observed between tissue calpain

activity and protein levels of all ion-channels examined (L-type Ca2+ channel, r=-0.73;

Kir3.1, r=-0.75; Kv1.5, r=-0.74 and minK, r=-0.79, all p<0.001, Figure 2). In contrast, no

correlation was found between GAPDH densities and tissue calpain activity (r=0.32,

p>0.05).

Structural changes

The amount of structural and ultra-structural changes in the atrial tissue was examined

by light microscopy and electron microscopy, respectively. Sections were scored for definite

signs of cellular degeneration or hibernation (see Materials and Methods). Atrial

myocardium of sinus rhythm patients showed mainly normal structured myocytes without

myolysis and degenerative features (Figure 3A). At the ultrastructural level, these control

patients showed a highly organized sarcomeric structure with mitochondria in between. A

129


typical distribution of heterochromatin in the form of clusters mainly at the cardiomyocyte

nuclear membrane was observed (Figure 3B). In contrast, contraction band necrosis was

abundantly present in patients with paroxysmal AF (Figure 3C). At the ultrastructural

level, their myocytes showed contraction band necrosis and pyknotic nuclei (Figure 3D).

Patients with persistent AF showed contracting myocytes with band necrosis or hibernation

(Figure 3E). Hibernating cells were only extensively present in patients with persistent AF

(Figure 3F).

The number of myocytes with contraction band necrosis was increased 3-fold both in

patients with paroxysmal and persistent AF compared to patients in sinus rhythm (Figure

4A). However, the amount of hibernating cells was only significantly increased in patients

with persistent AF compared to the other groups (Figure 4A). Furthermore, a good

correlation was found between the calpain activity and the total number of affected cells

(either contraction band necrosis or hibernation; r=0.71, p<0.001; Figure 4B). Thus,

atrialtissue of patients with increased calpain activity showed an increased number of

myocytes with structural changes.

When the type of structural change, i.e. contraction band necrosis or hibernation, in

patients with persistent AF was plotted against the duration of AF an intriguing relationship

was revealed (Figure 4C). Atrial myocytes of patients with the shortest duration of AF (<

10 months) showed high amounts of contraction band necrosis and low amounts of

hibernating cells, whereas the opposite pattern was found in patients with the longest

duration of AF (> 10 months).

AERP and Rate adaptation

AERPs were measured at five BCLs (600, 500, 400, 300 and 250 ms). Patients with

persistent and paroxysmal AF had significantly shorter AERPs than patients in sinus rhythm

(Table 2). Calpain activity showed significant negative correlations with AERP measured

at BCL 600, 500, 400 and 300ms (Figure 5A, Table 2).

Table 5. Ion-channel protein amounts in patients with paroxysmal (PAF) and chronic persistent AF (CAF)

and patients in sinus rhythm (SR).

Protein (ratio to GAPDH, arbitrary OD units) SR PAF CAF

L-type calcium channel 1.21 ± 0.13 0.63 ± 0.11* 0.55 ± 0.07*

Kv1.5 1.18 ± 0.14 0.67 ± 0.13* 0.60 ± 0.11*

Kir3.1 1.30 ± 0.15 0.80 ± 0.11* 0.63 ± 0.10*

MinK 0.83 ± 0.07 0.51 ± 0.03* 0.49 ± 0.03*

130

Chapter 9

Figure 4.

(A) Percentage structural changes of atrial cells affeted by contaction band necrosis or hibernation in patients

with sinus rhythm (SR), paroysmal AF (PAF) and chronic, persistent AF (CAF). (B) Significant correlation

between the total amount of affected cells and calpain activity. (C) Significant correlation between the type of

structural change and the duration of persistent AF. (•) contraction band necrosis and ( ) hibernation.

tissue calpain activity (nM AMC/mg/30 min)0 5 10 15 20 25 30 35 40

% affected cells

020406080

r=+/- 0.9, p<0.001

SR PAF CAF

% affected cells

0102030405060

* * #= contraction band necrosis= hibernation

duration of persistent AF (months)0 10 20 30 40 50

% affected

0

20

40

60

r=0.71, p<0.001

r= +/- 0.9, p<0.001

A

B

C

% a

ffecte

d c

ell

s%

aff

ecte

d c

ell

s%

aff

ecte

d c

ell

s

From the AERP obtained at different basal cycle lengths, the rate adaptation coefficient

was calculated. The rate adaptation coefficient was significantly reduced by 36% in

persistent AF compared to sinus rhythm (persistent AF: 95 ± 13, paroxysmal AF: 121 ± 9

and sinus rhythm: 148 ± 9), indicating a poorer adaptation to higher heart rates in patients

with AF. Furthermore, a significant negative correlation was observed between the calpain

activity and the rate adaptation coefficient (r=-0.80, p<0.001, Figure 5B).

131


Discussion

In this study we examined a variety of changes in atrial tissue from patients with

paroxysmal and persistent lone AF and related them to the level of calpain activity.

Immunohistochemical detection of calpain I demonstrated increased staining at the

intercalated disk and in the nucleus of atrial myocytes of AF patients, but not in the cytosol.

Accordingly, calpain activity was increased in patients with AF. Furthermore, an increased

number of degenerative myocytes was observed in both patient groups with AF. Hibernating

myocytes were only present in persistent AF and numbers increased with the duration of

AF. Finally, calpain activity correlated with the expression levels of ion-channel proteins,

the degree of structural changes, duration of AERP and the rate adaptation coefficient of

AERP. This study strongly indicates that calpain activation represents an important adaptive

molecular mechanism occurring in human AF. Consequently, interference with the calpain

pathway may both represent an important tool to investigate molecular events in AF, and

a possible future alternative for pharmacological intervention.

calpain activity (nM AMC/mg/30 min)0 10 20 30 40 50 60

adaptation coef

50

100

150

200

r=-0.80, p<0.001

calpain activity (nM AMC/mg/30 min)0 10 20 30 40 50 60

AERP (ms, BCL 500 ms)

150

200

250

300

r=-0.6, p<0.001

A

B

AE

RP

(m

s, B

CL

500 m

s)

adepta

tion c

oeff

icie

nt

Figure 5.

(A) Significant correlation between AERP

measured at BCL 500 ms and tissue calpain

activity. (B) Significant correlation between

rate adaptation coefficient and tissue

calpain activity. ( ) indicates patients with

PAF, (•) CAF and (o) SR.

132

Chapter 9

Localization of calpain I

Previously, we demonstrated induction of the calpain activity in atrial tissue of patients

with paroxysmal and persistent lone AF, mainly due to increased activation of the calpain

I protein (Brundel et al. submitted). Now we found that calpain I was localized in the

cytosol, intercalated discs and nucleus of atrial myocytes. Its intensity at the intercalated

discs and nucleus was increased in patients with AF, in accordance with the increased

calpain activity measured in AF patients. The change in cellular distribution in AF suggests

that calpain I activation mediates underlying molecular changes. Under normal conditions,

calpain is localized diffusely in the cytosol. After an increase in intracellular calcium,

calpain rapidly translocates to the inner surface of the plasma membrane, aggregates at the

intercalated discs, which is followed by its activation.20 At the intercalated discs, activated

calpain I may degrade important ion-channels, like the Na+-channel21 and Kv1.522, but

also proteins involved in excitation-contraction coupling23 and conduction.13 The increased

expression of calpain I at the nucleus is in agreement with its role in promoting necrosis

and apoptosis24, suggesting that calpain I could play a role in the degeneration observed in

AF.25 Thus, the increased calpain I expression in AF seems confined to cellular areas that

are vital for action potential conduction and structural integrity of the atrial myocytes.

Structural changes in myocytes

We systematically analyzed the amount of myocytes affected by degeneration and

hibernating in microscopic sections of right atrial appendages. Increased degeneration

was observed both in patients with persistent and paroxysmal AF, as evidenced by an

equal increase in contraction band necrosis in both groups. In contrast, hibernating cells

were only increased in patients with persistent AF. The latter is in agreement with

observations in an experimental goat model for AF, in which hibernation develops only at

a relatively late stage of the arrhythmia.18 Thus, these findings suggest that degeneration

represents an early structural abnormality in human lone AF, whereas hibernation is

dependent on protracted periods of sustained AF. Similar alterations in myocardial structure

were described in patients with atrial arrhythmias of various aetiology.26 In adition to

degenerative changes, part of the myocytes showed loss of myofibrils and presence of

glycogen granules as in our human study.

Further, in patients with persistent AF, hibernation increased with the duration of AF,

while degeneration decreased. Possibly, this is related tothis is due to a protection from

degeneration by the hibernation, as described in ischemic preconditioning.27 Taken together,

the observed structural changes are indicative of a substantial deterioration of normal

tissue architecture, likely to promote AF through heterogeneity of atrial refractoriness15,16

and slowed atrial conduction.3,12,14

133


Interpretation of correlation

Our data demonstrate that human paroxysmal and persistent lone AF is accompanied

by substantial electrical28, structural and ion-channel protein17 remodeling, as indicated by

shortening of AERP, decreased rate adaptation, increased numbers of cells showing

degeneration or hibernation and decreased expression of four ion-channel proteins. These

parameters all correlated significantly with calpain activity measured in the atrial tissue.

Accordingly, these parameters generally show a good correlation with each other. Therefore,

the question remains whether calpain activation causes the molecular changes as observed

in AF, or merely reflects gross cellular damage. There are strong indications that calpain

causes the molecular changes. First as cytosolic calcium is rapidly increased during AF11,

increased calpain activation may occur already early after AF. Secondly, calpain I and not

calpain II or proteasome showed increased activity (Brundel et al. submitted). Third, this

increased calpain activity was due to elevation of calpain I and not calpain II protein

expression. Finally we found increased staining intensity of calpain I at vital areas of atrial

myocytes. One can speculate that the rapidly induced calpain acitivity immediately

proteolyse the ion-channel proteins leading to the loss of physiological rate adaptation and

structural changes. In this way the fibrillating atria can no longer be expected to reverse

their action potentials when sinus rhythm is restored and this may explain the vulnerability

to AF.

Possible clinical relevance

Successful chemical cardioversion and maintenance of sinus rhythm after

cardioversion is dependent on the duration of AF.29 This clinically observed diminished

efficacy of chemical cardioversion after long term AF cannot only be explained by the

occurrence of electrical remodeling. The protein remodeling and structural remodeling

could also affect the success rate of cardioversion. In patients with persistent AF, there is

a correlation between the duration of AF before cardioversion and the time needed to

recover atrial contractile function thereafter.30 The increase in calpain activity and correlated

structural remodeling of the atrial myocytes might give an explanation for this delay in

recovery of atria contractile function after conversion to sinus rhythm. After restoration of

normal sinus rhythm it may take the cardiomyocytes a certain period to rebuild a normal

amount of sarcomeres, if this is possible at all.31 Our study implicates that it is necessary to

bring AF as soon as possible back to sinus rhythm, thereby preventing the continuation of

the atrial structural, ion-channel protein and electrical remodeling.

Acknowledgements

The authors would like to thank Bert Blaauw and Hans Duimel for their excellent

technical assistance.

134

Chapter 9

References




of Copenhagen. Thesis 1975;

3. Morillo CA, Klein GJ, Jones D, et al. Chronic rapid atrial pacing. Structural, functional, and

electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation 1995; 91:1588-

1595.

4. Daoud EG, Bogun F, Goyal R, et al. Effects of atrial fibrillation on atrial refractoriness in humans. Circu-

lation 1996; 94:1600-1606.



6. Van Wagoner DR, Pond AL, Lamorgese M, et al. Atrial L-Type Ca2+ Currents and Human Atrial Fibrilla-

tion. Circ Res 1999; 85:428-436.



454.











13. Van der Velden HMW, Ausma J, Rook MB, et al. Gap junctional remodeling in relation to stabilization of

atrial fibrillation in the goat. Cardiovasc Res 2000; 46:476-486.

14. Elvan A, Wylie K, Zipes DP. Pacing-induced chronic atrial fibrillation impairs sinus node function in dogs:

electrophysiological remodeling. Circulation 1996; 94:2953-2960.

15. Ramanna H, Hauer RNW, Wittkampf FHM, et al. Identification of the substrate of atrial vulnerability in

patients with idiopathic atrial fibrillation. Circulation 2000; 101:995-1001.

16. Fareh S, Villemaire C, Nattel S. Importance of refractoriness heterogeneity in the enhanced vulnerability

to atrial fibrillation induction caused by tachycardia-induced atrial electrical remodeling. Circulation 1998;

83:2202-2209.

17. Brundel BJJM, Van Gelder IC, Henning RH, et al. Alterations in potassium channel gene expression in

atria of patients with persistent and paroxysmal atrial fibrillation. J Am Coll Cardiol 2000; in press:



19. Borgers M, Guo Shu L, Xhonneux R, et al. Changes in ultrastructure and Ca2+ distribution in the isolated

working rabbit heart after ischemia. A time-related study. Am J Pathol 1987; 126:92-102.

20. De Tullio R, Passalacqua M, Averna, et al. Changes in intracellular localization of calpastatin during

calpain activation. Biochem J 1999; 343:467-472.

21. Cohen SA. Immunocytochemical localization of rH1 sodium channel in adult rat heart atria and ventricle.

Presence in terminal intercalated disks. Circulation 1996; 94:3083-3086.

22. Mays DJ, Foose JM, Philipson LH, et al. Localization of the Kv1.5 K+ channel protein in explanted

cardiac tissue. J.Clin.Invest. 1995; 96:282-292.

23. Laflamme MA, Becker PL. G(s) and adenylylcyclase in transverse tubules of heart: implications for cAMP-

dependent signaling. Am.J.Physiol. 1999; 277:H1841-H1848

24. Yuen PW, Wang KKW. Calpain inhibitors, novel neuroprotectants and potential anticataractic agents.

Drugs Future 1998; 23:741-749.

25. Aime-Sempe C, Folliguet T, Rucker-Martin, et al. Myocardial cell death in fibrillating and dilated human

right atria. J Am Coll Cardiol 1999; 34:1577-1586.

26. Mary-Rabine L, Pham TD, Hordof A, et al. The relationship of human atrial cellular electrophysiology to

clinical function and ultrastructure. Circ.Res. 1983; 52:188-199.

27. Tanaka M, Fujiwara H, Yamasaki K, et al. Expression of heat shock protein after ischemic precondition-

ing

135


in rabbit heart. Jpn Circ J 1998; 62:512-516.

28. Tieleman RG, Van Gelder IC, Tuinenburg AE, et al. Intra- and post-operative atrial refractory periods in

relation to atrial arrhythmia history and the presence of mitral regurgitation. Circulation 1999; 100:I-361



Arch Intern Med 1996; 156:2585-2592.



31. Ausma J, Duimel H, Wouters L, et al. Structural atrial changes induced in the goat by 16 weeks of atrial

fibrillation are still present 8 weeks after cardioversion. Europace 2000; 1:B12

136

137

General discussion

General Discussion

Calcium Homeostasis

Potassium Channels

L-type Calcium Channel

Neurohumoral Changes

Post-Transcriptional Regulations

Integrated Model for Remodeling Processes in AF

Evidence for calcium overload

Time course structural remodeling

Electrophysiological properties

Future Perspectives

Pre-conditioning

Possible clinical relevance

New experimental model for AF

Calcium Homeostasis

Studies showed that AF has the tendency to become more persistent over time. A

large percentage of patients with paroxysmal AF will develop persistent AF.1 Also,

pharmacological and electrical cardioversion and maintenance of sinus rhythm thereafter

become more difficult the longer the arrhythmia exists.2 Therefore it is important to study

the underlying mechanisms which play a role in the vulnerability to AF.

Experimental studies showed that electrical and contractile remodeling occur early

after the onset of AF (Figure 1A).3-5 Both remodeling processes were attenuated by blocking

of the L-type Ca2+ channel indicating that changes in the calcium homeostasis triggered by

tachycardia induced intracellular calcium overload, play a pivotal role in the induction of

atrial electrical remodeling and contractile dysfunction.

We first investigated the AF induced contractile dysfunction by studying the molecular

remodeling of proteins which influence calcium homeostasis (Chapter 2 and 3). The main

finding was reductions in mRNA and protein expression of the L-type calcium channel

and sarcoplasmatic reticulum Ca2+ ATPase (SR Ca2+ ATPase), predominantly in patients

with persistent AF. Also, an increased reduction in mRNA expression was found the longer

the duration of persistent AF existed. Patients with >6 months AF revealed reductions in

mRNA of L-type Ca2+ channel, in contrast to patients with <6 months duration of AF, in

whom no changes in mRNA expression were seen (Figure 1B and 2). This finding indicates

that changes in mRNA expression are the consequence rather than the cause of AF. The

results described in chapter 2 and 3 were confirmed by other studies which also showed

that the L-type Ca2+ channel6 and SR Ca2+ ATPase6,7 were both reduced in AF. Unfortunately,

138

Chapter 10

in one study no time dependent changes in mRNA expression could be investigated since

the duration of AF was not known.7 Both studies were limited by differences between AF

and controls with respect to the underlying heart disease, which could have influenced

mRNA expression. Another limitation was that in both studies only mRNA expression of

L-type Ca2+ channel and SR Ca2+ ATPase was measured and no protein levels, which are

anticipated to represent the amount of functional proteins more adequately.

Taken together these studies indicate that changes in gene expression of proteins

influencing the calcium homeostasis occur in persistent AF. These changes probably are a

contributory factor for the atrial contractile dysfunction in AF.

Figure 1A.

Overview AF induced adaptations in experimental studies

Figure 1B.

Overview AF induced adaptations in human AF

Electrical remodeling/contractile dysfunction

Changes in ion channel protein amounts

Structural changes (myolysis)

0 1 10 50 weeks

Start AF

Changes mRNA expression

Structural changes (degeneration)

Start induction calpain activity?

0 1 10 100 days

Start inductionof AF bypacing

Electrical remodeling / Contractile dysfunction

Functional changes ion channels

Changes in ion channel protein amounts

Calcium overload

Start induction calpain activity?

Structural changes (myolysis)

Changes in mRNA expression

Persistent AF

A

B

139

General discussion

Ion-channel remodeling related to atrial refractory periods

It is well known that an abrupt increase in heart frequency, like in AF, causes an

immediate (within one action potential) and then a gradual (reaching steady state over

several minutes) decrease in action potential duration (APD).8 These alterations in APD

reduce atrial effective refractory period (AERP) and shorten the wavelength for reentry,

which will facilitate the occurrence and maintenance of reentrant arrhythmias like AF.

The rapid nature of these changes suggests that the short-term APD adaptation to rate is

due to functional changes in ion channels. With longer periods of sustained atrial

tachycardia, changes develop over the course of hours to days.4,9,10 The latter alterations

appear to concern mainly ion channel density and are due to modified gene expression

(Figure 1A,B).11,12 Most studies in changes of ion-channel protein expression have been

performed in animal experimental settings and gradually, studies have revealed that the

rapid shortening of the AERP in animal experimental AF mainly involves functional changes

in the L-type Ca2+ channel.11,12 In human AF the relationship between changes in AERP

and ion channel gene expression has not been investigated previously. We studied the

regulation of L-type Ca2+ channel and K+ channels and their relation to AERP in patients

with persistent and paroxysmal AF (Chapter 5). We demonstrated a positive correlation

between the ion-channel protein expression of L-type Ca2+ channel, Kv4.3, Kv1.5, HERG,

minK and Kir3.1 and the AERP but also with the rate adaptation to AERP in patients with

persistent and paroxysmal AF. Low ion-channel protein levels were associated with short

AERP and poor rate adaptation. This indicates that electrical remodeling is paralleled by

general ion-channel protein reductions as part of the adaptation mechanisms during AF.

Since reduced ion-channel protein expression occurred due to AF we called this

phenomenon ion-channel remodeling. The ion-channel protein remodeling could play an

important role in the susceptibility to AF after restoration of sinus rhythm. Since shortening

of AERP can be explained by decrease in L-type Ca2+ channel and increase in K+ channel

gene expression or activity, the reductions in L-type Ca2+ channel could represent an

explanation for the electrophysiological changes during AF.

As noted the shortening in AERP can also be explained by an increase in K+ channel

activity and expression, we investigated the contribution of potassium channels in

paroxysmal and persistent AF (Chapter 4 and 5). Reductions in mRNA and protein levels

were found for several K+ channels in patients with persistent AF. In patients with

paroxysmal AF these reductions were observed predominantly at the protein level and not

at the mRNA level, suggesting the activation of a proteolytic system. The reductions in

mRNA and protein amount of K+ channels do not explain the shortening of AERP, however

other studies have also reported the decrease in K+ channels in AF.13 One study found

increased IKACh

and IK1

in isolated human atrial cells of patients with persistent AF due to

different underlying heart diseases.14 The apparent inconsistency between protein levels

140

Chapter 10

and current density can only be explained by assuming a change in single channel properties

in patients with persistent AF, such as an increase of mean open-time and increase in

channel conductance or a change in voltage dependency.

Unfortunately in human studies it is difficult to make a time course for remodeling

processes, but experimental data can elucidate the ion-channel remodeling in more detail.

Human and experimental studies showed that brief periods of experimental AF (<1 h)

abbreviate AERP and favor AF induction via functional changes, including Ca2+ overload

induced L-type Ca2+ current (ICaL

) inactivation, that cause APD shortening (Figure 1A,B).15,16

With longer periods of sustained atrial tachycardia adaptations appear to involve alterations

mainly in ion channel density that are due to modified gene expression (Figure 1A,B and

2).11,12 An examination of ionic current changes in atrial myocytes from dogs subjected to

rapid atrial pacing for 7 and 42 days 11,12 indicated that high rate stimulation of atrial

myocytes does not change a variety of currents, including inward and delayed rectifier K+

currents, T-type Ca2+ current and Ca2+ dependent Cl- current. Currents that show important

alterations were the transient outward K+ current (ITo

) and L-type Ca2+ current (ICaL

), both

of which are reduced by about 70% after 6 weeks of rapid atrial pacing due to reductions

in protein amount.12,17 Other properties of ICaL

and ITo

, like voltage, time and frequency

Figure 2. Overview L-type calcium channel alterations

141

General discussion

dependence are unchanged. This observation suggests that the changes observed are due

to a reduction in the number of functional channels in the membrane rather than to a

change in basic channel properties. The use of pharmacological probes to mimic the effects

of reduced ICaL

and ITo

on the action potential showed that reductions in ICaL

are likely to

play the central role in the APD alterations caused by atrial tachycardia with the changes

in ITo

being of much less importance,11 despite the quantitatively similar reduction.

Thus, experimental and human AF studies reported important observations concerning

ion-channel remodeling. Although a difference in time course was found between human

and experimental AF. Changes in mRNA expression were observed in animal experiments

around 1 week of AF, in human AF significant changes were observed only after > 6

months (Chapter 3) and >3 months6 (Figure 2), indicating that other factors play a role in

the adaptation mechanisms in human AF, for example preconditioning.

Furthermore in AF series of changes were found, involving rapid functional alterations

and slower changes in gene expression that cause APD reduction and reduced cellular

calcium loading. These changes can be considered to reduce ICaL

and thereby protect the

cell against potentially lethal Ca2+ overload resulting from an increase in rate of action

potential generation between resting sinus rhythm and AF. This protective effect occurs,

however, at the expense of electrophysiological changes that promote the maintenance of

AF.

Neurohumoral changes

The cardiac natriuretic peptide system and the endothelin system play an important

role in maintaining volume homeostasis especially in conditions that affect

hemodynamics.18,19 In Chapter 6 and 7 the local gene expression of these systems in atrial

tissue of patients with AF was studied. Persistent AF was associated with evident expression

of ANP and BNP mRNA and also endothelin-1 mRNA contents. The extent of these changes

was more pronounced in patients with concomitant valvular heart diseases, indicating that

these systems play a role in human AF and in particular in the presence of atrial pressure

or volume overload. Furthermore reductions were found in the protein expression of the

endothelin type A and B receptors during paroxysmal and persistent AF in contrast to

unchanged expression of mRNA amounts of these receptors suggesting post-transcriptional

regulation.

Post-transcriptional regulations

A remarkable finding during the study of mRNA and ion-channel protein remodeling

was a discrepancy between changes in mRNA and protein levels in patients with paroxysmal

AF (Chapter 4, 5 and 7). Whereas ion-channel protein levels of L-type Ca2+ channel, Kv1.5,

Kir3.1 and minK, but also ET-A were substantially decreased, the mRNA levels were

142

Chapter 10

essentially unaffected in paroxysmal AF. This discrepancy was also observed in other

studies20,21 and prompted us to explore the role of an adaptative mechanism of which the

influence in AF was unknown: the activation of a proteolytic system. Different proteolytic

pathways could be involved in AF. Since cytosolic calcium is increased during AF22,23,

proteolysis may be invoked by calcium dependent neutral proteases, calpain I and II.

Calpains are proteases which cleave mainly cytoskeletal and membrane-associated proteins

into ‘limited fragments’ without further degradation.24 In cardiac cells, calpains mediate

cell death in metabolically inhibited cultured rat cardiomyocytes and are involved in

troponin proteolysis and cross-linking following cardiac stunning and calcium overload.25-

27 In Chapter 8 an increased proteolytic activity in atrial tissue of patients with paroxysmal

and persistent lone AF was described. This increase was predominantly due to elevation

of calpain I activity and expression. Furthermore we observed that calpain I protein was

mainly localized at the nucleus and intercalated discs of atrial myocytes (Chapter 9). The

intensity of staining was low in sinus rhythm higher in paroxysmal AF and reached a

maximum in persistent AF. At the intercalated discs calpain can interact with Ca2+ and

thereby become an active proteinase and can degradate some important ion-channels like

the Na-channel28 and Kv1.519, but also several proteins directly involved in excitation-

contraction coupling.29 At the nucleus30 calpain can induce degenerative features leading

to apoptosis, which is observed in human AF (Figure 3).31

The role of calpain in cellular changes underlying the electrophysiological (Chapter

5), ion-channel (Chapter 5) and structural remodeling (Chapter 9) was examined. The

amount of structural and ultra-structural changes in the atrial tissue was examined by light

microscopy and electron microscopy, respectively. Calpain activity correlated with the

AF

calcium overload

induction calpain activity

degeneration proteins

necrosis/apoptosis

Figure 3.

143

General discussion

expression levels of ion-channel proteins, the degree of structural changes, measured AERP

and the rate adaptation coefficient of AERP. The results suggest that induction of calpain

activation represents a missing link between the calcium overload observed in AF22 and

remodeling of atrial myocytes during AF (Figure 3).

Integrated model for remodeling processes in AF

Evidence for AF induced Calcium overloadWhile studying the ion-channel gene expression adaptation mechanisms in human

AF, it became clear that the central feature in all these processes is calcium, in particular

calcium overload. Most investigations recognize the direct and indirect role of calcium

and our studies support the notion of calpain activation in AF. This notion is appealing

since it provides the molecular link between AF induced calcium overload and remodeling

processes, which was still not identified. Several lines of evidence for the crucial role of

AF induced calcium overload and subsequent calpain induction are reported below.

It has been proposed that atrial contractile dysfunction occurs after short-term and

chronic AF32-34 (Figure 1A,B). Contractile dysfunction after chronic AF is most likely

related to the cellular alterations in atrial myocytes35,36 reflected by structural alterations,

probably induced by proteolysis. The explanation for the atrial dysfunction after short-

term AF might be an increase in cytosolic Ca2+ due to the high rate of atrial activation. Fast

successive action potentials inhibit a proper sarcoplasmic reticulum Ca2+ re-uptake, resulting

in elevated cytosolic Ca2+, possibly impairing the excitation-contraction coupling and

contractile function.3,5,15,37,38 Ausma and coworkers showed that sarcolemma-bound Ca2+

and Ca2+ deposits in mitochondria increased markedly up to 2 weeks in experimental AF

and tends to regress towards normal levels at 4 and 8 weeks of AF (Figure 1A).22

Unfortunately, for Ca2+ localization they used antimonate based methods, which limit the

visualization of overall Ca2+ load at subcellular sites like the sarcoplasmic reticulum. Other

preliminary data showed that atrial tachycardia causes an immediate increase in cytoplasmic

Ca2+ concentration, which results in impaired Ca2+ release and cellular contractile

dysfunction after the cessation of tachycardia.23

Also other signaling pathway(s) by which AF leads to changes in atrial calcium

handling are involved. Recently completed experimental work suggests that T-type Ca2+

channels may mediate atrial tachycardia-induced electrical remodeling, because the T-

type Ca2+ channel blocker mibefradil limits both the ERP changes and AF promotion caused

by one week of rapid atrial pacing. Also in this case calcium overload would be prevented

by blocking a calcium channel.39

The similarity between the cellular ultrastructural changes caused by sustained AF

and those seen in hibernating myocardium40 have led to a suggestion that atrial ischemia

may play a role in triggering remodeling caused by AF. Whether ischemia occurs in AF is

144

Chapter 10

still debatable, but a reduced atrial blood flow in dogs with rapid pacing induced AF was

found and could result in atrial ischemia.41 A potential role for atrial ischemia is consistent

with the protective effect observed with blockade of the Na+/H+ exchanger in short term

tachycardia-induced atrial remodeling.42 In this model ischemia would give rise to a decrease

in intracellular pH, which leads to an exchange of intracellular hydrogen ions for

extracellular Na+ ions. Such an increase in intracellular Na+ results in a lower, or even

negative, equilibrium potential for the Na+/Ca2+ exchanger, thereby leading to a greater

magnitude of ‘reverse-mode’ functioning of the Na+/Ca2+ exchanger and therefore an influx

of Ca2+ ions.43 Alternatively, inhibition of the Na+/H+ exchanger may alter cellular ionic

homeostasis and combat calcium overload induced by ischemia.

In chapter 7 a mechanism that potentially modulates calcium overload in AF was

described. Studies on gene-expression of the endothelin system revealed that these systems

play a role in AF induced remodeling especially in patients with underlying valve disease.

We found that mRNA amounts of endothelin-1 are induced predominantly in persistent

AF with underlying valve disease. It is known that elevated endothelin-1 levels increase

intracellular calcium levels via the L-type Ca2+ channel44, indicating that calpain activation

also could play a role in AF with underlying valve disease, via different signal transduction

pathways. Moreover, increased amounts of BNP (Chapter 6) could be a compensatory

mechanism to reduce the intracellular calcium overload and thereby leading to relaxation

of the myocardial cell. BNP modulates cardiac calcium homeostasis via reduced intracellular

concentrations of cyclic adenosine monophosphate (cAMP), which inactivate the L-type

Ca2+ channel and activate the acetylcholine-dependent potassium channel leading to

repolarization of the action potential.18

Thus, several lines of evidence point to a central role of cellular calcium overload in

AF induced remodeling. Our work now reveals the potential role of calcium sensitive

processes that lead to changes in gene-expression and structural changes. Because elevated

levels of intracellular Ca2+ are known to activate proteolysis, this could result in increased

breakdown of myofilaments27,45 and ion-channel proteins (Chapter 9). In turn this could be

responsible for decreased contractility as well as for the vulnerability to AF.

Time course structural remodelingIn addition to electrophysiological, functional ion-current and ion-channel gene

expression changes, AF is associated with alterations in morphology. In Chapter 9 we

described an increase in degenerative contraction band necrosis observed in patients with

persistent and paroxysmal AF. Furthermore, we observed an increase in myocardial

hibernation (loss of sarcomeres and pale nuclei) only in patients with persistent AF, which

positively correlated with the duration of AF. This indicates that in human persistent AF

hibernation could be the specific structural change due to AF, in accordance with

145

General discussion

development of hibernation in the goat model for AF.40 We observed abundant degenerative

features in lone, paroxysmal AF. These could represent the prelude to the vulnerability to

AF by inducing dispersion of conduction (Figure 1B). Once persistent AF has been

developed, hibernation (which depends on prolonged periods of sustained AF) is more

abundant considered that cells liable to degeneration have now disappeared. These notions

are supported by the finding that in our patients with persistent AF, hibernation increased

with the duration of AF, while degeneration decreased. Possibly, hibernating myocardium

is protected against degeneration, as found after ischemic preconditioning.46 The reported

structural changes in human AF are in accordance with other studies. In humans structural

changes occur in atrial myocytes in patients with persistent AF.47 In patients with atrial

arrhythmias, myolysis and glycogen storage were only observed in a small number of

cells and that changes were frequently accompanied by lysosomal degeneration. In

experimental models these structural abnormalities appeared to be more pronounced when

the underlying pathology was aggravated by sustained AF.48,49 The occurrence of

degenerative myocardium could lead to increased dispersion of refractoriness and

conduction, which was found to enhance the inducibility and spontaneous occurrence of

idiopathic human AF.50 In addition to these defined changes in structural features, the

myocytes of AF patients displayed increased heterogeneity of cell size. Taken together,

the observed structural changes are indicative of a substantial deterioration of normal

tissue architecture, likely to promote AF through heterogeneity of atrial refractoriness50,51

and slowed atrial conduction.9,10,52 Since extreme physical stress, in combination with

sustained elevated cytosolic calcium levels, as in experimental AF22 often results in necrosis,

calpain could play an important role in this condition.53

In their model Ausma and coworkers noted mitochondrial enlargement, glycogen

accumulation, loss of sarcoplasmic reticulum and contractile elements in the atria of goats

subjected to chronic AF for upto 23 weeks.40 These changes resemble those observed in

the hibernating myocardium of the patients we investigated (Chapter 9). Recently the time

course of structural changes during AF in goats after 1-16 weeks of AF was studied.54

Here, the structural changes appeared to develop progressively, the earliest changes having

been noted after 1 week of AF and related to nuclear redistribution of heterochromatin

(Figure 1A). The nuclei showed a homogeneous distribution of chromatin, resembling

that found in embryonic/neonatal cardiomyocytes.40,55 After 4 weeks and at later times, AF

affected sarcomeres, glycogen, mitochondria and sarcoplasmatic reticulum simultaneously.

The loss of sarcoplasmatic reticulum and contractile proteins surely cause a decrease in

contractile force and hence atrial stunning54, which could be mediated by calpain.25,27

Since AF is promoted by slow conduction9,10,52,56 studies investigated the gap-junction

proteins, connexins, which play an important role in homogenous wavefront propagation

and conduction velocities in the heart.9,21,57,58 Gap-junctions are clusters of channels which

146

Chapter 10

span the closely apposed plasma membranes, forming cell-to-cell pathways. Connexins

are permeable to ions and small molecules up to 1 kDa in molecular mass, like second

messengers such as inositol triphosphate, cyclic AMP and calcium.

The initial data presented on changes in intercellular connexins were contradictory.

One study in the dog showed that AF increases connexin43 expression, the most abundant

connexin9 and another in the goat suggested that connexin43 is unaltered, but the distribution

of connexin40, mainly present in atrium was altered.57 In a recent study the gap junctional

changes in relation to stabilization of AF were studied.21 In goats that were in sinus rhythm

the distribution of connexin40, a connexin that gives high conductance, was homogeneous.

After 2 weeks in AF, which was the time associated with markedly increased intracellular

Ca2+ deposition22 and just before AF became sustained, heterogeneity in the connexin40

distribution was observed. The connexin40 distribution pattern correlated with the

occurrence of structural changes (myolysis) in atrial myocytes .

The structural changes, myolysis and heterogeneity of connexin40 distribution,

possibly relate to calcium induced calpain activity and explain the slow recovery (weeks

to months) after cardioversion of AF in patients33,59, the contractile dysfunction5,34 and the

electrophysiological changes during AF.4

Electrophysiological propertiesOver the past several years, AF-induced electrical remodeling and its underlying

mechanisms have been studied in substantial detail. In experimental studies part of the

underlying electrophysiological changes explaining the progressive nature of AF were

demonstrated.4,52 The increased tendency of the atria to fibrillate was paralleled by a

progressive shortening of the atrial effective refractory period (AERP) and loss of the

physiological rate adaptation of the refractory period which was termed atrial electrical

remodeling.4 The reduction in rate adaptation of the AERP is also observed in patients

with AF.60 All studies have shown that sustained atrial tachycardia decreases AERP and

changes occur over a period of days to weeks4,9,10,52, but AF can decrease AERP over a time

interval as short as several minutes (Figure 1A,B).15 Although the AERP reduction caused

by AF favors arrhythmia maintenance, it seems not be the only factor involved because

AF-induced AERP alterations become maximal well before AF-promoting effects

stabilize.4,10 One of the AF-promoting effects is tachycardia induced atrial conduction

slowing.9,10,52 It has a slower time course than AERP changes, probably due to delayed

onset of structural changes in the gap junctional remodeling21,57 and could account for at

least a part of the continued development of AF promotion after AERP changes have

stabilized. Whether gap junctional remodeling is caused by calpain induction is unknown,

but it is known that at least proteasome activity underlies a connexin43 degradation.61

147

General discussion

In addition to changes in the absolute value of AERP, atrial tachycardia alters the spatial

distribution of AERP. The spatial heterogeneity of AERP appears to be an important

determinant in the maintainance ot AF50,51,62,63 and there are indications that changed atrial

autonomic innervation, i.e. norephinephrine induced atrial sympathetic innervation, plays

an important role.64 Since norephinephrine causes elevation of the intracellular calcium

concentration in atrial myocytes, calpains might be activated and represent the causal link

to the maintenance of AF.

The combination of electrophysiological changes caused by sustained atrial tachycardia

i.e. reduced AERP, diminished or reversed adaptation to rate, slowed conduction and

increased spatial AERP heterogeneity, and the underlying structural changes caused by

calcium overload induced calpain ativity would be expected to promote AF maintenance

by enhancing the number of functional reentry circuits during AF.

Future perspectives

Pre-conditioningThe specific structural change induced by AF (Chapter 9) seemed to resemble chronic

hibernating myocardium.40 It is generally believed that by down-regulating their function,

cardiomyocytes adapt to a lowered oxygen availability and thereby restore the oxygen

supply/demand ratio. Part of the atrial cardiomyocytes acquired a dedifferentiated

phenotype, by re-expression of typical embryonic proteins. Furthermore, there is indirect

evidence that dedifferentiated,55 hibernating cardiomyocytes tolerate ischemia better than

non-dedifferentiated cardiomyocytes.65 It could be hypothesised that endogenous protective

mechanisms, such as an increased expression of certain heat shock proteins (stress induced

proteins which protects the cell against damage)66, are up-regulated in AF induced

hibernating myocytes. Although direct evidence of such up-regulation is missing, it is

known that ischemic preconditioning induces the efficient translation of stress proteins.46

Several of these proteins are subsequently translocated to the nucleus, possibly to protect

against degradation of DNA that has become more susceptible to degradation due to a

transformation of the chromatin organisation into a nuclease-sensitive conformation (as is

the case in apoptosis).67,68 Heat shock proteins, such as Hsp70, Hsp27 and αB-crystallin

are known to protect against ischemic cardiac damage. Unlike ischemic preconditioning,

which also attenuates apoptotic cell death induced by ischemia/reperfusion in a pig model

of short-term hibernation69, mRNA expression of Hsp70 and several other apoptosis-

modulating proteins was not altered in the ventricle during coronary stenosis nor during

subsequent stunning.70 Still it could be worthwhile investigating the role of protective heat

shock proteins in AF.

148

Chapter 10

Figure 4.

Example of an agarose gel. Here the L-type calcium channel α1 subunit and the GAPDH are given. H9c2 cells

were stimulated 0, 12, 24 and 36 hours and 36 hours stimuation combined with 48 hours recovery. The amount

of L-type Ca2+ channel increased during the different stimulation protocols.

0 12 24 36 recovery M

GAPDH

L-type Ca2+ channel

Figure 4

Possible clinical relevanceThe chance of successful chemical cardioversion and/or prevention of AF is dependent

on the duration of AF. This clinically observed diminished efficacy of cardioversion therapy

after long term AF cannot only by explained by the occurrrence of electrical remodeling.

The ion-channel protein remodeling and structural remodeling probably also affect the

electrophysiological function of the atrial myocardium.

In patients with persistent AF, there is a correlation between the duration of AF and

the time needed to recover atrial contractile function after cardioversion.33,71 The increase

in calpain activity which could lead to structural remodeling of the atrial myocytes might

give an explanation for the delay in recovery of contractile function in the atria after

conversion to sinus rhythm as seen in patients with persistent AF. Interference with the

calpain pathway by pharmacological intervention might represent an important new

therapeutic strategy to decrease protein degradation and thereby reduce the vulnerability

to AF. Calpain inhibitors as therapeutic agents are already used in nerve and muscle

degeneration72, but their potential benefit in heart diseases is not studied yet. After restoration

of normal sinus rhythm it may take the cardiomyocytes a certain period to rebuild a normal

amount of sarcomeres, if that is still possible at all.73 Data describing the recovery of ion-

channel protein expression are lacking, but a few reports describe, in contrast to the structural

remodeling, reversal of electrical remodeling in human AF after cardioversion.74,75 Since

AF induces remodeling in the atria it is essential to restore sinus rhythm as soon as possible,

thereby preventing the continuation of the atrial structural, ion-channel protein and electrical

remodeling.

Furthermore, differences in adaptation mechanisms which were found between patients

with lone AF and AF with underlying heart disease suggest the need for different

pharmacological treatment. For example, patients with elevated levels of endothelin could

be treated with an endothelin receptor antagonist, which has been shown to normalize

alterations in expression of various cardiac genes (like normalization of ryanodine receptor,

sarcoplasmic reticulum Ca2+ ATPase, angiotensin-converting enzyme, angiotensin II type

I receptor and prepro-endothelin 1) in failing myocardium.76

149

General discussion

Figure 5.

Calpain activation measurement in H9c2 cells.

The cells were stimulated 0, 20 and 30 hours.

An increase in calpain activity was found in

stimulated cells and inhibited by calpain I

inhibitor by 65%.hours of electrical stimulation0 5 10 15 20 25 30

calpain activation (arbitrary units)

0

100

200

300

400

500

calp

ain

acti

vati

on (

arb

itra

ry u

nit

s)

New Experimental model for AFTo mimic human AF experimental models were developed. AF is studied in more

detail in different animal models. The dog10,38,52,77 and goat models4 for AF are well studied

and characterized. For studying molecular and cellular mechanisms the animal models

have disadvantages. First, in the animal a lot of (unknown) parameters can interfere with

the results. Second, animal models take time and are expensive and the third important

disadvantage is for the animal itself. These are reasons to think of a different experimental

model for AF, especially for studying the fundamental calcium sensitive pathways. A pos-

sible new model could be the electrical stimulation of myocardium cells. Pilot experi-

ments have been done for H9c2 cells, which are rat myoblast cells and are able to differen-

tiate into myotubes. H9c2 cells appear to be unique in that they express the cardiac isoforms

of the L-type Ca2+ channel alpha 1-subunit mRNA (data not shown). Another good candi-

date cell line is the immortalised HL-1 cells. These are contracting mouse atrial cells and

when electrically stimulated could mimic human AF.

For the electrical stimulation experiment a special culture flask was developed by

Popta and Henning. Pilot experiments were done to investigate changes in L-type Ca2+

channel mRNA expression and for measuring the calpain activation during electrical stimu-

lation in H9c2 myotubes. Therefore, the cells were stimulated for 12, 24 and 36 hours

followed by three days recovery (0.5 Hz, 25V). Figure 4 shows that in stimulated H9c2

cells the mRNA expression of the L-type Ca2+ a1 subunit was increased compared to

unstimulated H9c2 cells. Thus the H9c2 cell line may prove to be useful when studying

the regulation of subtype-specific Ca2+ channel gene expression. The second set of experi-

ments were done to measure the calpain activity with a calpain specific fluorogenic sub-

strate. The calpain activity increased significantly after 20 and 30 hours of stimulation

(Figure 5). This activity could be reduced by 65% by a specific inhibitor of calpain I.

These two sets of pilot experiments reveal that these cell models could represent excellent

models for studying the signaling pathways activated by electrical stimulation and the

tential benefits of pharmacological intervention in AF.

150

Chapter 10

In conclusion, the combination of electrophysiological, ion-channel protein and struc-

tural remodeling caused by sustained AF would be expected to promote AF maintenance

by increasing the calcium overload induced calpain activity leading to induction of the

number of functional reentry circuits during AF. New cell models could be benificial for

studying the signaling pathways activated by electrical stimulation and pharmacological

intervention in AF.

Reference List


of Copenhagen. Thesis 1975;



Arch Intern Med 1996; 156:2585-2592.







6. Lai LP, Su MJ, Lin J, et al. Down-regulation of L-type calcium channel and sarcoplasmic reticular Ca2+ -

ATPase mRNA in human atrial firbillation without significant change in the mRNA of ryanodine receptor,

calsequestrin and phospholamban: an insight into the mechanism of atrial electrical remodeling. J Am Coll

Cardiol 1999; 33:1231-1237.

7. Ohkusa T, Ueyama T, Yamada J, et al. Alterations in cardiac sarcoplasmic reticulum Ca2+ regulatory

proteins in the atrial tissue of patients with chronic atrial fibrillation. J Am Coll Cardiol 1999; 34:255-263.

8. Boyett MR, Jewell BR. Analysis of the effects of changes in rate and rhythm upon electrical activity in the

heart. Prog Biophys Molec Biol 1980; 36:903-923.

9. Elvan A, Wylie K, Zipes DP. Pacing-induced chronic atrial fibrillation impaires sinus node function in

dogs: electrophysiological remodeling. Circulation 1996; 94:2953-2960.







13. Van Wagoner DR, Pond AL, McCarthy PM, et al. Outward K+ current densities and Kv1.5 expression are

reduced in chronic human atrial fibrillation. Circ Res 1997; 80:1-10.






dependency and effects of antiarrhythmic drugs. Circulation 1998; 97:2331-2337.

17. Gaspo R, Sun H, Fareh S, et al. Dihydropyridine and beta adrenergic receptor binding in dogs with

tachycardia-induced atrial fibrillation. Cardiovasc Res 1999; 42:434-442.

18. Clemo HF, Baumgarten CM, Ellenbogen K, et al. Atrial natriuretic peptide and cardiac electrophysiology:

autonomic and direct effects. J Cardiovasc Electrophysiol 1996; 7:149-162.

19. Mays DJ, Foose JM, Philipson LH, et al. Localization of the Kv1.5 K+ channel protein in explanted

cardiac tissue. J.Clin.Invest. 1995; 96:282-292.

20. Goette A, Arndt M, Röcken C, et al. Regulation of angiotensin II receptor subtypes during atrial fibrilla-


21. Van der Velden HMW, Ausma J, Rook MB, et al. Gap junctional remodeling in relation to stabilization of

atrial fibrillation in the goat. Cardiovasc Res 2000; 46:476-486.



151

General discussion


Circulation 1999; 100:I-200

24. Suzuki K, Imajoh S, Emori Y, et al. Calcium-activated neutral protease and its endogenous inhibitor.

Activation at the cell membrane and biological function. FEBS Letters 1987; 220:271-277.






Circ Res 1997; 80:393-399.

28. Cohen SA. Immunocytochemical localization of rH1 sodium channel in adult rat heart atria and ventricle.

Presence in terminal intercalated disks. Circulation 1996; 94:3083-3086.

29. Laflamme MA, Becker PL. G(s) and adenylylcyclase in transverse tubules of heart: implications for cAMP-

dependent signaling. Am.J.Physiol. 1999; 277:H1841-H1848

30. Lane RD, Allan DM, Mellgren R. A comparison of the intracellular distribution of mu-calpain, m-calpein

and calpastatin in proliferating human A431 cells. Exp Cell Res 1992; 203:5-16.

31. Aime-Sempe C, Folliguet T, Rucker-Martin, et al. Myocardial cell death in fibrillating and dilated human

right atria. J Am Coll Cardiol 1999; 34:1577-1586.

32. Manning WJ, Leeman DE, Gotch P, et al. Pulsed evaluation of atrial mechnical function after electrical

cardioversion of atrial fibrillation. J Am Coll Cardiol 1989; 13:617-623.



34. Daoud EG, Marcovitz P, Knight B, et al. Short-term effect of atrial fibrillation on atrial contractile func-


35. Van Gelder IC, Brundel BJJM, Henning RH, et al. Alterations in gene expression of proteins involved in

the calcium handling in patients with atrial fibrillation. J Cardiovasc Electrophysiol 1999; 10:552-560.

36. Brundel BJJM, Van Gelder IC, Henning RH, et al.Gene expression of proteins influencing the calcium

homeostasis in patients with persistent and paroxysmal AF. Cardiovasc Res 1999;42:443-454.

37. Tieleman RG, Van Gelder IC, Crijns HJ, et al. Early recurrences of atrial fibrillation after electrical

cardioversion: A result of fibrillation induced electrical remodeling of the atria? J Am Coll Cardiol 1998;

31:167-173.



39. Fareh S, Bénardeau A, Thibault, et al. The T-type Ca2+ channel blocker mibefradil prevents the develop-

ment of a substrate for atrial fibrillation by tachycardia-induced atrial remodeling in dogs. Circulation

1999; 100:2191-2197.



41. Jayachandran JV, Winkle W, Sih HJ, et al. Chronic atrial fibrillation from rapid atrial pacing is associated

with reduced atrial blood flow: a positron emission tomography study. Circulation 1998; 98:I-209

42. Jayachandran JV, Zipes DP, Weksler J, et al. Role of the Na+/H+ exchanger in short-term atrial electro-

physiological remodeling. Circulation 2000; 101:1861-1866.

43. Levi AJ, Dalton GR, Hancox JC, et al. Role of intracellular sodium overload in the genesis of cardiac

arrhythmias. J Cardiovasc Electrophysiol 1997; 8:700-721.

44. He JQ, Pi Y, Walker J, et al. Endothelin-1 and photoreleased diacylglycerol increase L-type Ca2+ current

by activation of protein kinase C in rat ventricular myocytes. J Physiol (Lond) 2000; 524:807-820.

45. Gorza L, Menabo R, Di Lisa F, et al. Troponin T cross-linking in human apoptotic cardiomyocytes. Am

J Pathol 1997; 150:2087-2097.

46. Tanaka M, Fujiwara H, Yamasaki K, et al. Expression of heat shock protein after ischemic precondition-

ing in rabbit heart. Jpn Circ J 1998; 62:512-516.

47. Mary-Rabine L, Pham TD, Hordof A, et al. The relationship of human atrial cellular electrophysiology to

clinical function and ultrastructure. Circ.Res. 1983; 52:188-199.

48. Boyden PA, Tilley LP, Albala A, et al. Mechanisms for atrial arrhythmias associated with cardiomyopa-

thy: a study of feline hearts with primary myocardial disease. Circulation 1984; 69:1036-1047.

49. Boyden PA, Tilley LP, Pham T, et al. Effects of left atrial enlargement on atrial transmembrane potentials

and structure in dogs with mitral valve fibrosis. Am J Cardiol 1982; 49:1896-1908.

50. Ramanna H, Hauer RNW, Wittkampf FHM, et al. Identification of the substrate of atrial vulnerability in

patients with idiopathic atrial fibrillation. Circulation 2000; 101:995-1001.

51. Fareh S, Villemaire C, Nattel S. Importance of refractoriness heterogeneity in the enhanced vulnerability

152

Chapter 10

to atrial fibrillation induction caused by tachycardia-induced atrial electrical remodeling. Circulation 1998;

83:2202-2209.

52. Morillo CA, Klein GJ, Jones D, et al. Chronic rapid atrial pacing. Structural, functional, and electrophysi-

ological characteristics of a new model of sustained atrial fibrillation. Circulation 1995; 91:1588-1595.

53. Majno G, Joris I. Apoptosis, oncosis and necrosis. An overview of cell death. Am J Pathol 1995; 146:3-

15.

54. Ausma J, Lenders MH, Duimel H, et al. Time course of structural changes in atria after atrial fibrillation

in goats: existence of hibernating myocardium. in press 2000;

55. Ausma J, Wijffels M, Van Eys G., et al. Dedifferentiation of atrial cardiomyocytes as a result of chronic

atrial fibrillation. Am J Pathol 1997; 151:985-997.



57. Van der Velden HMW, Van Kempen MJA, Wijffels MCEF, et al. Altered pattern of connexin40 distribu-

tion in persistent atrial fibrillation in the goat. J Cardiovasc Electrophysiol 1998; 9:596-607.

58. Britz-Cunningham SH, Shah MM, Zuppan CW, et al. Mutations of the connexin43 gap-junction gne in

patients with heart malformations and defects of laterality. N Engl J Med 1995; 332:1323-1329.

59. Van Gelder IC, Crijns HJ, Blanksma PK, et al. Time course of hemodynamic changes and improvement of

exercise tolerance after cardioversion of chronic atrial fibrillation unassociated with cardiac valve disease.

Am J Cardiol 1993; 72:560-566.

60. Attuel P, Childers R, Cauchemez B, et al. Failure in the rate adaptation of the atrial refractory period: its

relationship to vulnerability. Intern J Cardiol 1982; 2:179-197.

61. Laing JG, Tadros PN, Saffitz J, et al. Proteolysis of connexin43-containing gap junctions in normal and

heat-stressed cardiac myocytes. Cardiovasc Res 1998; 38:711-718.

62. Liu L, Nattel S. Differing sympathetic and vagal effects on atrial fibrillation in dogs: role of refractoriness

heterogeneity. Am J Physiol 1997; 273:H805-H816

63. Wang J, Liu L, Feng, et al. Regional and functional factors determining induction and maintenance of

atrial fibrillation in dogs. Am J Physiol 1996; 271:H148-H158

64. Jayachandran JV, Sih HJ, Winkle W, et al. Atrial fibrillation produced by prolonged rapid atrial pacing is

associated with heterogeneous changes in atrial sympathetic innervation. Circulation 2000; 101:1185-

1191.

65. Vanoverschelde JLJ, Wijns W, Depré, et al. Mechanisms of chronic regional postischemic dysfunction in

humans: new insights from the study on non-infarcted collateral dependent myocardium. Circulation 1993;

87:1513-1523.

66. Martin JL, Mestril R, Hilal-Dandan R, et al. Small heat shock proteins and protection against ischemic

injury in cardiac myocytes. Circulation 1997; 96:4343-4348.

67. Arrigo AP, Suhan JP, Welch WJ. Dynamic changes in the structure and intracellular locale of the mamma-

lian low-molecular-weight heat shock protein. Mol Cell Biol 1988; 8:5059-5071.

68. McMillan DR, Xiao X, Shao L, et al. Targeted disruption of heat shock transcription factor 1 abolished

thermotolerance and protection against heat-inducible apoptosis. J Biol Chem 1998; 273:7523-7528.

69. Maulik N, Yoshida T, Engelman RM, et al. Ischemic preconditioning attenuates apoptotic cell death

associated with ischemia/reperfusion. Mol Cell Biochem 1998; 186:139-145.

70. Bartling B, Hoffmann J, Holtz J, et al. Quantification of cardioprotective gene expression in porcine short-

term hibernating myocardium. J Mol Cell Cardiology 1999; 31:147-158.

71. Van Gelder IC, Crijns HJ, Van Gilst WH, et al. Prediction of uneventful cardioversion and maintenance of

sinus rhythm from direct-current electrical cardioversion of chronic atrial fibrillation and flutter.

Am.J.Cardiol. 1991; 41

72. Stracher A. Calpain inhibitors as therapeutic agents in nerve and muscle degeneration. Ann N Y Acad Sci

1999; 28:52-59.

73. Ausma J, Duimel H, Wouters L, et al. Structural atrial changes induced in the goat by 16 weeks of atrial

fibrillation are still present 8 weeks after cardioversion. Europace 2000; 1:B12

74. Hobbs WJC, Fynn S, Todd D, et al. Reversal of atrial electrical remodeling after cardioversion of persis-

tent AF in humans. Circulation 2000; 101:1145-1151.


cardioversion of persistent atrial fibrillation. Circulation 1998; 98:2860-2865.

76. Sakai S, Miyauchi T, Yamaguchi. Long-term endothelin receptor antagonist administration improves al-

terations in expression of various cardiac genes in failing myocardium of rats with heart failure. Circula-

tion 2000; 101:2849-2853.

77. Yue L, Feng J, Li GR, et al. Transient outward and delayed rectifier currents in canine atrium: properties

and role of isolation methods. Am J Physiol 1996; 270:H2157-H2168

153

Summary

Summary

Clinical and experimental studies showed that electrical and contractile remodeling

occurred early after onset of atrial fibrillation. Both processes could be reduced by blocking

the L-type Ca2+ channel suggesting the notion that changes in the calcium homeostasis

triggered by tachycardia induced intracellular calcium overload, play a pivotal role in the

induction of these remodeling processes. To obtain insight in the underlying molecular

mechanisms we first studied the molecular remodeling of proteins, which influence the

calcium homeostasis in a heterogeneous group of AF patients (Chapter 2 and 3). We found

that reductions in mRNA and protein expression of the L-type Ca2+ channel and sarcoplasmic

reticulum Ca2+ ATPase occurred predominantly in patients with persistent AF. Furthermore,

mRNA expression was found to be dependent on the duration of persistent AF. Patients

with >6 months duration of AF revealed reductions in mRNA in contrast to patients with

<6 months duration of AF. In these patients no changes in mRNA expression were found.

Secondly, we investigated molecular changes in ion-channels that contribute

importantly to the action potential duration. Apart from the L-type Ca2+ channel, we

investigated the contribution of potassium channel gene expression in right atrial appendages

in paroxysmal and persistent AF, since an increase in potassium channel amount or activity

could explain the electrical remodeling (Chapter 4). Reductions in mRNA and protein

levels were found for several K+ channels in patients with persistent AF. In patients with

paroxysmal AF these reductions were observed predominantly at the protein level and not

at the mRNA level. In addition, the regulation of L-type Ca2+ channel and several K+

channels and its relation to AERP in patients with persistent and paroxysmal AF was

studied (Chapter 5). We demonstrated a positive correlation between the ion-channel protein

expression of L-type Ca2+ channel, Kv4.3, Kv1.5, HERG, minK and Kir3.1 and the AERP

but also with the rate adaptation of AERP in patients with persistent and paroxysmal AF.

The correlation between ion-channel protein amounts and AERP indicate that ion-channel

protein remodeling, beside the electrical remodeling plays an important role in the

vulnerability to AF. The reductions in L-type Ca2+ channel could represent a possible

explanation for the electrophysiological changes during AF. Furthermore, the data indicate

that reduced ion-channel protein expression occurres due to high atrial rate. We called this

phenomenon ion-channel remodeling to describe the AF induced changes in ion-channel

protein expression.

The impact of other compounds such as the natriuretic peptide system (Chapter 6)

and the endothelin system (Chapter 7) were also studied. These studies revealed the

influence of concomitant valvular disease in patients with persistent AF on mRNA

expression of neurohumones. The right atrial appendage of these patients showed increased

levels of ANP, BNP and endothelin-1 in combination with reduced expression of their

receptors. Possibly the increased ANP and BNP mRNA enhance the myocyte relaxation

properties by combatting the calcium overload. An increase in endothelin-1 might generate

an additional increase in intracellular calcium concentrations by activation of the L-type

154

Ca2+ channel.

A remarkable finding during the study of mRNA and ion-channel protein remodeling

was the discrepancy between changes in mRNA and protein levels in patients with

paroxysmal AF (Chapter 4 and 5). Whereas ion-channel protein levels of the examined L-

type Ca2+ channel, Kv1.5, Kir3.1 and minK were substantially decreased, the mRNA

levels were essentially unaffected in paroxysmal AF. This discrepancy prompted us to the

role of an adaptive mechanism which influence in AF was previously unknown, i.e. the

activation of a proteolytic system. Different proteolytic pathways could be involved in AF.

Since cytosolic calcium is increased during AF, proteolysis may be invoked by calcium

dependent neutral proteases, calpain I and II. In Chapter 8 the increased proteolytic activity

during paroxysmal and persistent lone AF due to activation of the calpain pathway was

described. This increase seemed to be predominantly due to elevated activation and

expression of calpain I.

In Chapter 9 we examined a variety of molecular changes in atrial tissue from patients

with paroxysmal and persistent lone AF and related them to the level of calpain activity.

Immunohistochemical detection of calpain I demonstrated increased staining at the

intercalated disk and in the nucleus of atrial myocytes of AF patients. Accordingly, calpain

activity was increased in patients with AF. Furthermore, an increased number of

degenerative myocytes was observed in both patient groups with AF. Hibernating myocytes

were only present in persistent AF and numbers increased with the duration of AF. Finally,

calpain activity correlated inversely with the expression levels of ion-channel proteins,

the degree of structural changes and the rate adaptation coefficient of AERP. These results

strongly suggest that induction of calpain activation represents the missing link between

the calcium overload observed in AF and remodeling of atrial myocytes during AF.

The incidence of AF increases due to ageing of the population and AF has the tendency

to promote itself. Currently the arrhythmogenic electrophysiological changes (electrical

remodeling) are well known, but can not explain by themselves that ‘AF begets AF’. This

thesis shows significant molecular and ultrastructural changes related to cellular hibernation.

The latter support the notion of a second factor in AF promotion, since these - in part -

form the basis for conduction slowing and dispersion of conduction and refractoriness.

This thesis also shows a novel mechanism of protein remodeling taking place very early

after AF onset. This mechanism is driven by calpain I activation. Future studies on

pharmacological interventention in AF induced protein remodeling, like calpain activation,

may prove effective in preventing atrial damage after AF onset. Such interventions may

also break the chain of events by which AF tends to beget AF.

155

Samenvatting

Samenvatting

Boezemfibrilleren is een hartritmestoornis welke resulteert in een sterke toename

van de hoeveelheid contracties in de boezemcel. Dierexperimentele en klinische studies

hebben laten zien dat snel na de inductie van boezemfibrilleren verkorting van de atriale

effectieve refractaire periode (AERP) en ook de actie potentiaal duur plaatsvindt. Daarnaast

treedt verminderde contractiliteit van de boezem op nadat boezemfibrilleren is teruggebracht

in normaal sinus ritme. Beide processen zijn te verbeteren door toediening van een L-type

calciumantagonist voorafgaand aan boezemfibrilleren, deze voorkomt namelijk dat cal-

cium de cel instroomt en kan leiden tot contractie van de boezem. Dit effect van een L-

type calciumantagonist suggereert dat een verstoring in de calciumhuishouding,

waarschijnlijk een verhoging van de hoeveelheid calcium in de boezemcel, ten grondslag

ligt aan beide processen.

Als eerste wilden we inzicht verkrijgen in de moleculaire mechanismen welke ten

grondslag liggen aan een verstoorde calcium huishouding. Daarvoor hebben we

veranderingen in genexpressie van eiwitten, die een belangrijke rol in de

calciumhuishouding spelen bestudeerd in hartoren van patiënten met paroxysmaal en

persisterend boezemfibrilleren (Hoofdstuk 2 en 3). We beschrijven dat de mRNA

(boodschapper RNA welke de informatie bevat voor de volgorde van de aminozuren in

het te vormen eiwit) en de eiwit expressie van het L-type calcium kanaal en van het

sarcoplasmatisch reticulum calcium ATPase zijn verminderd. Beide verminderingen zijn

het meest uitgesproken in patiënten met persisterend boezemfibrilleren (Hoofdstuk 2).

Daarnaast zien we een relatie tussen de mate van reductie in mRNA expressie en de duur

van boezemfibrilleren. Patiënten met boezemfibrilleren van langer dan 6 maanden geven

de meeste mRNA reducties en deze nemen af naarmate de duur van boezemfibrilleren

korter is (Hoofdstuk 3).

Ten tweede hebben we de moleculaire veranderingen in genexpressie bestudeerd van

ion-kanalen welke een belangrijke rol spelen in het bepalen van actie potentiaal duur.

Naast het L-type calcium kanaal hebben we diverse kalium kanalen betudeerd, aangezien

activatie van deze kanalen resulteert in repolarisatie en dus verkorting van de actiepotentiaal.

In hoofdstuk 4 beschrijven we reducties in mRNA en eiwitexpressie van verschillende

kalium kanalen in patiënten met persisterend boezemfibrilleren. In patiënten met

paroxysmaal boezemfibrilleren zagen we ook vermindering in kalium kanaal expressie

maar dan alleen in eiwit en niet mRNA hoeveelheid. Verder zijn in een groep patiënten

met boezemfibrilleren de AERP’s gemeten en deze correleren positief met de hoeveelheden

eiwit van het L-type calcium kanaal en de diverse kalium kanalen. Dit houdt in dat patiënten

met boezemfibrilleren minder ion-kanaal eiwitexpressie hebben gecombineerd met een

verkorting van de AERP. Dit resultaat laat zien dat boezemfibrilleren niet alleen leidt tot

verkorting van de AERP, maar parallel hieraan ook leidt tot vermindering in ion-kanaal

eiwitexpressie.

Naast de genexpressie van de calcium huishouding eiwitten en ion-kanaal eiwitten

156

hebben we ook gekeken naar de genexpressie van andere eventueel belangrijke systemen

in boezemfibrilleren; het natriuretische systeem (ANP en BNP) (Hoofdstuk 6) en het

endotheline systeem (Hoofdstuk 7). Deze studies laten zien dat de genoemde systemen in

patiënten met boezemfibrilleren in combinatie met hartfalen een rol spelen. De hartoren

van deze groep patiënten gaven hogere expressie hoeveelheden van ANP, BNP en

endotheline-1, maar de corresponderende receptoren waren verminderd in expressie. Een

verhoging in ANP en BNP hoeveelheden zullen leiden tot relaxatie van de boezemcel. Een

verhoging van de hoeveelheid endotheline-1 in de boezemcel zal leiden tot meer contractie,

waarschijnlijk door een stimulerende werking op het l-type calcium kanaal.

Een in het oog springende bevinding tijdens deze studies is de discrepantie tussen

mRNA- en eiwitexpressie van ion-kanaal eiwitten in patiënten met paroxysmaal

boezemfibrilleren (Hoofdstukken 4 en 5). We zien dat de eiwitexpressie van het L-type

calcium kanaal en de diverse kalium kanalen was verminderd in deze patiënten. Echter de

hoeveelheid mRNA van deze ion-kanalen bleef onveranderd. Deze discrepantie zorgde

ervoor dat we de rol van een tot dan toe onbekend adaptatie mechanisme, namelijk het

eiwit-afbraak (proteolyse) systeem, in boezemfibrilleren zijn gaan bestuderen. We hebben

de mogelijke invloed van verschillende proteolytische systemen in boezemfibrilleren

onderzocht (Hoofdstuk 8). We vonden dat de calcium afhankelijke protease, calpaine I, is

geactiveerd tijdens idiopatisch paroxysmaal en persisterend boezemfibrilleren. De

verhoogde activiteit in patiënten met persisterend boezemfibrilleren kan verklaard worden

door een toename in hoeveelheid van het calpaine I eiwit.

In hoofdstuk 9 laten we zien waar in de boezemcel het calpaïne I voorkomt. We zien

dat calpaïne I zich bevindt in de zeer nauwe opening tussen boezemcellen (de intercalated

disk) en in de kern van de boezemcel en de hoeveelheden zijn het hoogst in patiënten met

persisterend boezemfibrilleren. Daarnaast beschrijven we de relaties tussen de verhoogde

calpaïne I activiteit en verschillende moleculaire veranderingen veroorzaakt door

boezemfibrilleren. De mate van structurele verandering, degeneratief en hibernerend (in

‘winterslaap’), is gequantificeerd en nam toe in hartoren van patiënten met idiopatisch

paroxysmaal en persisterend boezemfibrilleren. Daarnaast observeerden we dat

hibernerende boezemcellen alleen voorkomen in persisterend boezemfibrilleren en dat de

hoeveelheid toeneemt met de duur van het boezemfibrilleren. Verder correleerde de mate

van structurele, maar ook de mate van ion-kanaal veranderingen en de duur van de AERP

met de calpaïne activiteit. Samengevat laten deze resultaten zien dat het calpaïne systeem

de ontbrekende schakel kan zijn tussen de calcium overmaat en de verschillende moleculaire

veranderingen tijdens boezemfibrilleren in de boezemcel.

Kortom, de gepresenteerde studies laten zien dat naast de electrische veranderingen

er zich ook belangrijke veranderingen voordoen op eiwit en structureel niveau in

boezemcellen van patiënten met boezemfibrilleren. Deze veranderingen kunnen de

waargenomen verminderde contractiliteit van de boezem verklaren. Er zijn duidelijke

aanwijzingen dat het calpaïne systeem hier een essentiële rol in speelt.

157

Dankwoord

Dankwoord

Ik hoor het mezelf nog regelmatig zeggen: “Ik ga nooooooit naar Groningen”. Dat was

ongeveer vijf jaren geleden in Amsterdam. Een paar maanden later waren we spullen aan

het inpakken vanwege de verhuizing naar Groningen. Ik had een onderzoeksplaats

geaccepteerd waar ik nooit direct op had gesolliciteerd. Ik ging werken bij de Klinische

Farmacologie. De afdeling was juist oververhuist van de oudbouw naar de nieuwbouw en

niet 1 of 2 maar 3 nieuwe hoogleraren (Wiek van Gilst, Dick de Zeeuw en Pieter de Graaff)

waren er aangesteld. Maar eigenlijk werkte ik niet voor de klinische farmacologie maar

voor de Cardiologie waar ook net een nieuwe hoogleraar was aangesteld (Harry Crijns).

Daarnaast had ik niet één maar twee begeleiders (Isabelle van Gelder en Rob Henning).U

kunt zich waarschijnlijk voorstellen dat dit wat verwarrend was in de beginfase.

Ik ging werken in een moleculair lab waar nog een andere AIO rondhing. Leo Deelman,

King of the Lab, vastgeplakt aan zijn 486-er zat hij dagelijks computerspelletjes te spelen

met Jepe, een vervangende dienstplicht vervullende bioloog, of illegale

computerprogramma’s van het internet te plukken. Isabelle van Gelder kwam toen nog

wekelijks langs om te kijken of er niet gelanterfant werd en de experimenten wel de gewenste

resultaten gaven. En resultaten waren er (gelukkig). Dit is mede mogelijk gemaakt doordat

er een enorme humane hartoor vooraad was aangelegd en ik meteen aan de slag kon.

Isabelle heeft in deze begin fase een belangrijke en essentiële rol gespeelt. Isabelle, ik wil

je dan ook hartelijk danken voor je enorme inzet om het onderzoek zo goed mogelijk uit te

laten voeren. Als er problemen waren dan reageerde je snel en accuraat, factoren welke

onmisbaar zijn voor het nabehoren uitvoeren van onderzoek. Verder was je een stimulerende

factor om resultaten op papier te krijgen in de vorm van een abstract dan wel een artikel.

Dit laatste heeft weleens tot strubbelingen geleid, maar toch heeft dat onze verdere

samenwerking niet kunnen schaden. Verder heb ik onze persoonlijke gesprekken erg kunnen

waarderen.

Aangezien de experimenten naar behoren verliepen kwam er plaats voor analytische

ondersteuning en wel eerst in de vorm van Marion Franke later Cecile Driessen en nog

later Mirian Wietses. Ik wil jullie alle drie ontzettend bedanken voor jullie inzet en

gezelligheid. Mirian en Simone Gschwend wil ik nog extra bedanken niet alleen omdat ze

mijn paranimfen wilden zijn, maar ook vanwege de prettige omgang en hun organisatie

vermogen.

Regelmatig had ik promotie besprekingen met Harry Crijns, Wiek van Gilst, Rob Henning

en Isabelle van Gelder. Harry, ik wil je bedanken voor je steun en inlevingsvermogen.

Moest ik tijdens de eerste promotiebesprekingen nog uitleggen wat DNA, RNA en eiwitten

waren, je hebt altijd de moeite genomen om mee-te-denken in het onderzoek ondanks je

niet zo rooskleurige visoenen (de eerste twee jaren van het onderzoek). Wiek van Gilst zei

nooit erg veel tijdens promotie- en werkbesprekingen, maar wat hij zei heb ik altijd als

waardevol ervaren. Verder heb ik me verbaasd over de danskwaliteiten van Wiek (nachtclub

158

in Stockholm tijdens congres 1997). Ik hoop dat ik tijdens mijn promotiefeest er nog een

keer van mag genieten!

Dan hebben we nog Rob Henning, ja ja ja ja. Rob heeft een gave en dat is om de meest

negatieve resultaten om te buigen naar iets briljants. Een andere niet geheel onbelangrijke

kwaliteit is de positieve instelling en de respectvolle benadering van ieder individu. Ik heb

met name in de laatste twee jaren van mijn promotie onderzoek dankbaar gebruik mogen

maken van je kwaliteiten. Ik besef me dat jij de succesfactor bent achter een aantal

belangrijke artikelen. Bedankt Rob en ik hoop in de toekomst weer met je samen te werken.

Naast de directe medewerkers in het lab waren er ook buiten het lab mensen aan het werk

voor dit onderzoek. Bij de cardiologie was vooral Anton Tuinenburg druk om nog meer

hartoren te verzamelen. Dankzij de door Robert Tieleman en Anton gemeten ‘actie potentiaal

duur’ waarden in de hartoren van patiënten kon er een brug geslagen worden tussen klinische

en laboratorium meetwaarden. Heren bedankt voor jullie inzet en gezelligheid. Robert wil

ik nog bedanken voor mijn inwijding in het mega congres gebeuren en de gezelligheid

aldaar (American College of Cardiology, Los Angeles, 1997).

Verder wil ik de mensen van het mollab bedanken voor de gezelligheid welke ertoe heeft

bijgedragen voor het succesvol afronden van dit boekje: Marry, bedankt voor de goede

gesprekken; Leo, bedankt voor je kennis maar zeker ook voor je humor en het grote avontuur

in Keulen (we weten nu waar de het gezegde ‘Het in Keulen horen donderen’ vandaan

komt); Marion; Cecile; Mirian; Edith; Marjolein; Rudolf (nooit gedacht dat jongens zo

goed konden roddelen); Soesja; Anton; Ulricke; Sara en Kaska.

Daarnaast wil ik alle medewerk(st)ers van de Klinische Farmacologie bedanken voor de

prettige sfeer. Dick de Zeeuw, je manier van benaderen heeft me gescherpt, dedankt

daarvoor. Ad Nelemans, mocht je ooit ons ‘cannabis’ receptor artikel afkrijgen, je weet me

te vinden! Ook de secretaresses van de klinische farmacologie (Alexandra, Ardy en Paula),

maar vooral ook van de Cardiologie (Gretha) wil ik bedanken voor hun onmisbare

ondersteuning.

Tijdens mijn promotie onderzoek heb ik samengewerkt met een paar andere vakgroepen.

Met regelmaat liep ik rond in het Laboratorium voor Celbiologie en Electronenmicroscopie.

Ik wil alle medewerkers van deze afdeling bedanken maar expliciet de vakkundige analist

EHB Blaauw (waar stond EHB ook alweer voor?). Bert heeft menig stukje weefsel van

patiënten en geiten in superdunne plakjes gesneden. Bert ik wil je nog hartelijk danken

voor je geweldige inzet en interesse in het onderzoek, maar ook voor de prettige sfeer bij

jullie op de afdeling. Dan kom ik meteen bij Han van der Want terecht. Han ook jij bedankt

voor je inzet, openhartigheid en je creatieve oplossingen. Verder wil ik nog Peter van der

Syde bedanken voor de opmaak van het proefschrift. Ik hoop in de toekomst ook weer met

jullie samen te werken.

In de eindfase van mijn onderzoek kwam ik Jannie Ausma tegen tijdens een congres in

Atlanta U.S.A.. Ik vertelde Jannie wat voor experimenten ik graag zou willen doen en

Jannie stelde meteen voor om eens in Maastricht te komen. Ik ben nog nooit iemand

tegengekomen met zo’n toewijding aan het onderzoek. Jannie, het was voor mij dan ook

erg prettig dat je me zo goed hielp met de experimenten, maar vooral ook met het schrijven

van het artikel (hoofdstuk 9). Ik ben me ervan bewust dat dit artikel zonder je kennis nooit

het huidige niveau zou hebben gehaald. Bedankt Jannie! Ik ben tweemaal in Maastricht

geweest en daar heeft Hans Duimel me geholpen bij de uitvoer van de experimenten. Hans

hartelijk bedankt voor je inzet en deskundige begeleiding gecombineerd met je (Brabantse)

gezelligheid.

Tsja, en dat ik na mijn promotie-onderzoek, het ter wereld brengen en (op)voeden van drie

kinderen en de hectiek waarmee dit gepaard gaat nog niet overspannen ben komt vooral

door mijn fantastische partner Marcus. Marcus, Myrthe, Joachim & Jona, ik wil jullie

bedanken voor het feit dat jullie me iedere dag laten zien waar het leven eigenlijk om

draait......

160

Documents

University of Groningen Molecular adaptations in human