From the Center for Cardiovascular Research, Washington University School
of Medicine, St Louis, Mo.
Correspondence to Jeffrey E. Saffitz, MD, PhD, Department of Pathology, Box 8118, Washington University School of Medicine, 660 S Euclid Ave, St Louis, MO 63110. E-mail saffitz{at}patholology.wustl.edu
Myocardial
stunning1 and hibernation2
have been subjects of intense laboratory and clinical investigations to
elucidate mechanisms responsible for contractile dysfunction after
transient ischemia or chronic hypoperfusion. Considerable
evidence has implicated the generation of oxygen-derived free radicals
and derangements in myocardial energy metabolism and
excitation-contraction coupling as major contributors to the
pathogenesis of myocardial stunning and
hibernation.3 4 Despite enormous progress in this
area, however, our understanding of the cellular pathophysiology of
contractile dysfunction in ischemic heart disease remains
incomplete.
In this issue of Circulation, Kaprielian et
al5 provide evidence for a novel mechanism
involving alterations in intercellular coupling that might contribute
to the pathogenesis of contractile dysfunction in hibernating
myocardium. Using digital image processing techniques and
confocal immunofluorescence microscopy, they
measured the amount of the major cardiac gap junction protein
connexin43 (Cx43) in left ventricular samples obtained from
patients at the time of coronary artery bypass graft surgery.
Biopsies were taken from well perfused, normally contracting wall
segments (identified with preoperative thallium scans and magnetic
resonance imaging studies) and from "reversibly ischemic"
segments (showing improved thallium uptake on stress/redistribution
images but no improvement in contractile function after
revascularization) or "hibernating" segments
(showing improved contractile performance in postoperative MRI
studies). Mean gap junction area, determined by measuring the amount of
Cx43 immunoreactive signal at intercellular junctions and expressed per
unit of intercalated disk area, was reduced by
Kaprielian et al5 developed and extensively
validated sophisticated analytical techniques to measure the amount and
spatial distribution of Cx43 in the heart. Despite the quantitative
rigor of the analytical methods used to measure Cx43 immunoreactive
signal, however, the data do not establish a causal role for diminished
intercellular coupling in arrhythmogenesis or contractile dysfunction.
No relation was shown between the extent of Cx43 downregulation and the
development of arrhythmias. In addition, the amount of Cx43
signal in intercalated discs was not significantly different in
reversibly ischemic versus hibernating tissues. Thus, claims
that pathophysiologically important differences
in these two types of contractile dysfunction may be attributable to
different degrees of Cx43 downregulation are not supported strongly by
the data. Nevertheless, a clear difference was observed between levels
of Cx43 expression in normal versus abnormal segments. This observation
focuses attention on the possibility that alterations in intercellular
coupling at gap junctions could directly or indirectly contribute to
the pathogenesis of hibernating myocardium. This idea is
worthy of further consideration.
Intercellular Coupling at Gap Junctions
Nearly all eukaryotic cells communicate with their
nearest neighbors by means of gap junctions, specialized regions of the
cell surface containing densely packed arrays of membrane-spanning
channels that directly link the cytoplasmic compartments of adjacent
cells.6 Ubiquitous among metazoans, gap junctions
play fundamental roles in mediating intercellular communication from
the early stages of embryogenesis to the establishment and
maintenance of coordinated multicellular functions in diverse
tissues. Gap junction channels are composed of subunit proteins called
connexins.7 More than a dozen individual
connexins (each identified numerically on the basis of its predicted
molecular mass in kilodaltons) have been cloned and sequenced. In most
differentiated tissues, individual parenchymal cells have been found to
express multiple connexins. Three different connexins have been
identified in cardiac myocytes: Cx43, connexin45 (Cx45), and connexin40
(Cx40).8 These cardiac connexins are expressed in
different amounts and combinations in different regions of the heart.
For example, ventricular myocytes express Cx43 and Cx45,
whereas atrial myocytes express Cx43, Cx45, and
Cx40.9 When expressed individually in
"communication-deficient" cells, each of the cardiac connexins
forms channels with distinct unitary conductances, voltage
sensitivities, and different capacities to pass anions, cations, or
fluorescent dyes of varying sizes and charge
densities.10 These observations suggest that the
number, spatial distribution, and connexin composition of gap junction
channels are important determinants of the conduction properties of a
given cardiac tissue. Regional alterations in connexin expression
phenotypes or gap junction distribution induced under a variety
of pathophysiological conditions could produce
considerable functional heterogeneity within the
ventricle.
Relatively little is known about the composition and molecular
regulation of the "flow of information" across gap junctions and
how this information coordinates the activities of groups of cells that
act together as a functional unit. Many low-molecular-weight (<1000 D)
regulatory and signaling molecules probably traverse gap junctions to
coordinate multicellular events involved in fundamental biological
processes, such as cell metabolism, signal transduction,
secretion, cell division, cell migration, and
morphogenesis.6 In electrically excitable tissues
such as smooth muscle and cardiac muscle, gap junction channels are
critical determinants of passive conduction
properties.11 It is important to remember,
however, that in addition to subserving an obvious role in
intercellular ionic conductance, gap junctions in the heart probably
facilitate the exchange of molecular information that regulates
metabolic activity and, possibly, contractile activity
among groups of coupled cells.
How Might Reduced Coupling Affect Contractile Function?
Because gap junctions play a central role in conduction,
alterations in connexin expression and gap junction distribution have
been implicated in the pathogenesis of reentrant arrhythmias in
patients with anatomic substrates characterized by slow,
heterogeneous conduction or conduction block. Kaprielian et
al5 speculated that by unmasking local
inhomogeneities in individual cell action potential durations, reduced
gap junction coupling in areas of hibernating myocardium
could disrupt wavefront propagation, and this might interfere with
local coordination of myocyte contraction. Although localized
uncoupling clearly contributes to the pathogenesis of conduction
abnormalities and unidirectional conduction
block,12 it is not so apparent how uncoupling
might contribute to contractile dysfunction. Before considering
potential mechanisms, it should be stressed that immunohistochemical
data reflect the amount but not necessarily the functional state of
Cx43 located in gap junctions. The amount of a connexin is certainly an
important determinant of total junctional conductance, but other
factors play important roles. For example,
immunostaining may not distinguish
phosphorylated and nonphosphorylated
isoforms of Cx43, although the extent of
phosphorylation can have a major impact on channel
conductance properties13 and turnover
dynamics.14 Most of the Cx43 in the normal heart
appears to be phosphorylated, but under
pathophysiological conditions, the relative
proportions of phosphorylated and
nonphosphorylated Cx43 could change with resultant
changes in channel function. Gap junction channel function also is
modulated by changes in intracellular Ca2+ and
H+ concentrations and by the lipid composition of
the membrane in which the channels reside.6 7 15
It is possible, therefore, that changes in the intracellular milieu of
dysfunctional myocytes could affect junctional coupling independent of
changes in the amount of Cx43. Intracellular acidosis and
Ca2+ overload may also affect
contractility directly, which further complicates
elucidation of potential causal relationships between changes in
coupling and contractility. Other mechanisms (both
direct and indirect) could link development of derangements in passive
electrical properties and the development of alterations in active
electrical properties, excitation-contraction coupling, and contractile
performance. For example, recent evidence indicates that some
voltage-sensitive potassium channels,16
voltage-gated sodium channels,17 and inositol
trisphosphate receptors18 are concentrated at
sites of intercellular junctions. Alterations in the number or
distribution of gap junctions could affect the distribution and
possibly even the function of sarcolemmal ion channels or the functions
of Ca2+ regulatory proteins in the sarcoplasmic
reticulum that would affect excitation-contraction coupling and
mechanical function.
Although Cx45 is clearly expressed in ventricular myocytes,
Kaprielian et al5 did not detect an
immunoreactive signal for Cx45, the other ventricular
connexin, in formalin-fixed tissues from patients undergoing bypass
surgery. They suggest that because no detectable changes in the levels
of other connexin isoforms occur in myocytes of hibernating tissues,
potential pathophysiological effects are likely
related to alterations in expression levels of Cx43 per se. However,
when a tissue expresses multiple connexins, the functional consequences
of a change in the expression level of a single connexin may be more
complex than simply reducing one type of channel. There is convincing
evidence from experiments with transfected cell lines that Cx43 and
Cx45 can combine to form hybrid gap junction channels composed of both
proteins.19 Channels composed exclusively of Cx45
have a lower unitary conductance than Cx43 channels and fail to pass
detectable amounts of the fluorescent dye Lucifer yellow (which
passes readily through Cx43 channels). Hybrid channels composed of Cx45
from one member of a cell pair and Cx43 from the other member have low
unitary conductance and do not pass Lucifer yellow, biophysical
properties that resemble those of pure Cx45 channels more closely than
Cx43 channels.19
This raises the possiblity that ventricular myocytes may be
coupled by a heterogeneous population of gap junction
channels, including homotypic Cx43 and Cx45 channels and hybrid
channels composed of both proteins. If the relative proportions of
homotypic and hybrid channels are related to the stoichiometries of
Cx43 and Cx45, then downregulation of Cx43 expression could affect the
numbers of both homotypic and hybrid channels even if the total level
of Cx45 were unchanged. Thus far, there has been no direct proof of the
existence of hybrid channels in the heart. Nevertheless, there is
emerging concensus among gap junction biologists that hybrid channels
can form and, if so, vastly extend the range of possible channel
properties. The known differences in channel properties of individual
connexins suggest that specific channel subtypes could subserve
specific functions, including electrical coupling,
metabolic coupling and coordination of contractile
activity.
New Tools for Elucidating Pathophysiological
Effects of Altered Connexin Expression
One promising approach to the study of the
physiological and
pathophysiological roles of different connexins in
the heart is to investigate mice in which expression of a specific
connexin has been manipulated genetically. Reaume et
al20 created a Cx43 knockout mouse through
homologous recombination techniques in embyronic stem cells. Mice that
are homozygous for the null mutation develop to term but die shortly
after being born. These animals consistently have a
malformation of the right ventricular outflow tract, which
has been implicated in their failure to survive once the maternal-fetal
circulation has been disrupted and the neonate becomes reliant on its
own cardiac function. Interestingly, a similar type of cardiac
malformation has been observed in mice expressing a transgene encoding
Cx43 under the control of a cytomegaloviral promoter that results in
targeted overexpression of Cx43 in neural crest cells that ultimately
participate in cardiac conotruncal development.21
Thus, both deletion and overexpression of Cx43 can be associated with
similar types of cardiac malformations, suggesting that either gain or
loss of intercellular coupling may have deleterious morphogenetic
consequences.
Recent studies have shown that mice heterozygous for the Cx43 null
mutation exhibit slow ventricular
conduction.22 Although these animals express only
50% of the wild-type level of Cx43 (an even greater degree of
downregulation than that seen by Kaprielian et
al5 in hibernating myocardium), they
show no outward signs of ventricular dysfunction. It should
be emphasized, however, that detailed studies of contractile function
have not been reported in Cx43 knockout mice. Furthermore, reduced
intercellular coupling in Cx43 heterozygous mice is global in
distribution. Contractile dysfunction in hibernating
myocardium could be a consequence, in part, of regional
loss of coordinated conduction.
Kaprielian et al5 report that myocardial
biopsies of both normally contracting and hibernating regions exhibited
ultrastructural abnormalities characterized by marked loss of
myofibrils. Depletion of myofibrils may not, by itself, account for
contractile dysfunction in hibernating myocardium, but the
combination of this structural alteration and a defect in coupling
could contribute to contractile dysfunction in hibernating
myocardium. It is possible that downregulation of Cx43
expression in chronically hibernating myocardium is only
one component of a generalized disassembly of subcellular organelles.
It is noteworthy, however, that although the control and hibernating
tissues demonstrated similar degrees of myofibrillar depletion, only
hibernating tissues showed downregulation of Cx43. This suggests that
diminished expression of Cx43 is not an obligatory feature of
degenerative changes in myocytes in hearts of patients with complex
ischemic heart disease. Further studies will be required to
elucidate mechanisms of downregulation of Cx43 expression in selected
regions of the heart and understand the potential
pathophysiological roles that uncoupling might play
in contractile dysfunction.
In conclusion, the time has come to broaden our thinking about
myocardial gap junction channels beyond their electrical coupling role
and to consider their possible roles in coordinating
metabolic and contractile functions as well.
Ventricular remodeling has both mechanical and
electrophysiological consequences. These
two major types of pathophysiological changes may
interact in complex ways to contribute to both contractile dysfunction
and enhanced arrhythmogenesis in patients with chronic ischemic
heart disease.
Acknowledgments
We thank Dr Edward Geltman for a helpful review of the
manuscript.
Footnotes
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
References
© 1998 American Heart Association, Inc.
Editorials
Do Alterations in Intercellular Coupling Play a Role in Cardiac Contractile Dysfunction?
Key Words: Editorials • myocardium • connexin43 • ischemia • proteins
23% in reversibly
ischemic segments and by 33% in hibernating tissue compared
with normally perfused regions. The average size of an individual gap
junction was also reduced by 13% and 30% in reversibly
ischemic and hibernating regions, respectively, due mainly to
loss of the largest gap junctions in myocytes of dysfunctional regions.
The authors suggest, on the basis of these observations, that reduction
of Cx43 content in gap junctions may contribute not only to
arrhythmogenesis by creating anatomic substrates of abnormal conduction
but also to wall motion impairment in hibernating
myocardium.
This article has been cited by other articles:
 |
|
 |
|
  T. Betsuyaku, N. S. Nnebe, R. Sundset, S. Patibandla, C. M. Krueger, and K. A. Yamada Overexpression of cardiac connexin45 increases susceptibility to ventricular tachyarrhythmias in vivo Am J Physiol Heart Circ Physiol, January 1, 2006; 290(1): H163 - H171. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Chu, A. N. Carr, K. B. Young, J.W. Lester, A. Yatani, A. Sanbe, M. C. Colbert, S. M. Schwartz, K. F. Frank, P. D. Lampe, et al. Enhanced myocyte contractility and Ca2+ handling in a calcineurin transgenic model of heart failure Cardiovasc Res, April 1, 2002; 54(1): 105 - 116. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Gutstein, G. E. Morley, D. Vaidya, F. Liu, F. L. Chen, H. Stuhlmann, and G. I. Fishman Heterogeneous Expression of Gap Junction Channels in the Heart Leads to Conduction Defects and Ventricular Dysfunction Circulation, September 4, 2001; 104(10): 1194 - 1199. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||