Biomedical Engineering Reference
In-Depth Information
expected to suppress excitability. We thus explored the possibility that dominant-
negative suppression of Kir2-encoded inward rectifier potassium channels in the
ventricle would suffice to produce spontaneous, rhythmic electrical activity.
Replacement of three critical residues in the pore region of Kir2.1 by alanines
(GYG144-146 AAA, or Kir2.1AAA) creates a dominant-negative construct [26]. The
GYG motif plays a key role in ion selectivity and pore function [67]. Kir2.1AAA and
GFP were packaged into a bicistronic adenoviral vector (AdEGI-Kir2.1AAA) and
injected into the left ventricular cavity of guinea pigs during transient cross-clamp of
the great vessels [48]. This method of delivery sufficed to achieve transduction of
~20% of ventricular myocytes. Myocytes isolated 3-4 days after in vivo transduction
with Kir2.1AAA exhibited ~80% suppression of
I
K1
, but the L-type calcium current
was unaffected.
Non-transduced (non-green) left ventricular myocytes isolated from AdEGI-
Kir2.1AAA-injected animals, as well as green cells from AdEGI-injected hearts,
exhibited no spontaneous activity, but fired single action potentials in response to
depolarizing external stimuli (Fig. 1a). In contrast, Kir2.1AAA myocytes exhibited
either of two phenotypes: a stable resting potential from which prolonged action
potentials could be elicited by external stimuli (7 of 22 cells, not shown) or
spontaneous activity (Fig. 1b). The spontaneous activity, which was seen in all cells
in which
I
K1
was suppressed below 0.4 pA/pF (at -50 mV; cf. >1.5 pA/pF in controls,
or 0.4-1.5 pA/pF in non-pacing Kir2.1AAA cells), resembles that of genuine
pacemaker cells; the maximum diastolic potential (- 60.7 ± 2.1 mV,
n
= 15 of 22
Kir2.1AAA cells, P < 0.05 t test) is relatively depolarized, with repetitive, regular and
incessant electrical activity initiated by gradual ''phase 4'' depolarization and a slow
upstroke [9, 32]. Kir2.1AAA pacemaker cells responded to
ȕ
-adrenergic stimulation
(isoproterenol) just as SA nodal cells do, increasing their pacing rate [10, 32].
Electrocardiography revealed two phenotypes in vivo. What we most often
observed was simple prolongation of the QT interval (not shown). Nevertheless, 40%
of the animals exhibited an altered cardiac rhythm indicative of spontaneous
ventricular foci. In normal sinus rhythm, every P wave is succeeded by a QRS
complex (Fig. 1c). In two of five animals after transduction with Kir2.1AAA,
premature beats of ventricular origin can be distinguished by their broad amplitude,
and can be seen to ''march through'' to a beat independent of, and more rapid than,
that of the physiological sinus pacemaker (Fig. 1d). In these proof-of-concept
experiments, the punctate transduction required for pacing occurred by chance rather
than by design, in that the distribution of the transgene throughout the ventricles was
not controlled. Nevertheless, ectopic beats, arising from foci of induced pacemakers,
cause the entire heart to be paced from the ventricle.
Our findings provide new insights into the biological basis of pacemaker activity.
The conventional wisdom postulates that pacemaker activity requires the highly
localized expression in nodal cells of ''pacemaker genes'', such as those of the HCN
family [64], although an importance of scarce
I
K1
density has also been recognized
[32]. Exposure to barium induces automaticity in ventricular muscle and myocytes
because of its time-and voltage-dependent block of
I
K1
[27, 31]. However, barium also
permeates L-type calcium channels in mixed solutions of Ca
2+
(4 mM) and Ba
2+
(1 mM) [62] and slows their inactivation [12], effects which make it difficult to
interpret barium effects strictly in terms of
I
K1
. Our dominant-negative approach is
durable and regionally specific; the barium effect is not.
