Cardiac-restricted angiotensin-converting enzyme overexpression causes conduction defects and connexin dysregulation

Abstract

Renin-angiotensin (RAS) system activation is associated with an increased risk of sudden death. Previously, we used cardiac-restricted angiotensin-converting enzyme (ACE) overexpression to construct a mouse model of RAS activation. These ACE 8/8 mice die prematurely and abruptly. Here, we have investigated cardiac electrophysiological abnormalities that may contribute to early mortality in this model. In ACE 8/8 mice, surface ECG voltages are reduced. Intracardiac electrograms showed atrial and ventricular potential amplitudes of 11% and 24% compared with matched wild-type (WT) controls. The atrioventricular (AV), atrio-Hisian (AH), and Hisian-ventricular (HV) intervals were prolonged 2.8-, 2.6-, and 3.9-fold, respectively, in ACE 8/8 vs. WT mice. Various degrees of AV nodal block were present only in ACE 8/8 mice. Intracardiac electrophysiology studies demonstrated that WT and heterozygote (HZ) mice were noninducible, whereas 83% of ACE 8/8 mice demonstrated ventricular tachycardia with burst pacing. Atrial connexin 40 (Cx40) and connexin 43 (Cx43) protein levels, ventricular Cx43 protein level, atrial and ventricular Cx40 mRNA abundances, ventricular Cx43 mRNA abundance, and atrial and ventricular cardiac Na(+) channel (Scn5a) mRNA abundances were reduced in ACE 8/8 compared with WT mice. ACE 8/8 mice demonstrated ventricular Cx43 dephosphorylation. Atrial and ventricular L-type Ca(2+) channel, Kv4.2 K(+) channel alpha-subunit, and Cx45 mRNA abundances and the peak ventricular Na(+) current did not differ between the groups. In isolated heart preparations, a connexin blocker, 1-heptanol (0.5 mM), produced an electrophysiological phenotype similar to that seen in ACE 8/8 mice. Therefore, cardiac-specific ACE overexpression resulted in changes in connexins consistent with the phenotype of low-voltage electrical activity, conduction defects, and induced ventricular arrhythmia. These results may help explain the increased risk of arrhythmia in states of RAS activation such as heart failure.

Fig. 1

Electrophysiological comparison of 6- to 8-wk-old wild-type (WT) mice, heterozygous (HZ) mice, and a homozygous mouse model of renin-angiotensin system activation that uses cardiac-restricted angiotensin-converting enzyme (ACE 8/8 mice). A: representative standard surface ECG limb leads I and II in WT, HZ, and ACE 8/8 mice. QRS amplitude of ACE 8/8 mice was markedly reduced. ACE 8/8 mice ECGs showed no evidence of P waves and irregular R-R intervals. All traces are presented at the same sweep speed and amplification. B: representative surface lead II and intracardiac (His lead) recordings from a WT mouse. There are distinct, high-amplitude atrial (A), Hisian (H), and ventricular (V) electrograms. Note low-amplitude and fractionated atrial activity in the ACE 8/8 mouse and prolonged AH and HV intervals compared with the WT mouse. Sweep speed and amplifier gain are the same in both cases.

Fig. 2

Evidence of arrhythmic risk in ACE 8/8 mice. A: ventricular tachyarrhythmic induction by burst pacing. An 18-beat burst of extrastimuli at a 50-ms cycle length was unable to induce ventricular tachycardia in a WT mouse, whereas a 13-beat burst of extra-stimuli at the same cycle length was sufficient to induce polymorphic ventricular tachycardia in an ACE 8/8 mouse. Representative surface ECG leads I and the intracardiac right atrial (RA) and His leads are shown. Only ACE 8/8 mice were inducible. B: representative sequential recordings of standard surface ECG limb lead II at the time of death of an ACE 8/8 mouse. a: Baseline ventricular rate was 420 min⁻¹, and QRS complexes are indicated by solid arrows. A premature beat (open arrow) initiates ventricular tachycardia at >1,200 min⁻¹. High-frequency bursts on the ECG, seen most clearly during the initial rhythm, represent diaphragmatic activity. b: Continuation of ventricular tachycardia with a “torsades de pointes” morphology. c: A more regular ventricular tachycardia that culminates in asystole, as recorded in d.

Fig. 3

Reduced connexin protein expression in ACE 8/8 mice. A, C, and E: Western blot analysis of atrial Cx40 and atrial and ventricular Cx43 expression in WT, HZ, and ACE 8/8 mice. B, D, and F: quantification of results from analysis in A, C, and E, respectively, expressed relative to GAPDH. Values are means ± SE (n = 6). *P < 0.05. G and H: anti-Cx43 antibody-stained sections of cardiac ventricles from WT and ACE 8/8 mice, respectively. Sections are representative of 4 similar results. Cx43 staining is shown as brown pigment. Arrows, Cx43 expression in the intercalated disk regions in WT and ACE 8/8 ventricles. Cx43 expression in ACE 8/8 ventricles remains in the intercalated disks, but expression level is markedly decreased.

Fig. 4

Relation between Cx43 and surface QRS amplitude in ACE 8/8 mice. A: Western analysis of Cx43 expression in the ventricle of WT, HZ, and ACE 8/8 mice as a function of the corresponding QRS complexes (surface lead II). B: ratio of phosphorylated (P1 + P2 + P3) Cx43 to the nonphosphorylated form (P0) for each genotype. *P < 0.05. C: ratio of phosphorylated (P1 + P2 + P3) to nonphosphorylated Cx43 plotted against the QRS amplitude (composite frontal vector calculated from leads I and aVF). QRS amplitude refers to frontal-plane amplitude of the QRS vector, calculated as √I² + aVF².

Fig. 5

Changes in cardiac Na⁺ current in ACE 8/8 cardiac myocytes. A: current-voltage curves for cardiac myocytes derived from WT and ACE 8/8 mice. Values are means ± SE (n = 3 animals and 10 cells for each group). There are no statistically significant differences in slope conductance. V_m, membrane potential. B: activation kinetics in ACE 8/8 and WT myocytes. Data are fitted to the Boltzmann activation curve: I/I_max = 1/[1 + e^{{V_h − V}/k} ], where V_h is the gating potential, k is the slope, and I/I_max is the ratio of the current at each holding potential to the maximum current calculated using a linear fit of the reversal region. Activation curve for ACE 8/8 mice is shifted to the left (V_h = −37 mV, n = 10) compared with WT cells (V_h = −32 mV, n = 10, P < 0.001). C: alterations in inactivation kinetics in ACE 8/8 compared with WT myocytes. Data are fitted to the Boltzmann inactivation curve described in B. Similar to activation kinetics, inactivation curve for ACE 8/8 cardiac myocytes is shifted to the left (V_h = −65 mV in ACE 8/8 myocytes, n = 11) compared with WT cells (V_h = −55 mV, n = 13, P < 0.001).

Fig. 6

Alterations in connexin mRNA abundances with cardiac ACE expression. Relative atrial and ventricular mRNA abundances for Cx40 (A), Cx43 (B), Cx45 (C), Kcnd2 (D), Cacna1 (E), and Scn5a (F) determined by quantitative real-time PCR are shown for WT (n = 4), HZ (n = 4), and ACE 8/8 (n = 4) mice. Relative gene expression was assayed and normalized to endogenous β-actin. *P < 0.01 vs. WT.

Fig. 7

Effect of 1-heptanol perfusion on electrical activity. A: representative example of effect of perfusion with the gap junction blocker 1-heptanol on pseudo-ECG recording from an isolated arterially perfused WT mouse heart. Abscissa reflects time from the moment perfusion started. au, arbitrary units. B: 1 mM heptanol-induced reduction of QRS amplitude in WT mouse hearts was consistent and statistically significant (n = 5, P = 0.001). C: 1 mM heptanol does not affect other ECG parameters, such as basal cycle length (BCL), P-R interval (PR), and QRS width (QRSw).