G_o but not G_i2 or G_i3 is required for muscarinic regulation of heart rate and heart rate variability in mice

Abstract

Muscarinic receptor-mediated cardiac parasympathetic activity is essential for regulating heart rate and heart rate variability (HRV). It has not been clear which G_i/G_o protein is responsible for these effects. We addressed this question using knockout mice that lack G protein α_i2, α_i3, or α_o specifically. Unlike previously reported, our α_o-null mice had significantly more survivors with normal life span. Isolated hearts from α_o-null mice demonstrated much less sensitivity to the negative chronotropic effects of muscarinic agonist carbachol to lower heart rate at baseline and more profound under the stimulation of β-adrenergic agonist isoproterenol. In the presence of parasympathetic activation indirectly produced by methoxamine, an α₁-adrenergic agonist, α_o-null mice showed markedly decreased HRV compared with wild type control mice. These differences in heart rate and HRV were not observed in α_i2-null or α_i3-null mice. Our findings establish an essential role for α_o G protein in the anti-adrenergic effect of carbachol on heart rate regulation.

Introduction

Normal heart rate is controlled by cardiac sympathetic and parasympathetic nerves to achieve homeostasis [1, 2]. Cardiac sympathetic system is mediated by β-adrenergic receptors that are coupled primarily to α_s G protein and responsible for cardiac activation including positive chronotropic, dromotropic, lusitropic, and inotropic effects. Cardiac parasympathetic system is mediated by muscarinic acetylcholine (M2) receptor that is coupled to α_i/α_o G proteins and responsible for negative chronotropic and dromotropic effects. In sinoatrial node pacemaker cells, α_s and α_o act simultaneously, with α_o being more potent, suggesting that parasympathetic inhibition of heart rate is much greater on a background of sympathetic stimulation [3]. Mutations in either α_i2 or α_o to disrupt binding to regulator of G protein signaling results in enhanced muscarinic M2 receptor–mediated bradycardic responses in cardiocytes derived from embryonic stem cells [4]. However, how different α_i/α_o G proteins (α_i2, α_i3, and α_o) control whole heart rate has not been determined.

The opposing effects of sympathetic and parasympathetic systems cause beat-to-beat alterations in heart rate, a phenomenon called heart rate variability (HRV) [5]. Because HRV is a reflection of parasympathetic tone it predicts survival after heart attack [6]. It is not clear how different α_i/α_o G proteins control HRV.

To investigate the specificity of α_i/α_o G proteins in controlling heart rate and HRV, we created α_o-null mice and used an isolated heart perfusion method to study the alteration in muscarinic-induced chronotropic effects under basal conditions and in the presence of prior β-adrenergic stimulation, as well as telemetry implants to analyze their EKG. The results were compared with those from α_i2-null mice and α_i3-null mice previously reported [7, 8].

Materials and Methods

Animal Model

Production of gene targeted mice: Using previously published constructs for gene targeting [7, 8], the α_o gene was inactivated in J1 cells cultured on mouse embryo fibroblasts. Targeted lines identified as previously described by Southern analysis [8], were injected into Balb/c blastocysts. Resulting chimeras passed the targeted mutation in the germline bred to Balb/c or 129SvEv. Heterozygotes were mated to obtain littermates that were either wild type or homozygous for the gene inactivation in order to control for genetic background.

Western analysis

Plasma membranes from mouse cardiac ventricles were prepared, subjected to electrophoresis, and transferred to PVDF membranes as described previously [9]. Membranes were incubated with primary antibodies specifically recognizing α_o, α_i2, α_i3, β_comm, and α_s. The membranes were then incubated in secondary antiserums conjugated with horseradish peroxidase and detected as before[9].

Isolated Heart Preparation

Hearts were excised, the aortic root cannulated to a 20-gauge needle and perfused retrograde with a modified Krebs-Henseleit buffer (in mM): 120 NaCl, 4.5 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 2.5 CaCl₂, 0.5 EDTA, 25 NaCO₃, and 10 dextrose (pH 7.4 when gassed with 95%O₂/5%CO₂ at 37°C). The entrance of the pulmonary veins into the left atrium was then cannulated and the perfusion was subsequently switched to the working mode [10]. The pulmonary artery was cut to allow measurement of coronary flow. Left ventricular pressures were monitored continually through indwelling cannulae coupled to pressure transducers and linked to a MacLab A/D system and a Macintosh computer with the sampling rate set at 2 kHz. LVP and heart rate were monitored in the working heart mode for 5–10 minutes to ensure stability of the preparation before initiation of the experiment. After a 10-minute stabilization period, the dose-response to carbachol was determined in the absence or presence of 1 μM isoproterenol; stable beating rate was then quantified after each drug dose addition.

In Vivo EKG

Telemetry monitor implants (Data Science, model TA10ETA-F20) were placed in mice. After recovery for at least 24 hours, heart rate was monitored by telemetry using a MacLab digitization and a Power PC computer. Heart rate variability (HRV) was analyzed using MacLab software. Data was collected at baseline for at least 15 minutes and continuously following administration of methoxamine. Data were analyzed for HRV and heart rate when stabilized 5–15 minutes after addition of methoxamine.

Statistics

Mean ± standard error (SE) values were analyzed using Prism (GraphPad Software Inc.). Statistical comparisons between groups were performed by Student’s t test. Dose response curves were compared by Two-way ANOVA and Bonferroni post-tests. Groups were considered significantly different if P values were ≤ 0.05.

RESULTS

Generating α_o-null mice

Inactivation of the α_o gene was confirmed by western blot (Figure 1A). There was no apparent change of other G-protein subunits, including α_i2, α_i3, β_comm, and α_s, in response to α_o inactivation (Figure 1A). Mice homozygous for the α_o targeted gene exhibited perinatal mortality, but in contrast to previous reports [11, 12], a significant percentage in the present study had normal life spans (Figure 1B). We were therefore able to study healthy, fertile, adult mice for alterations in heart rate control. Improved animal husbandry in decreasing stress is likely the reason for better survival rate, although difference in genetic background can not be ruled out.

Figure 1. Inactivation of α_o in mice.

A. Western blots of hearts from α_o-null mice and control wild type mice. Antibodies against α_o, as well as α_i2, α_i3, β_comm, and α_s were used to verify the deletion of α_o and no compensatory changes in other G proteins. B. Survival of α_o-null mice. Mice that survived the critical period lived nearly 2 years. The numbers of animals were n=25 at 100 days, 18 at 150 days, and 12 at 200 days. The decrease in number of animals was due to sacrifice for experiments.

Response to carbachol in heart rate is significantly reduced in mouse heart lacking α_o

At baseline, carbachol dose response curve in α_o-null mouse hearts demonstrated a shift to the right for a half log unit when compared with WT (Figure 2A, lower two curves), showing decreased response to carbachol. When isoproterenol was present, the extent of slowing was much reduced and the effect of lacking α_o was not overcome even by very high doses of carbachol (Figure 2A, upper two curves). These results indicate that α_o-mediated pathways are critical to negative chronotropic effects of carbachol, particularly in the presence of isoproterenol. To establish whether this is specific to α_o, we tested α_i2-null and α_i3-null mice using the same set up. Neither α_i2-null nor α_i3-null mouse heart displayed significantly different responses to carbachol than WT at baseline or with isoproterenol stimulation (Figure 2B, C). Mutants lacking both α_i2 and α_i3 were not born implying embryonic lethality.

Figure 2. Hearts from α_o-null mice show decreased sensitivity to muscarinic agonist.

A. Negative chronotropic responses to carbachol in isolated hearts from α_o-null and littermate control (WT) mice. Hearts were perfused ex vivo in working or ejecting mode as described in methods. Carbachol dose response curves in the absence and presence of isoproterenol (1 μM, squares) were determined by increasing addition of carbachol. n = 5 for each group. By Two-Way ANOVA, P = 0.0071 WT vs. null; P < 0.0001 WT+ISO vs. null+ISO. Bonferroni post-tests, * P < 0.05 vs. WT; ** P < 0.01 vs. WT+ISO; *** P < 0.001 vs. WT+ISO. B, C. Negative chronotropic responses to carbachol in isolated hearts from α_i2-null and α_i3-null mice respectively. No statistical significance was detected.

Decreased HRV in mice lacking α_o

HRV was determined from at least 1000 R-R intervals between successive beats before and after administration of methoxamine to stimulate the parasympathetic system. This system was chosen in order to compare results with the study on I_KACh inactivation [13]. As an α₁-adrenergic agonist methoxamine causes vasoconstriction and increases blood pressure leading to activation of the carotid sinus baroreceptor and hence reflex activation of the parasympathetic system. The basal heart rate was not significantly different between α_o-null animals and their WT littermates and methoxamine lowered the heart rate in both genotypes (Figure 3A), consistent with low parasympathetic tone in mice [14]. There was a markedly diminished response to methoxamine in HRV in the α_o-null animals, but not in α_i2-null or α_i3-null mice (Figure 3B). Spectral analysis of HRV data indicated that the difference between α_o-null and WT was significant after administration of methoxamine, in all 3 components, high frequency (HF), low frequency (LF), and very low frequency (VLF) (Figure 3C).

Figure 3. Diminished HRV in α_o-null mice responding to methoxamine.

A. In vivo heart rates before and after methoxamine. Heart rates were measured using implantable EKG telemetry devices at baseline and after the injection of 6 mg/kg methoxamine. n=4 for each group. ** P < 0.01. B. Decreased HRV in the α_o-null mice compared to their WT littermates after methoxamine stimulation. Mice were chronically maintained EKG telemetry implants as described in methods. Data of the R-R intervals between successive beats before and after addition of methoxamine were collected continuously. C. Quantification of HRV data in α_o-null and WT mice. Measurement of the power spectra was broken into three components: high frequency (HF) 1.5–4 Hz; low frequency (LF) 0.4–1.5 Hz; and very low frequency (VLF) <0.4 Hz. * P < 0.02 vs. WT VLF; ** P < 0.002 vs. WT LF; *** P < 0.0001 vs. WT HF.

Discussion

Gene inactivation studies have established the requirement for the I_KACh channel in parasympathetic slowing heart rate and increases in HRV. It was estimated that about 50% of the negative chronotropic effect required an intact I_KACh channel [13]. We now show that α_o-mediated pathways, which are not required for I_KACh activation [7, 12] but do regulate I_f, are also required for normal negative chronotropy in heart rate and parasympathetic stimulated increases in HRV. These results indicate that activation of I_KAch through the remaining pertussis toxin-sensitive G-proteins is not capable of maintaining normal negative chronotropic responses on its own. Therefore, at least two pathways are required, one leading to activation of I_KACh mediated by α_i2 and α_i3 released βγ [7] and the other likely leading to inhibition of hyperpolarization-activated current I_f, which could require α_o but not α_i2 or α_i3.

The exact contributions of these two pathways may differ depending on the system studied. In the intact heart, both pathways appear to be important in regulating heart rate and HRV. Both I_KACh and α_o inactivation (and hence effects of I_f), markedly affected the muscarinic negative chronotropic effects. Here, we show the importance of α_o under ex vivo conditions where complicating effects on blood pressure and sympathetic or parasympathetic nerve activity have been eliminated. A detailed interpretation of the meaning of the whole animal experiments is difficult in that, like the previous study [13], alterations in blood pressure or effects of knockouts on central nervous system function were not evaluated. However, there is a clear decrease in the HRV in α_o-null animals. Inactivation of I_KACh nearly eliminated HRV at baseline whereas α_o inactivation had no effect. When stimulated with methoxamine, both I_KACh and α_o inactivation caused marked decreases in HRV with I_KACh effects being slightly more pronounced. The effect on HRV is consistent with the decreased sensitivity to muscarinic stimulation by carbachol seen in the isolated perfused hearts. The contribution of the α_o dependent pathway may increase with high sympathetic activity or with other β-adrenergic stimulation compared to resting conditions.

These results establish a critical role for α_o-containing heterotrimers in the parasympathetic regulation of heart rate and HRV.