Lipid Lowering by Pravastatin Increases Parasympathetic Modulation of Heart Rate
Gαi2, a Possible Molecular Marker for Parasympathetic Responsiveness
Background— We have previously demonstrated in an in vitro model for lipid lowering that lipoprotein depletion resulted in a marked increase in the negative chronotropic response to the acetylcholine analogue carbamylcholine. In this study we used heart rate variability analysis to determine the effect of lipid lowering by statins on the response of the heart to parasympathetic stimulation. In parallel, we examined whether changes in parasympathetic responsiveness correlated with changes in the expression of Gαi2, a molecular component of the parasympathetic signaling pathway in the heart.
Methods and Results— Patients were randomized in a crossover study of pravastatin and simvastatin. R-R interval analysis of Holter monitor studies demonstrated that in patients treated initially with pravastatin, the peak high-frequency power fraction during sleep, which reflects parasympathetic modulation of heart rate, increased by 24.0±5.02% (SEM, n=13, P<0.001) compared with the untreated control value. Simvastatin had no significant effect. Western blot analysis of lymphocytes from patients treated with pravastatin demonstrated a 90.1±27.3% (n=10, P=0.009) increase in Gαi2 expression, whereas simvastatin had no effect. Relative changes in Gαi2 correlated significantly with the changes in the fraction of high-frequency power (ρ=0.574, P=0.016).
Conclusions— Taken together with our in vitro data, these data are the first to suggest that cholesterol lowering by pravastatin might increase the response of the heart to parasympathetic stimulation and that changes in Gαi2 expression might serve as a molecular marker for this effect.
Received March 11, 2003; de novo received August 8, 2003; revision received October 2, 2003; accepted October 6, 2003.
Using an in vitro cell culture model for lipid lowering, we previously demonstrated that growth of atrial cells in the absence of lipoproteins resulted in a 10-fold increase in their negative chronotropic response to the acetylcholine analogue carbamylcholine, in parallel with an increase in the expression of the α-subunit of the heterotrimeric G-protein, Gαi2.1 The relationship between the response of the heart to parasympathetic stimulation and the expression of Gαi2 is supported by recent studies of the effect of viral expression of Gαi2 in the porcine atrioventricular (AV) node, in which increased expression of Gαi2 decreased AV-nodal conduction and the response of the heart to β-adrenergic stimulation.2 HMG-CoA reductase inhibitors (statins) are now widely used in the treatment of hyperlipidemia.3 However, the effect of lipid lowering by statins on the parasympathetic response of the heart has not been studied. This study uses analysis of R-R intervals from Holter monitor studies of patients randomized to pravastatin and simvastatin to determine the effect of statins on the response of the heart to parasympathetic stimulation.4 In parallel, we examine whether changes in parasympathetic responsiveness correlate with changes in the expression of Gαi2 in lymphocytes from these patients. This study might have important implications for understanding the relationship between lipid lowering and the response of the heart to parasympathetic stimulation. Furthermore, it might support a possible role for Gαi2 expression as a molecular marker for the parasympathetic response of the heart.
Thirty patients were enrolled who met the following criteria for statin therapy: LDL ≥160 mg/dL with no risk factors, ≥130 mg/dL and 2 risk factors, or ≥100 mg/dL with coronary artery disease (CAD) or the equivalent. Patients with a myocardial infarction within the last 4 months, diabetes mellitus, congestive heart failure, or hypertension and patients taking antiarrhythmic agents, phenothiazines, and other antidepressants were not considered for enrollment. Thirteen women and 8 men ages 34 to 68 years (mean, 55.2±2.1) completed the entire study. None of the patients had a history of CAD. Three of the patients were smokers.
Of the 9 patients who did not complete the study, 2 were initially treated with pravastatin and 7 with simvastatin. Three had significant elevations of creatine phosphokinase, 4 developed myalgias, 1 discontinued the study for personal reasons, and data from 1 were not analyzed for reasons described below.
The study was a single-blind, randomized crossover study of the effects of pravastatin and simvastatin on parasympathetic responsiveness of the heart in which each patient served as his or her own control. Blinding was maintained throughout the analysis of the Holter monitor data and the Western blot analysis of Gαi2 in lymphocyte extracts. After giving informed consent, blood was drawn for fasting baseline lipid profiles and liver enzymes. An ambulatory 24-hour surface ECG was recorded using a solid state Holter monitor device (Seer MC ambulatory digital recorder, GE Marquette Medical Systems). Total cholesterol and triglycerides were determined enzymatically (Determiner L TC-II, Determiner L TG-II diagnostic kits, Kyowa Medex), and HDL was measured using a Determiner L HDL-C diagnostic kit from Kyowa using an AU 5400 chemistry analyzer (Olympus). LDL cholesterol was calculated as total cholesterol−(triglycerides/5+HDL) for triglycerides lower than 400 mg/dL. Patients were randomized to either 20 mg pravastatin or 20 mg simvastatin for 8 weeks followed by crossover to the second drug for 10 weeks. Lipid profiles and Holter monitor studies were obtained at baseline, crossover, and completion of the study. The second treatment period was prolonged to minimize the effects of the first treatment drug.
The Holter monitor recordings were uploaded to a Marquette Mars workstation for R-wave detection and beat annotation, which were both manually reviewed. Spectral analysis was performed after constructing an instantaneous RR interval time series, resampled at 3 Hz. The power spectrum densities of consecutive 5-minute segments were computed for the frequency ranges between 0.03 and 0.15 Hz, designated as low-frequency power (LF), and between 0.15 and 0.4 Hz, designated as high-frequency power (HF). Total power (TP) was defined as LF+HF. HF has been shown to result predominantly from parasympathetic modulation of heart rate, whereas LF reflects both sympathetic and parasympathetic modulation of heart rate.4 The 2-hour means of the maximum HF fraction (HF/TP) and minimum LF/HF ratio during sleep were used for statistical analysis. If no such point could be identified, the data for that patient were excluded, resulting in the 1 exclusion noted above. Lymphocytes were prepared from freshly drawn blood by Ficoll gradient centrifugation. Gαi2 levels were determined on equally loaded samples of lymphocyte homogenates by Western blot analysis using a polyclonal Gαi2 antibody (Santa Cruz). The blots were incubated using anti-rabbit IgG conjugated to horseradish peroxidase as a secondary antibody and visualized using Super Signal West Pico Chemiluminescent Substrate (Pierce). Equal loading and transfer was determined by reprobing the blot with an anti-tubulin antibody (Santa Cruz). An aliquot of purified recombinant Gαi2 (NEN) was run in parallel to help identify the Gαi2 band in cell extracts. The relative intensity of the bands was quantified by densitometry scanning using NIH Image Pro.
Parameters were calculated either as the difference between treatment values and each patient’s own baseline or as the relative change between treatment values. Student’s t test was used to determine whether the mean changes from baseline in HF fraction were different from zero after treatment. The correlation between changes in HF fraction and Gαi2 expression between treatments was computed using Spearman’s ρ. A P value of <0.05 was considered significant. All significance levels are computed as 2-tailed. Statistical results are given as mean±SEM.
Effect of Statin Treatment on Lipid Levels
LDL levels before and after treatment are summarized in the Table. Pravastatin decreased LDL by 32.7±2.3% (n=21, P<0.001) and simvastatin decreased LDL by 40.6±3.9% (n=20, P<0.001) compared with baseline independent of the order in which drugs were administered.
Effect of Statin Treatment on Parasympathetic Modulation of Heart Rate
In patients initially treated with pravastatin, HF fraction increased by 0.096±0.019 (n=13, P=0.001), which is equivalent to a 24.0±5.02% (n=13, P=0.001) increase in parasympathetic responsiveness (Figure 1A). After crossover to simvastatin, these patients demonstrated no significant change in HF fraction (0.009±0.038, n=12, P=0.806, Figure 1A). In patients treated initially with simvastatin, there was no significant effect on HF fraction (0.004±0.04, n=8, P=0.933, Figure 1B). However, after crossover to pravastatin, HF fraction increased by 0.097±0.038 (n=8, P=0.037, Figure 1B), an increase similar to that seen when pravastatin was given first (Figure 1A). There was no correlation between changes in HF fraction and changes in LDL cholesterol.
In patients initially treated with pravastatin, LF/HF ratio decreased by 0.556±0.139 (n=13, P=0.002), which is equivalent to a 30.7±6.18% (n=13, P<0.001) decrease. After crossover to simvastatin, these patients demonstrated no significant change in LF/HF ratio (−0.003±0.241, n=12, P=0.992). In patients treated initially with simvastatin, there was no significant effect on LF/HF ratio (0.397±0.439, n=8, P=0.396). However, after crossover to pravastatin, LF/HF ratio decreased by 0.448±0.214 (n=8, P=0.075), which is equivalent to a 30.6±8.71% (n=8, P = 0.010) decrease. Thus, heart rate variability analysis demonstrated that pravastatin increased the response of the heart to parasympathetic stimulation whereas simvastatin had no effect, irrespective of the order in which the drugs were administered.
Effect of Statin Therapy on Gαi2 Expression: Correlation With Changes in HF Fraction
Western blot analysis of extracts of lymphocytes from patients randomized to pravastatin demonstrated a significant increase in Gαi2 expression compared with prior values (Figure 2A) and increased by a mean of 90±27% (n=10, P=0.009), whereas simvastatin had no significant effect on Gαi2 expression (−14±11%, n=8, P=0.248, Figure 2B). These changes in Gαi2 level after treatment with either pravastatin or simvastatin correlated significantly with changes in HF fraction, with a Spearman’s ρ correlation coefficient of 0.574 and P=0.016.
This study supports the unique observation that pravastatin significantly increases parasympathetic modulation of heart rate as measured by changes in peak HF fraction during sleep. These data additionally support the conclusion that pravastatin stimulates a marked increase in the expression of Gαi2 in the lymphocytes of these patients, which significantly correlates with changes in HF fraction. Hence, these data suggest the intriguing hypothesis that changes in Gαi2 expression might constitute a molecular marker for changes in parasympathetic responsiveness.
In the analysis of the 24-hour Holter monitor recordings, we limited the presentation of the HRV data to the most interpretable HRV measure, fluctuations of HF fraction, which is most directly related to parasympathetic response.4 To minimize the effects of physical activity on the ambulatory recordings, we chose the 2-hour segment with the highest mean HF fraction during sleep as the single segment for comparing the results between each phase of the study.
The hypothesis that changes in Gαi2 expression may be a marker for parasympathetic responsiveness is supported by several studies. Viral overexpression of Gαi2 in the porcine AV node has been shown to be associated with an increase in the sensitivity of the heart to parasympathetic stimulation.2 Also, treatment of rats with pertussis toxin, which inactivates Gαi2, increased isoprenaline-induced arrhythmias.5 Furthermore, data from an in vitro model for lipid lowering demonstrated that culture of atrial cells in the absence of lipoproteins resulted in an increase in the negative chronotropic response to muscarinic stimulation in parallel with an increase in Gαi2 expression.1,6 Although direct measurements of Gαi2 expression in the heart were not possible, these studies support the conclusion that the correlation between the increase in the parasympathetic response of the heart and Gαi2 expression in lymphocytes may be physiologically relevant and suggest that measurements of Gαi2 expression made in lymphocytes reflect changes in the heart. Changes in cardiac and lymphocyte physiology have been correlated with changes in Gαi2 function and expression. Biopsy specimens of a focal region of the right ventricular outflow tract in a patient shown to develop idiopathic ventricular tachycardia were shown to express a Gαi2 with a point mutation in its GTP binding domain, which rendered it incapable of regulating inhibiting cAMP production levels, whereas pertussis toxin treatment of lymphocytes interfered with their ability to migrate in response to the chemoattractant fMLP.7,8
The differences between the effects of simvastatin and pravastatin on the response of the heart to parasympathetic stimulation may be attributable to differences in hydrophobicity. It has been demonstrated that in the liver hydrophilic statins are taken up primarily via an active transporter whereas hydrophobic statins are taken up by passive diffusion as a function of relative hydrophobicity.9 Because of the absence of a transporter in nonliver cells, hydrophilic statins are taken up at much lower levels in nonliver tissues.9,10 If, as suggested by our in vitro data, lipoprotein depletion increases the parasympathetic responsiveness of atrial cells,1 both classes of statins might increase parasympathetic responsiveness because of their lipid-lowering effects on the liver. However, one might speculate that only the more hydrophobic simvastatin would diffuse into the heart, reversing the effect of lipid lowering on cholesterol biosynthesis and parasympathetic responsiveness. As a result, only pravastatin treatment would increase parasympathetic responsiveness. Such an effect is supported by in vitro data that demonstrate that the hydrophobic statin lovastatin but not pravastatin reversed the effect of lipid lowering on parasympathetic response and Gαi2 expression in cultured atrial cells1 (Galper and Gadbut, unpublished observation, 2001). Although presently there are no data available comparing the effects of hydrophilic and hydrophobic statins on intracellular cholesterol in the heart, our in vitro data support the conclusion that cholesterol lowering by the hydrophilic statin pravastatin significantly increases parasympathetic modulation of heart rate.
Data have been presented that support the relationship between parasympathetic stimulation of the heart and protection of the heart from the genesis of arrhythmias.11 A recent retrospective study of the effects of statin therapy on outcomes after coronary artery bypass grafting12 and a study of the recurrence of life-threatening arrhythmias in patients with CAD treated with implantable defibrillators13 suggested that patients treated with lipid-lowering agents might demonstrate a decrease in the incidence of arrhythmias. Furthermore, the decrease in lipid levels observed in the immediate post–myocardial infarction period has been shown to correlate with a decrease in the incidence of cardiac arrhythmias, and the recovery of lipids to preinfarction levels has been associated with an increase in the incidence of arrhythmias.14 The clinical significance of the increase in parasympathetic responsiveness in patients treated with pravastatin is not clear.
The authors would like to acknowledge Dr Eugene Braunwald for insightful discussions of this study.
↵*These authors contributed equally to this study.
Haigh LS, Leatherman GF, O’Hara DS, et al. Effects of low density lipoproteins and mevinolin on cholesterol content and muscarinic cholinergic responsiveness in cultured chick atrial cells: regulation of levels of muscarinic receptors and guanine nucleotide regulatory proteins. J Biol Chem. 1988; 263: 15608–15618.
Grundy SM. Statin trials and goals of cholesterol-lowering therapy. Circulation. 1998; 97: 1436–1439.