|Home | About | Journals | Submit | Contact Us | Français|
The small GTPase RhoA and its associated kinase ROCKII are involved in vascular smooth muscle cell contraction and endothelial nitric oxide synthase (eNOS) mRNA destabilization. Overactivation of the RhoA/ROCKII pathway is implicated in a number of pathologies including chronic heart failure (CHF) and may contribute to the enhanced sympathetic outflow seen in CHF due to decreased nitric oxide (NO) availability. Thus, we hypothesized that central ROCKII blockade would improve the sympatho-vagal imbalance in a pacing rabbit model of CHF in an NO-dependent manner. CHF was induced by rapid ventricular pacing and characterized by an ejection fraction of ≤45%. Animals were implanted with an ICV cannula and osmotic minipump (rate: 1 µL/h) containing sterile saline, 1.5 µg/kg/day fasudil (Fas, a ROCKII inhibitor) for 4 days or Fas + 100 µg/kg/day L-NAME, a NOS inhibitor. Arterial baroreflex control was assessed by IV infusion of sodium nitroprusside and phenylephrine. Fas infusion significantly lowered resting HR by decreasing sympathetic and increasing vagal tone. Furthermore, Fas improved baroreflex gain in CHF in an NO-dependent manner. In CHF Fas animals, the decrease in HR in response to IV metoprolol was similar to sham and was reversed by L-NAME. Fas decreased AT1R and phospho-RhoA protein expression and increased eNOS expression in the brainstem of CHF animals. These data strongly suggest that central ROCKII activation contributes to cardiac sympatho-excitation in the setting of CHF and that central Fas restores vagal and sympathetic tone in an NO-dependent manner. ROCKII may be a new central therapeutic target in the setting of CHF.
Hallmarks of chronic heart failure (CHF) include increased circulating Angiotensin II, upregulation of the Angiotensin II type 1 Receptor (AT1R), and an imbalance in the autonomic nervous system1. In CHF, there is heightened sympathetic outflow and a decrease in vagal tone, two driving forces in the progression of the disease2. This sympatho-vagal imbalance may be mediated, in part, by decreased nitric oxide (NO) bioavailability. Synthesized by endothelial nitric oxide synthase (eNOS), NO can diffuse to nearby cells to activate guanylyl cyclase to increase intracellular cyclic guanosine monophosphate and activate downstream signaling cascades leading to vasodilation and other anti-inflammatory and antiproliferative effects3.
The sympatho-vagal imbalance seen in CHF may also reflect a blunted baroreflex, an integral mechanism of maintaining cardiovascular homeostasis4,5. A decrease in baroreceptor sensitivity may participate in hyperactivity of cardiovascular control centers of the brain, including the major center of pre-sympathetic neurons that project to the spinal cord, the rostral ventrolateral medulla (RVLM)6, 7.
Activation of the RhoA/ROCKII pathway can occur as a downstream consequence of activation of Gα protein subunits via GPCR signaling. For example, Angiotensin II signaling through the AT1R has been shown to indirectly activate the RhoA/ROCKII pathway in cardiac myocytes8–10. RhoA/ROCKII pathway hyperactivity is seen in a number of pathophysiological states including stroke, hypertension and chronic heart failure3, 11, 12. Following cysteine modification by superoxide, RhoA protein can exchange GDP for GTP and can subsequently bind ROCKII13. In addition to destabilizing eNOS mRNA3, activation of the RhoA/ROCKII signaling cascade leads to upregulation of NF-κB, increased myosin light chain kinase (MLCK) activation, and phosphorylation of ezrin/radixin/moesin proteins11, 14, leading to actin cytoskeletal rearrangement and reduction in dendritic spine formation15, and a shift towards a pro-inflammatory state8.
While the role of RhoA/ROCKII in peripheral vasoconstriction and eNOS activation has been well established16, 17, 18, there is still far less of an understanding of how central RhoA/ROCKII exerts its effects under both physiological and pathophysiological conditions. The RhoA/ROCKII pathway has previously been shown to be hyperactive in the NTS of spontaneously hypertensive rats (SHR), and that the RhoA/ROCKII pathway in the NTS contributes to neurogenic hypertension in part by chronic NOS inhibition19, 20. Given its ability to disrupt tight junction proteins, central RhoA/ROCKII pathway activation has also been implicated in blood brain barrier dysfunction in experimental models of multiple sclerosis and in some animal stroke models3, 21. However, the contribution of the central RhoA/ROCKII pathway to the blunted baroreflex and sympatho-vagal balance seen in CHF has yet to be determined. Therefore, the goal of this study was to determine the role of the central Rho Kinase pathway in the sympatho-vagal imbalance seen in CHF. The RhoA/ROCK II pathway destabilizes eNOS mRNA thereby reducing NO availability. This decrease in NO prevents the inhibition of sympathetic neurons in the RVLM, which contributes to the increase in sympathetic outflow in CHF and ultimately leads to a decrease in baroreflex sensitivity. We therefore hypothesized that blockade of the RhoA/ROCKII pathway with the specific ROCKII inhibitor Fasudil (Fas) would restore cardiac sympatho-vagal balance and baroreflex sensitivity in CHF rabbits in an NO-dependent manner.
Experiments were carried out on a total of 36 New Zealand White male rabbits weighing 3.0–4.5 kg (Charles River Laboratories, Wilmington, MA). All experiments were reviewed and approved by the University of Nebraska Medical Center Institutional Animal Care and Use Committee. Animals were randomly placed into one of five experimental groups: Sham Vehicle (Sham Veh), Sham Fasudil (Sham Fas), Chronic Heart Failure Vehicle (CHF Veh), Chronic Heart Failure Fasudil (CHF Fas), or Chronic Heart Failure Fasudil + L-NAME (CHF Fas + L-NAME).
Rabbits were instrumented as described previously22. In brief, animals were given a pre-anesthetic cocktail of 35 mg/kg ketamine, 5.8 mg/kg xylazine, and 0.01 mg/kg atropine in 1 cc of lactated Ringer’s solution. Following intubation, anesthesia was maintained using a low level of isofluorane (0.5–1.0%). A radiotelemetry transducer (Data Sciences Internationial (DSI), Minneapolis, MN) was implanted in the right femoral artery to monitor mean arterial blood pressure (MAP) and heart rate (HR) using the DSI telemetry and ADInstruments (Colorado Springs, CO) Powerlab data acquisition systems. During the same surgical procedure, a left thoracotomy was carried out and a reference electrode was attached to the left atrium and a platinum pacing electrode was sutured to the left ventricle. Pacing wires were tunneled beneath the skin and exited in the mid-scapular area.
At least two weeks following thoracic surgery, an intracerbroventricular (icv) cannula was inserted and reinforced with dental cement to the surface of the skull. An osmotic minipump (Alzet 2001, rate: 1 µL/hr, Cupertino, CA) containing sterile normal saline, 1.5 µg/kg/day Fas (Tocris, Minneapolis, MN), or 1.5 µg/kg/day Fas plus 100 µg/kg/day Nω-Nitro-L-arginine methyl ester hydrochloride (L-NAME, Sigma Aldrich, St. Louis, MO) was attached to the brain cannula using a Microrenathane (size, manufacturer, etc) catheter. Based on rat studies reported in the literature we scaled the dose of L-NAME to be used in rabbits23.
Animals were allowed to acclimate by sitting in a Plexiglass box in a dimly lit laboratory for 20 minutes with the pacemaker turned off for CHF rabbits prior to experimentation. Following 5 minutes of baseline MAP and HR recording, the maximal changes in HR and MAP were assessed by the response to nasopharyngeal stimulation with 60 mL of cigarette smoke. This maneuver has been used previously by our laboratory as a means of measuring vagal tone24,25. After HR and MAP returned to baseline, baroreflex curves were generated by IV infusion (marginal ear vein) of 100 µg/kg sodium nitroprusside (SNP) and 80 µg/kg phenylephrine (PE) at a rate of 0.5 mL/min. SNP was infused until MAP was decreased to approximately 40–45 mm Hg at which point the SNP infusion was replaced with PE until MAP reached ~100 mm Hg. Following generation of a control baroreflex curve, a second curve was constructed after IV bolus administration of 0.2 mg/kg atropine methylbromide or 1 mg/kg metoprolol bitartrate. After 10 minutes, a second 10 minute baseline was recorded, and baroreflex experiments were repeated. Atropine and metoprolol experiments were done on separate days.
In CHF vehicle treated rabbits, 24 hours following completion of all baroreflex experiments, the saline containing pump was replaced with a pump filled with Fas or Fas + L-NAME. Three days following pump implantation, baseline recordings and baroreflex assessments were repeated.
Arterial baroreflex curves were constructed as described previously25. Briefly, curves were constructed by sampling data points for HR every 5 seconds from the lowest to highest MAP following SNP and PE infusions. Individual logistic regression curves were fit to the data points as described previously25.
Data are expressed as mean ± standard error of the mean. All statistical analyses were performed with GraphPad Prism Software (GraphPad, La Jolla, CA). Differences between groups was assessed with one-way Analysis of Variance with a Bonferroni correction when appropriate. P<0.05 was considered statistically significant.
Echocardiographic profiles of pre- and post-pace animals are shown in Table S1. Pacing animals had a significantly lower ejection fraction and fractional shortening and significantly higher left ventricular systolic volume and left ventricular end systolic diameter compared to pre-pace animals. Table 1 summarizes resting HR and MAP. Baseline HR was higher in CHF Veh animals compared to sham groups, and HR was significantly lowered by icv Fas treatment in CHF rabbits. The decrease in HR seen in CHF Fas animals was reversed by co-administration of icv L-NAME. MAP was unchanged between groups.
To determine if icv Fas had an effect on autonomic balance, we recorded baseline HR before and following IV bolus administration of atropine, a muscarinic receptor antagonist or metoprolol, a β-1 adrenergic receptor antagonist (Figure 1). The tachycardia seen in response to atropine is indicative of the degree of vagal tone; conversely, the extent of bradycardia seen upon metoprolol administration reflects sympathetic tone. Fas restored the blunted tachycardia in response to atropine seen in CHF Veh animals (Figure 1A). Simultaneous icv administration of Fas + L-NAME prevented the improvement in vagal tone. Similarly, Fas also normalized the exaggerated bradycardic response to metoprolol in CHF Veh animals (Figure 1B). This protective effect seen in CHF Fas animals was also reversed by L-NAME, suggesting that the Fas-mediated restoration of the sympatho-vagal imbalance may be due, in part, to increased NO production via NOS.
The bradycardic response to nasopharyngeal smoke is vagally mediated25; thus, we examined the change in resting HR following administration of smoke across all groups of animals (Figure 2). Given that Fas normalized the tachycardic response to atropine in CHF animals, we hypothesized that the response to smoke in CHF Fas animals would be as robust as Sham controls. Indeed, there was a blunted response to nasopharyngeal smoke in CHF Veh animals as compared to sham, and this response was restored with icv Fas treatment. L-NAME administration partially reversed the improved vagal tone in the CHF Fas animals.
To further assess the protective effect of icv Fas on the autonomic profile in CHF animals, we constructed baroreflex curves in all five groups of animals (Figure 3). As expected, in CHF, there was a significantly blunted baroreflex curve compared to controls. CHF Fas animals exhibited significantly improved baroreflex sensitivity as compared to CHF Veh. The normalized baroreflex in CHF Fas animals was prevented by icv L-NAME, suggesting Fas restores autonomic balance in an NO-dependent manner. Quantification of baroreflex curve parameters are shown in supplemental Table S2. Peak slope (maximal gain) was also calculated from the baroreflex curves under normal conditions, and following administration of either atropine or metoprolol (Figure 4). Peak slope was significantly lower in CHF Veh animals, indicating blunted baroreflex sensitivity compared to Sham animals. Peak slope was higher in CHF Fas animals compared to CHF Veh and comparable to Sham animals. The baroreflex sensitivity of CHF Fas + L-NAME animals was also blunted compared to Sham animals, further suggesting that Fas is restoring autonomic balance in an NO-dependent manner.
To assess if the protective effect of Fas in CHF also correlated with a change in protein expression, we performed western blot analysis of RVLM micropunches (Figure 5). There was a significant decrease in eNOS protein expression in CHF Veh compared to Sham animals (Figure 5A), which was normalized by Fas treatment. The Fas-mediated increase in eNOS was prevented by L-NAME. Conversely, a hallmark of CHF is an upregulation in AT1R 22. CHF Veh animals had significantly higher AT1R compared to Sham animals (Figure 5B). Fas treatment dramatically decreased AT1R protein in CHF animals to below Sham levels. CHF Fas + L-NAME treated animals maintained an increase in AT1R compared to Sham. Finally, to confirm that Fas had an effect on downstream RhoA/ROCKII targets, we examined changes in ROCKII and phosphorylated ezrin/radixin/moesin proteins, a readout of RhoA/ROCKII activity (Figure 5C, 5D). There was an increase in both of these proteins in CHF Veh compared to control, and these values were normalized with Fas treatment, confirming that the Fas was indeed having an effect RhoA/ROCKII targets. Interestingly, L-NAME did not fully reverse the increase in CHF Veh animals compared to CHF Fas. This may imply that the hyperactivation of the RhoA/ROCKII pathway in CHF has both NO-dependent and NO-independent components.
The key findings of this study are that central Fas treatment restores baroreflex sensitivity, eNOS and AT1R protein imbalance in an NO-dependent manner in a rapid ventricular pacing rabbit model of CHF. To our knowledge this is the first demonstration that central ROCKII blockade has been shown to alter autonomic function in the setting of CHF. Furthermore, these data show a correlation between AT1R and eNOS expression in one potent cardiovascular regulatory area, the RVLM. Previous work from this laboratory has shown that both peripheral and central administration of chronic simvastatin to conscious animals with CHF reduces renal sympathetic nerve activity and restores both cardiac and peripheral baroreflex function27, 28. In a recent study carried out in humans with CHF, one month of simvastatin therapy reduced muscle sympathetic nerve activity and was correlated with a reduction in global oxidative stress29. While statin therapy primarily blocks HMG Co-A reductase activity, one consequence of this is indirect inhibition of the RhoA/ROCKII pathway30, 31
A large body of literature points to the fact that central NO acts as a sympatho-inhibitory molecule through a GABA mechanism32–34. While multiple pathways have been shown to regulate NOS in both the periphery and in the CNS, one pathway that has been implicated in abnormal NOS signaling is the RhoA/ROCKII pathway35. Most of the previous work on the central RhoA/ROCKII pathway relates to changes in neuronal plasticity, axonal and dentritic spine formation, etc.15, 36. There is little evidence to suggest that this pathway directly alters neuronal activity. The implications of the current work suggest that inhibition of this pathway in cardiovascular disease reduces autonomic outflow and may constitute a new and novel therapeutic target in those disease states characterized by increased sympathetic outflow and decreased vagal outflow. Indeed, studies by Hirooka and co-workers have clearly shown that both statins and ROCKII inhibitors given directly into the NTS in SHR and WKY restored glutamate sensitivity, reduced blood pressure and in some cases, increased NOS expression and no changes in ROCKII expression 16, 17, 36,20. However, in the current study, ROCKII protein expression was increased in the RVLM of CHF animals, and the decrease in ROCK II protein seen in CHF Fas treated animals was only partially increased by L-NAME. Although most of the data in this study suggest that the beneficial effects of Fas are largely NO dependent, it is certainly possible that there are other mediators like Angiotensin II that can lead to the hyperactivity of the RhoA/ROCKII pathway in CHF.
Several studies have now shown that arterial baroreflex function can be impacted by the balance between NO generation and Angiotensin II signaling37–40. The finding in the present study that Fas evokes a decrease in AT1R expression in the RVLM may suggest a mechanism by which this pathway regulates both baroreflex function and global autonomic outflow. Indeed, it is well established that activation of the RhoA/ROCKII pathway can upregulate the transcription factor, NFkB; similarly Angiotensin II activation also increases both NFkB and RhoA/ROCKII41,42. Several areas of the CNS have been shown to exhibit increases in oxidative stress in the CHF state43–45. In this regard Angiotensin II signaling would contribute to oxidative stress. Inhibition of the RhoA/ROCKII pathway can contribute to a reduction in central oxidative stress by lowering AT1R expression and by increasing the bioavailability of NO. In the current study we only examined the effects of Fas on eNOS protein. Previous studies have shown that eNOS upregulation is an important mechanism in the protective effects of statin therapy in a murine stroke model and that this is dependent on the RhoA/ROCKII pathway46. However, the connection between the RhoA/ROCKII pathway and other NOS isoforms has not been well defined. It will be necessary in future work to establish the interaction between the RhoA/ROCKII pathway, central AT1R and reactive oxygen species.
Arterial baroreflex function has been shown to be depressed in several cardiovascular diseases including CHF and hypertension47, 48. The regulation of cardiac baroreflex sensitivity is mediated by both sympathetic and vagal outflow. The results of the present study suggest that central administration of Fas impacts both arms of autonomic outflow to the heart. Because Fas was chronically administered by the icv route it is not possible to determine the precise location of its action. Icv fas administration would have immediate access to several areas that have been shown to be integral in Angiotensin II and oxidative stress signaling. One such area is the subfornical organ (SFO) which is well endowed with AT1Rs49. This area of the midbrain is also accessible by the systemic route, thus promoting the possibility that RhoA/ROCKII inhibition globally may have effects on autonomic outflow. The present study focused on proteins in the RVLM, given that there was a pronounced effect on vagal tone it is possible that both the Nucleus Ambiguus and the Dorsal Motor Nucleus of the Vagus may also be a target for RhoA/ROCKII inhibition. Additionally, future studies addressing the effects of systemic Fas on sympathetic and vagal outflow as well as changes in both the peripheral and central RhoA/ROCKII pathways would be of interest.
It is of interest that although baroreflex function was improved, we did not examine cardiac function following Fas treatment. There is limited evidence that central manipulation of sympathetic outflow in CHF impacts cardiac function per se. One study carried out in a mouse myocardial infarction model suggests however that reducing oxidative stress in the SFO does indeed increase myocardial performance50. On the other hand, global reduction in sympathetic outflow and increases in vagal outflow are antiarrythmogenic51, 52and reducing peripheral sympathetic outflow in CHF may benefit renal function, exercise tolerance and quality of life53, 54.
This study demonstrates that central blockade of the RhoA/ROCKII pathway with Fas can restore blunted baroreflex sensitivity seen in CHF animals in an NO-dependent manner. Furthermore, central Fas can also restore the increase in AT1R and decreased eNOS expression that is a hallmark of CHF. Taken together, the RhoA/ROCKII pathway may be a novel therapeutic target for the treatment of CHF.
The authors would like to thank Johnnie Hackley for technical assistance.
Sources of Funding
This research was supported by National Heart, Lung, and Blood Institute Grants PO1-HL62222 (I.H.Z.) and F32HL116172-01 (K.K.V.H).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
There are no disclosures to report.