Altered With Left Ventricular Hypertrophy Induced Heart Failure

30 31 Hypertension, cardiac hypertrophy and heart failure (HF) are widespread and 32 debilitating cardiovascular diseases that affect nearly 23 million people worldwide. A distinctive 33 hallmark of these cardiovascular diseases is autonomic imbalance, with increased sympathetic 34 activity and decreased parasympathetic vagal tone. Recent device-based approaches, such as 35 implantable vagal stimulators that stimulate a multitude of visceral sensory and motor fibers in 36 the vagus nerve, are being evaluated as new therapeutic approaches for these and other diseases. 37 However little is known about how parasympathetic activity to the heart is altered with these 38 diseases and this lack of knowledge is an obstacle in the goal of devising selective interventions 39 that can target and selectively restore parasympathetic activity to the heart. To identify the 40 changes that occur within the brainstem to diminish the parasympathetic cardiac activity left 41 ventricular hypertrophy was elicited in rats by aortic pressure overload using a transaortic 42 constriction approach. Cardiac vagal neurons (CVNs) in the brainstem that generate 43 parasympathetic activity to the heart were identified with a retrograde tracer and studied using 44 patch clamp electrophysiological recordings in-vitro. Animals with left cardiac hypertrophy had 45 diminished excitation of CVNs which was mediated both by an augmented frequency of 46 spontaneous inhibitory GABAergic neurotransmission, (with no alteration of inhibitory 47 glycinergic activity), as well as diminished amplitude and frequency of excitatory 48 neurotransmission to CVNs. Opportunities to alter these network pathways and 49 neurotransmitter receptors provide future targets of intervention in the goal to restore 50 parasympathetic activity and autonomic balance to the heart in cardiac hypertrophy and other 51 cardiovascular diseases. 52 53 keywords: Heart failure, cardiac hypertrophy, parasympathetic, autonomic, vagal 54 55 56 New and Noteworthy 57 58 A common hallmark of cardiovascular diseases is autonomic imbalance. Left ventricular 59 hypertrophy (LVH) altered both excitatory and inhibitory neurotransmission to cardiac vagal 60 neurons (CVNs) that generate parasympathetic activity to the heart. These preferentially altered 61 network pathways and neurotransmitter receptors provide future targets to restore 62 parasympathetic activity in these diseases. 63 64 65 66 67 Introduction 68 69 Hypertension, cardiac hypertrophy and heart failure (HF) are widespread and 70 debilitating cardiovascular diseases that affects nearly 23 million people worldwide with 71 approximately 2 million new patients diagnosed annually (15). Cardiac rhythm disturbances 72 lead to sudden cardiac death in 40-50% of advanced heart failure patients with a 1 year 73 mortality rate greater than 50% (10). A distinctive hallmark of cardiac hypertrophy, heart failure 74 and accompanying cardiac conduction abnormalities is autonomic imbalance, particularly 75 increased sympathetic activity and decreased parasympathetic tone. 76 77 Parasympathetic cholinergic activity to the heart plays a major role in cardiac function, 78 and is often cardioprotective, suppressing the endogenous high rate of firing of pacemaker cells 79 in the sinoatrial node and maintaining heart rate at normal levels. Cardioinhibitory 80 parasympathetic activity to the heart arises from the preganglionic cardiac vagal neurons 81 (CVNs) located in the nucleus ambiguus (NA), dorsal motor nucleus of the vagus (DMNX), and 82 intermediate zone of the medulla oblongata (4, 5, 34). Vagal efferent axons from these cell 83 bodies terminate upon the postganglionic intracardiac ganglia neurons located near the 84 sinoatrial and atrioventricular nodes of the heart (2). Resting heart rates, and changes in 85 response to challenges, are mediated to a large extent by alterations in parasympathetic vagal 86 outflow originating from CVNs in the brainstem. CVNs exhibit tonic firing activity that is cardiac 87 pulse synchronous, and also inhibited during each inspiration (14, 23, 28, 33). 88 89 While parasympathetic activity to the heart is absent or diminished in many 90 cardiovascular diseases, including hypertension, cardiac hypertrophy and heart failure, 91 augmentation of diminished cardiac vagal activity prevents arrhythmias, decreases risk of 92 sudden death, and protects against ischemia/reperfusion injury (7, 12, 20, 24, 25, 31, 36, 37, 39, 93 41). However there are few, if any selective methods to increase parasympathetic activity to the 94 heart in patients. Recent device-based approaches, such as implantable vagal stimulators that 95 stimulate a multitude of visceral sensory and motor fibers in the vagus nerve, are being 96 evaluated as new therapeutic approaches for these and other diseases. However little is known 97 about how parasympathetic activity to the heart is altered with these diseases and this lack of 98 knowledge is an obstacle in the goal of devising selective interventions that can target and 99 selectively restore parasympathetic activity to the heart. To provide this foundation and identify 100 the changes that occur within the brainstem to diminish parasympathetic cardiac activity left 101 ventricular hypertrophy was elicited in rats by aortic pressure overload using a transaortic 102 constriction approach and the neurobiology of cardiac vagal neurons was studied with patch 103 clamp electrophysiological approaches. The goal of this study was to identify the changes within 104 the brainstem that are likely responsible for diminishing parasympathetic activity to the heart, 105 and by doing so provide future targets of intervention to restore parasympathetic activity and 106 autonomic balance to the heart in cardiac hypertrophy and other cardiovascular diseases. 107 108 109 Methods 110 Ethical approval 111 All animal procedures carried out were in accordance with The George Washington University 112 Institutional Guidelines and in compliance with the recommendations of the panel of 113 Euthanasia of the American Veterinary medical association and the NIH publication (85-23, 114 revised 1996) “Guide for the care and use of laboratory animals’. The minimal number of 115 animals was used and care was taken to reduce any possible discomfort. 116 117 Aortic banding to Induce Heart Failure 118 Left ventricular hypertrophy secondary to pressure overload was produced in male Sprague119 Dawley rats using a minimally invasive transaortic constriction (TAC) approach previously 120 established for mice (21, 35). Animals at 4-5 weeks of age were subjected to either TAC or a 121 sham aortic banding surgery. A 1-1.5 cm skin incision was made at the level of the suprasternal 122 notch and the thymus was retracted to reveal the aortic arch. A 4-0 silk suture was passed 123 around the aortic arch between the origin of the right innominate and left common carotid 124 arteries and, with a 20 gauge needle temporarily placed adjacent to the aorta, the suture was 125 tied around the aorta and needle. The needle was then removed which produced a chronic 126 partial aortic constriction. Successful aortic banding was confirmed by increased flow through 127 the right carotid artery at the surgical session and upon examination of the aorta after the 128 animal was sacrificed. Sham animals underwent the same surgical procedure except the suture 129 was not tied. 130 131 After 6-8 weeks animals were euthanized and the degree of left ventricular hypertrophy was 132 measured. Hearts were rapidly excised and retrogradely perfused with isotonic media 133 containing heparin to wash out the blood. Fat, vessels, and connective tissue were trimmed from 134 the base. Hearts were sliced into four cross-sections and these sections of tissue were 135 photographed and analyzed using the software ImageJ (NIH). The "free wall" measurement was 136 defined as the distance from the closest point on the inside of the left ventricle to the edge of the 137 slice opposite the right ventricle. The "septum" measurement was defined as the distance from 138 the closest point on the inside of the left ventricle to the closest point on the inside of the right 139 ventricle. The weight of the whole heart and left ventricle were recorded. Heart dimensions and 140 weight were divided by animal weight to normalize for variations in the size of the animals. 141 142 Labeling of CVNs 143 Animals in which electrophysiological recordings from CVNs were obtained underwent an 144 additional surgery when the animals were at postnatal days 2-5 (Sprague-Dawley, Hilltop 145 Laboratory animals Inc, Scottdale, PA, USA). Animals were anesthetized using hypothermia by 146 cooling to approximately 4oC. A right thoracotomy was performed and retrograde tracer X147 Rhoda-mine-5-(and-6)-isothiocyanate (Invitrogen, USA) was then injected into the fat pads at 148 the base of the heart to retrogradely label CVNs (26). The animals were then allowed to recover 149 until they were 4-5 weeks of age and then underwent either sham or TAC surgery. 150 151 In vitro brainstem slice preparation 152 We adopted the methodology from Ye and colleagues (45) to obtain viable brainstem slices from 153 10-12 week old animals. According to this method, glycerol base artificial cerebrospinal fluid 154 (aCSF) was used for cardiac perfusion and brainstem slicing. Glycerol-based aCSF contained (in 155 mM): 252 glycerol, 1.6 KCl, 1.2 NaH2PO4, 1.2 MgCl, 2.4 CaCl2, 26 NaHCO3, and 11 glucose. Rats 156 were anaesthetized using isoflurane and placed on ice. Glycerol aCSF (4oC, pH-7.4, bubbled 157 with 95% O2-5%CO2) was perfused transcardially at a speed of ~ 10 ml/min after which the 158 brain was quickly removed, glued on to a stage using 2% low melt agarose and placed in a 159 vibrotome containing glycerol aCSF. Brainstem slices (330μm thickness), containing either 160 DMNX or NA, were obtained and briefly placed in a solution with following composition (in 161 mM): 110 N-methyl-d-glucamine (NMDG), 2.5 KCl, 1.2 NaH2PO4, 25 NaHCO3, 25 glucose, 110 162 HCl, 0.5 CaCl2, and 10 MgSO4 equilibrated with 95% O2 and 5% CO2 (pH 7.4) at 34°C for 15 min. 163 NMDG based aCSF was used to help slices recover and to maintain viable brainstem slices for 164 electrophysiological recordings (47). The slices were then mounted in a recording chamber 165 constantly perfused with a normal aCSF with following composition (in mM): 125 NaCl, 3 KCl, 2 166 CaCl2, 26 NaHCO3, 5 glucose and 5 HEPES; oxygenated with 95% O2–5% CO2 (pH-7.4) and 167 allowed to recover for at least 30 minutes before an experiment was performed. 168 169 Electrophysiological recordings 170 CVNs in NA and DMNX were identified by the presence of fluorescent tracer rhodamine and 171 imaged using differential interference contrast optics and infrared illumination. Whole cell 172 voltage clamp recordings from CVNs were done using Axopatch 200B and pClamp 8 software 173 (Axon Instruments, Union city, USA), at a holding voltage of -80 mV at room temperature. The 174 patch pipettes (2.5–5 MΩ) were filled with a solution consisting (in mM) of KCl (150), MgCl2 175 (4), EGTA (10), Na-ATP (2) and HEPES (10) or K-gluconic acid (150), HEPES (10), EGTA (10), 176 MgCl2 (1) and CaCl2 (1) at a pH of 7.3 for recording inhibitory or excitatory events respectively. 177 GABAergic inhibitory post synaptic currents (IPSCs) were isolated by application of a solution 178 containing strychnine (1μM, glycine receptor antagonist), 6-cyano-7-nitroquinoxaline-2, 3-dione 179 (CNQX, 50μM, non-NMDA receptor antagonist) and D-2-amino-5-phosphonovalerate (AP5, 180 50μM, NMDA receptor antagonist). Glycinergic IPSCs were isolated by including gabazine 181 (25μM, GABA-A receptor antagonist), CNQX, and AP5 in the perfusate. The perfusate included 182 gabazine and strychnine (25μM and 1μM, respectively) to isolate glutamatergic excitatory 183 postsynaptic currents . Gabazine, strychnine, CNQX, and AP5 were obtained from Sigma 184 Aldrich (St. Louis, MO, USA). 185 186 Data analysis 187 Synaptosoft software (version 6.0.3; Synaptosoft, Decatur, GA) was used to analyze the synaptic 188 events recorded from CVNs. Threshold value was set to the root mean square of noise levels 189 multiplied by 5. The data were presented as mean ± SEM. Students unpaired t-test was used to 190 compare statistical significance between sham and cardiac hypertrophy groups using Graphpad 191 Prism 5 software (La Jolla, CA, USA). Data with p<0.05 was considered significant; in the 192 figures, * denotes p<0.05, ** denotes p<0.01, *** denotes p<0.001. 193 194 195 Results 196 197 Transaortic constriction (TAC) resulted in left ventricular hypertrophy (LVH), consistent with 198 previous reports in rats (44). Hearts from TAC animals were significantly heavier than hearts 199 from control animals and had thicker left ventricles, (Figure 1A). Animal weight did not differ 200 between TAC and control (Figure 1B), but the ratios of LV free wall thickness to body weight, LV 201 weight to body weight, and heart weight to body weight were all significantly higher in TAC 202 animals compared to control (Figure 1C). 203 204 Actions of LVH on excitatory glutamatergic neurotransmission to CVNs 205 Spontaneous excitatory synaptic currents (EPSCs) were different in CVNs from the NA and 206 DMNX. DMNX CVNs had EPSCs with a larger amplitude than CVNs in the NA (amplitude of 207 EPSCs in DMNX CVNs 32.3 ± 1.1 pA, NA CVN amplitude 17.6 ± 2.8 pA, p<0.001), and DMNX 208 CVNs had a higher frequency of EPSCs (3.4 ± 0.3 Hz in DMNX CVNs compared to 1.1 ± 0.3 Hz 209 in NA CVNs, p<0.001). 210 211 Animals with LVH had a significantly diminished frequency of EPSCs, compared to sham 212 animals, in CVNs from both the NA and DMNX. In the DMNX CVNs EPSC frequency was 213 significantly diminished from 3.4 ± 0.3 to 0.8 ± 0.1 Hz (p<0.01). In CVNs from the NA EPSC 214 frequency was blunted from 1.1 ± 0.3 to 0.4 ±0.07 Hz (p<0.05), see figure 2. Animals with LVH 215 also had significantly decreased amplitudes of EPSCs in CVNs in the DMNX from 32.3 ± 1.1 pA 216 to 24.3 ± 1.7 pA, but the amplitude of EPSCs in the NA were not different in LVH and sham 217 animals. 218 219 Actions of LVH on inhibitory neurotransmission to CVNs 220 In sham animals spontaneous GABAergic inhibitory post-synaptic currents (IPSCs) in CVNs 221 from the NA had significantly lower amplitudes (28 ±2 pA in NA, 47 ±5 pA in DMNX, p<0.001) 222 but a higher frequency of GABAergic IPSCs (7.3 ± 1.3 Hz in NA, 2.7 ± 0.5 Hz in DMNX, p<0.01), 223 than CVNs from the DMNX. Spontaneous glycinergic IPSCs were not significantly different, in 224 amplitude or frequency, in CVNs from the NA and DMNX in sham animals, see figure 3. 225 226 LVH animals did not differ from sham animals in the frequency of IPSCs in NA CVNs. However 227 in CVNs in the DMNX LVH animals had significantly augmented frequency of inhibitory events 228 (from 6.0 ± 0.7 Hz in the sham animals to 8.4 ± 1.1 Hz in the LVH animals, p<0.05), see figure 229 4. To identify whether these changes were due to alterations in GABAergic or glycinergic 230 neurotransmission, or both, glycinergic and GABAergic IPSCs were isolated for study in 231 additional studies in LVH animals. LVH animals had no significant difference compared to 232 sham animals in the frequency of glycinergic IPSCs in either the NA or DMNX CVNs. While 233 LVH animals had no significant change in the amplitude, LVH animals had significantly 234 (p<0.05) elevated frequency of GABAergic inhibitory events (from 2.7 ± 0.5 Hz in the sham 235 animals to 5.3 ± 1.5 Hz in the LVH animals) in CVNs in the DMNX, see figure 4. 236 237 238 Discussion 239 In this study we identified two major changes that occur in the function of 240 parasympathetic cardiac vagal neurons in animals with left ventricular hypertrophy secondary 241 to pressure overload. As other studies have shown, this animal model elicits progressive 242 hypertrophy 2 weeks after aortic banding, within 6 weeks there are increases in lung-to-body 243 weight ratio (11), an index of HF, and this model progresses to end-stage heart failure with 244 hypertrophy, dilation and systolic dysfunction, within 20 weeks (11, 32). The 2 key findings of 245 this study are that within 6-8 weeks after aortic banding there is decreased excitation of CVNs 246 by 1) elevated frequency of inhibitory GABAergic neurotransmission to CVNs in the DMNX, and 247 2) diminished frequency of excitatory neurotransmission to CVNs in both the NA and DMNX. 248 As LVH animals possessed an opposing alteration in IPSC and EPSC frequencies (increasing 249 and decreasing, respectively), it is highly likely that in LVH animals the activity of preceding 250 excitatory and inhibitory neurons and pathways that synapse upon CVNs are altered. While we 251 cannot rule out postsynaptic changes in CVN membrane properties in the LVH animals, our 252 results indicate the LVH animals have significantly altered excitatory and inhibitory GABAergic 253 (but not glycinergic) pathways to CVNs. 254 255 The sites of action of LVH on preceding neurons and pathways to CVNs are not known, however 256 there are two known major origins of inhibitory inputs to CVNs, one originates from the locus 257 coeruleus, while the other originates from inspiratory neurons in the brainstem. As shown 258 previously, CVNs are inhibited during each inspiratory burst, and this cardiorespiratory network 259 function is the likely substrate for respiratory sinus arrhythmia as heart rate increases with each 260 inspiration (28). The second major source of inhibitory activity to CVNs originates in the locus 261 coeruleus, a nucleus involved in inducing cortical arousal and orchestrating changes in 262 accompanying autonomic system function that compliments increased attention, such as during 263 stress, excitation and/or exposure to averse or novel stimuli. Locus coeruleus noradrenergic 264 neurons depress the activity of cardioinhibitory parasympathetic cardiac vagal neurons by 265 polysynaptic activation of inhibitory neurotransmission within this brainstem autonomic and 266 attentiveness circuitry(38). This network interaction is dependent upon activation of α1 267 receptors that mediate increases in both GABAergic and glycinergic neurotransmission, while β1 268 receptor activation increases glycinergic, but not GABAergic neurotransmission to CVNs upon 269 locus coeruleus photoactivation (38). 270 271 Other work has shown there are four specific areas that contain GABAergic cells that 272 monosynaptically project to CVNs; 3 of the 4 loci are in close apposition to the CVNs in the NA 273 (200 microns medial, 400 microns medial, 200 microns ventral to the CVNs in the NA) and the 274 fourth locus is in the nucleus of the solitary tract region close to CVNs in the DMNX (13). These 275 4 populations of GABAergic neurons were retained in the brainstem slice utilized in this study 276 and are the probable source of the GABAergic neurons that directly project to CVNs that are 277 facilitated in LVH animals. 278 279 There are 2 known major sources of excitatory input to CVNs, one originates from neurons in 280 the nucleus tractus solitaries (NTS) (27), and the other is an excitatory pathway from neurons in 281 the hypothalamic paraventricular nucleus (PVN) (9, 29, 30). The NTS receives primary 282 information from cardiorespiratory sensory neurons and the excitatory pathway from the NTS 283 to CVNs likely plays an essential role in the chemoreceptor and baroreceptor reflex control of 284 heart rate (1). Retrograde tracing studies have demonstrated direct projections from the PVN to 285 the NA (6, 22). This long range neurotransmission from the PVN is excitatory with the 286 endogenous release of oxytocin facilitating glutamatergic neurotransmission and excitation of 287 CVNs (30). This oxytocin pathway is most likely involved in the slowing of heart rate during 288 periods of low vigilance as activation of oxytocin receptors reduces the adverse cardiovascular 289 consequences of anxiety and stress (8, 16, 17). 290 291 While no other work has, to the best of our knowledge, examined how LVH induces blunting of 292 brainstem parasympathetic activity, there has been considerable advances in understanding 293 how LVH and HF augments neurons involved in sympathetic activity. The discharge rates of 294 neurons likely involved in sympathetic activity within the dorsolateral periaqueductal gray 295 neurons are augmented in HF as compared with control rats (42). Other work has shown 296 changes in sympathetic neurons in the rostral ventrolateral medulla with upregulation of AT1R, 297 GRK5, and NF-κB expression (18), as well as increased firing activity of PVN neurons that were 298 antidromically activated from the sympathetic target in the rostral ventrolateral medulla 299 (RVLM) (43). The increased activity of PVN pre-sympathetic neurons with HF may be due to 300 many causes including diminished endothelial NOS expression and NOS-derived NO availability 301 in the PVN (3), enhanced expression of chemokines (such as chemokine stromal cell-derived 302 factor 1) (40), proinflammatory cytokines and angiotensin II type-1 receptors in the PVN in HF 303 animals (46), as well as a reduction in the frequency of spontaneous inhibitory postsynaptic 304 currents in RVLM-projecting PVN neurons (19). 305 306 How LVH augments the inhibitory pathways to CVNs, either originating from the preceding 307 locus coeruleus or inspiratory neurons, or via the inhibitory GABAergic neurons themselves, is 308 not known. Likewise, LVH may inhibit the excitatory neurotransmission originating from the 309 NTS, PVN or both pathways that provides excitation to CVNs. Potential candidates for 310 selectively restoring parasympathetic activity to the heart include identifying parameters to 311 selectively stimulate parasympathetic cardiac vagal fibers with vagal nerve stimulators, blunting 312 the activity of neurons in the locus coeruleus that inhibit CVNs, and preferentially augmenting 313 the excitatory oxytocin/glutamate pathway from the PVN to CVNs. These sites and 314 neurotransmitter receptors provide future targets of intervention in the goal to restore 315 parasympathetic activity to the heart in left ventricular hypertrophy, heart failure, hypertension 316 and other cardiovascular diseases with an imbalance in autonomic activity. 317 318 319 320 Acknowledgements: Support from NIH grants HL49965, HL59895 and HL72006 to D.M., NIH 321 grant HL095828 to M.K., and an American Heart Association Postdoctoral Fellowship 322 14POST20490181 to S.K.G. 323 324 325 326 327