Cardiac TRPV1 afferent signaling promotes arrhythmogenic ventricular remodeling after myocardial infarction

Abstract

Chronic sympathoexcitation is implicated in ventricular arrhythmogenesis (VAs) following myocardial infarction (MI), but the critical neural pathways involved are not well understood. Cardiac adrenergic function is partly regulated by sympathetic afferent reflexes, transduced by spinal afferent fibers expressing the transient receptor potential cation subfamily V member 1 (TRPV1) channel. The role of chronic TRPV1 afferent signaling in VAs is not known. We hypothesized that persistent TRPV1 afferent neurotransmission promotes VAs after MI. Using epicardial resiniferatoxin (RTX) to deplete cardiac TRPV1-expressing fibers, we dissected the role of this neural circuit in VAs after chronic MI in a porcine model. We examined the underlying mechanisms using molecular approaches, IHC, in vitro and in vivo cardiac electrophysiology, and simultaneous cardioneural mapping. Epicardial RTX depleted cardiac TRPV1 afferent fibers and abolished functional responses to TRPV1 agonists. Ventricular tachycardia/fibrillation (VT/VF) was readily inducible in MI subjects by programmed electrical stimulation or cesium chloride administration; however, TRPV1 afferent depletion prevented VT/VF induced by either method. Mechanistically, TRPV1 afferent depletion did not alter cardiomyocyte action potentials and calcium transients, the expression of ion channels, or calcium handling proteins. However, it attenuated fibrosis and mitigated electrical instability in the scar border zone. In vivo recordings of cardiovascular-related stellate ganglion neurons (SGNs) revealed that MI enhances SGN function and disrupts integrated neural processing. Depleting TRPV1 afferents normalized these processes. Taken together, these data indicate that, after MI, TRPV1 afferent-induced adrenergic dysfunction promotes fibrosis and adverse cardiac remodeling, and it worsens border zone electrical heterogeneity, resulting in electrically unstable ventricular myocardium. We propose targeting TRPV1-expressing afferent to reduce VT/VF following MI.

Keywords: Arrhythmias; Cardiology; Cardiovascular disease; Innervation; Neuroscience.

Figure 1. Chemical depletion of cardiac TRPV1 afferents by resiniferatoxin.

(A) Schematic representation of the central projection of cardiac sympathetic afferent fibers. (B) Representative images of immunoreactivity (IR) to protein gene product 9.5 (PGP9.5), a panneuronal marker; tyrosine hydroxylase (TH, adrenergic marker); and calcitonin gene–related peptide (CGRP), a sensory peptide coexpressed by most TRPV1 afferents, are shown in LV epicardium of control, myocardial infarction (MI), and resiniferatoxin–treated (RTX-treated) MI subjects (MI + RTX). Scale bar: 50 μm. (C) Relative to nerve fibers seen, RTX depleted sensory but not adrenergic fibers (n = 4–6/group). (D–F) Epicardial depletion of TRPV1 afferents is further evidenced functionally by a lack of response to epicardial application of TRPV1 agonists, capsaicin (D and E) and bradykinin (F). Arrow in D indicates time of capsaicin application. Representative hemodynamic tracings of responses to capsaicin 20 μg/mL are shown in D. MI subjects showed a sensitization of TRPV1 afferents compared with controls; however, responses were abolished in RTX-treated subjects. n = 6–8/group. *P < 0.05; **P < 0.01; ***P < 0.001 by Kruskall-Wallis test.

Figure 2. Cardiac sympathetic afferent depletion reduces ventricular arrhythmogenesis.

(A) Representative tracing of ventricular tachycardia (VT) induced by programmed electrical stimulation in myocardial infarction (MI) with 2 extrastimuli (S3) but not in resiniferatoxin-treated (RTX-treated) MI subjects (MI + RTX) with 4 extrastimuli (S5). (B) Across subjects, VT induction was substantially reduced in MI + RTX subjects compared with MI (P = 0.005). (C–E) Arrhythmogenicity index, calculated from the drive cycle length and number of extrastimuli, was reduced in MI + RTX versus MI (n = 13 and 7 for MI and MI + RTX, respectively; P = 0.002). // for MI + RTX Subject 1 indicates ventricular effective refractory period. Ventricular arrhythmogenesis was also examined by bolus administration of cesium chloride (80 mg/kg). (F) Representative tracings showing sinus rhythm, frequent premature ventricular contractions (PVCs), and nonsustained ventricular tachycardia (NSVT) from an MI subject. (G–I) Reduction in number of PVCs, NSVT episodes, and number of beats of each NSVT episode in MI vs. MI + RTX animals (n = 12 and 8 for MI + RTX and MI, respectively). χ² test (B) and Mann-Whitney U test (C and G–I) were used.

Figure 3. Cardiomyocyte electrophysiologic properties after cardiac TRPV1 afferent depletion.

(A) Representative action potential (AP) tracing from untreated myocardial infarction (MI) and RTX-treated MI subjects (MI + RTX). (B–F) No differences were observed in resting membrane potential (B), AP amplitude (C), rate of AP rise (D), and APD₉₀ (E) and APD₅₀ (F). (G) Representative tracing of calcium (Ca²⁺) transients recorded from isolated border zone MI and MI + RTX cardiomyocytes. (H and I) There were no differences in time to peak Ca²⁺ fluorescence (H) and half maximal relaxation of the Ca²⁺ transient (τ₅₀) (I). n = 15 cells from 3 animals/group. Mann-Whitney U test (B–D), 2-way ANOVA (H and I) were used.

Figure 4. Cardiac TRPV1 afferent depletion ameliorates adverse postinfarction border zone remodeling.

(A) Trichrome-stained transmural sections from the left ventricular (LV) border zone (BZ) of myocardial infarct (MI) subjects and the same region in resiniferatoxin-treated (RTX-treated) MI subjects (MI + RTX) show less scar (blue) in MI + RTX compared with MI. Scale bar: 2 mm. (B) Percent scar transmurality and wall thickness (top and bottom, respectively) in the dense scar region (apex), BZ, and remote myocardium (basal-lateral LV) from MI and MI + RTX animals show that the BZ, not dense scar or remote myocardium, is impacted by cardiac TRPV1 afferent depletion. (C) Several genes involved in fibrosis are differentially expressed in MI + RTX hearts compared with MI, as revealed by RNA sequencing of BZ myocardium. (D) Representative epicardial activation map in sinus rhythm from the scar BZ of MI and MI + RTX subjects showing significant spatial activation heterogeneity in MI compared with MI + RTX. (E) Mean activation time, activation dispersion, activation delay, and maximal activation time in sham, MI, and MI + RTX subjects show reduced electrical heterogeneity in MI + RTX subjects. n = 10–11/group. *P < 0.05; ***P < 0.001. ANOVA with Welch’s test was used.

Figure 5. Cardiac TRPV1 depletion normalizes cardiac function after MI.

(A) Representative left ventricular short axis echocardiographic images from MI and MI + RTX subjects. (B–E) Quantified data across subjects showing left ventricular (LV) ejection fraction (LVEF), LV end-systolic volume (LVESV), LV internal dimension in diastole (LVIDd), and LV anterior wall thickness, respectively (n = 5–6/group). *P < 0.05; **P < 0.01. Kruskall-Wallis test was used.

Figure 6. Amelioration of infarct-induced sympathetic neuronal network dysfunction by cardiac TRPV1 afferent depletion.

(A) Location of the recording electrode (left). DESC, descending; LUE, left upper extremity. Examples of spike-sorted neuronal action potential morphologies (right). Scale bar: 1 ms. (B) Basal activity of cardiovascular-related stellate ganglion neurons recorded from control, myocardial infarction (MI), and resiniferatoxin-treated (RTX-treated) MI subjects (MI + RTX). MI subjects exhibited higher basal activity compared with control and MI + RTX (n = 78, 137, and 108 for control, MI, and MI + RTX, respectively. ***P < 0.001. ANOVA was used. (C) Of neurons recorded throughout the experimental protocol, more were active at baseline in MI subjects, compared with control and MI + RTX. P < 0.0001. χ² test was used; n = 22, 90, and 39, respectively. (D) Responses of recorded neurons to cardiovascular stressors: rapid ventricular pacing (RVP), inferior vena cava occlusion (IVC), and aortic occlusion (AO). Responses in MI subjects deviated from those in controls. However, responses in MI + RTX subjects were restored toward that seen in controls. (E) Conditional probability maps (1, 2) of the 3 stressors in D, showing reduced network interdependence in MI compared with control, and reversion toward normal physiology in MI + RTX subjects.