Fibrinogen was used to label scar (magenta) and Tyrosine Hydroxylase (TH) was used to label sympathetic neurons (cyan). Example image of (A) denervated infarct in non-targeting control-treated animal versus (B) normal density of TH + fibers in the peri-infarct region adjacent to the scar. Systemic delivery of silencing RNA (siRNA) against Chst15 post-MI restored TH + fibers in the cardiac infarct (C) without altering nerve density outside the infarct (D). Scale bar, 100 μm. (E) Quantification of TH innervation density 10 days after MI. n=4 animals for non-targeting controls and n=5 for Chst15 siRNA-treated animals. Data are mean percent TH ± SD. Two-way ANOVA, Tukey’s post-test to compare all groups, select comparisons shown; siChst15 infarct vs. peri-nfarct ns – not significant p-value = 0.668, non-targeting infarct vs. peri-nfarct ****p-value < 0.0001, non-targeting infarct vs. siChst15 infarct ****p-value < 0.0001. (F) Arrhythmia scores based on the most severe arrhythmia observed in each heart after injection of isoproterenol and caffeine (0=no PVCs, 1=single PVCs, 2=bigeminy or salvos, 3=non-sustained ventricular tachycardia). See Materials and methods for details. Treatment with siRNA against Chst15 reduced arrhythmias compared to non-targeting controls. n=4 animals for non-targeting controls and n=5 animals for Chst15 siRNA. Data are arrhythmia score for each animal. Statistics: Student’s t-test (Mann-Whitney test), *p<0.047. (G) Cardiac scar size assessed as a percent area of total left ventricle (LV) area was unaltered in Chst15 siRNA-treated animals compared to non-targeting controls. Quantification of n=4 animals for non-targeting controls and n=5 for Chst15 siRNA-treated animals. Data are mean percent area scar ± SD. Statistics: Student’s t-test (Welch’s test), ns – not significant p-value = 0.283. (H) Example 2× image of cardiac scar in Chst15 siRNA-treated heart using autofluorescence, absence of autofluorescence indicates region of the infarct, outlined in white. Scale bar, 500 μm.