Ischemia-reperfusion myocardial infarction induces remodeling of left cardiac-projecting stellate ganglia neurons

Abstract

Neurons in the stellate ganglion (SG) provide sympathetic innervation to the heart, brown adipose tissue (BAT), and other organs. Sympathetic innervation to the heart becomes hyperactive following myocardial infarction (MI). The impact of MI on the morphology of cardiac sympathetic neurons is not known, but we hypothesized that MI would stimulate increased cell and dendritic tree size in cardiac neurons. In this study, we examined the effects of ischemia-reperfusion MI on sympathetic neurons using dual retrograde tracing methods to allow detailed characterization of cardiac- and BAT-projecting neurons. Different fluorescently conjugated cholera toxin subunit B (CTb) tracers were injected into the pericardium and the interscapular BAT pads, respectively. Experimental animals received a 45-min occlusion of the left anterior descending coronary artery and controls received sham surgery. One week later, hearts were collected for assessment of MI infarct and SGs were collected for morphological or electrophysiological analysis. Cardiac-projecting SG neurons from MI mice had smaller cell bodies and shorter dendritic trees compared with sham animals, specifically on the left side ipsilateral to the MI. BAT-projecting neurons were not altered by MI, demonstrating the subpopulation specificity of the response. The normal size and distribution differences between BAT- and cardiac-projecting stellate ganglion neurons were not altered by MI. Patch-clamp recordings from cardiac-projecting left SG neurons revealed increased spontaneous excitatory postsynaptic currents despite the decrease in cell and dendritic tree size. Thus, increased dendritic tree size does not contribute to the enhanced sympathetic neural activity seen after MI.NEW & NOTEWORTHY Myocardial infarction (MI) causes structural and functional changes specifically in stellate ganglion neurons that project to the heart, but not in cells that project to brown adipose fat tissue.

Keywords: morphology; neural plasticity; sympathetic; tract tracing.

Graphical abstract

Figure 1.

Cardiac infarct size assessed by fibrinogen labeling of heart sections was greater after myocardial infarction (MI) than sham surgery. A: representative images of fibrinogen labeling in sections through the left ventricle (LV) of the heart from a sham (top) and MI (bottom) mouse. B: infarct size quantified as percent area of left ventricle. Bars represent means ± SD and open circles represent individual values for each heart (n = 4 sham, 5 MI mice). Scale bars = 1 mm. MI vs. sham, Mann–Whitney rank sum test: *P = 0.02.

Figure. 2.

Representative confocal micrographs of cardiac-projecting (orange) and brown adipose tissue (BAT)-projecting (green) stellate ganglia (SG) neurons from the left (A and C) and right (B and D) SG of sham (A and B) and myocardial infarction (MI; C and D) mice. Scale bars = 100 µm.

Figure 3.

Representative confocal micrographs and Imaris three-dimensional (3-D) renderings of a brown adipose tissue (BAT)-projecting (A and B) and a cardiac-projecting (C and D) stellate ganglia (SG) neuron. A: BAT-projecting neurons in SG (green) were labeled with 0.1% CTb-488 tracer. C: cardiac-projecting neurons in SG (orange) were labeled with 0.1% CTb-555 tracer. Three-dimensional renderings of projection neurons were created using Imaris surfaces (somas), filaments (axons, dendrites, branch points), and spots (VAChT puncta) segmentation tools (B and D). Scale bars = 10 µm for all images. CTb, conjugated cholera toxin subunit B; VAChT, vesicular acetylcholine transporter.

Figure 4.

Myocardial infarction (MI) had a unilateral impact on soma sizes in mouse stellate ganglia (SG). A: soma volumes of cardiac neurons in left SGs were significantly smaller 1 wk after MI as compared with somas in the left SG of sham controls. There were no changes in right SG after MI or in brown adipose tissue (BAT) soma sizes across groups. B: soma cross-sectional areas for cardiac-projecting neurons in left SGs from MI-treated animals were also significantly smaller than sham controls. Bars represent means ± SD, and open circles represent individual soma volumes in each SG; n = 4 sham, 5 MI mice. Left cardiac MI vs. left cardiac sham; Kruskal–Wallis one-way ANOVA on ranks with Dunn’s post hoc pairwise comparisons (A) and unpaired Welch’s t test (B). *P < 0.05.

Figure 5.

Myocardial infarction (MI) increased the distances measured between cardiac-projecting neurons in the left stellate ganglia (SG) compared with sham left and right and MI right. Bars represent means nearest neighbor distance measurement ± SD; n = 4 sham, 5 MI mice; cardiac left and right sham vs. cardiac left and right MI comparisons, Kruskal–Wallis one-way ANOVA on ranks with Dunn’s post hoc pairwise comparisons: *P < 0.05.

Figure 6.

Cardiac-projecting stellate ganglia (SG) neurons have an overall higher percentage of neurons in direct contact (“coupled”) compared with brown adipose tissue (BAT)-projecting SG neurons (inset). There was no difference found between treatment groups or left/right SGs. Percentage of coupled projection neurons was calculated from the total number of projection neurons in each ×20 low-magnification confocal scan. Myocardial infarction (MI) did not impact the number of projection neurons in contact with like projection neurons in either population. Bars represent means ± SD, individual dots are means percentage per SG examined; n = 4 sham, 5 MI mice. Cardiac vs. BAT, three-way ANOVA, comparing cell type (cardiac/BAT), treatment (MI/sham), and side (left/right), Holm–Sidak post hoc comparisons: *P = 0.002.

Figure 7.

Sham stellate ganglia (SG) demonstrate a difference in the length of the dendritic tree between left and right sides and myocardial infarction (MI) impacts the dendritic arborization on both sides compared with sham. Bars represent the means ± SD, and the open circles represent individual values for each projection neuron in that group; n = 4 sham, 5 MI mice. Two-way ANOVA with Holm–Sidak post hoc comparisons, sham vs. MI, *P < 0.05 vs. corresponding sham SG on the same side; left vs. right, #P < 0.05 between sides within a treatment group.

Figure 8.

Representative cardiac-projecting neuron in the stellate ganglia (SG) apposed by vesicular acetylcholine transporter (VAChT)-labeled puncta. A: confocal micrograph of a cardiac-projecting neuron in SG (orange) labeled with 0.1% CTb-555 tracer surrounded by VAChT-labeled puncta (red). B: three-dimensional renderings of the soma volume (yellow surface) and the VAChT puncta (red spots) allow for measurements of neuronal size and the number of puncta apposed to the cell surface. Scale bars = 5 µm.

Figure 9.

Bone morphogenetic protein (BMP) genes are suppressed in the left ventricle (LV) 24 h after myocardial infarction (MI). Gene expression in unoperated control and MI LVs were assayed 24 h after ischemia-reperfusion. BMP mRNA was normalized to actin mRNA in the same sample. Bars represent the means ± SD, and circles represent individual values for each mouse; n = 4 mice/treatment group (2 males, 2 females). Control vs. 24 H post-MI, unpaired Welch’s t test: BMP2, *P = 0.03; BMP4, **P = 0.002; BMP5, ***P = 0.0002; BMP6, ***P = 0.0007; BMP7, ***P = 0.0003. nd, not detected.

Figure 10.

Neuronal activity in cardiac-projecting neurons was impacted by myocardial infarction (MI). A: representative recordings from sham and MI ganglia. B: MI caused a significant increase in the frequency of spontaneous excitatory postsynaptic currents (sEPSCs). C: amplitude trended higher but was not significantly different (P = 0.07). All cells were recorded from the left stellate ganglia (SG). Bars are means ± SD, filled circles represent cells back labeled from the heart and open circles represent cells located in the cardiac pole of unlabeled SG. MI, n = 6 cells from 4 mice; sham, n = 10 cells from 7 mice. MI vs. sham, unpaired Welch’s t test. *P = 0.03.