SPARC: Distribution of Intrinsic Cardiac Neurons in 3D Reconstructed Hearts of F344 Rat

Abstract

The intrinsic cardiac nervous system (ICN) is involved in the autonomic control of the heart and functions as an integrative center for local sensory and extrinsic autonomic inputs. There has been increasing interest in the functions of ICN neurons as well as their ablation as a therapeutic intervention for atrial fibrillation. The goal of the present study is to assess the distribution of ICN neurons between individuals with rat as a model using a novel data acquisition pipeline combining Knife Edge Scanning Microscope (KESM) with the custom software developed by MBF (Tissue Mapper) to perform single cell mapping of ICN neurons in 3D reconstructed rat hearts. Fischer 344 male rats (n = 3) were perfused through the abdominal aorta and whole‐mount perfusion stained with Cresyl Echt Violet. The hearts were then embedded into paraffin and mounted onto a robotic XYZ stage for simultaneous sectioning and imaging by a KESM. Individual section images that consisted of 1400, 1570, and 2580 sections for each heart that was cut at a 5 μm thickness were assembled into an image stack for anatomical cardiac digitization and single cell mapping of ICN neurons in Tissue Mapper for 3D heart reconstruction. The distribution of ICN neurons between individuals was assessed in two different planes of section: parallel sections showing 4‐chambers, and cross sections perpendicular to the long axis of the heart. The number of neurons counted in each of the heart samples were 2676, 2973, and 2885, with an average of 2844/rat. ICN neurons mainly localized around the posterior surface of the left and right atria and the hilum ‐ pulmonary veins, interatrial septum, along the pulmonary artery, and the root of the superior vena cava. We performed computational analysis of ICN neuron distribution using a partial projection approach based on principal component analysis of the 3D coordinates. We analyzed the first two principal components to compute a distribution of packing density in the 2D projection corresponding to maximum variability in the spatial location of single neurons. In parallel, we conducted unsupervised clustering of 3D coordinates of single neurons using partitioning around medoids algorithm and annotated the packing density distributions and 3D ICN neuron maps with the resultant clusters. Our analysis revealed that the density distribution, numbers of clusters and orientation of ICN neurons were similar across males with 7–10 clusters within each ICN. This demonstrates the robustness of the technique where single ICN neurons were well‐preserved and they localized around the same regions despite variation in section orientation. Thus, this technique can potentially be applied to study the distribution of ICN neurons in 3D reconstructed hearts of large animals and humans as well as to examine the sex differences and diseased‐induced remodeling of the ICN for future studies.Support or Funding InformationThis work is supported by NIH SPARC Program, grant #3OT2OD023848‐01 (PD/PI: K. Shivkumar; Subaward PIs: J. S. Schwaber, R. Vadigepalli, and Z. J. Cheng)