Cardiovascular diseases are the primary cause of death with ischemic heart diseases being the most abundant one often caused by myocardial infarction (MI). Zebrafish and neonatal mice are able to regenerate their heart after injury unlike adult mammals which have limited regenerative capacities. Interestingly, the autonomic nervous system (ANS) contributes to this regenerative capacity. Surgical impairment of the vagus nerve limited cardiac regeneration and resulted in incomplete healing. How MI impacts cardiac innervation and whether the ANS supports cardiac repair in the adulthood remains elusive. Therefore, we aim to provide an in-depth characterization of ANS patterns after MI and identify the role of nervous fibers in the adult MI. To assess the cardiac ANS after infarction, we performed histological profiling of hearts from mice 3, 7, 14 and 28 days after left anterior descending artery ligation. Non-infarcted hearts served as control group. Tyrosine hydroxylase staining revealed a gradual decrease in sympathetic innervation of the infarct and border zones (IZ and BZ) in the acute phase after MI, peaking at 14 days post-MI (5.25-fold (BZ, p=0.03) and 22.06-fold (IZ, p=0.02) decrease compared to control). Interestingly, we observed a re-innervation of these areas at day 28 (3.65-fold (BZ) and 8.35-fold (IZ) increase compared to day 14). The remote zone in the left ventricle showed no decline in innervation but a tendency towards sympathetic hyperinnervation at day 28 (1.87-fold increase compared to homeostasis). To study the mechanisms underlying this dynamic innervation kinetics, scRNA sequencing data from murine hearts at day 0, 1, 3, 5, 7, 14 and 28 post-MI were analyzed. The data revealed that axon guidance cues are dysregulated with an initial increase followed by a decline in the chronic phase post-MI. This regulation was particularly observed in vascular cells. As endothelial cells (EC) were shown to contribute to the cardiac neurovascular unit, we assessed capillary EC, where the expression of neuro-protective factors (such as Ntf3 and Vegfb) decreased in the acute phase of MI and increased again at days 14 and 28. Axon repelling factors (such as Sema3a and Flrt2) showed an inverse correlation. Flrt2 was of particular interest as its expression correlates with incidences for cardiovascular diseases according to the UK Biobank. Following MI, Flrt2 expression increased and peaked at day 7 post-MI. The expression decreased after but plateaued on day 14 and 28 post-MI. Additionally, we analyzed spatial transcriptomics data at day 1, 7, 14 and 28 post-MI. Flrt2 expression transiently increases after MI specifically in the IZ and peaked at day 7. This suggests that Flrt2 might be involved in cardiac ANS patterning post-MI. Indeed, EC specific deletion of Flrt2 in mice revealed an increased cardiac innervation compared to control mice (1.3-fold, p=0.07). In contrast to Flrt2, the expression of the neurotrophic factor Ntf3 transiently decreased particularly in the IZ post-MI. However, the remote zone revealed an increase in Ntf3 expression at day 28 fitting to the re-innervation observed at late time points. In conclusion, these data suggest that MI results in transient denervation of the infarcted heart with a re- and partial hyperinnervation at late time points which might be mediated by vascular axon guidance cues. Whether restoring physiological cardiac innervation attenuates cardiac remodeling, will be of therapeutic interest.
