Immune-Electrical Connection

Abstract

Macrophages interact with cardiomyocytes in the AV node via Cx43 gap junctions and assist normal AV nodal conduction, and macrophage depletion promotes AV block. Macrophages interact with cardiomyocytes in the AV node via Cx43 gap junctions and assist normal AV nodal conduction, and macrophage depletion promotes AV block. CITATION Hulsmans M, Clauss S, Xiao L, et al. Macrophages facilitate electrical conduction in the heart. Cell 2017; 169: 510–522. While the phagocytic role of macrophages in fighting pathogens and foreign environmental antigens is well known, the function of tissue-resident macrophages is less well understood. Cardiac-resident macrophages are yolk sac–derived rather than circulating monocyte–derived, and they self-renew via proliferation of local progenitors. They have been shown to play a role in tissue repair following heart tissue injury, but it is not clear whether they have a function at steady state. Using elegant genetic mouse models, single-cell RNA sequencing (RNA-Seq) and mathematical modeling, Hulsmans and colleagues demonstrate that cardiac-resident macrophages form gap junctions with cardiomyocytes in the atrioventricular (AV) node and help electrical conduction independent of the autonomic nervous system. The authors first demonstrated that the distal AV node, but not the left ventricle, is rich in macrophages, both in mice (CD11b+ F4/80+ Ly6Clo CD64+ CX3CR1+cells) and in humans (CD68+ CD163+ cells). Use of Cx3cr1GFP reporter mice enabled the investigators to visualize the cells’ elongated form in close contact with cardiomyocytes, whereas single-cell RNA-Seq demonstrated their enrichment in genes associated with conduction. Immunofluorescence and electron microscopy revealed an average of three punctate gap junctions/interactions between macrophages and cardiomyocytes, composed of the connexin 43 (Cx43) protein. Using whole-cell patch clamp, the authors showed that cardiomyocytes were depolarized when cultured in the presence of cardiac macrophages compared with cultured alone, whereas a fraction of the cocultured macrophages showed rhythmic depolarization and had a more negative resting membrane potential compatible with electrical coupling. Pharmacological blockade of Cx43 reversed the depolarization of cardiomyocytes cocultured with macrophages. Mathematical modeling recapitulated the observed more positive resting membrane potential of cardiomyocytes when associated with macrophages, which should facilitate myocyte depolarization with less stimulation; moreover, increasing coupling of macrophages/cardiomyocytes predicted accelerated cardiomyocyte repolarization, which should result in a shorter refractory period. Together, these alterations should assist AV conduction at higher frequencies. Indeed, forced depolarization of macrophages via photogenetics (in hearts from mice with tamoxifen-induced expression in macrophages of the photoactivatable cation channelrhodopsin 2, such that illuminating the AV node resulted in a Na+ influx) improved AV node conduction during rapid pacing, as measured by electrocardiogram. Finally, the authors used complementary mouse models to address the necessity of macrophage/cardiomyocyte electrical coupling for normal AV nodal conduction. Depletion of Cx43 in macrophages (Cx3cr1 Cx43−/−) promoted first- and second-degree AV block, as did use of CSF1op mice lacking tissue macrophages in many organs. Importantly for mouse immunologists who often use this genetic approach, diphtheria-induced depletion of macrophages in Cd11bDTR mice resulted in first- to third-degree AV block starting one day after diphtheria toxin (DT) injection, in a DT dose–dependent manner. In contrast, macrophage ablation following clodronate liposome injection had no detectable effect, presumably due to the very partial deletion effected by this approach. Thus, overall, macrophages in the distal AV node associate with cardiomyocytes via Cx43 gap junctions and are required for normal AV electrical conduction (Figure 1). These unanticipated results have implications for cardiac transplantation. Ischemia/reperfusion injury and inflammation are likely to change the phenotype and function of resident macrophages and thus, perhaps, alter their contribution to electrical conduction, which might result in threatening arrhythmias. Indeed, IL-1β has been associated with arrhythmias in diabetes-associated macrophage activation. More research needs to be devoted to the outcome of donor resident macrophages after transplantation. Do they continue to self-renew from local precursors, or are they replaced by host macrophages and, if so, can they establish gap junctions with donor cardiomyocytes? Certainly, arrhythmias and AV block are not uncommon after cardiac transplantation. Mouse experiments suggest that, after transplantation, donor class II disappears over time and T cells with direct reactivity to donor class II cease to be activated after a few weeks. This may suggest that donor resident macrophages are eliminated and replaced by host cells or that their expression of class II quiesces sufficiently to avoid direct T cell activation. A closer look at whether donor and/or host macrophages connect with donor myocytes at different times after transplantation and how they influence AV conduction may help optimize the monitoring of heart transplant recipients.