The enigmatic role of macrophages in cardiac electrical signalling

This editorial refers to ‘Novel insights into the electrophysiology of murine cardiac macrophages: relevance of voltage-gated potassium channels’ by A Simon-Chica et al., pp. 798–813. ‘Yet, taught by time, my heart has learned to glow for other’s good, and melt at other’s woe’.—The Odyssey, Homer, 7th–8th century BCE The prominence of the macrophage in infected tissues has long been recognized.1 As are the frontline contributions of these versatile cells to the inflammatory response, including directed migration to sites of infection and injury, phagocytosis, and antigen presentation.2 These functions are further complemented by a strong body of evidence for allied roles in tissue repair and regeneration, the foreign body response, and the killing of tumour cells—amongst other related assignments. In 2017, a study in Cell by Hulsmans et al.3 reported a novel and unexpected role for macrophages in the electrical function of the heart. The electrophysiological and genetic data from mouse models provided therein were a tour-de-force. Most especially, the paper described loss-of-function experiments involving macrophage ablation and selective deletion of the gap junction protein Cx43 in macrophages that resulted in varying levels of atrioventricular block. Nevertheless, the morphological conclusions of this paper, based in part on glowing signals from lineage-tagged Green Fluorescent Protein positive (GFP+) macrophages, were less persuasive. The authors interpreted their results as being explained by Cx43-based couplings between macrophages and myocytes in the atrioventricular node—a component of the pacemaking and conduction system (PCS) thought responsible for generating the delay that separates activation of the atrial and ventricular chambers.4 However, an open-minded inspection of the first figure of the article suggests that the sole PCS component with a notable presence of GFP+ macrophages is the atrioventricular bundle—a specialized structure contributing to expedited conduction of action potentials between the node and the ventricles. The atrioventricular bundle is also called the bundle of His (after its discoverer Wilheim His, 1836–1934) and would be grouped by most anatomists with the specialized fast-conducting tissues of the His-Purkinje system, not the node. Indeed, the one figure in the article that shows the atrioventricular node proper, an atrial structure classically defined by Sunao Tawara more than 100 years ago,4 indicates that GFP-expressing macrophages are comprehensively absent from this tissue. These comments notwithstanding, the paper by Hulsmans et al.3 is a pivotal contribution. However, it is my view that its significance may relate as much to further our understanding of working myocardial tissues as it does to the PCS. Lineage-tagged macrophages could be seen to be abundant in non-PCS myocardial tissues in figures within the article. Moreover, experiments on co-cultures of macrophages and neonatal ventricular myocytes carried out by the authors convincingly showed Cx43-based contacts were common between the two cell types and that these contacts were associated with significant effects on the resting membrane potential (RMP) of working myocardial cells. Electrical interactions between macrophages and working myocytes are explored further in a new paper in this issue of Cardiovascular Research by Simon-Chica et al.5 In wet-lab studies, macrophages lineage-tagged with Yellow Fluorescent Protein (YFP) were isolated from 8 to 12 weeks old transgenic mice and probed using single-cell patch-clamp, RNA-seq, immunocytochemistry and high-resolution 3D fluorescence and morphometric imaging. The authors also used computational modelling to reproduce electrophysiological signatures observed from murine cardiac macrophages in vitro, and together with a previously reported model of the electrophysiology of ventricular myocytes from adult mice, the effects of macrophage-myocyte coupling on both cells types were simulated. A number of fascinating insights were provided by these models, including modulatory effects on the duration of action potential, with implications for understanding normal and disturbed cardiac electrical activity. However, in line with the 2017 Cell paper,3 the authors of this new paper concluded that the governing effect of macrophages likely related to their influence on myocyte RMP—with the −83.5 mV resting potential of myocytes being increased by more than 10% at the highest levels of heterocellular coupling (8 nS) modelled.5 It was furthermore proposed that these effects on RMP were determined primarily by the absence or presence of the inward-rectifier potassium channel Kir2.1. Consistent with this, the gene encoding the Kir2.1, Kcnj2, was robustly expressed by YFP+ macrophages isolated from mouse hearts and this channel was also prominent in functional studies of the cells. Kir2.1 is one of the usual suspects in the cohort of ion channels present in ventricular myocytes. The authors provided a helpfully comprehensive table of other ion channel transcripts identified by RNA-seq. This table indicated elevated transcript read counts in cardiac macrophages for various types of gap junctional (mostly Gja1/Cx43) calcium, sodium, and chloride channels. The presence of some of these channels, including the Ca2+ currents, were not confirmed by single-cell electrophysiology. However, what was surprising was the extent of the package—with evidence of other stalwarts of cardiac electrogenesis such as the gene encoding Nav1.5, and the confirmed presence of the channels responsible for the outward potassium currents Kv1.3 and Kv1.5 (Figure 1). One of the main parameters varied in computational models of the study was gap junctional coupling between macrophages and myocytes. However, as noted by the authors an additional type of electrical interaction—ephaptic coupling—may also need to be considered. Over the last decade, we have identified that Cx43, Kir2.1 Scn5a/Nav1.5, and Scn1b/β1 converge in a structure called the perinexus6,7—a narrow cleft (<30 nm) of extracellular space formed by cell membranes at the edge of gap junctions. Evidence continues to build that the perinexus contributes to propagation of excitation between myocytes.8,9 Given that the necessary components are present both in myocytes and cardiac macrophages, assembly of heterocellular perinexi between these two cell types seems a distinct possibility. This being said, the most interesting questions stemming from the study by Simon-Chica et al. may relate not to howness, but to whatness. Conceivably, either electrotonic or ephaptic coupling might give an account of HOW macrophages influence electrophysiological properties such as RMP. However, WHAT explains the presence of key parts of the machinery of action potential generation and propagation in macrophages remains enigmatic. Importantly and in this vein, towards the end of the discussion the authors share that, based on their analysis, macrophages seem likely to have, ‘…a very limited effect on cardiac electrophysiology in intact (healthy) mouse ventricles’. However, they go on to note that the effects of macrophages on electrophysiology may differ in injured hearts. For example, the well-established behaviours elicited by macrophages such as journeying to and proliferation at sites of injury may well provide high densities of these cells at myocardial infarctions. This same group has done stellar past work supportive of the possibility that Cx43-expressing non-myocytes may serve as something like electrical bridges in cardiac scar tissue, reducing the potential of block and arrhythmia.10 On the other hand, the authors concluding comments raise the prospect of turning full circle to canonical assignments of macrophages in immune and injury responses. What better way for the macrophage to detect, and even participate in the repair of injuries to an excitable tissue like the heart than to be able to directly sample electrical activity at cellular resolution. Thus, could the mirrored presence of channels such as Cx43, Kir2.1, Kv1.5, and possibly Nav1.5, also be consistent with a mechanism that enables macrophages to sense and assist troubled cells—in the working myocardium, as well as the atrioventricular bundle? A little empathy can go a long way for a broken heart. Simon-Chica et al.5 shine new light on characteristics of electrical signalling between cardiac myocytes and macrophages, including evidence of mirrored expression of ion channels subunits including Kir2.1, Kv1.5, Kv1.3, Cx43, and possibly Nav1.5. However, the purpose of this facility for heterocellular electrical interaction remains to be uncovered. The opinions expressed in this article are not necessarily those of the Editors of Cardiovascular Research or of the European Society of Cardiology. Many thanks to Ms Linda Collins for editorial assistance. RGG acknowledges support from NIH/NHBLI RO1 HL56728-20 and NIH/NHBLI RO1 HL141855-4. Conflict of interest: none declared.