Indirect CLEM Identifies Nanoscale Remodeling Associated with Atrial Fibrillation in Diverse Etiologies, Enabling a Unified Therapeutic Approach

Atrial fibrillation (AF) is the most common form of irregular cardiac rhythm (arrhythmia), affecting ∼3% of the US population. It is widely associated with vascular leak and inflammation and several risk factors are known to increase the incidence of new onset AF, including hypertension, diabetes, myocardial infarction, and many other pathologies associated with vascular dysfunction [1]. Despite its high prevalence and poor outcomes, the mechanisms underlying AF is not well understood and current treatments are limited to managing this progressive disease, rather than arresting the underlying pathology. AF patients have elevated levels of the inflammatory cytokine vascular endothelial growth factor (VEGF-A; 90–580 pg/ml), which promotes vascular leak and edema [2–4]. We investigated the structural underpinnings of AF using a novel approach we developed called indirect correlative light and electron microscopy (iCLEM). By exploiting structural landmarks identifiable via both light and electron microscopy, we were able to correlate transmission electron microscopy (TEM) –derived cardiac myocyte ultrastructure with molecular organization, assessed using sub-diffraction confocal imaging (sDCI) and stochastic optical reconstruction microscopy (STORM). Using this approach, we recently demonstrated that the vascular leak-inducing cytokine vascular endothelial growth factor (VEGF; 90–580 pg/ml – levels found in AF patients) prompts acute remodeling (30–60 minutes) of sodium channel (NaV1.5) –rich intercalated disk (ID) nanodomains, disrupting their ultrastructure and prompting translocation of NaV1.5 from these sites [5]. This in turn disrupted atrial impulse propagation (high speed optical mapping) and promoted atrial arrhythmias (electrocardiography). These results highlight inflammation-induced vascular leak as a novel AF mechanism, and further, characterizes a subpopulation of AF patients who present with inflammation and vascular dysfunction. To investigate if this novel vascular leak-induced arrhythmia mechanism has implications in other diseases associated with inflammation, we examined ID nanodomain protein distribution in cardiac tissue samples from patients experiencing two different pathologies associated with an increased risk of developing new onset AF: acute myocardial infarction (MI) and severe Coronavirus disease 2019 (COVID-19). First, we examined myocardial samples from five human MI patients. VEGF was overexpressed (IHC) in both cardiomyocytes and vascular endothelium in the border zone surrounding <6-month-old infarcts. Notably, co-localization analysis showed significantly reduced NaV1.5 near both connexin43 and N-cadherin within the border zone in 1-, 3-, and 9-day-old infarcts, paralleling our observations in mouse atria. Next, we compared the histologic and molecular findings in human cardiac tissue from fatal COVID-19 cases and age-matched controls. Hematoxylin and eosin (H&E) stains revealed histologic changes in the COVID-19 samples, most notably multifocal endothelial cell swelling and degeneration plus microvascular thrombi. Trace amounts of viral RNA and no nucleocapsid protein were detected in the COVID-19 hearts (qRT-PCR and in situ hybridization [ISH]). However, 7/11 COVID-19 hearts were positive for viral spike protein. Further validation by immunohistochemistry (IHC) showed that viral spike protein was detected in all 11 COVID-19 hearts, with increased inflammatory cytokine expression. Both Cx43 and NaV1.5 showed marked lateral migration in myocytes of COVID-19 hearts, consistent with previous experiments with inflammatory insults in murine hearts. These findings suggest that vascular-leak inducing inflammatory insult from viral spike protein, and not direct infection by SARS-CoV2, results in the pathophysiology of cardiac dysfunction in fatal COVID-19. We then tested the hypothesis that protecting the vascular endothelial barrier may prevent arrhythmias following an acute inflammatory insult by preserving cardiac ID nanodomains. We used our murine model of AF induced by acute inflammatory insult (100 pg/ml VEGF for 60 minutes) to test the antiarrhythmic efficacy of protecting the vascular endothelial barrier. We identified two molecular targets for vascular barrier protection, connexin43 (Cx43) and pannexin-1 (Panx1) [6,7]. In vascular endothelial cells, the opening of connexin43 hemichannels and pannexin channels, providing electrical and chemical coupling between cells and extracellular space, has been implicated in cytokine-induced vascular leak. Median in vivo arrhythmia burden was higher in VEGF–treated mice relative to vehicle controls (7.5±11 vs. 0±6 s/hr). VEGF shifted NaV1.5 signal to longer distances from Cx43 gap junctions (GJs), measured by a distance transformation-based spatial analysis of 3D confocal images of IDs. Further, VEGF decreased non-random attraction of NaV1.5 to Cx43 GJs (0.43 vs. 0.58 in vehicle controls), where 0 indicates random distribution of NaV1.5 in relation to Cx43 GJs and 1 indicates total superposition. Similar effects were observed with NaV1.5 localized near mechanical junctions (MJs) composed of N-cad. Blocking connexin43 hemichannels (αCT11 peptide) decreased in vivo arrhythmia burden to 0 ± 6.07 s/hr. PxIL2P (a peptide inhibitor of Panx1 channels) treatment also decreased in vivo arrhythmia burden ((0 ± 6.09 s/hr with 1.6 µM PxIL2P). Concurrently, both peptide therapies preserved NaV1.5 distance from GJs to control levels (0.50 with PxIL2P and 0.53 with αCT11) in these hearts. Once again, similar effects were observed near N-cad MJs. Notably, similar antiarrhythmic efficacy was also achieved with clinically relevant small molecule inhibitors of Cx43 and Panx1. In summary, these results highlight inflammation-induced vascular leak as a novel arrhythmia mechanism and present a novel approach to treating early-stage AF regardless of etiology with robust mechanistic validation.