Cardiomyocyte β-arrestins mediate inflammation and cGAS-STING activation in CVB3 viral myocarditis

Abstract

Viral myocarditis is a major cause of sudden cardiac death and can lead to dilated cardiomyopathy in adults. However, effective treatments remain elusive due to an incomplete understanding of its molecular drivers. Here, we investigate the role of β-arrestins (βarrs), scaffolding proteins that regulate GPCR signaling, in acute viral myocarditis. Using global βarr1 and βarr2 knockout (KO) mice, we assessed immune cell infiltration and apoptosis as markers of cardiac inflammation under Coxsackievirus (CVB3) infection. CVB3-infected βarr1 and 2 KO mice exhibited suppressed recruitment of NK cells, monocytes, macrophages, dendritic cells, and T cells over a broad range of viral titers at 7 days post-infection along with reduced cardiac apoptosis. At 4 days post-infection, immune cell expansion in secondary lymphoid organs, including B cells, CD8+ T cells, CD64+ myeloid progenitors, and monocyte/macrophages was also impaired in βarr KO mice. Importantly, cardiomyocyte-specific βarr1 and 2 dual deletion mirrored the attenuated inflammatory response and apoptosis observed in global βarr KO mice. Mechanistically, cardiomyocytes lacking βarr1 or βarr2 displayed defective cGAS-STING pathway activation, with impaired STING, TBK1, and IRF3 phosphorylation and inhibited IFNβ production at 24 hours post-CVB3 infection. These data highlight βarrs as critical mediators of the inflammatory response in the heart and secondary lymphoid organs during viral myocarditis and demonstrate that cardiomyocyte βarrs play a fundamental role in the inflammatory response to CVB3 viral myocarditis.

Graphical Abstract

New & Noteworthy

Viral myocarditis can lead to dilated cardiomyopathy in adults, yet the molecular mechanisms that induce cardiac inflammation in viral myocarditis remain unclear. β-arrestins are scaffolding proteins that mediate intracellular signaling during viral infection. We demonstrated that cardiomyocyte β-arrestins are necessary for immune cell recruitment, induction of cardiac apoptosis, and cGAS-STING pathway activation during acute coxsackievirus B3 infection. Targeting β-arrestins appears to decrease the inflammatory response in acute viral myocarditis.

INTRODUCTION

Myocarditis is defined as an inflammatory disease of the myocardium, predominantly triggered by a viral pathogen, that can lead to dilated cardiomyopathy and heart failure. One of the most common cardiotropic viruses in viral myocarditis is the positive single-stranded RNA enterovirus, coxsackievirus B3 (CVB3). During infection, CVB3 replicates in cardiomyocytes leading to acute and chronic immune-mediated damage (1). After acute CVB3 infection, cardiomyocytes release danger-associated molecular patterns (DAMPs) and proinflammatory cytokines, which serve to recruit and activate immune cells, amplifying the inflammatory response (2, 3). The innate immune detection of CVB3 is primarily mediated by the MDA5–MAVS signaling pathway, a branch of the RIG-I–like receptor (RLR) system (4). During viral replication, CVB3 generates long double-stranded RNA intermediates, which are sensed by MDA5 (Melanoma Differentiation-Associated protein 5), leading to the activation of MAVS (Mitochondrial Antiviral Signaling protein). MAVS activation induces the expression of type I interferons and proinflammatory cytokines through TBK1, IRF3, and NF-κB (5). In addition to RNA-sensing pathways, recent studies have identified the DNA-sensing cGAS–STING pathway as a critical contributor to host defense against RNA viruses (6). Although cyclic GMP-AMP synthase (cGAS) is not activated by viral RNA directly, virus-induced cellular stress leads to mitochondrial or nuclear DNA release into the cytosol, which is then sensed by cGAS, triggering stimulator of interferon genes (STING) activation and further amplification of inflammatory signaling (7). Several RNA viruses have evolved mechanisms to evade or antagonize STING-mediated immunity, underscoring the importance of the cGAS–STING pathway in antiviral immunity (8).

In humans, viral myocarditis morbidity and mortality correlate with the degree of local immune cell infiltration (9, 10). Monocytes, macrophages, and T cells (11) predominate in the immune cell infiltrates that contribute to the severity of viral myocarditis (12). Consistent with this, both T cell- and CD4+ T cell-deficient mice develop minimal cardiac inflammation (13). while CD8+ T cell-deficient mice display increased disease severity during acute CVB3 infection (14). This inflammatory process is further modulated by the recruitment of both anti- and pro-inflammatory monocytes that attenuate (15, 16) or aggravate CVB3-induced myocarditis, respectively (17-19). Thus, understanding the molecular mechanisms involved in the inflammatory response and recruitment of specific immune cell types to the heart could be used to determine the outcome of myocarditis progression versus resolution.

β-arrestins (βarrs), scaffolding proteins that mediate G protein-coupled receptor (GPCR) internalization, desensitization, and signaling, are also involved in viral cell entry and tissue infection (20, 21) as well as in regulation of immune responses. These responses include chemotaxis and activation of T cells (22-24), dendritic cells (25), macrophages (26, 27), and NK cells (28). Moreover, βarrs are known positive regulators of the DNA sensor cGAS which activates the STING pathway, a critical innate immune anti-viral pathway (29). Thus, βarrs regulate multiple features of the inflammatory response to infectious agents, including viruses associated with viral myocarditis. The precise role of βarrs on the cardiac inflammatory response induced by CVB3 infection is not known.

Here we examined the acute inflammatory response in CVB3-induced myocarditis using global βarr1 and βarr2 Knockout (KO) mice and conditional cardiomyocyte βarr1/2 KO mice (αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox) to determine if cardiomyocyte βarrs contribute to cardiac immune response during CVB3 myocarditis. We demonstrate that cardiomyocyte βarrs are necessary for a robust cardiac inflammatory response in response to CVB3 infection, possibly through activation of the cGAS-STING pathway in cardiomyocytes.

METHODS

Animals.

All animal studies were handled according to the approved protocol and accepted by the Duke University Animal Care and Use Committees. All mice had C57BL/6J genetic background. Global βarr1 KO, and global βarr2 KO mice have been previously described.(30) Breeding of α-myosin heavy chain-cre (C57BL:αMyHC-Cre) mice with βarr1/2 flox (C57BL: Arrb1^flox/flox/Arrb2^flox/flox) mice led to the generation of cardiomyocyte-specific deletion of both βarr1 and βarr2 (αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox). Littermate βarr1/2 flox mice without Cre were used as WT controls. Genetically modified mice did not result in a detectable cardiac phenotype in the steady state.

Coxsackievirus strain B3 (CVB3) generation.

The plasmid encoding the cDNA for the heart-specific CVB3 H3 strain was obtained from J. Lindsay Whitton, MD, PhD (Department of Immunology and Microbiology, Scripps Research). The CVB3 virus was propagated in HeLa cells and purified through ultracentrifugation over a sucrose cushion (31). The virus was generated in the laboratory as previously described (32), and stored in a −80°C freezer.

Animal infection and tissue harvesting.

Animals were assigned to experimental groups using simple randomization. Male and female 5 to 8-week-old mice were inoculated intraperitoneally (i.p.) with 1.5 to 2x10⁴ PFU of CVB3 in 100 μl of sterile distilled water. 100 μl of isotonic saline solution (0.9%) was used in control mice. Mice were killed at day 4 or 7 after infection, or earlier if they met the humane endpoint. Viral titers in the hearts of mice that lost 5% or less of body weight were not detectable and, therefore, were excluded from data analysis. The heart was perfused with 600 μl of PBS and harvested thereafter together with the spleen and mandibular lymph nodes from control and infected mice. Tissues were placed in 500 μl of Hanks' Balanced Salt Solution (HBSS) with 5% Fetal Bovine Serum (FBS) for further processing. Blood was also collected for flow cytometric analysis. The apexes of hearts were cut, weighed, suspended into 1 ml of Minimum Essential Medium (MEM), and stored in −80°C freezer for viral titration measurements.

Plaque assay.

The amount of virus was assessed by plaque assays as previously described (33). Briefly, samples were thawed and homogenized in a Beadbug^™ Microtube Homogenizer with beads (Benchmark Scientific) at 4000 RPM for 90 seconds before centrifugation at 5000g for 5 minutes. To determine viral titers, supernatants of the samples were serially diluted in 7B MEM (5% FBS, 1X penicillin/streptomycin [P/S], 1X non-essential amino acids [NEAA]; Corning), incubated in 70% confluent HeLa cells for 60 minutes at room temperature, and overlayed with 1:1 mixture of 1% agarose and 2X MEM containing 4% FBS, 2% P/S, and 2% NEAA. Plaques were visualized and counted 45 hours post-infection by 0.05% crystal violet staining in 10% ethanol (Sigma).

Histological and immunofluorescence assessment.

For histology analysis, hearts were fixed in formalin (10% formaldehyde) overnight. Routine hematoxylin and eosin staining was performed per manufacturer’s instructions (Electron Microscopy Sciences) and imaged under an Axio Imager Z2 upright microscope (Zeiss). For immunofluorescence assessment, hearts were fixed in 4% paraformaldehyde (PFA) overnight and immersed for 24 hours in a 30% sucrose solution. The hearts were embedded in OCT (Tissue-Tek) using an ethanol-dry ice bath and sectioned at 5μm using a cryostat (Leica). Sections were blocked with blocking buffer (1x PBST, 10% NCS, 1% dimethyl sulfoxide [DMSO]) and incubated overnight at 4°C with anti-CD45 primary antibody (Invitrogen, RRID: AB_467251, Cat # 14-0451-82, 1:200). The secondary antibody Alexa Fluor 594 donkey anti-rat (1:200) and 4’,6-diamidino-2-phenylindole (DAPI; 100 ng/ml) were added the next day before imaging at 40X on a CSU-X1 spinning disk confocal microscope (Zeiss). Images were quantified using ImageJ software (RRID:SCR_003070) by calculating the ratio of blue (DAPI, all cell nuclei) and red (positively stained cells) pixels to determine the percentage of inflammation.

TUNEL assay.

TUNEL staining on the cryo-sectioned slides were performed with in situ cell death detection kit (Sigma-Aldrich, Roche) according to the manufacturer’s protocol. Slides were then counterstained with DAPI and mounted in ProLon Diamond Antifade Mountant (Thermo Fisher Scientific). Slides were recorded with CSU-X1 spinning disk confocal microscope (Zeiss) with 20X objective. In each heart, 3 areas were selected for maximum TUNEL-positive cells, and at least 3 sections were quantified for each heart. The total number of positive TUNEL cells and DAPI-stained nuclei were counted using ImageJ software.

Single-cell isolation and flow cytometry.

Solid tissues were digested in digest media (HBSS with Collagenase A [10 mg/ml; Millipore Sigma], DNase I [20mg/ml; Millipore Sigma], 5% FBS, and 10 mM HEPES) for 30 mins at 37°C under constant shaking (120 rpm). The tissues were then mechanically dissociated, and 4 ml of fresh digest media was added for another 45 minutes of incubation vortexing every 8 minutes until tissue was completely digested. The digest media was deactivated with phosphate buffered saline (PBS) buffer and the solution was filtered with a 70 μM cell strainer to obtain single-cell suspensions. For blood, a single wash with PBS and filtering was used to obtain single cells. Red blood cells were lysed with ammonium-chloride-potassium (ACK) buffer (Gibco^™) for 3 mins. Single cell suspensions were washed with PBS and stained with Zombie Aqua^™ fixable viability kit (Biolegend, cat # 423101) for 25 minutes at 4°C to detect live and dead cells. After centrifugation, samples were blocked with 1% anti-mouse CD16/32 monoclonal antibody (Biolegend, RRID: AB_312801, cat # 101302), normal rat and mouse serum (5% each) in FACS buffer (PBS with 3% FBS, 10 mM Ethylenediaminetetraacetic acid [EDTA]). Cell suspensions containing 2 x 10⁶ cells were labeled with the antibodies listed in Table 1 to detect multiple immune cell populations (Supplemental Figures 2 and 4). Flow cytometric compensation was performed at each session using UltraComp eBeads^™ (Invitrogen) individually stained with every antibody/fluorochrome in the panel to compute a run-specific compensation matrix. After compensation, residuals and spillover-spreading were inspected using single-stain overlays. Debris, dead cells, and doublets were excluded (FSC/SSC, FSC-A vs. FSC-H, viability dye). Approximately 3 x 10^5 events per heart were acquired. Fluorescence-minus-one (FMO) controls for all dim/critical markers were used to set gates, which were verified by back-gating into parent populations.

TABLE 1.

Antibodies used for flow cytometry to label distinct immune cell populations

Antibodies	RRID	Catalog #	Company	Dilution
FITC anti-mouse Ly-6G	AB_1236494	127606	Biolegend	1:50
PE anti-mouse/human CD11b	AB_312791	101208	Biolegend	1:200
PE-Cyanine5.5, CD19 Monoclonal	AB_891395	35-0193-82	Invitrogen	1:100
PE/Cyanine7 anti-mouse CD3ε	AB_312685	100320	Biolegend	1:100
APC anti-mouse CD11c	AB_313778	117309	Biolegend	1:800
APC/Cyanine7 anti-mouse CD8a	AB_312753	100714	Biolegend	1:100
Pacific Blue™ anti-mouse CD4	AB_493646	100427	Biolegend	1:100
Brilliant Violet 605™ anti-mouse CD45	AB_2562342	103140	Biolegend	1:100
Brilliant Violet 650™ anti-mouse I-A/I-E	AB_2565975	107641	Biolegend	1:200
Brilliant Violet 711™ anti-mouse CD64 (FcγRI)	AB_2563846	139311	Biolegend	1:50

Cardiomyocyte isolation.

Ventricular cardiomyocytes were isolated using standard Langendorff perfusion procedures. 24 hours after infection, mice were injected i.p. with 200U of heparin and euthanized 10 minutes later by an overdose of isoflurane administered via inhalation in a closed chamber. Dissected whole hearts were placed immediately into perfusion buffer (120 mM NaCl, 14.8 mM KCl, 0.6 mM KH₂PO₄, 0.6 mM Na₂HPO₄, 1.2 mM MgSO₄×7 H₂O, 10 mM HEPES, 4.6 mM NaHCO₃, 30 mM taurine, and 5.6 mM glucose, pH 7.3) and cannulated through the aorta. Hearts were perfused in retrograde for 3 minutes with oxygenated perfusion buffer and then for ~8 minutes with digestion buffer containing 2.4 mg/ml collagenase (Worthington) at 37°C. To terminate enzymatic digestion, ventricular tissues were transferred to perfusion buffer containing 10% calf serum with 12.5 μM CaCl₂. Myocytes were dissociated by trituration and brought to 0.4 mM Ca²⁺ concentration.

Western blot analysis.

Lysed cardiomyocytes were centrifuged at 500g for 10 minutes, the supernatant was collected, and the pellet with nuclei and cell debris was discarded to exclude nuclear cGAS detection (34). Total protein concentration of the cytosolic and membrane fractions of cardiomyocyte lysates was determined by BCA Protein Assay (Thermo Fisher), and 100 μg of total protein was loaded on the gel. Samples were separated with SDS-PAGE, transferred to nitrocellulose membranes, and immunoblotted. Primary antibodies and dilution used are as follow: monoclonal phospho-STING (Ser365) antibody (1:5000) (72971, Cell Signaling, RRID: AB_2799831), monoclonal phospho-IRF-3 (Ser396) (1:1000) (29047, Cell Signaling, RRID: AB_2773013), monoclonal phospho-TBK1/NAK (Ser172) (1:1000) (5483, Cell Signaling, RRID: AB_10693472), monoclonal cGAS antibody (1:5000) (703149, Thermo Fisher, RRID: AB_2809227), monoclonal IFN- β1 antibody (1:5000) (73671, Cell Signaling, RRID: AB_2799843), monoclonal HRP conjugated GADPH antibody (1:1000) (3683, Cell Signaling, RRID:AB_1642205). Secondary antibodies included HRP conjugated anti-rabbit (NA934, Cytiva Life Sciences, RRID: AB_772206). Western Blot images were taken using BioRad ChemiDoc system and the densitometry analysis was performed by ImageLab v6.1.

Statistical analysis.

Data were analyzed using unpaired t-test, one- or two-way ANOVA with Dunnet post-hoc test for multiple comparisons when appropriate. Y-intercepts were used to compare linear regression analysis. Mortality analysis was done using Kaplan-Meier survival curve. All analyses were performed using GraphPad Prism software (version 10.4, RRID:SCR_002798). A P value of ≤0.05 was considered statistically significant.

RESULTS

A lack of β-arrestins impairs the cardiac acute inflammatory response to CVB3 infection.

We examined acute CVB3 viral myocarditis in the previously validated mouse strain C57BL/6J (35) as Wild Type (WT) controls, together with global βarr1 KO, and βarr2 KO mice with the same genetic background 7 days after an i.p. injection with CVB3 virus (~1.5x10⁴ PFU). The influence of βarrs on the acute cardiac inflammatory response was assessed by measuring immune cell infiltrates. Immunofluorescence studies using the pan-leukocyte marker CD45 in CVB3-infected heart sections of male WT mice showed extensive leukocyte infiltration that was significantly blunted in the hearts of global βarr1 KO and βarr2 KO mice (Figure 1A). A positive correlation between leukocyte infiltrates and cardiac viral titers was observed in WT mice, which was absent in βarr1 KO and βarr2 KO mice (Figure 1C). This finding was confirmed using flow cytometric analysis in mice of both sexes (Figures 1B and 1D). Compared to male mice, an attenuated inflammatory phenotype was observed in female WT mice (bottom graph, Figure 1D) as previously described (36, 37). To determine if the reduced inflammatory response in βarr1 KO and βarr2 KO mice during viral myocarditis led to reduced cardiac injury, we measured apoptotic cells in heart tissues. Cardiac tissues of infected WT mice had ~6% TUNEL-positive cells, while both βarr1 KO and βarr2 KO mice showed reduced cardiac apoptosis with only ~2% TUNEL-positive cells (Figures 1E and 1F). These data indicate that βarrs are necessary to induce an inflammatory response and cardiac injury to CVB3 infection.

Figure 1.

The cardiac inflammatory response by CVB3 infection at 7 days is dampened in global β-arrestin 1 and 2 KO mice. (A) Cross-sections of infected hearts with similar viral titers (left; WT=3.87; βarr1 KO=4.16; βarr2 KO=4.62 pfu/mg of heart). Scale bars are 500 μm or 100 μm for full cross-sections or zoomed-in sections, respectively. (B) Representative single-cell flow panels of uninfected (saline; left) and CVB3-infected (right) hearts for all groups. (C) Quantification of inflammation against viral titers (right) shows reduced immune infiltration (CD45+; pink) in global βarr1 KO (Red) and βarr2 KO (Green) mice (7.8% and 9.3%, respectively) when compared to WT (54.4%; Black) by immunofluorescence. (D) Quantification of the CD45+ population in the heart by flow cytometry analysis is positively correlated to the cardiac viral titers in WT (Black; n=13 males, n=13 females) for both male (full circles) and female (empty circles), whereas βarr1 KO (Red; n=8 males, n=9 females), and βarr2 KO (Green; n=9 males, n=9 females) mice show marked reduction of the CD45+ population across a broad range of viral titers. Mice injected with saline are shown as uninfected controls (Males: WT n=6, βarr1 KO n=5, βarr2 KO n=5; Females: WT n=5; βarr1 KO n=5, βarr2 KO n=5). (E) Representative images of cardiac tissue (left) showing reduced apoptosis (TUNEL+ cells; white arrows) in global βarr1 KO and βarr2 KO mice during acute CVB3-induced myocarditis. Scale bars are 20 μm. (F) Quantification of cardiac apoptosis (right) measured as TUNEL+ cell nuclei over nuclei number (DAPI), n=6 mouse hearts per group. Statistical test in C and D is linear regression analysis comparing Y-intercepts, and in F is one-way ANOVA with Dunnett for multiple comparisons. Data expressed as mean ± S.E.M. Significance was *p≤0.05, **p ≤0.01; ****≤0.0001 vs WT.

Survival rate, cardiac and hepatic viral titers as well as body weight were also analyzed. Mortality at 7 days post-infection (dpi) was reduced by ~40% in global βarr1 KO male mice when compared to WT controls while the survival rate for females was similar among all groups (Supplemental Figure S1A). Early mortality during the first 4 dpi in CVB3-infected mice was associated with increased hepatic viral titers (Supplemental Figure S1B) and macroscopic hepatic damage (not shown). The hepatic viral titers during late mortality at 5 to 7 dpi were significantly lower when compared with early mortality (Supplemental Figure S1B). This suggests an early survival effect of βarr1 in male mice that is not directly associated with myocarditis. All CVB3-infected groups showed similar weight loss (~20%; Supplemental Figure S1C) and a broad range of cardiac viral titers at 7 dpi (Supplemental Figure S1D). When analyzed by sex, cardiac viral titers in female βarr1 KO mice were significantly higher than their WT controls and βarr2 KO mice, whereas male mice had similar cardiac viral titers among all groups (Supplemental Figure S1A and S1D). The sex-specific effect in βarr1 KO mice was not explored further. Hematoxylin & Eosin staining showed extensive cellular infiltrates across the infected hearts of WT mice that were not visible in either global βarr1 KO or βarr2 KO mice, also confirming our initial findings (Supplemental Figure S1E).

The innate and adaptive immune response elicited by CVB3-induced myocarditis is dampened in global βarr1 KO and βarr2 KO mice.

The pathophysiology of viral myocarditis involves multiple immune cell types that infiltrate the heart during infection in a sequential manner based on their specific function (12, 38, 39). Once recruited, each immune cell subset contributes to the inflammatory process in the heart that ultimately determines the progression of the infectious disease (12). Therefore, we assessed whether βarrs played a role in mediating the infiltration of a specific subset of the innate and adaptive immune cell populations in CVB3-infected hearts at 7 dpi. We developed a gating strategy based on the flow cytometric analysis of Yu et al. (40), previously used to identify major immune cell populations (Supplemental Figure S2). At 7 dpi, the innate immune populations of cDCs, CD11b- DCs, NK cells, macrophages, and monocytes showed increased cardiac infiltration that was positively correlated with viral titers in WT mice (Figure 2, Supplemental Figure 3A). In contrast, cardiac infiltration of innate immune cell populations in βarr1 KO and βarr2 KO mice was blunted even at high viral titers (Figure 2). This inflammatory pattern in both βarr KO mice was also consistent for the adaptive immune populations of CD4+ T cells and CD8+ T cells, while B cell infiltrates were only decreased in βarr2 KO mice (Figure 3). Cardiac infiltrates were less prominent in WT females than WT males, as previously shown (36, 37). These data indicate that βarrs are required for the inflammatory response over a broad range of viral titers in CVB3 myocarditis.

Figure 2.

Global β-arrestin 1 or 2 deletion abrogates the innate immune response in the heart elicited after 7 days of CVB3 infection. (A and B) Representative single-cell flow panels of saline and CVB3-infected hearts with similar viral titers (3.5 - 4 Log₁₀[PFU/mg heart]) across groups show decreased innate immune cell populations in βarr1 KO and βarr2 KO mice when compared to WT mice. WT mice (Black; n=13 males, n=13 females) show increased immune cell recruitment to the heart directly proportional to the viral titers of the CVB3-infected hearts. Conversely, infected βarr1 KO (Red; n=8 males, n=9 females) and βarr2 KO (Green; n=9 males, n=9 females) mice show a marked reduction in the recruitment of multiple innate immune cell populations to the heart, including (C) conventional CD11b- Dendritic cells, (D) NK cells, (E) Macrophages, and (F) Monocytes. Mice injected with saline are shown as uninfected controls (Males: WT n=6, βarr1 KO n=5, βarr2 KO n=5; Females: WT n=5; βarr1 KO n=5, βarr2 KO n=5). Linear regression comparing Y-intercepts was used to assess statistical differences among infected mice. One-way ANOVA and Dunnett for multiple comparisons was used among saline controls. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001 vs WT mice.

Figure 3.

Global β-arrestin 1 or 2 deletion abrogates the adaptive immune response in the heart elicited after 7 days of CVB3 infection. (A and B) Representative single-cell flow panels of saline and CVB3-infected hearts with similar viral titers (3.5 - 4 Log₁₀[PFU/mg heart]) across groups show decreased adaptive immune cell populations in βarr1 KO and βarr2 KO mice when compared to WT mice. WT mice (Black; n=13 males, n=13 females) show increased immune cell recruitment to the heart directly proportional to the viral titers of the CVB3-infected hearts. Conversely, infected βarr1 KO (Red; n=8 males, n=9 females) and βarr2 KO (Green; n=9 males, n=9 females) mice show a marked reduction in the recruitment of multiple adaptive immune cell populations to the heart, including (C) B cells, (D) CD4+ T cells, and (E) CD8+ T cells. Mice injected with saline are shown as uninfected controls (Males: WT n=6, βarr1 KO n=5, βarr2 KO n=5; Females: WT n=5; βarr1 KO n=5, βarr2 KO n=5). Linear regression comparisons were used to assess statistical differences among infected mice. One-way ANOVA and Dunnett for multiple comparisons was used among saline controls. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001 vs WT mice.

Neutrophils, one of the responders during infection (39), and eosinophils showed a similar number of cardiac infiltrates in infected hearts in all groups (Supplemental Figure S4). Measurement of professional antigen presenting cells showed fewer CD11b+ conventional dendritic cells (cDCs) in uninfected βarr1 KO and βarr2 KO mice when compared to WT mice (Supplemental Figure S3A). For CD11b- DCs, only βarr1 KO mice showed a decreased cell number in uninfected hearts when compared to WT mice (Figure 2A and 2C). Further analysis by sex showed that such significant decrease was driven by male mice for cDCs and female mice for CD11b- DCs (Supplemental Figure S4B). Taken together, these results suggest that the innate and adaptive immune responses elicited by CVB3 myocarditis are dampened in global βarr1- and βarr2-KO mice.

β-arrestin deficiency impedes immune cell expansion in secondary lymphoid organs after CVB3 infection.

We determined if immune cell expansion was altered in βarrs KO mice by measuring immune cell populations in lymph nodes and the spleen before and 4 and 7 days after CVB3 infection. We used an alternative gating strategy to account for the high heterogeneity in myeloid populations among lymphoid tissues (41, 42). The CD11c marker was used to detect DC subsets, whereas CD11c-, CD11b+ populations were subdivided into monocyte/macrophages (type II major histocompatibility complex [MHC II]+) or CD64+ myeloid progenitors (MHC II-). This allowed us to comprehensively assess immune progenitors and specific immune cells relevant to viral myocarditis (Supplemental Figure S5). The lymphoid organs of female mice during CVB3 infection were not analyzed as the cardiac inflammatory response was less severe.

An increased number of monocytes/macrophages, B cells, and CD8+ T cells was observed in the lymph nodes of CVB3-infected WT mice at 4 dpi, which returned to basal values at 7 dpi (Figure 4A to 4C). In contrast, this expansion of lymph node monocytes/macrophages, B cells, and CD8+ T cell numbers at 4 dpi did not occur in either βarr1 KO or βarr2 KO mice, and numbers of monocytes/macrophages and CD8+ T cells further decreased at 7 dpi (Figure 4A and 4C). In the spleen, expansion of B cells and CD64+ myeloid progenitors was also detected at 4 dpi and returned to basal values at 7 dpi in CVB3-infected WT mice (Figure 4D and 4E). B cells of βarr1 KO and βarr2 KO mice were unchanged across 4 and 7 dpi when compared to their uninfected controls (Figure 4D). CVB3 infection increased CD64+ myeloid progenitors in βarr1 KO mice at 4 dpi and reached significance at 7 dpi while CD64+ myeloid progenitors remained unchanged in βarr2 KO mice (Figure 4E).

Figure 4.

Immune cell expansion is inhibited at secondary lymphoid organs during acute CVB3 infection in global βarestin1 or 2 KO mice. Immune populations (A) Monocytes/Macrophages, (B) B cells, and (C) CD8+ T cells from lymph nodes of CVB3-infected WT mice (Black; n=6 controls [C]) increase at 4 dpi (n=6) and recede at 7 dpi (n=7) whereas the same populations show limited expansion in βarr1 KO (Red; n=4 at 4dpi, n=7 at 7dpi) and βarr2 KO (Green; n=6 at 4dpi, n=6 at 7dpi) mice compared to their uninfected controls (βarr1 KO n=5; βarr2 KO n=5). Similarly, (D) B cells and (E) CD64+ myeloid progenitors in the spleen of CVB3-infected WT mice increase at 4 dpi and diminish at 7 dpi. These immune cell subsets are unaffected at 4 dpi in βarr1 KO and βarr2 KO mice compared to their uninfected controls. Data is shown as the mean ± S.E.M. Two-way ANOVA and Dunnett for multiple comparisons within groups was used for statistical analysis. *p<0.05, **p<0.01, and ***p<0.001, vs uninfected controls. C - uninfected day zero; d4, d7 – 4 and 7 days after CVB3 infection.

During the uninfected steady state, both DC subsets in lymph nodes and spleen were decreased in βarr1 KO mice and βarr2 KO mice when compared to WT controls and seemed unresponsive to viral infection in lymph nodes (Supplemental Figure S6A and S6B). Additionally, DCs in blood were reduced in uninfected βarr2 KO mice when compared to WT mice, although statistical significance was reached only for cDCs (Supplemental Figure S7). These findings suggest a compromised DC surveillance system (38) in mice lacking βarrs under the context of CVB3 viral myocarditis.

Lack of β-arrestins in cardiomyocytes recapitulates the blunted inflammatory response and reduced apoptosis observed in global β-arrestin KO mice.

A fundamental role for cardiomyocytes in the activation of the immune response during infectious (43, 44) and non-infectious diseases (45, 46) is increasingly being recognized. Therefore, we investigated whether cardiomyocyte βarrs are critical mediators of the cardiac inflammatory response during CVB3 viral infection. Mice with cardiomyocyte-specific deletion of βarr1 and 2 (αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox) were used to test whether cardiac inflammation to CVB3 infection was also blunted at 7 dpi. CVB3-infected WT littermates (FLOX βarr1/2) showed a positive correlation between cardiac viral titers and leukocyte infiltration at 7 dpi (Figure 5A). This pattern was also clearly observed when CVB3-infected hearts of FLOX βarr1/2 mice were analyzed for monocytes, macrophages, CD4+ T cells, and CD8+ T cells (Figures 5B to 5E). In marked contrast, infiltration of immune cell subsets was significantly abrogated across a wide range of viral titers in αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox mice consistent with our findings in global βarr1 KO and βarr2 KO mice. Both DC subsets were comparable between αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox mice and their littermate controls (Figures 5F and 5G) suggesting that DCs are not affected by the absence of cardiomyocyte βarrs during viral myocarditis. The mortality rate and cardiac viral titers were similar between groups at 7 dpi (Supplemental Figure S8A and S8B).

Figure 5.

Cardiomyocyte deletion of β-arrestin1 and 2 abrogates the cardiac inflammatory response in acute CVB3-induced myocarditis. Flox βarr1/2 WT mice (Grey; n=16 in all populations) show increased (A) immune cell recruitment to the heart at 7 dpi, directly proportional to the viral titers of the CVB3-infected hearts. Conversely, infected αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox mice (Purple; n=11 for lymphoid populations, n=9 for myeloid populations) show a marked reduction in the recruitment of adaptive immune cell populations to the heart, including (B) monocytes, (C) macrophages, (D) CD4+ T cells, and (E) CD8+ T cells. (F and G) Both DC subset counts are similar between αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox mice and their WT littermates. (H) Representative images of cardiac tissue showing reduced apoptosis (TUNEL+ cells; white arrows) in αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox mice during acute CVB3-induced myocarditis. Scale bars are 20 μm. (I) Quantification of cardiac apoptosis measured as TUNEL+ cell nuclei over nuclei number (DAPI; n=8 gray, n=5 purple). Statistical tests in A to G are linear regression analysis comparing Y-intercepts. (I) was compared using unpaired T-test. Data expressed as mean ± S.E.M. *p≤0.05 and ****≤0.0001 vs FLOX βarr1/2 controls.

Cardiac apoptosis was reduced in αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox mice (~3%) when compared to WT controls (~6%), even though there was an average viral titer 5-fold higher in αMyHC-Cre:Arrb1^flox/flox/Arrb2^flox/flox hearts compared to the WT FLOX βarr1/2 hearts (5.61x10⁴ vs 1.16x10⁴ PFU/mg, respectively; Figures 5H and 5I). Taken together, these data indicate that cardiomyocyte βarrs are major molecular contributors to the cardiac inflammatory response against CVB3 viral infection.

Activation of the cGAS-STING pathway after CVB3 infection is impaired in cardiomyocytes of mice lacking βarr1 or βarr2.

The cGAS-STING pathway is a critical component of the innate immunity that detects foreign and native dsDNA fragments in the cytoplasm to then trigger the synthesis and secretion of type I interferons (IFN) and other proinflammatory cytokines (47, 48). Activation of this pathway induces the phosphorylation of downstream proteins such as STING itself, TANK-binding kinase 1 (TKB1), and IFN regulatory factor 3 (IRF3; Figure 6A) (49). Additionally, cGAS-STING pathway activation has been detected in macrophages from murine ischemic hearts and during viral myocarditis (50, 51). Since βarrs promote the activation of the cGAS-STING pathway to enhance an anti-viral response (29), we explored the possibility of cGAS-STING activation within cardiomyocytes after CVB3 infection. We first assessed the time course of IRF3 phosphorylation, a hallmark of pathway activation, in freshly isolated cardiomyocytes from WT mice infected with CVB3 for 24, 48, and 96 hours. Phosphorylated IRF3 (pIRF3) was detectable as early as 24 hours but was no longer observed by 2 days post-infection, prompting us to select the 24-hour time point for subsequent experiments (Supplemental Figure S9A). Accordingly, cardiomyocytes isolated from WT, βarr1 KO, or βarr2 KO mice were analyzed at 24 hours post-infection for expression of cGAS, phosphorylated STING (pSTING), pTKB1, pIRF3, and IFNβ. Cytosolic and membrane fractions of infected WT cardiomyocytes showed a similar expression of cGAS when compared to cardiomyocytes devoid of βarr1 or 2 (Figure 6B, Supplemental Figure S9B). In contrast, both βarr1 and βarr 2 KO cardiomyocytes showed decreased pSTING and pTKB1 (Figures 6C and 6D, Supplemental Figure S9B) when compared to WT cardiomyocytes. For IRF3, only βarr2 KO cardiomyocytes showed decreased phosphorylation when compared to WT controls (Figure 6E, Supplemental Figure S9B). Importantly, IFNβ was decreased in both βarr1 and βarr2 KO cardiomyocytes compared to WT cardiomyocytes consistent with the inhibition of the upstream components of the cGAS-STING pathway (Figure 6F, Supplemental Figure S9B). Taken together, these data suggest that βarrs are important mediators of the activation of the cGAS-STING pathway within cardiomyocytes at the onset of CVB3 viral infection to induce inflammation.

Figure 6.

β-arrestins are necessary for the activation of the cGAS-STING pathway in cardiomyocytes at 24 hours of CVB3 infection. Cardiomyocytes isolated from WT (Grey; n=6), βarr1 KO (Red; n=6), and βarr2 KO (Green; n=6) after 24 hours of CVB3 infection. (A) Representative illustration of the innate immune cGAS-STING pathway activation in cardiomyocytes after viral infection. Illustration was created in BioRender (Lucero, E. ((2025)) https://BioRender.com/1yyft09). All groups show similar expression of (B) cGAS. Cardiomyocytes of βarr1 and βarr2 KO mice show reduced phosphorylation of (C) STING, (D) TBK1, as well as reduced expression of (F) IFNβ when compared to WT controls. (E) IRF3 phosphorylation was reduced only in βarr2 KO cardiomyocytes when compared to WT controls. One-way ANOVA with Dunnett for multiple comparisons was used for statistical analysis. Data expressed as mean ± S.E.M. *p≤0.05, **p ≤0.01; ***≤0.001 vs WT. CAR: Coxsackievirus Adenovirus receptor

DISCUSSION

βarrs are highly versatile proteins that are critical in modulating distinct immune responses against viral infection; thus, we investigated their role in CVB3-induced viral myocarditis. Herein, we describe systemic and cardiomyocyte-specific βarrs as necessary mediators to induce a robust cardiac inflammatory response by both the innate and adaptive immunity under acute CVB3 infection. In secondary lymphoid organs, βarrs are required to trigger B cell, CD8+ T cell, monocyte/macrophages, and CD64+ myeloid progenitor expansion. Deletion of cardiomyocyte βarrs phenocopied the blunted inflammatory response and reduced apoptosis observed in CVB3-infected global βarr KO mice. These effects were correlated with an impaired activation of the cardiomyocyte cGAS-STING pathway at 24 hours of CVB3 infection.

The extent of the cardiac inflammatory response and mortality rate was sex-dependent impacting WT male mice more severely (36, 52). Early mortality in male mice was correlated with high hepatic viral titers, consistent with the concept of the liver as a first filter for CVB3 infection (53).

Reduced recruitment of most immune cells to the heart in both global βarr KO mice included the major contributors to the viral myocarditis phenotype, namely, monocytes, macrophages, and CD8+ T cells. βarrs are necessary for macrophage and T cell chemotaxis (23, 26). The degree to which each immune population relies on βarrs for migration may help explain the differential recruitment phenotypes in βarr1 and 2 KO mice during acute viral myocarditis. The reduced recruitment of most immune cell populations correlated with an earlier inhibition of immune cell expansion of such immune cells in the lymph nodes or spleen during antigen presentation. During viral infection, antigen presentation in secondary lymphoid organs drives immune cell proliferation, contributing to a robust cardiac inflammatory response (38, 54). Moreover, antigen presentation requires functional DCs that survey the heart for the detection of viral particles and DAMPs (2, 38, 55). Once antigens are detected, DCs travel to lymph nodes for antigen presentation, immune cell activation, and expansion (56, 57). Alternatively, viral infection triggers extramedullary myelopoiesis and B cell expansion in the spleen as we observed in WT mice (15, 58). Unexpectedly, βarr1 and βarr2 KO mice exhibited abnormal DC counts in secondary lymphoid organs and the heart, as well as in the blood during the steady state. Moreover, CD11b- DCs, the subset predominantly reactive to viral infection, were unresponsive in lymph nodes of both βarr KO mice. Some of these findings have been previously reported in uninfected mice lacking βarr2 where splenic DC accumulation was observed due to increased migration (25). This affected DC functionality when animals were challenged with antigens to induce autoimmune diseases (25). Therefore, it is possible that, in uninfected mice, βarrs regulate DC homeostasis across multiple organs impacting the inflammatory response to viral infection. The optimal antiviral response to infection might be dependent on a functional DC surveillance system for cell expansion to occur at secondary lymphoid organs during antigen presentation.

βarrs are scaffolds for multiple signaling pathways involved in cell proliferation (59, 60). In the immune system, βarr1 overexpression promotes, while its deletion impairs, T cell proliferation following antigen stimulation (61). This resembles the null CD8+ T cell expansion in lymph nodes observed in virally infected global βarr1 and 2 KO mice. Interestingly, altered cell counts were also observed in other immune populations, including macrophages/monocytes in lymph nodes (Figure 4A) and of CD4+ T cells in the hearts (Figure 3D) of uninfected global βarr1 KO and βarr2 KO mice, respectively. Since βarrs are involved in developmental signaling pathways for stem/progenitor cell differentiation (62), we postulate that the development of the myeloid and lymphoid lineages is dependent on βarrs for differentiation and distribution throughout lymphoid and non-lymphoid organs. Future studies will explore the role of βarrs in the development and proliferation of these immune cell populations before and after viral infection. Taken together, the effects contributing to the abrogated cardiac inflammatory response during viral myocarditis come from disturbing multiple signaling networks within immune cells when βarrs are absent. At a cellular level, the shifts in the distribution of immune populations in various organs, the impaired responsiveness of DCs to viral infection, and the loss of cell expansion in lymphoid organs contribute to the blunted immune infiltration in cardiac tissue of mice devoid of βarrs. An immune cell type that predominantly orchestrates the blunted inflammation in viral myocarditis was not clearly identified likely because of the global loss of βarrs.

β-arrestins mediate many cellular processes with βarr1 and βarr2 promoting different outcomes depending on the βarrestin isoform, cellular context, and disease being studied (63). In the context of viral myocarditis, the lack of either βarr1 and/or 2, globally or in cardiomyocytes, reduced inflammation and apoptosis to a similar extent. In models of chronic heart failure, βarr1, the predominant isoform in cardiomyocytes, aggravates heart failure after myocardial infarction, as its deletion reduces apoptosis, infarct size, and remodeling while improving cardiac function (64). In contrast, βarr2 overexpression improves cardiac function and decreases apoptosis, inflammation, and remodeling (65) and in a model of polymicrobial sepsis (66). The differences observed by each βarr isoform is likely due to the complex biology of βarrs (63) and the nature of the cardiac insult triggering alternative molecular mechanisms in cardiomyocytes during sterile injury when compared to viral injury (67, 68). Therefore, signaling pathways involved in viral infection might require both βarrs in cardiomyocytes for their proper activation.

The development of the viral myocarditis phenotype initially requires CVB3 internalization and replication within cardiomyocytes (69). Virally infected cardiomyocytes are key contributors to the immune response by releasing alarmins and cytokines necessary for immune cell recruitment, ultimately leading to cardiac injury (2, 70). The widespread immunosuppression in lymphoid organs and cardiac tissue under CVB3 infection suggests that βarrs participate in the cardiac inflammatory response. Importantly, the blunted immune cell recruitment and anti-apoptotic effect observed in global βarr1 and βarr2 KO mice after CVB3 infection were largely recapitulated in mice devoid of βarrs in cardiomyocytes. Moreover, our data show that the loss of cardiomyocyte βarrs did not affect viral titers (Supplemental Figure 8), indicating that CVB3 retained its infectivity and replication in cardiac tissue (44, 71). Although we could not positively identify all apoptotic cells as cardiomyocytes, the parallel reduction in apoptosis and immune recruitment in both global and cardiomyocyte-specific βarrs KO mice, together with the cardiotropism of CVB3, suggests cardiomyocytes are likely the primary apoptotic cells. Cardiomyocyte apoptosis has been reported during acute CVB3 myocarditis (72, 73) and can be triggered by two main mechanisms: 1) immune-mediated apoptosis and 2) direct virus-induced cardiomyocyte injury (67). The reduced cellular apoptosis in cardiac tissue likely reflects diminished immune-mediated injury, consistent with impaired immune cell recruitment. Whether the diminished cardiac inflammation is a direct consequence of reduced release of DAMPs from apoptotic cardiomyocytes remains to be elucidated. Importantly, the correlation between higher viral titers and reduced inflammatory response and apoptosis observed in global βarr1 KO mice raises the possibility that reduced inflammation in βarr1 KO mice, while limiting immune-mediated cardiac injury, may also weaken antiviral defenses and thereby increase infection risk. Taken together, the diminished inflammatory response observed in CVB3-infected global βarr KO mice was primarily mediated by cardiomyocyte βarrs and supports our hypothesis that cardiomyocyte βarrs are signaling scaffolds in pathways activated by CVB3 viral infection. Since we did not test the effect of Cre overexpression alone in CVB3 infection, we cannot exclude any potential deleterious effects produced by introducing the Cre recombinase as a possible confounder in our model of viral myocarditis. Of note, mice lacking cardiomyocyte βarrs showed a similar mortality rate as WT controls (Supplemental Figure S8A), suggesting that the higher survival in the global βarr1 KO mice compared to the global βarr2 KO mice is likely related to a viral-induced inflammatory response extrinsic to the heart. Due to the difficulty of C57BL/6J mice in developing chronic cardiomyopathy after viral myocarditis, as previously described (35), we were unable to assess the long-term effects of reducing cardiac inflammation by deletion of βarrs. Therefore, caution should be taken regarding extrapolation of our data to chronic viral myocarditis.

The CVB3 virus contains a positive-sense single-stranded RNA, which efficiently triggers a type I IFN response through multiple pathways. While RNA viruses like CVB3 are primarily detected by the MDA5–MAVS signaling pathway, deletion of its main components after acute CVB3 infection shows the opposite disease phenotype observed in βarr KO mice, namely, increased viral titers, systemic inflammation, and mortality (74, 75). Therefore, we focused on accumulating evidence underscoring the role of the cGAS–STING pathway in antiviral defense against RNA viruses (76). During infection, mitochondrial stress and nuclear envelope rupture can lead to the release of dsDNA into the cytoplasm, which is sensed by cGAS. This activation results in the production of cyclic GMP–AMP (cGAMP), a second messenger that is detected by STING for activation. Once activated, STING recruits and activates TBK1, which in turn phosphorylates IRF3, promoting its nuclear translocation and the transcription of interferon-stimulated genes (Figure 6A) (47). Notably, this STING/NF-κB axis has been identified as a critical contributor to the development of CVB3-induced viral myocarditis (60). Our findings further support a key role for the cGAS–STING pathway in the pathogenesis of viral myocarditis and demonstrate that cGAS-STING activation in cardiomyocytes requires βarrs. cGAS-STING activation was detected as early as 24 hours post-infection, consistent with prior in vitro findings in primary cardiomyocyte cultures, where prompt cGAS-STING activation was observed (77). By contrast, later activation of this pathway has been reported at 7 days post-infection in infiltrated cardiac macrophages and selective deletion of STING in macrophages impairs the cardiac inflammatory response to CVB3 infection (78). As previously shown, βarr2 directly binds cGAS and enhances its ability to interact with dsDNA, amplifying STING pathway signaling, at least in macrophages (29). Our study extends this observation to cardiomyocytes, identifying βarrs as important drivers of cGAS–STING pathway activation during viral infection. Since CVB3 induces myocarditis by first infecting cardiomyocytes, it suggests that cGAS-STING activation in cardiomyocytes is a proximal mechanistic event in the pathogenesis of acute CVB3-induced myocarditis. This aligns with the detection of viral capsid protein VP1 at 24 hours post-CVB3 infection in cardiomyocytes, whereas it remains undetectable in macrophages, the predominant cardiac immune population (79). However, the precise molecular mechanisms through which βarrs facilitate cGAS activation remain to be elucidated. One possibility is that βarrs promote the assembly or stability of the cGAS–DNA complex by acting as scaffolds or by inducing conformational changes in cGAS that enhance its affinity for DNA. Alternatively, βarrs may regulate the intracellular trafficking of cGAS between the nucleus and cytosol and thereby increase its cytosolic concentration. It is also currently unknown whether βarr-mediated activation of cGAS–STING is driven by GPCRs or occurs via GPCR-independent mechanisms. Further structural and functional studies will be necessary to define the architecture and dynamics of the βarr–cGAS complex and to determine the signaling context in which βarrs recruit and activate cGAS.

Our findings show decreased inflammatory response in mice devoid of cardiac βarrs during acute viral myocarditis. In this regard, βarrs show a complex behavior in the regulation of inflammation that is dependent on their molecular interactions and activation of a wide variety of signaling pathways (80). While it has been demonstrated that carvedilol, a weak βarr-biased non-selective β-adrenergic receptor (βAR) antagonist, can reduce inflammation by decreasing pro-inflammatory cytokines, fibrosis, and CD4+ T cell infiltration during acute viral myocarditis (81), metoprolol, a selective β₁AR blocker with no βarr activity enhances long-term inflammation and reduces survival (82). To determine whether inhibiting βarrs in acute CVB3 myocarditis can be beneficial will require the development of direct βarrs inhibitors (83).

Study limitations

This study establishes a key role for βarrs during CVB3 infection in many features of the pathophysiology of myocarditis. However, our primary focus was limited to the influence of βarrs on the cardiac inflammatory response and other pathologic features of viral myocarditis were not studied. These include a) sex-dependent differences in survival and viral titers that might point towards hormonal regulation of viral replication/clearance mediated by βarrs; b) the role of βarrs in intrinsic mechanisms governing the development and/or function of selective immune populations recruited to the heart during viral myocarditis, c) the molecular mechanisms by which βarrs promote cGAS activation, including potential cross-talk with the MDA5-MAVS antiviral signaling axis in cardiomyocytes, d) the role βarrs on long-term outcomes including fibrosis, adverse remodeling, and cardiac function in a mouse model of chronic myocarditis susceptible to chronic cardiomyopathy, and e) any potential deleterious effects produced by introducing the Cre recombinase as a possible confounding factor in our model of viral myocarditis. Given the remaining uncertainties regarding immune cell–intrinsic functions, cGAS–STING regulation, long-term cardiac remodeling and survival, and complex behavior of βarrs during infection, our findings do not establish a cardioprotective role for βarrs in acute viral myocarditis. Despite these limitations, the findings lay a strong foundation for future investigations that focus on inhibiting β-arrestins to reduce acute cardiac inflammation that occurs with viral myocarditis.

CONCLUSION

In summary, our data demonstrate that both βarrs are essential for cardiac inflammation and proper immune cell expansion in secondary lymphoid organs. Early activation of the cardiomyocyte cGAS-STING pathway relies on the expression of both βarrs to trigger CVB3-induced inflammation in the heart. Targeting βarrs at an early stage of viral myocarditis appears to diminish the immune infiltration in the heart in acute viral myocarditis.

Supplementary Material

Supplemental Figs. S1-S9 are available at: 10.6084/m9.figshare.30598013