In vivo visualization of cardiac extracellular adenosine dynamics and its pharmacological modulation in zebrafish heart failure models
Phurpa Phurpa, Magdeline E. Carrasco Apolinario, Ryohei Umeda, Shinichiro Kume, Hitoshi Teranishi, Nobuyuki Shimizu, Mengting Shan, Kenshiro Shikano, Tatsuki Kurokawa, Takatoshi Hikida, Toshikatsu Hanada, Yulong Li, Reiko Hanada
TL;DR
This study uses a zebrafish model to visualize and modulate extracellular adenosine dynamics in heart failure, identifying a new therapeutic target.
Contribution
The study provides in vivo evidence linking extracellular ATP/Ado signaling to heart failure and introduces VNUT as a novel therapeutic target.
Findings
Zebrafish with heart failure showed increased extracellular cardiac adenosine, indicating elevated ATP release.
Pharmacological inhibition of VNUT reduced extracellular ATP/Ado levels and improved cardiac dysfunction.
Combining VNUT inhibitors with purinergic modulators had additive therapeutic effects in heart failure models.
Abstract
Heart failure (HF) remains a major clinical challenge with poor prognosis despite extensive research and established treatments. Extracellular ATP and adenosine (Ado), as damage-associated molecular patterns (DAMPs), contribute to HF pathogenesis through pro-inflammatory effects. However, in vivo evidence of ATP/Ado dynamics in failing hearts remains elusive. In this study, we investigated the spatiotemporal dynamics of extracellular Ado in a zebrafish HF model using a genetically encoded heart-specific extracellular Ado sensor (GRABAdo). Terfenadine-induced HF zebrafish showed increased extracellular cardiac Ado, indirectly reflecting elevated extracellular ATP release. These findings demonstrate a strong correlation between extracellular ATP/Ado dynamics and HF progression. Furthermore, pharmacological inhibition of vesicular nucleotide transporter (VNUT), a key regulator of ATP…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8- —Cooperative Research Project Program of the Institute for Protein Research at Osaka University
- —Grant-in-Aid for JSPS Fellows
- —https://doi.org/10.13039/100007428Naito Foundation
- —https://doi.org/10.13039/100007449Takeda Science Foundation
- —https://doi.org/10.13039/501100001691Japan Society for the Promotion of Science
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdenosine and Purinergic Signaling · Cardiac Ischemia and Reperfusion · Calcium signaling and nucleotide metabolism
Introduction
Heart failure (HF) is a leading cause of morbidity and mortality worldwide and imposes a substantial clinical and economic burden^1^. Despite advances in pharmacological and device-based therapies, the prognosis of HF remains poor, largely because of its complex and heterogeneous pathogenesis involving neurohormonal dysregulation, metabolic stress, and chronic inflammation^1^. Sterile inflammation driven by damage-associated molecular patterns (DAMPs) has emerged as a key contributor to HF progression^2,3^. Among the DAMPs, extracellular ATP (eATP) and extracellular Ado (eAdo) released from stressed or injured cells play pivotal roles in modulating inflammatory responses and cardiac remodeling^4,5^.
In HF, elevated eATP levels activate the P2X7 receptor, leading to the formation of the NLRP3 inflammasome and the release of proinflammatory cytokines, which exacerbate cardiac dysfunction^6,7^. eAdo, which is derived from the rapid enzymatic degradation of eATP, exerts diverse effects through its receptors and contributes to both protective and pathological outcomes in the cardiovascular system^5^. However, therapeutic interventions targeting purinergic signaling in HF have yielded limited success in clinical settings, highlighting the need for a deeper understanding of extracellular ATP/Ado dynamics in vivo^5^.
Recent studies have identified the vesicular nucleotide transporter (VNUT), which mediates ATP storage in secretory vesicles, as a critical regulator of purinergic signaling^8^. VNUT-dependent ATP exocytosis has been implicated in the amplification of inflammatory responses in various diseases^8–10^. Pharmacological inhibition of VNUT has shown anti-inflammatory and antifibrotic effects in liver disease models, including steatohepatitis^10,11^. Nevertheless, neither the in vivo dynamics of cardiac extracellular ATP/Ado nor the effect of VNUT inhibition in HF have been studied to date, leaving a critical gap in the understanding of purinergic regulation in cardiac pathology.
To address this gap, we employed genetically encoded G protein-coupled receptor activation-based (GRAB) sensors, a class of fluorescent biosensors that enhance green fluorescent protein (GFP) fluorescence upon binding to specific ligands such as neurotransmitters and nucleotides^12^. GRAB sensors enable real-time cell-specific visualization of extracellular molecules, including ATP and Ado, in vivo. In this study, we generated a transgenic zebrafish model expressing cardiomyocyte-specific GRAB_Ado_, which selectively detects cardiac eAdo through changes in GFP fluorescence intensity.
Using this model, we visualized the spatiotemporal dynamics of cardiac extracellular Ado during HF progression induced by terfenadine (TFD), a well-known cardiotoxic agent that potently blocks the human ether-à-go-go-related gene (hERG) potassium channel—thereby inhibiting I_Kr, prolonging QT intervals, and causing ventricular arrhythmias^13^, which may contribute in ventricular remodeling with the consequent development of cardiac dysfunction. In addition, TFD causes cardiomyocyte apoptosis which most likely reduces cardiac function^14^. We confirmed that eAdo levels were markedly elevated, as indicated by increased GFP fluorescence. Based on this observation, we investigated whether the pharmacological inhibition of VNUT could modulate purinergic signaling and improve HF pathology. VNUTi treatment markedly suppressed eAdo accumulation and ameliorated HF by reducing inflammation and apoptosis, restoring calcium homeostasis, and preserving cardiac function. These findings provide the first in vivo evidence that eAdo dynamics are directly linked to the pathophysiology of HF and demonstrate that VNUT inhibition effectively modulates purinergic signaling, thereby establishing VNUT as a promising therapeutic target for HF-associated inflammation and dysfunction.
Results
Establishment of zebrafish HF model
We used TFD to induce HF in zebrafish larvae^15,16^, followed by in vivo assessment of cardiodynamics and cardiac morphology (Fig. 1a). TFD exerts dose-dependent cardiotoxicity (Fig. 1b). High TFD doses induced severe HF (Movie. 3) and reduced survivability (Supplementary Fig. S1). We mainly used 20 μM TFD to induce HF in zebrafish larvae for subsequent experiments as it caused morphologically clear HF (Movie. 2; Movie. 5) and was less lethal (Supplementary Fig. S1). HF was determined based on heart rate, abnormal cardiac morphology and cardiodynamics, pericardial edema, venous congestion (ascites), and decreased myocardial contractility in live cardiac videos (Fig. 1c, d; Movie. 2; Movie. 3; Movie. 5) compared to control (Movie. 1; Movie. 4). TFD significantly reduced heart rate (Fig. 1e) and caused dilatation of the ventricle, marked by an increase in end-diastolic volume (EDV) (Fig. 1f) with thinning of the ventricle walls (Fig. 1d). Functionally, it decreased myocardial contractility (Fig. 1d), quantified in terms of fractional shortening (FS) (Fig. 1g). To evaluate exercise intolerance due to HF, we examined locomotion. TFD-induced HF exhibited a remarkable reduction in the distance traversed by zebrafish larvae (Fig. 1h). These data provide clear evidence of the development of HF in zebrafish larvae due to TFD treatment.Fig. 1TFD-induced HF in zebrafish larvae. (a) Experimental protocol for TFD treatment of zebrafish larvae. (b) Heart rate of zebrafish treated with different doses of TFD. n = 9–13 per group. (c) Representative images of morphology of heart, pericardium, and peritoneum in zebrafish larvae at 6 days post-fertilization (dpf) treated with TFD. a atrium, v ventricle, b bulbus arteriosus, L liver, PE pericardial edema (blue dotted line). Scale bar 100 μm. (d) Representative images of the heart and pericardium during different phases of the cardiac cycle in TFD-treated zebrafish larvae at 3 dpf. White dotted lines: ventricle walls. Inner line: inner ventricular wall; outer line: outer ventricular wall. Orange dotted double-headed arrow: ventricular short axis. Blue dotted line: PE. Scale bar 50 μm (magnified images) and 100 μm (left image). (e) Heart rate, (f) EDV, and (g) FS of TFD-treated zebrafish and controls. n = 14–20 per group. (h) Total distance moved by zebrafish treated with or without TFD. n = 17–22 per group. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. CTL: control; TFD: terfenadine; EDV: end-diastolic volume; FS: fractional shortening; bpm: beats per minute; nL: nanoliter; m: meter.
Ado content in zebrafish with HF
To determine the whole-body Ado content of zebrafish larvae with HF, we performed liquid chromatography–mass spectrometry (LC-MS/MS) on 15 zebrafish larvae treated with 20 or 30 μM TFD and morphologically confirmed to have HF. Simultaneously, 15 zebrafish larvae were treated with DMSO (0.02%) as a control. The total Ado content in HF induced with 20 μM and 30 μM TFD was 404.1 and 448.7 ng/mg protein, respectively. In the control group, the Ado content was 288.1 ng/mg protein. This suggested that an elevated extracellular ATP/Ado is involved in the pathophysiology of HF. This prompted us to study in vivo cardiac eAdo dynamics in HF. We established novel transgenic zebrafish expressing cardiomyocyte-specific GRAB_Ado_ sensors to detect cardiac eAdo to understand the role of purinergic signaling in HF, as previously described^10^. GRAB_Ado_, a protein sensor fused to eGFP, was inserted into the zebrafish genome using the Tol2 transposon system (Fig. 2a). Subsequently, we exposed cardiomyocyte-specific GRAB_Ado_ zebrafish to exogenously administered Ado to corroborate the association between eAdo and GFP fluorescence intensity. Unlike control, GRAB_Ado_ zebrafish treated with Ado strongly increased cardiac GFP fluorescence (Fig. 2b). Next, GRAB_Ado_ zebrafish were treated with TFD to induce HF, which caused a marked increase in cardiac GFP fluorescence (Fig. 2b). Upon quantification of change in GFP fluorescence intensity, administration of Ado and HF induced by TFD caused significant increase in cardiac GFP fluorescence intensity compared to the control (Fig. 2c). Based on these data, we conclude that extracellular ATP/Ado substantially contributes to HF pathophysiology.Fig. 2HF increases extracellular adenosine in zebrafish. (a) Schematic representation of the construction of pDestTol2-CC2-cmlc2-GRAB_Ado_ using in-fusion cloning and the establishment of transgenic zebrafish expressing GRAB_Ado_ in cardiomyocytes. (b) Upper panel: representative images of baseline heart GFP fluorescence intensity in cmlc2:GRAB_Ado_ zebrafish 5 days postfertilization (dpf). Scale bar 100 μm. Lower panel: representative images of heart GFP fluorescence intensity in cmlc2:GRAB_Ado_ zebrafish treated with or without 6 mM Ado and after inducing heart failure with TFD. Scale bar 100 μm. (c) Quantification of changes in heart GFP fluorescence intensity in cmlc2:GRAB_Ado_ zebrafish treated with or without Ado, and before and after inducing HF with TFD. n = 9–25 per group. **P < 0.01; ****P < 0.000; CTL: control; TFD: terfenadine; Ado: adenosine.
Vesicular nucleotide transporter inhibition-mediated alleviation of HF
We explored the cardioprotective effect of a vesicular nucleotide transporter inhibitor (VNUTi) in HF, which blocks VNUT-dependent exocytosis of ATP into the extracellular space. We first tested different doses of VNUTi (clodronate) in wild-type zebrafish larvae and found no adverse effects on cardiac function (Supplementary Fig. S2). Next, we treated zebrafish larvae with TFD and various doses of VNUTi simultaneously (Fig. 3a). VNUTi robustly preserved both whole body morphology (Fig. 3b) and cardiac morphology (Fig. 3b). VNUTi significantly reduced EDV (Fig. 3d) and increased heart rate in TFD-treated zebrafish (Fig. 3e). Among different doses of VNUTi, we used 170 μM in the subsequent experiments as it showed consistent cardioprotective effect. Comparable to control (Fig. 3f; Movie. 6; Movie. 12), VNUTi distinctly preserved cardiac morphology and cardiodynamics, and prevented pericardial edema (Fig. 3f; Movie. 8; Movie. 14), whereas, zebrafish treated with TFD showed all the features of HF (Movie. 7; Movie. 13). In addition, it maintained almost normal peripheral blood flow velocity compared to that of TFD-induced HF (Movie. 9; Movie. 10; Movie. 11). VNUTi improved cardiac functions in TFD-treated zebrafish larvae, namely heart rate, FS, ejection fraction (EF), stroke volume (SV), and cardiac output (CO) (Fig. 3g–k). Next, we used GRAB_Ado_ zebrafish to study the effect of TFD and TFD + VNUTi on cardiac eAdo. We treated a group of GRAB_Ado_ zebrafish with TFD to induce HF and another group was treated with TFD with VNUTi. Unlike in TFD-induced HF, a marked decrease in cardiac GFP fluorescence was noted in zebrafish treated with TFD + VNUTi (Fig. 3l). VNUTi significantly reduced cardiac GFP fluorescence intensity compared to TFD-induced HF (Fig. 3m). Based on these data, we conclude that VNUTi provides cardioprotection by modulating extracellular ATP/Ado in HF through VNUT inhibition.Fig. 3VNUTi alleviates HF in zebrafish larvae. (a) Experimental protocol for simultaneous treatment of zebrafish larvae with TFD and different doses of VNUTi. (b) Representative morphological images of wild-type zebrafish larvae at 4 days post-fertilization (dpf) treated with different doses of VNUTi. Scale bar 200 μm. (c) Representative images of cardiac morphology in cmcl2:eGFP zebrafish 4 dpf treated with different doses of VNUTi. White dotted line: outer ventricle wall. Scale bar 50 μm. (d) EDV and (e) Heart rate of TFD-treated zebrafish treated with or without different doses of VNUTi. n = 10 per group. (f) Representative images of heart and pericardium morphology of cmcl2:eGFP zebrafish (4 dpf) treated with TFD plus VNUTi (170 μM) or without VNUTi. a atrium, v ventricle, PE, pericardial edema. Orange arrow indicating PE. Scale bar 100 μm. (g) Cardiac functions (heart rate, FS, EF, SV, and CO) of TFD-induced HF treated with or without VNUTi. n = 14–15 per group. (e) Upper panel: representative baseline cardiac GFP fluorescence images of cmlc2:GRAB_Ado_ zebrafish at 5 dpf. Lower panel: representative images of cardiac GFP fluorescence of cmlc2:GRAB_Ado_ zebrafish treated with TFD plus VNUTi or without VNUTi. Scale bar 50 μm. (f) Quantification of change in cardiac GFP fluorescence intensity cmlc2:GRAB_Ado_ zebrafish treated with TFD plus VNUTi or without VNUTi. n = 16–25 per group. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns: not significant; CTL: control; TFD: terfenadine; VNUTi: vesicular nucleotide transporter inhibitor; bpm: beats per minute; nL: nanoliter.
Vesicular nucleotide transporter inhibition-mediated alleviation of acute HF
We treated adult zebrafish with 10 μM TFD, and they showed signs of physical distress due to HF (surface scratching, swimming sideways, and erratic movement) after 2 h (Fig. 4a; Movie 15), mimicking acute HF. Histologically, TFD-treated adult zebrafish heart showed thinning of compact myocardium with increased intertrabecular spaces (Supplementary Fig. S3). We confirmed HF by performing high-frequency echocardiography, followed by measurement of cardiac functions, including heart rate, fractional area contraction (FAC), EF, SV, and CO, based on cardiodynamics recorded through echocardiography. TFD-induced HF showed reduced ventricular contractility (Fig. 4b; Movie 16; Movie 17; Movie 18) and altered pulse wave morphology with abnormal rhythm (Fig. 4b; Movie 19; Movie 20; Movie 21) compared to control and VNUTi-treated zebrafish. Adult zebrafish with TFD-induced HF showed a significant decrease in heart rate, FAC, EF, SV, and CO (Fig. 4c–g), which were clearly preserved upon treatment with VNUTi (Fig. 4c–g). These data indicate that VNUTi has a profound cardioprotective effect against acute HF.Fig. 4VNUTi alleviates acute HF in adult zebrafish. (a) Experimental protocol for TFD and VNUTi (200 μM) treatment in adult wild-type zebrafish. (b) Representative echocardiographic images of adult zebrafish (6–12 months post-fertilization) treated with TFD plus VNUTi or without VNUTi, showing cardiac morphology during different phases of the cardiac cycle with pulse wave (PW). White dotted lines: outer ventricle wall. Green dotted arrow: ventricle short axis. Orange dotted arrow: ventricle long axis. (c) Heart rate (d) FAC (e) EF (f) SV, and (g) CO of TFD-treated adult zebrafish (6–12 months post-fertilization) with or without VNUTi. n = 15 per group. EF, ejection fraction; SV, stroke volume; CO, cardiac output; FAC, fractional area of contraction. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant, CTL: control; TFD: terfenadine; VNUTi: vesicular nucleotide transporter inhibitor; bpm: beats per minute; nL: nanoliter.
Effect of VNUTi on cardiac adenosinergic system in HF
With the evident involvement of eAdo in the pathophysiology of HF, as described in the previous section, we studied the effects of Ado, an A_2A_R antagonist (Istradefylline; Istr), and an A_1A_R antagonist (PSB-36) in HF (Fig. 5a). We exposed TFD-treated zebrafish larvae to different doses of Ado, with or without VNUTi (Fig. 5b). Exogenous Ado had no protective effect against HF (Fig. 5b). Interestingly, VNUTi significantly improved cardiac function, even in the presence of Ado (Fig. 5b). Both Istr (Fig. 5c) and PSB-36 (Fig. 5d) provided significant cardioprotection, with the stronger effects observed in the presence of VNUTi. This suggests that VNUTi preserves cardiac function by antagonizing the cardiac adenosinergic system in HF.Fig. 5VNUTi dampens cardiac adenosinergic effect in HF. (a) Experimental protocol for treatment of zebrafish larvae with TFD, VNUTi (170 μM), Ado (2.5–30 μM), Istr (0.01 μM), and PSB36 (0.01 μM). (b) Heart rate of TFD-treated zebrafish larvae treated with or without VNUTi and/or Ado. n = 7–8 per group. (c) Heart rate of TFD-treated zebrafish larvae treated with or without VNUTi and/or Istr. n = 9–10 per group. (d) Heart rate of TFD-treated zebrafish with or without VNUTi and/or PSB36. n = 10 per group. **P < 0.01; **P < 0.01; ****P < 0.0001; ns: not significant; CTL: control; TFD: terfenadine; VNUTi: vesicular nucleotide transporter inhibitor; Ado: adenosine; Istr: istradefylline; PSB36: A1 adenosine receptor antagonist; bpm: beats per minute.
Interplay of anti-purinergic pathways in HF
The activation of P2X7R by excess eATP plays an integral role in the pathophysiology of heart diseases ^7^. Therefore, we explored the effects of different doses of a P2X7R antagonist (brilliant blue G; BBG), with or without VNUTi, in TFD-treated zebrafish larvae (Fig. 6a). BBG showed a strong cardioprotective effect, with an enhanced effect when combined with VNUTi (Fig. 6b, Supplementary Fig. S4a, b). Subsequently, to evaluate cardiac functions in detail, we used 0.1 mg/mL BBG, with or without VNUTi, to treat TFD-treated zebrafish larvae. Both BBG and VNUTi, and their combination, remarkably preserved cardiac morphology and myocardial contractility, and prevented pericardial edema in TFD-treated zebrafish larvae (Fig. 6c). In parallel, BBG, VNUTi, and their combination significantly increased cardiac function (heart rate, FS, EF, SV, and CO) in TFD-treated zebrafish larvae, with an enhanced cardioprotective effect when BBG was combined with VNUTi (Fig. 6d,h). Next, to evaluate the role of passive release of ATP through a hemichannel (PANX1) into the extracellular spaces in the pathogenesis of HF, we treated TFD-treated zebrafish larvae with different doses of probenecid (Prob), a PANX1 inhibitor (Fig. 6a, Supplementary Fig. S4c). Among these doses, 30 μM Prob significantly increased heart rate in TFD-treated zebrafish (Supplementary Fig. S4c). Furthermore, the combination of Prob and VNUTi caused a significantly greater increase in heart rate in TFD-treated zebrafish (Supplementary Fig. S4d). This prompted us to explore the benefits of 30 μM Prob with or without VNUTi in HF. TFD-treated zebrafish exposed to 30 μM Prob alone or in combination with VNUTi showed significant preservation of cardiac morphology and myocardial contractility, and prevention of pericardial edema (Fig. 6i). On further analysis of cardiac functions (heart rate, FS, EF, SV, and CO), 30 μM Prob alone significantly improved these parameters (Fig. 6j–n). Notably, VNUTi significantly enhanced the improvements in heart rate and CO caused by 30 μM Prob (Fig. 6j, n). Overall, these data highlight the cardioprotective effects of suppressing different purinergic pathways and their facilitatory interactions in HF.Fig. 6VNUTi enhances the cardioprotective effects of other purinergic antagonists in zebrafish HF. (a) Experimental protocol for treatment of zebrafish larvae with TFD, VNUTi (170 μM), BBG (0.05–0.15 mg/ml), and Prob (30 μM). (b) Representative images of cardiac and pericardial morphology in 4 days post-fertilization of wild-type zebrafish treated with TFD along with different doses of BBG. All TFD-treated zebrafish without BBG died. (c) Representative images of cardiac and pericardial morphology of TFD-treated cmlc2:eGFP zebrafish treated with or without BBG (0.1 mg/ml) or VNUTi, and their combinations during different phases of the cardiac cycle. White dotted line: outer ventricle wall. White dotted double-headed arrow: ventricular short axis. Scale bar 100 μm (upper panel). Scale bar 100 μm (middle and lower panel). (d-h) Cardiac functions (heart rate, FS, EF, SV, and CO) of TFD-treated cmcl2:eGFP zebrafish treated with or without BBG or VNUTi, and their combinations. n = 9–13 per group. T terfenadine, +V terfenadine + VNUTi, +B terfenadine + BBG, +VB terfenadine + VNUTi + BBG. (i) Representative images of cardiac and pericardial morphology of TFD-treated cmcl2:eGFP zebrafish treated with or without Prob or VNUTi, and their combinations during different phases of the cardiac cycle. White dotted line: outer ventricle wall. White dotted double-headed arrow: ventricular short axis. Scale bar 100 μm (upper panel). Scale bar 100 μm (middle and lower panel). (j-n) Cardiac functions (heart rate, FS, EF, SV, and CO) of TFD-treated cmcl2:eGFP zebrafish treated with or without Prob or VNUTi, and their combinations. n = 9–13 per group. T terfenadine, +V terfenadine + VNUTi, +P terfenadine + Prob, +VP terfenadine + VNUTi + Prob. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; CTL: control; TFD: terfenadine; VNUTi: vesicular nucleotide transporter inhibitor; BBG: brilliant blue G; Prob: probenecid; bpm: beats per minute; nL: nanoliter.
Effect of VNUTi on calcium homeostasis
TFD is known to disrupt intracellular calcium handling of cardiomyocytes^16^. This intrigued us to evaluate if VNUTi had any impact on calcium homeostasis. First, we tested toxicity of different doses of CaCl_2_ on zebrafish larvae. The lower doses of CaCl_2_ did not cause impairment of cardiac morphology, myocardial contractility and heart rate (Supplementary Fig. 5a, b). Simultaneously, we exposed TFD-treated zebrafish larvae with different concentrations of CaCl_2_ with or without VNUTi (Fig. 7a, b). Except for a very high dose, CaCl_2_ per se significantly improved heart rate (Fig. 7b). Subsequently, we performed a comprehensive assessment of cardiac functions by using 9 mM CaCl_2_ alone or in combination with VNUTi in TFD-treated zebrafish (Fig. 7a). Both CaCl_2_ and VNUTi, and their combination preserved cardiac morphology and ventricle contractility (Fig. 7c). On performing further assessment of cardiac functions (HR, FS, EF, SV and CO), we noted a significant increase of these cardiac functions with CaCl_2_ treatment per se (Fig. 7d - h). Interestingly, VNUTi significantly increased the cardioprotective effect of CaCl_2_ as it can be noted with HR (Fig.7d) and CO (Fig. 7h). In order to further corroborate the effect of VNUTi on cardiomyocyte calcium handling, we induced HF in zebrafish larvae with verapamil, a well-known calcium channel blocker of cardiomyocyte (Fig. 7i). Verapamil, significantly decreased heart rate (Fig. 7k) and triggered development of abnormal cardiac morphology and pericardial edema (Movie. 22; Movie. 23; Fig. 7j). Verapamil-treated zebrafish larvae treated with VNUTi clearly preserved cardiac morphology and prevented pericardial edema (Fig. 7l), and significantly increased heart rate (Fig. 7h). Collectively, these findings underscore the role of VNUTi in maintaining calcium homeostasis in cardiomyocytes in HF.Fig. 7VNUTi ameliorates calcium dyshomeostasis in zebrafish HF. (a) Experimental protocol for treatment of zebrafish larvae with TFD, VNUTi (170 μM), and CaCl_2_. (b) Heart rate of TFD-treated zebrafish larvae treated with or without VNUTi and/or different doses of CaCl₂. n = 10 per group. (c) Representative images of cardiac morphology of TFD-treated zebrafish larvae with or without CaCl₂ (9 mM) and/or VNUTi during different phases of the cardiac cycle. White dotted line: outer ventricle wall. White double-headed arrow: ventricle short axis. Scale bar 50 μm. (d-h) Cardiac functions (heart rate, FS, EF, SV and CO) of TFD-treated zebrafish treated with TFD with or without CaCl₂ (9 mM) and/or VNUTi. n = 10–13 per group. T: terfenadine; +V: terfenadine + VNUTi; +C: terfenadine + CaCl_2_; +VC: terfenadine + VNUTi + CaCl_2_. (i) Experimental protocol for treatment of zebrafish larvae with verapamil, with or without VNUTi. (j) Representative images of cardiac and pericardial morphology in zebrafish exposed to verapamil, with or without VNUTi. White dotted line: ventricle wall; blue dotted line: pericardial edema (PE); orange double-headed arrow: ventricle short axis. (k) Heart rate of zebrafish larvae treated with verapamil. n = 9–10 per group. (l) Heart rate of verapamil-treated zebrafish treated with or without VNUTi. n = 10 per group. *P < 0.05; **P < 0.01; ****P < 0.0001; ns: not significant; CTL: control; TFD: terfenadine; VNUTi: vesicular nucleotide transporter inhibitor; PE: pericardial edema; Ver: verapamil; bpm: beats per minute; nL: nanoliter.
VNUTi ameliorates apoptosis and inflammation
To understand the underlying pathophysiology of TFD-induced HF in zebrafish and the cardioprotective mechanism of VNUTi, we performed cardiomyocyte apoptosis staining using acridine orange (AO) (Fig. 8a). TFD clearly caused apoptosis of cardiomyocytes, marked by innumerable GFP fluorescence spots (representing condensed chromosomes), which were not observed in the control and VNUTi-exposed TFD-treated zebrafish larvae (Fig. 8b). The GFP fluorescence intensity due to apoptosis was significantly higher in TFD-induced HF than in those treated with VNUTi and the control (Fig. 8c). Next, we measured the mRNA levels of the pro-apoptotic (Baxb) and anti-apoptotic (Bcl2) markers. Baxb expression levels significantly increased in TFD-induced HF, accompanied by a downregulation of Bcl2, whereas in TFD-induced HF treated with VNUTi, Baxb was significantly downregulated and Bcl2 was upregulated (Fig. 8d). On evaluating the mRNA levels of key pro-inflammatory markers (nlrp3, caspase 1, and il-1β), TFD-induced HF clearly showed upregulation of these markers; in contrast, VNUTi significantly downregulated caspase 1 and il-1β with a strong tendency to downregulate nlrp3 (Fig. 8e). These data suggest that VNUTi plays an intricate role in dampening apoptosis and inflammation in HF.Fig. 8VNUTi ameliorates apoptosis and inflammation in zebrafish HF. (a) Acridine orange (AO) staining of zebrafish larvae with HF, treated with or without VNUTi (170 μM). (b) Representative cardiac images of zebrafish (5 days post-fertilization) after AO staining. Red arrows: apoptotic spots. Scale bar 50 μm. (c) Quantification of AO staining cardiac GFP fluorescence intensity of TFD-treated zebrafish larvae with or without VNUTi. n = 14–15 per group. (d) Expression levels of the pro-apoptotic gene Baxb and anti-apoptotic gene Bcl2 in TFD-treated zebrafish larvae with or without VNUTi. n = 7–9 per group. Each dot represents 15 zebrafish larvae. (e) Expression levels of inflammatory genes (il-1β, caspase 1, and nlrp3) in TFD-treated zebrafish larvae with or without VNUTi. n = 7–10 per group. Each dot represents 15 zebrafish larvae. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns: not significant; CTL: control; TFD: terfenadine; VNUTi: vesicular nucleotide transporter inhibitor.
Discussion
This study highlights the central involvement of extracellular ATP/Ado signaling in the pathogenesis of non-ischemic HF, using a transgenic zebrafish model with cardiac-specific visualization of purinergic activity. By inducing HF with a cardiotoxic drug, we successfully demonstrated, for the first time in vivo, the accumulation of eAdo in the stressed hearts. This purinergic shift is associated with impaired cardiac function, calcium dysregulation, inflammation, and apoptosis. Targeted inhibition of VNUT, a key mediator of ATP release, significantly ameliorates these pathological changes and preserves cardiac performance. The model enabled the real-time tracking of purinergic dynamics in living organisms, providing mechanistic insights into the contribution of VNUT to disease progression. Furthermore, combinatorial approaches with other purinergic modulators have suggested potential additive effects, underscoring the relevance of purinergic signaling as a promising therapeutic axis in HF.
Previous studies have implicated eATP and eAdo signaling in cardiovascular pathologies, particularly through the activation of the P2X7 receptor and the propagation of sterile inflammation through the NLRP3 inflammasome^7,17^. Despite the growing recognition of purinergic signaling in HF, the role of VNUT-mediated ATP release remains largely unexplored in cardiac contexts, with most previous evidence confined to hepatic and neuronal inflammation^8,10,18^. Our study addresses this gap by providing the first in vivo evidence that VNUT inhibition suppresses purinergic activation in the heart and mitigates pathological outcomes in non-ischemic HF.
Consistent with previous studies, we confirmed that blockade of P2X7R by BBG preserved cardiac function in zebrafish HF, reflecting the therapeutic relevance of this receptor in mammalian models^19^. Similarly, the inhibition of PANX1 hemichannels with probenecid yielded beneficial effects, consistent with reports implicating PANX1-mediated ATP release in cardiac dysfunction^20,21^. Interestingly, combined treatment with VNUTi and either BBG or Prob showed only modest additive effects. This finding supports the notion that eATP and eAdo may exert both protective and deleterious effects, depending on the receptor subtype and tissue-specific context^4,5^.
Beyond receptor-level modulation, our study highlights the complexity of Ado signaling in HF. The lack of therapeutic effect following systemic Ado administration reflects previously reported challenges, including rapid degradation and variable receptor affinities^22^. Despite these limitations, VNUTi effectively improved cardiac function, even at elevated Ado levels, suggesting beneficial reprogramming of purinergic dynamics. Furthermore, inhibition of A_1A_R and A_2A_R by PSB-36 and Istr, respectively, corroborated the involvement of these receptor subtypes in HF pathogenesis. Notably, our findings align with an earlier study showing that A_2A_R antagonism preserves ventricular function in arrhythmogenic cardiomyopathy^23^, thereby reinforcing the relevance of adenosine receptor targeting.
Another key observation of our study was the link between purinergic signaling and calcium homeostasis. In both terfenadine- and verapamil-induced HF models, VNUT inhibition significantly improved cardiac performance, underscoring its capacity to stabilize intracellular calcium dynamics. This finding supports previous evidence that calcium dyshomeostasis is a key contributor to arrhythmogenesis and myocardial dysfunction^24,25^. The enhancement of cardiac function by calcium chloride supplementation, and its further improvement when combined with VNUTi, suggests that purinergic signaling and calcium handling intersect at multiple regulatory nodes.
At the molecular level, VNUTi suppressed canonical inflammatory and apoptotic pathways activated downstream of ATP-P2X7R-NLRP3 signaling. Reductions in nlrp3, caspase-1, and il-1b mRNA levels, along with the shift in apoptotic markers—downregulation of Baxb and upregulation of Bcl2—highlight the anti-inflammatory and cytoprotective effects of VNUT inhibition. These results echo earlier findings in hepatic tissues, where VNUT blockade reduced cytokine production and fibrosis^10,11^. In vivo imaging of GRAB_Ado_ also confirmed reduced eAdo levels, providing direct evidence for the role of VNUTi in modulating extracellular purinergic tone.
Mechanistically, VNUT inhibition disrupts ATP vesicle exocytosis, limiting extracellular purine availability and preventing overactivation of the P2X7R and PANX1 signaling cascades. This mechanism likely contributed to the normalization of calcium flux, reduction of cell death, and attenuation of inflammatory signaling. Importantly, the zebrafish model enabled real-time visualization of these dynamic changes, underscoring its value as a platform for translational cardiovascular research.
Taken together, our findings suggest that VNUTi is a multitarget modulator of purinergic signaling that exerts cardioprotective effects through anti-inflammatory, anti-apoptotic, and calcium-stabilizing mechanisms. These results suggest a novel conceptual shift in HF therapy, from hemodynamic support to modulation of purinergic pathways. The zebrafish model, due to its optical transparency and high-throughput compatibility, provides a valuable platform for dissecting spatiotemporal purinergic dynamics in vivo. Its utility in screening pharmacological modulators and visualizing mechanistic responses underscores its suitability for translational cardiovascular research.
Nevertheless, our study had some limitations that warrant considerations. Although the zebrafish model provides valuable insights due to its genetic and physiological similarities to mammals, certain species-specific cardiac features, such as a single-chambered heart and regenerative capacity, should be considered when assessing its applicability to human HF. In particular, differences in the distribution and expression profiles of purinergic receptors between zebrafish and humans can influence pharmacodynamic responses, complicating the direct extrapolation of drug efficacy. The pharmacokinetic characteristics of VNUTi in zebrafish remain undefined, raising uncertainty regarding its systemic exposure and dosing relevance for human applications. Furthermore, the experimental HF model used in this study reflects an acute pathological state that may not fully recapitulate the complex and progressive nature of human HF, particularly long-term remodeling and functional deterioration. Although VNUTi consistently improved cardiac performance in pharmacologically induced HF models, we did not directly assess intracellular calcium dynamics, which is critical for understanding the mechanistic underpinnings of its therapeutic effects. Future studies employing calcium imaging and electrophysiological approaches are necessary to clarify the effects of VNUTi on calcium regulation in cardiomyocytes.
In conclusion, our study provides the first in vivo evidence of dynamic changes in extracellular Ado levels during HF progression, establishing a direct link between purinergic signaling and HF pathophysiology. Through pharmacological modulation, VNUT inhibition ameliorates HF by reducing eAdo levels, suppressing inflammation and apoptosis, and restoring calcium homeostasis. Although these findings identify purinergic signaling as a key therapeutic target for HF and VNUT inhibition may represent a novel strategy to treat HF, further preclinical validation is warranted. Moreover, further preclinical studies are needed to clarify the precise molecular mechanisms involved and assess the translational potential of targeting this pathway in the clinical settings.
Methods
Zebrafish husbandry
All animal experiments were performed according to institutional and national guidelines and regulations. The ARRIVE guidelines were implemented in this study. The study protocol was approved by the Institutional Review Board of Oita University (approval numbers: 230301, 2–18). All zebrafish (AB strain; ZFIN, Eugene, OR, USA, and Tg, cmlc2:eGFP) were maintained on a 10/14-h dark/light cycle at 28 ± 1 °C and fed twice daily. Embryos were collected and incubated at 28.5°C. After completion of each experiment, zebrafish larvae were euthanized by immersion, along with the net, into chilled egg water (5 parts ice to 1 part water) at 0–4°C for 20 minutes^26^. Similarly, adult zebrafish were euthanized by immersion in the breeding tank containing chilled system water for 10 minutes^26^.
Drug-induced HF model
Zebrafish larvae at 2-5 dpf (wild type or Tg[cmlc2:eGFP]) were treated with different doses (5, 10, 20, 30, 50, and 100 μM) of terfenadine (Sigma-Aldrich; T9652) to induce HF, as described in previous studies^15,16^. A fluorescence microscope (Keyence, Osaka, Japan) was used to assess the HF. Terfenadine-treated larvae were placed in a glass-bottom dish in the right lateral position to visualize cardiodynamics under a Keyence fluorescence microscope (Osaka, Japan). Video recordings of bright-field or dark-field GFP fluorescence were acquired for 15 s. The heartbeat was visually counted for 15 s and multiplied by four to determine the heart rate in 1 min. These videos were analyzed in ImageJ (National Institute of Health, Bethesda, MD, USA) to capture ventricle size in the diastolic and systolic phases, followed by measurement of the long and short axes of the ventricle^28^. Based on the short- and long-axis diameters of the ventricle in the diastolic and systolic phases, EDV, ESV, FS, EF, SV, and CO were calculated, as described previously^27^. Data acquisition and analysis were performed automatically in a blinded manner using a Keyence fluorescence microscope and ImageJ software.
Echocardiography
Echocardiographic imaging of adult zebrafish (6–12 months old) was performed using the Prospect A70 ultrasound system (S-Sharp, Taiwan) with an ultrasound probe (P/N: PB406e) at a frequency of 40 MHz, as previously described^28^. Zebrafish were lightly anesthetized with 2-phenoxyethanol (0.06%) for 30 s and then placed on the ventral side of a slit sponge soaked in maintenance water. The ultrasound transducer was positioned parallel to the anterior–posterior axis of the zebrafish in the longitudinal view and perpendicular in the cross-sectional view. After clear visualization of the heart, recording of echocardiography and images were by using B-mode and pulsed-wave Doppler. The images were analyzed using the measurement tools available on the Prospect Ultrasound System. The systolic ventricle area (Vas), diastolic ventricle area (VAd), systolic ventricle long axis (Ls), systolic ventricle short axis (Ds), diastolic ventricle long axis (Ld), and diastolic ventricle short axis (Dd) were also measured. Heart rate (HR) was determined by tracing three consecutive inflow peaks (A wave). Cardiac function was assessed as previously described^29,30^: EDV = 8 × (Vad)^2^/3π × Ld, ESV = 8 × (Vas)^2^/3π × Ls, and FAC (%) = ((Vad – VAs)/Vad) × 100. Ejection fraction EF (%) = (EDV – ESV)/EDV x 100. SV = EDV-ESV, CO = HR × SV. Data acquisition and image analyses were performed using automated quantification software (the Prospect A70 ultrasound system) in a blinded manner.
Locomotion analysis
For locomotion assessment, each wild-type zebrafish larva (12 dpf) was placed in each well of 24 well plate and treated with 20 μM terfenadine for 8 h, followed by 1 h of locomotion recording using a Zebrafish Chamber Behavioral Analysis System (Zantik, Cambridge, UK).
Quantitation of Ado using LC-MS/MS
Wild-type zebrafish larvae (4 dpf), both control and those treated with different doses of terfenadine (20, and 30 μM) were washed in phosphate-buffered saline, snap-frozen in liquid nitrogen, and stored at –80°C until LC-MS/MS analysis. Samples for LC-MS/MS were prepared as previously described^10^. The pellets from the supernatants were dissolved in 20 μL of deionized water. Three microliters of each sample were injected into the LC-MS/MS system (LCMS-8040, Shimadzu). Internal standards (50, 100, and 200 mg/mL 2-isopropylmalate) and external standards (50, 100, 500, and 100 ng/mL Ado solution) were injected together. Protein concentrations were determined using the DC Protein Assay Kit II (Bio-Rad Laboratories, Inc., CA, USA).
Establishment of transgenic cardiomyocyte-specific GRABAdo Zebrafish
The plasmid of interest was constructed by PCR using KOD Plus Neo (Toyobo) and In-Fusion Snap Assembly Master Mix (Takara) to generate a vector with Tol2 transposon sites. Recombination was performed by restriction digestion, ligation, and infusion using two plasmids: pDestTol2-insulin-GRABAdo-PolyA-cmlc2-mCherry and Tol2-FABP2-eGFP-PolyA-Tol2. Alkaline phosphatase was used to generate free 5′-hydroxyl groups after digestion. PCR of pDestTol2-insulin-GRABAdo-PolyA-cmlc2-mCherry and pDest-fabp10-mCherry-tol2 was performed to obtain cmlc2 and fabp10 promoter elements, respectively. The DNA sequence of the plasmid was confirmed using BigDye Terminator v3.1 (Applied Biosystems, Foster City, CA, USA) and a DNA sequencer (SeqStudio; Applied Biosystems). The final plasmid (50 ng) was injected with Tol2-mRNA (25 ng) into wild-type zebrafish embryos at the one-cell stage. The insertion of the plasmid was confirmed by observing liver mCherry in 5 dpf larvae. The injected embryos were raised and mated with adult wild-type zebrafish to obtain the next generation. Furthermore, to confirm GFP expression in the GRAB_Ado_ zebrafish hearts, they were embedded at 5 or 6 dpf in a low-melting agarose gel and stimulated with Ado (6 mM) solution as described previously^10^. Live images were captured as described below.
In-vivo imaging of GRABAdo zebrafish
Larvae at 5 or 6 dpf were embedded in 1.5% low-melting-point agarose in a glass-bottom dish in the right lateral position for live imaging with a confocal microscope (FluoView FV3000; Olympus, Tokyo, Japan) using an NA 0.5/20× water immersion objective lens. Image acquisition and analysis of GFP fluorescence using Olympus cellSens were performed in a blinded manner as described previously^10^.
Pharmacological interventions
Zebrafish larvae at 2, 3, 4, or 5 dpf (wild type or Tg, cmlc2:eGFP ) were treated simultaneously with terfenadine, VNUTi, probenecid, BBG, Ado, PSB36, istradefylline, calcium chloride, and verapamil, with or without VNUTi (Table 1). Cardiac morphology and function were assessed as described in the previous section.Table 1. Details of drugs used in the study.DrugSourceIdentifier****EffectDisodium Clodronate TetrahydrateTokyo Chemical IndustryLot. VCXME-WDVNUT inhibitionProbenecidSigma AldrichP86761-25GPANX1 inhibitionBrilliant blue GSelleck ChemicalsLot: s321701P2X7R blockerAdenosineNacalai TesqueLot: M9T9486Adenosinergic functionsPSB36Sigma AldrichCAS:524944-720-7A_1_AR antagonistIstradefyllineSigma AldrichLot: 0000333704A_2_AR antagonistCalcium chlorideWako039–00475Ionotropic effectVerapamil hydrochlorideSigma AldrichSML0212-50MGL-type calcium channel blockerThe table lists the drug name, commercial source, product identifier (lot number or CAS number), and the primary pharmacological effect or molecular target investigated in the study. VNUTi: Vesicular Nucleotide Transporter; PANX1: Pannexin 1; P2X7R: P2X Purinoceptor 7; A1AR: Adenosine A1 Receptor; A2AR: Adenosine A2A Receptor.
Histological analysis
Adult zebrafish were dissected, and their hearts were extracted. They were then fixed overnight in 4% paraformaldehyde at 4 °C, embedded in paraffin, and processed according to standard protocols. Thereafter, 4-μm sections were stained with HE. All images were obtained using the color bright-field mode of a Keyence fluorescence microscope (Osaka, Japan).
Apoptosis staining
Wild-type zebrafish larvae (5, 6, or 7 dpf) treated with 0.0015% 1-phenyl 2-thiourea (Sigma Aldrich) were exposed to terfenadine with or without VNUTi and incubated at 29 °C, followed by the addition of 1 μL of 1 mg/mL acridine orange (Sigma Aldrich, CAS: 10127-02-3) to the zebrafish medium. The larvae were then washed three times with the zebrafish medium. Live imaging, image acquisition, and analysis (Olympus cellSens) of cardiomyocyte apoptosis spots emitting GFP fluorescence in zebrafish hearts were performed using a confocal microscope (FluoView FV3000; Olympus, Tokyo, Japan) with an NA 0.5/20× or 0.5/40× water-immersion objective lens, as described in the previous section.
mRNA analysis
Total RNA was isolated from pools of 15 zebrafish larvae (4 dpf) using a PureLink^TM^ RNA Mini Kit (Thermo Fisher Scientific) and a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). qPCR was performed for genes of interest by using their primers (Table 2) with the KAPA SYBR® Fast qPCR Kit (Kapa Biosystems, Woburn, MA, USA) on a LightCycler 96 System (Roche Diagnostics, Basel, Switzerland) using the following protocol: denaturation at 95 °C for 3 min, followed by 45 cycles of denaturation at 95 °C for 10 s, annealing at 63 °C for 30 s, and extension at 72 °C for 10 s. For data analysis, the mRNA levels of the target genes were normalized to glyceraldehyde-3-phosphate dehydrogenase using the comparative threshold cycle method.Table 2. Primers used in RT-qPCR of zebrafish gene.**Gene**Forward primer (5’−3’)**Reverse primer (5’−3’)**gapdhCTGGTGACCCGTGCTGCTTTGTTTGCCGCCTTCTGCCTTAnlrp3TCAGCTCTGAGTTCAAACCCCCACCCATAGGATCAGTTTTGAGTGcaspase 1TTCTCTGATGTCGTGCACCCATGTGATCCTCATGTGCGCAil1bACTGTTCAGATCCGCTTGCATCAGGGCGATGATGACGTTCbaxbGGTGTTCACAGATGGCCAGACAATTCTGGGTAGGCGAGCbcl2GGGCCACTGGAAAACTGGATCCAAGCCGAGCACTTTTGTTList of target genes with their corresponding forward and reverse primer sequences. The housekeeping gene gapdh was used for normalization of expression data.
Statistical analysis
All results are reported as mean ± standard error of the mean. Statistical analyses were performed using GraphPad Prism 9.5.0 (730) software (GraphPad Software, San Diego, CA, USA). Unpaired two-tailed sample t-tests were used to assess significance when comparing two groups. Statistical significance among three or more groups was determined by using one-way analysis of variance (ANOVA). Differences were considered statistically significant at *p-*value < 0.05.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary material 1.Supplementary material 2.Supplementary material 3.