Zebrafish heart regenerates after chemoptogenetic cardiomyocyte depletion

Abstract

Background

Zebrafish can regenerate adult cardiac tissue following injuries from ventricular apex amputation, cryoinjury, and cardiomyocyte genetic ablation. Here, we characterize cardiac regeneration from cardiomyocyte chemoptogenetic ablation caused by localized near-infrared excited photosensitizer-mediated Reactive Oxygen Species (ROS) generation.

Results

Exposure of transgenic adult zebrafish, Tg(myl7:fapdl5-cerulean), to di-iodinated derivative of the cell- permeable Malachite Green ester fluorogen (MG-2I) and whole-body illumination with 660nm light resulted in cytotoxic damage to about 30% of cardiac tissue. After chemoptogenetic cardiomyocyte ablation, heart function was compromised, and macrophage infiltration was detected, but epicardial and endocardial activation response was much muted when compared to ventricular amputation. The spared cardiomyocytes underwent proliferation and restored the heart structure and function in 45-60 days after ablation.

Conclusions

This cardiomyocyte ablation system did not appear to activate the epicardium and endocardium as is noted in other cardiac injury models. This approach represents a useful model to study specifically cardiomyocyte injury, proliferation and regeneration in the absence of whole organ activation. Moreover, this system can be adapted to ablate distinct cell populations in any organ system to study their function in regeneration.

INTRODUCTION

Cardiac diseases are a leading cause of death in US. Many approaches have been pursued to boost the poor regenerative ability of human hearts, yet an efficient treatment is still lacking. Zebrafish offers a model to understand how cardiac regeneration is naturally optimized. It has been shown that zebrafish can efficiently regenerate the heart after ventricular apex amputation ¹, cryoinjury ^2–4,5, puncture ⁶, cauterization ⁷, and cardiomyocyte genetic ablation ^8,9. After cardiac injury, multiple cell types are activated to respond. The epicardium, the outside layer of the heart that is dormant in normal conditions, re-activates genes that are involved in cardiac development ¹⁰. Epicardial cells proliferate and migrate to the injured area ¹¹, undergo epithelial to mesenchymal transition (EMT) and promote the formation of new vasculature ¹². Neoangiogenesis takes place as early as 15 h post cryoinjury and supports the replenishment of the damaged area with healthy tissue ¹³. Blocking fast angiogenic revascularization inhibits cardiomyocyte proliferation and impairs regeneration, suggesting that neoangiogenesis is necessary for cardiac regeneration. In the injured area, fibroblasts are re-activated into myofibroblasts and produce extra cellular matrix (ECM) components including fibrin, fibronectin, collagen and hyaluronic acid ^1,12,14. A key characteristic of zebrafish cardiomyocytes is that they are mononucleated and can proliferate to respond to injury ^1,15–17. Many growth factors have been identified as necessary for cardiomyocyte proliferation either directly or indirectly including FGFs ¹⁰, IGF ¹⁸, Neuregulin ¹⁹, BMP ²⁰, NGF ^21,22, Notch ^23,24, and PDGF-ββ ²⁵. Genetic loss-of-function or chemical suppression of those pathways impairs regeneration, showing that heart regeneration involves multiple players. Recently, the Ras/MAPK pathway has been shown to be involved in cardiomyocyte proliferation ²⁶. In particular, suppressing Dusp6, a feedback attenuator of the Ras/MAPK pathway, enhances cardiomyocyte proliferation, coronary angiogenesis and regenerative ability. Injury also causes inflammation and activation of the immune system and the role of immune cells has been shown to be critical for regeneration ^27–29. Here, we further characterized a light activated photoablation model to damage cardiomyocytes and compared the regenerative response to ventricular apex amputation. In this model, transgenic zebrafish express a fluorogen-activating protein (FAP) in cardiomyocytes. FAP binds with high affinity to heavy atom-substituted fluorogenic dyes such as a di-iodinated derivative malachite green-ester fluorogen (MG-2I) and upon illumination with 660nm light can generate singlet oxygen as a means to induce cellular death. Using this chemoptogenetic approach, we observed consistent compromised heart function and activated macrophage migration into the damaged heart. Unablated cardiomyocytes responded to injury through proliferation and restored lost tissue. This chemoptogenetic ablation system offers a cardiomyocyte-specific, or other lineage-specific model of damage and regeneration.

RESULTS

Cardiomyocyte chemoptogenetic ablation damages zebrafish hearts

Previously, we generated a genetically targetable near-infrared photosensitizer to specifically ablate cardiomyocytes in living larval and adult zebrafish⁹. We showed that the Fluorogen Activating Peptide Targeted and Activated Photosensitizer (FAP-TAPs) generated Singlet Oxygen and induced photoablation of cardiac tissue in Tg(myl7:fapdl5-cerulean) zebrafish, without causing damage in other organs. In particular, injecting the Tg(myl7:fapdl5-cerulean) zebrafish with MG-2I, caused ablation of cardiomyocytes via increasing cell death, as shown by increased TUNEL staining. Here, we aimed to characterize the extent of the damage caused by this approach and to compare this model with the better characterized ventricular apex amputation method. In the ventricular amputation model, a resection of about 20% of the ventricular apex is performed using scissors, as previously described¹. Zebrafish recover from ventricular apex amputation in about 30-60 days. In addition, here, we wanted to determine if zebrafish were able to fully recover from chemoptogenetic ablation damage as this was not investigated previously. To address these questions, we performed AFOG staining at multiple time points after amputation (Fig. 1A–C) and after MG-ester (control) (Fig. 1D, E), or MG-2I (Fig. 1F–H) injection and near-infrared light exposure. After amputation, fibrin and collagen clots were visible in all the hearts (Fig. 1A, B), from 3 to 15 days post amputation (dpa). By 30 dpa, 50% of the hearts completed regeneration (Fig. 1A, C) and by 60 dpa, almost all the hearts showed complete lack of collagen deposits (Fig. 1A, C), confirming previous observations¹. Tg(myl7:fapdl5-cerulean) hearts treated with control MG-ester and irradiated with near-infrared light appeared normal, with no obvious loss of myocardium at each time point collected (Fig. 1D, E). In contrast, MG-2I and near infrared light treatment caused cardiac damage in 50% of the zebrafish by 3 days post injection (dpi) and in all hearts by 7 dpi, suggesting that cardiomyocyte chemoptogenetic ablation was effective (Fig. 1F, G). This was quantified by measuring loss of myocardial tissue in AFOG stained hearts showing clear spaces appearing in the MG-2I treated and near infrared light treated animals (Fig. 2). To further determine that chemoptogenetic ablation injury, expression of myl7 mRNA was used to visualize myocardial damage at early injury states. MG-2I and near infrared light treatment at 3 dpi and 5 dpi Tg(myl7:fapdl5-cerulean) hearts caused a punctate damage which supports the pattern revealed through AFOG staining (Fig. 3). Distinct pockets of space within the compact myocardium were noted at 3dpi and 5dpi (compare Fig. 3A with 3M and 3Q). By 60 dpi almost all the photo-ablated hearts exhibited recovery of cardiac muscle (Fig. 1F, H), suggesting that zebrafish are able to regenerate after ROS-induced cardiomyocyte injury.

Figure 1. Zebrafish regenerate after cardiomyocyte chemoptogenetic ablation.

(A) AFOG staining in WT adult zebrafish hearts before and after ventricular apex amputation. (dpa) = days post amputation. Black dashed lines delineate injured area. (B) Graph representing the percentage of hearts that were injured (orange) at multiple time points. Percentages were calculated upon sectioning and AFOG staining. 100% of the hearts undergoing amputation were injured, showing the efficiency of the surgery. (C) Graph reporting the percentage of hearts that were completely (blue) or partially (red) regenerated at 30-, 45- and 60 dpa. At 30 dpa, 50% of the fish were completely regenerated, while at 60 dpa almost all the fish showed complete regeneration. (D) AFOG staining of Tg(myl7:fapdl5-cerulean) hearts after injection of control MG-ester and IR light treatment at several time points. (dpi)=days post injection. Hearts appeared normal at every time point studied. Damage was determined as fraction of myocyte area of the total ventricular area. (E) Graph showing that upon injection of control MG-ester and IR light treatment. (F) AFOG staining of Tg(myl7:fapdl5-cerulean) hearts after injection of MG-2I and IR light treatment at several time points. Visible loss of cardiomyocytes was detected as early as 3dpi. (G) Graph representing the percentage of hearts that were injured (orange) or not injured (light blue) at several time points after MG-2I injection. At 3 dpi only 50% of the hearts showed injury, but by 7 dpi, all the hearts showed injury. (H) Graph representing the percentage of hearts that show complete or partial regeneration at later time points. Total number of hearts analyzed are indicated. Scale bars = 100 μm.

Figure 2. Chemoptogenetic damage causes loss of myocardium after MG-2I treatment.

Regions of Interest were taken from WT uninjured, Tg(myl7:fapdl1-cerulean) MG-ester-, and Tg(myl7:fapdl1-cerulean) MG-2I-treated hearts after AFOG staining and intact myocardium was measured using Threshold Particle Analysis (Image J). Myocardial area from MG-ester and MG-2I hearts was normalized to WT uninjured hearts to generate the amount of tissue lost as a percentage. MG-2I-treated hearts showed significant loss of myocardium at 3 dpi and 5 dpi (compared to MG-ester controls), and also regenerated this damage by 45–60dpi. A minimum of 3 hearts were used per condition with the following as the total n values: WT uninjured (n = 6), MG-ester (3 dpi (n = 5), −5 dpi (n = 5), −7 dpi (n = 5), −15 dpi (n = 5), −30 dpi (n = 5), −45 dpi (n = 3)), and MG-2I (3 dpi (n = 5), −5 dpi (n = 5), −7 dpi (n = 6), −15 dpi (n = 6), −30 dpi (n = 5), −45 dpi (n = 3)). *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001; Two-way ANOVA using Šidák’s multiple comparisons test. Values without asterisks are not statistically different from MG-ester control samples.

Figure 3. Myocardium integrity was reduced after MG-2I treatment.

Fluorescent in situ hybridization was used to assess the integrity of the myocardium after different types of injury. A representative image of uninjured heart (A-D; n=5), 3 dpa (E-H; n=3), control PBS treated Tg(myl7:fapdl1-cerulean) at 3 dpi (I-L and U, U’; n=4), MG-2I treated Tg(myl7:fapdl1-cerulean) at 3 dpi (M-P and V, V’; n=3) and 5 dpi (Q-T and W, W’; n=4) are shown. Uniform distribution of myl7 were noted in uninjured and PBS control treated hearts at 3dpi (A-D and E-H). Absence of myl7 were detected in 3dpi (M-P, V) and 5dpi (Q-T, W) hearts treated with MG-2I and infrared light that indicated loss of trabeculated myocardium and damage to the compact myocardium as marked by asterisks. Scale bar = 50 μm.

Cardiomyocyte chemoptogenetic ablation causes arrhythmia

Various cardiac damage models used in zebrafish compromise cardiac structure and cardiac function ^26,30, and may model the progressive cardiac injury and chronic degradation of cardiac tissue important in a number of prevalent cardiovascular diseases. Here, we compared the cardiac function after ventricular apex amputation and cardiomyocyte chemoptogenetic ablation, using ECG. One day prior to injury, ECG readings from WT and Tg(myl7:fapdl5-cerulean) were comparable and did not show arrhythmic events (Fig. 4A-Left Panel, & 4B). At 7 dpa, or in MG-2I 7 dpi, arrhythmic events were detected and were higher when compared to fish injected with control MG-ester (Fig. 4A-Center Panel, & 4B), showing that both amputation and chemoptogenetic ablation caused arrhythmia. We next investigated if Tg(myl7:fapdl5-cerulean) zebrafish could regenerate and show normal cardiac function after ROS induced injury. We recorded ECG activity in both ventricular amputation and chemoptogenetic models at 45 days after the injury event (Fig. 4A-Right Panel, & 4B), and found that cardiac function in both models were comparable to control uninjured hearts. These results show that zebrafish hearts regenerated from both injury methods and critically, restoration of cardiac conduction was observed.

Figure 4. Chemoptogenetic cardiomyocyte ablation causes arrhythmia.

(A) Representative electrocardiograms (ECG) of adult zebrafish before and after amputation or ablation. 1 day prior amputation or ablation (left panels), hearts showed normal heart beat rhythm without arrhythmic events. 7 days post amputation or MG-2I injection and IR light treatment, several zebrafish showed arrhythmic events (central panels) Red double heads arrows indicate longer or shorter R-R intervals. Importantly, injection and irradiation of control MG-ester did not cause arrhythmia (central panels). 45 days post injury (right panels), ECG were comparable to uninjured hearts, suggesting that all the hearts have recovered. (B) Graph quantifying the ECG. At least n=7 for each group. *p<0.05; ****p<0.0001 One-way ANOVA. Each ECG trace revealed at least 250 heartbeats.

Macrophages are activated after cardiomyocyte chemoptogenetic ablation

The immune system rapidly responds upon injury to clear dead cells and cellular debris^28,29. To address if cardiac injury activated an immune response, we performed a time course immunostaining for the presence of Mpeg positive cells, (macrophage marker), after ventricular resection (Fig. 5A, B), after MG-ester (control) injection and near-IR light exposure (Fig. 5C, D) and after ROS induced injury (Fig. 5E, F). We found that upon ventricular resection, macrophages were present in the injured area by 3 dpa and remained in the injury until 15 dpa (Fig. 5A, B). By 30 dpa, when most of the hearts are completely regenerated, the macrophage number is comparable to uninjured hearts, suggesting that the immune response goes back to baseline after recovery from the injury. In Tg(myl7:fapdl5-cerulean) zebrafish injected with MG-ester (control) followed by near-IR irradiation, a similar number of Mpeg positive cells were detected in the apex of the heart as noted in control uninjured hearts (Fig. 5C, D). In contrast upon cardiomyocyte chemoptogenetic ablation (Fig. 5E, F), the number of macrophages was significantly increased, but the number of macrophages was much reduced compared to amputation. This likely reflects the diffuse ROS-induced injury and that the total number of dead cells and debris produced was not as traumatic as ventricular amputation.

Figure 5. Macrophages are activated after cardiomyocyte chemoptogenetic ablation.

(A) Mpeg and DAPI staining before and at multiple time points after ventricular apex resection. White dashed lines indicate injury area. (B) Quantification of mpeg⁺ cells inside the injury after amputation. Between 3-and 15 dpa, macrophages were detected inside the injured area. By 30 dpa, the number of macrophages were similar to the uninjured hearts. At least n=3 for each group. ****p<0.0001 One-way ANOVA. (C) Mpeg and DAPI staining in Tg(myl7:fapdl5-cerulean) hearts before and after injection of control MG-ester and near IR light exposure. (D) Quantification of Mpeg positive cells in the ventricle. No differences were noted at different time points after MG-ester injection. (E) Mpeg and DAPI staining before and after cardiac chemoptogenetic ablation. (F) Graph with the number of Mpeg⁺ cells per ventricular area. Chemoptogenetic ablation causes accumulation of macrophages inside the ventricle between 3-and 7 dpi. At 15-and 30 dpi, the number of macrophages is comparable to control. At least n=4 for each group. *p<0.05; ***p<0.001; ****p<0.0001 One-way ANOVA. Scale bar = 100 μm.

Cardiomyocyte proliferate after cardiomyocyte chemoptogenetic ablation

The ability of zebrafish to regenerate the heart is dependent on cardiomyocyte proliferation upon injury ^15,16,31. Here, we quantified the proliferative response following cardiomyocyte ablation. We performed Mef2c (cardiomyocyte nuclear marker) and Pcna (proliferation marker) immunostaining before and after injury, and compared amputation with cardiomyocyte chemoptogenetic ablation (Fig. 6). Cardiomyocytes proliferation after ventricular apex resection was detected by 3 dpa (Fig 6A, B) and had a peak at 7 dpa as previously reported¹. No differences in cardiomyocyte proliferation were found in Tg(myl7:fapdl5-cerulean) zebrafish injected with MG-ester (control) and irradiated with near-IR light (Fig. 6C, D). In contrast, cardiomyocyte chemoptogenetic ablation induced cardiomyocyte proliferation by 3 dpi and lasted until 7 dpi (Fig 6E, F), suggesting that after ablation the spared cardiomyocytes were activated to undergo proliferation. In the chemoptogenetic ablation model the regeneration was homogenously distributed in the ventricular area and we did not observe a different response in different region of the heart. This observation could be explained by the fact that myl7 is not chamber specific.

Figure 6. Cardiomyocyte proliferate after chemoptogenetic ablation.

(A) Mef2c and Pcna staining in amputated ventricles. White dashed lines indicate injury area. Amputation induced cardiomyocyte proliferation. (B) Graph showing the percentage of proliferating cardiomyocytes before and after amputation. Cardiomyocyte proliferation started at 3 dpa and peaked at 7 dpa. **p<0.01; ****p<0.0001 One-way ANOVA. (C) Tg(myl7:fapdl5-cerulean) hearts injected with control MG-ester and treated with near IR-light, stained for Mef2c and Pcna. (D) Quantification of cardiomyocyte proliferation in hearts treated with MG-ester. (E) Mef2c and Pcna staining before and after cardiomyocyte chemoptogenetic ablation. (F) Graph showing the percentage of proliferating cardiomyocytes before and after ablation. MG-2I induced cardiomyocytes proliferation from 3 to 7 dpi. At least n=6 for each group. **p<0.01; ****p<0.0001 One-way ANOVA. Scale bar = 100 μm.

Reduced activation of Myofibroblasts after cardiomyocyte chemoptogenetic ablation

Upon ventricular apex resection, the blood clot is rapidly formed to prevent the fish from exsanguination ¹. This blood clot undergoes remodeling, and it gets replaced by an accumulation of fibrin and collagen that are produced by activated fibroblasts named myofibroblasts^6,12,14,32. We immunostained cryosectioned hearts at multiple time points after ventricular apex resection and visualized myofibroblasts inside the injured area from 3 dpa until 15 dpa (Fig. 7A). At 30 dpa, when hearts were completely regenerated from amputation, myofibroblasts were not detected (Fig. 7A). Tg(myl7:fapdl5-cerulean) zebrafish injected with MG-ester (control) and irradiated with near-IR light showed no activated fibroblast staining at any of the time points studied (Fig. 7B), In contrast to ventricular amputation, myofibroblasts deposit after cardiomyocyte chemoptogenetic ablation was much reduced (Fig. 7C), suggesting that fibroblasts are not reactivated. Additionally, qRT-PCR was performed to measure expression of cytochrome b (cybb), collagen (col12a1a, col1a1b), fibronectin (fn1b), and periostin a (postna) in uninjured and injured hearts (Fig. 7D–E). cybb (FC: 4.11 vs. 25.1), col12a1 (FC: 3.21 vs. 17.2), and col1a1b (FC: 1.78 vs. 5.60) were mildly induced at 5dpi compared to 7 dpa (Fig. 7D). At 20dpi, increases in fibrosis-associated gene expression were not observed in ROS-mediated cardiomyocyte ablation compared to 20 dpa (Fig. 7E). Next, we performed qPCR for genes that have been implicated in cardiac regeneration and as expected, after ventricular amputation, expression of mps1, hmmr, plk1 and cxcl12a were all markedly increased at 7dpa (Fig. 8A). On the contrary, significant increase in these genes were not detected upon cardiomyocyte chemoptogenetic ablation (Fig. 8A). These observations suggest that the injury caused by cardiomyocyte ablation results in a diminished regenerative response when comparted to ventricular amputaiton.

Figure 7. Myofibroblasts are mildly activated upon cardiomyocyte chemoptogenetic ablation.

(A) Activated fibroblasts (αSMA) and MHC (cardiac muscle) immunostaining before and after ventricular amputation at several time points. In uninjured ventricles αSMA staining is not detectable. After amputation, αSMA staining is visible inside the injury but by 30 dpa activated fibroblasts were not detected. (B) Tg(myl7:fapdl5-cerulean) hearts injected with control MG-ester and treated with near IR-light, stained for αSMA and MHC. αSMA was not detected at any time point. (C) In chemoptogenetic ablated hearts, myofibroblasts are weakly activated starting at 5 dpi, until 15 dpi. At least n=3 for each time point. Q-PCR of cytochrome b, (cybb), collagen (col12a1a and col1a1b), fibronectin (fn1b), and periostin a (postna) in amputated and ablated hearts at 7 days and 5 days (D) and at 20 days (E). Scale bar = 100 μm

Figure 8. Several cardiac regeneration markers are not reactivated upon cardiomyocyte chemoptogenetic ablation.

(A) Q-PCR for several cardiac regeneration markers (mps1, hmmr, plk1, cxcl12a) at 3 days after amputation or genetic ablation. Increased expression of regeneration genes was noted in amputation model but not upon ROS-induced cardiomyocyte ablation. (B) Q-PCR for epicardial markers (snai1b, twist1a, aldh1a2, and tbx18) at 3 days after injury. Amputation induced epicardial gene expression at 3 days, but ROS-induced ablation did not. *p<0.05; **p<0.01; ***p<0.001; One-way ANOVA. Values without asterisks are not statistically different from uninjured control samples using One-way ANOVA.

The epicardium is not activated upon cardiomyocyte chemoptogenetic ablation

Several studies highlighted the importance of the epicardium during zebrafish heart regeneration^10,15,16. Here, we tested if cardiomyocyte genetic ablation induced epicardium activation. We performed qPCR for snai1b, twist1a, aldh1a2 and tbx18, well established markers for activated epicardium (Fig. 8B)^6,12,32. 3 days after injury, all four genes were increased upon amputation, but no changes were seen upon cardiomyocyte ablation, suggesting that the cardiomyocyte ablation model is an injury that specifically affect cardiomyocytes. Next, we used fluorescent in situ probes to detect tcf21 and aldh1a2 to determine if epicardial cells are activated after MG-2I treatment (Figure 9). Uninjured hearts showed aldh1a2 restricted to the trabeculated myocardium and tcf21 expressed in the epicardium at low levels (Fig. 9A–C, M–O). After ventricular amputation, aldh1a2 expression greatly increased in the epicardium and in the endocardium within the border zone at 3dpa with increased tcf21 present in epicardium (Fig. 9P–R). This pattern was present in distal regions (Fig. 9D–F) and confirms the entire epicardium becomes activated after ventricular amputation. In contrast, MG-2I and infrared light-treated hearts at 3 dpi (Fig. 9G–I, S–U) and 5 dpi (Fig. 9J–L, V–X) displayed only minimal increase in tcf21 and aldh1a2 expression in the epicardium. Thus, chemoptogenetic ablation causes minimal activation of epicardial cells after injury that is distinct from the ventricular amputation model.

Figure 9. Chemoptogenetic ablation of cardiomyocytes causes a mild activation of canonical epicardial markers.

Fluorescent in situ hybridization for epicardial markers tcf21 and aldh1a2 were used in both the distal ventricle (A-L) and the ventricular apex (M-X) to identify changes in epicardial cell activation after injury. 3 dpa ventricles (D-F, P-R; n=3) displayed increased tcf21 and ald1a2 in the clot and epicardium compared to uninjured ventricles (A-C, M-O; n=6). 3dpi (G-I, S-U; n=3) and 5dpi (J-L, V-X; n=4) hearts injected with MG-2I showed reduced expression of these markers in the epicardium compared to 3dpa hearts. Scale bar = 50 μm.

DISCUSSION

The discovery of the natural ability zebrafish has to regenerate the heart in 2002 has stimulated intense interest into understanding the molecular processes that enables an adult organ to replenish lost tissue¹. However, the regenerative response in the heart is dependent on the injury model⁶. Here, we present a detailed analysis of the regenerative response following cardiomyocyte chemoptogenetic ablation by inducing ROS production only in cardiomyocytes. The first cardiac injury model to be characterized in adult zebrafish was ventricular apex resection¹ This technique involves the surgical amputation of about 20% of the ventricle apex and leads to the formation of a fibrin clot. Using this model, apoptosis is not triggered, and regeneration is achieved after 30-60 days. In 2011, three independent groups established a cryoinjury model with the aim to better mimic myocardial infarction^2–4. Cryoinjury involves the application of a liquid nitrogen-cooled cryoprobe to induce local necrotic and apoptotic cell death to up to 30% of the ventricle. Regeneration is prolonged to 130 days since cellular debris must be cleared before regeneration can occur. Both models implicate that multiple cell types present in the ventricle are injured, including cardiomyocytes, endothelial cells, and epicardial cells. Another injury model available in zebrafish is cardiomyocyte genetic ablation⁸. In this model, inducible cardiomyocyte expression of the cytotoxic diphtheria toxin-A chain selectively ablates up to 60% of the cardiomyocytes, while preserving the remaining cell types. After cardiomyocyte ablation, other cells types within the heart are activated in response to the loss of cardiac muscle and these include the endocardium and epicardium. The response results in full recovery of the heart by 30 days. The cardiac injury model characterized here involves chemoptogenetic ablation of cardiomyocytes through near-infrared ROS generation⁹. This technique specifically ablates 25-30% of the cardiomyocytes. Since cardiomyocyte apoptosis plays an important role in numerous pathologic conditions involving the heart (reviewed by³³), this model offers a new tool to understand cardiomyocyte damage and regeneration. After ROS-induced cardiomyocyte ablation, heart function was compromised, and macrophage infiltration was detected, presumably to remove cardiomyocyte debris. Subsequently, cardiomyocyte proliferation is noted, and structural and functional regeneration is achieved in 45-60 days. In other zebrafish models of cardiac injury, the epicardium is activated. However, we failed to detect robust induction of snai1b, twist 1a, aldh1a2, tbx18, and tcf21, known markers for activated epicardium. Compared to cardiomyocyte genetic ablation using diphteria toxin-A, the near-infrared ROS-induced cardiomyocyte genetic ablation appears to be gentler since it ablates only 25-30% of the cardiomyocytes, but regeneration is achieved over a longer period (45-60 days). One possible explanation could be that since other cell types are not reactivated upon ablation, regeneration is slower. Our observations are in line with a previous study that showed that cardiac injuries in the zebrafish heart such as puncturing or scratching the ventricular surface, produced a much weaker response when compared to amputation⁶. This chemoptogenetic ablation system offers a cardiomyocyte-specific model of damage and regeneration. This injury model did not cause significant fibronectin or collagen accumulation as shown in the cryoinjury and amputation models ^14,34. This lack of induction of fn1 transcripts, which was shown to be important for heart regeneration could be a reason for a prolonged restoration of function. A limitation of this approach is that this model produces milder injury and appear to only be restricted to cardiomyocytes without large loss of other cell types that can generally occur in myocardial infarction. In conclusion, the FAP-TAP system is a practicable approach to target and damage sub-populations of cells, within an organ in a lineage-specific.

Experimental Procedures

Zebrafish maintenance, ventricular amputation, and retro-orbital injections

The zebrafish experiments were performed according to protocol approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Pittsburgh that conforms to the NIH guidelines. Adult (6-18 months old) wild type AB* and Tü, and transgenic, Tg(myl7:fapdl5-cerulean)^pt23 were maintained at 28°C. Ventricle apex amputation was performed as described previously ¹. Approximately 20% of the ventricle apex was resected and zebrafish were returned to the aquatic facility. Survival rate of ventricular apex amputation was greater than 80%. Retro-orbital injections were performed as previously described ³⁵. Zebrafish were injected with 3μl of MG-ester or Phosphate Buffered Saline (PBS) as control or with 3 μl of MG-2I to induce cardiomyocyte ablation as previously described ⁹. 20-30 minutes after injection, fish were exposed to IR light (660nm wavelength), from either a light box that illuminated from below (power > 100 mW/cm² for total 30 minutes) ⁹ or from a custom infrared light box designed by the University of Pittsburgh’s Cell Biology Machine Shop. Water was changed every 10 minutes to avoid overheating.

Immunostaining and cell counting

Immunostaining in zebrafish hearts was performed as previously described ¹². 14 µm cryosections were collected and consecutive sections were used for immunostaining and Acid Fuchsin Orange G (AFOG). AFOG staining was performed as previously described ¹. Primary antibody used for immunostaining were: rabbit polyclonal anti Mef2c (Santa Cruz Biotechnology, Sc313) (1:500); mouse monoclonal anti Pcna (Sigma, P8825) (1:1000); mouse anti MHC (DSHB, F59) (1:50); rabbit anti αSMA (Genetex, GTX100034) (1:200), rabbit anti Mpeg (GeneTex, GTX54246) (1:200). Secondary antibodies were diluted 1:1000 Alexa Fluor 488 goat anti rabbit IgG Peroxidase Conjugate (Invitrogen, A11008); Alexa Fluo 594 goat anti mouse IgG (H+L) (Invitrogen, A11005). Slides were mounted with Vectashield mounting medium with DAPI (Vector Laboratories, H-1200). Images were taken with Zeiss LSM 700 confocal microscope. For each experiment, at least four sections were analyzed for each heart. Cardiomyocyte proliferation index was calculated as percentage of number of Mef2c⁺-Pcna⁺ cells divided by the number of Mef2c⁺ cells. Cardiomyocyte proliferation was measured in the ventricle. To cover the full area of the ventricle, at least 4 images were taken for each heart and values were averaged. Macrophages were counted using ImageJ (NIH) in sections stained with mpeg antibody and DAPI. For ventricular apex amputation, all the mpeg+ cells were counted despite their location in the ventricle. For chemoptogenetic ablation, images were taken in the ventricles. To cover the full area of the ventricle, at least 4 images were taken, and values were averaged. To assess cardiac damage, AFOG staining was performed in sectioned hearts and the areas of these sections were measured by ImageJ (NIH) using two different methods. The first method measures areas of AFOG staining (normalized to the average total area of control hearts) for each condition, and the entire area of the ventricles in the four largest sections of each heart were measured and averaged. For the second method, the myocardial area was quantified by creating three Regions of Interests (ROI) (defined by a 150 x 150 pixel square) per heart and this was measured using the Thresholding and Particle Analysis tools. The percentage of lost myocardial tissue was quantified by dividing the myocardial area by total ROI area and was normalized to the myocardial area of WT uninjured hearts. A minimum of 3 hearts were used per condition and the averages of lost myocardium were calculated and graphed.

Fluorescent in situ hybridization via RNAscope

Fluorescent in situ hybridization was performed on Tg(myl7:fapdl5-cerulean)uninjured, 3 days post-amputation (dpa), 3 days post-injury (dpi), and 5 dpi hearts. 3 dpi and 5 dpi hearts were injected with 3 μl of PBS or 12.5 μM MG-2I as previously described. Hearts were fixed and sectioned at 14 μm as previously described ¹⁰. RNAscope [Advanced Cell Diagnostics (ACD)] RNAscope probe hybridization, amplification and immunostaining were performed following the protocol provided in the RNAscope Multiplex Fluorescent Reagent Kit v2 user manual (ACD). myl7, tcf21, and aldh1a2 RNAscope hybridization probes were designed and are available from ACD. After hybridization, slides were treated with DAPI (1:500, 3 min R.T.). Images were taken on a Zeiss 700 confocal microscope using a 40X water objective. Auto-fluorescent background was used to visualize the myocardium in the 488nm channel; all in situ probes were imaged in the Cy3 and Cy5 channels. Image analysis was performed using ImageJ Fiji (NIH).

RNA extraction, cDNA synthesis and quantitative PCR

Total RNA was extracted from uninjured hearts and hearts at 3- and 20 days post amputation (dpa), post MG-ester or MG-2I injection (dpi) using TRIzol (Invitrogen), and RNeasy Micro kit (Qiagen). The 3 days timepoint was chosen to test epicardial genes expression. The 20 days timepoint was chosen to test collagen and fibronectin deposition. Eight hearts were pooled together for each condition. One µg of RNA was reverse transcribed to cDNA with SuperScript (Invitrogen) using random hexamers. Quantitative PCR (Q-PCR) was performed as described previously ¹². β-Actin and RNA polymerase were used to normalize gene expression. The primer sequences for Q-PCR are listed in Table 1. Two independent biological replicates were done.

Table 1.

Primer sequence for Q-PCR

Gene	Sequence 5’ to 3’	Amplicon Length (bp)
aldh1a2-F	ACCTCCAGTGAAGTTGAACTGC	149
aldh1a2-R	CTCACAGAATCATGCCATTCAT
β -actin -F	CGTGCTGTCTTCCCATCCA	86
β -actin -R	TCACCAACGTAGCTGTCTTTCTG
col12a1a-F	GGTGAAAGAGGAGACACTGCGT	124
col12a1a-R	AGTTGCTGGGGATCTGGTT
col1a1b-F	GCGTAGAAGTTGGCCCAGTCT	179
col1a1b-R	TGGTTTAAGCACATGGACCGATTG
cxcl12a-F	ATTCTACACAGTGCGGATCTCTTC	112
cxcl12a-R	GGCTTGGCGTTGGAAATC
cybb-F	AACACCCTGGTACAAAAGTAGGT	102
cybb-R	ACTCAGTTCCGCCCTCAG
fn1b-F	ATTCCTGCCATTGGTACTGGATC	97
fn1b-R	AGTTGCCCTCTTCGATAAGTTCAT
hmmr-F	GAATAAAGACCTTGAGGACCTACATC	146
hmmr-R	CTTCATTCGCTGAGTCCAAGTAAC
mps1-F	TGTTCAATGAAGATGACACGGA	114
mps1-R	GGTCTGCTTGAGGATTGCC
plk1-F	CTCAGTTCCTAGACACATTAATCCAG	150
plk1-R	ACAGTGAGACAGGAAGTGGGC
RNA pol-F	CCAGATTCAGCCGCTTCAAG	149
RNA pol-R	CAAACTGGGAATGAGGGCTT
snai1b-F	GACTGACCGGCGATATGC	193
snai1b-R	ATCCTCCTCTCCACTGCT
tbx18-F	CGAAGTGCCAAGAATGACAGA	106
tbx18-R	AGGCTCCGGGGATTCGT
twist1a-F	TTGATGCAGCATGATTCTCG	191
twist1a-R	ACATTCACCGTCACAACAG

Electrocardiogram (ECG)

ECG were performed in adult zebrafish one-day prior ventricle amputation or injection to set a baseline. At day 0, surgery or injection were performed, and ECG were recorded after 7- and 45 days using iWorx ECG system, as previously described ²⁶. Time points to assess ECG were chosen according to the AFOG staining results. ECG was performed at 7 days to assess if the cardiac damage present was enough to cause arrythmia. ECG was performed at 45 days to test if hearts were regenerated. Data were analyzed with Labscribe3 software and the R-R interval were measured as previously described ³⁶. Data were expressed as SEM of delta R-R values. Typically, individual ECG traces consisted of 120 seconds of data from each zebrafish and represented about 250 heartbeats for analysis of the R-R interval values.

Statistical analysis

All statistical analyses were performed with GraphPad Prism version 7.0 (GraphPad Software, San Diego CA, USA). Statistical significance was analyzed by one-way ANOVA and shown as mean ± SEM. p-values were considered significant when < 0.05.