Assessing the Metabolic Activity of Whole Regenerating Zebrafish Ventricles Ex Vivo Using an Extracellular Flux Assay

Summary

In this article, we describe a detailed protocol for quantifying the oxygen consumption rate and extracellular acidification rate of ex vivo adult zebrafish ventricles to characterize their oxidative and glycolytic metabolic capacity.

Abstract

The ability of zebrafish to regenerate their hearts throughout adulthood is partially attributed to metabolic adaptations. Although it has been hypothesized that the fish heart relies significantly on glycolysis, recent studies have uncovered a more complex metabolic profile in which oxidative metabolism arises as an essential component of cardiomyocyte redifferentiation and successful late-stage heart regeneration. In 2025, we adapted a high-throughput method to assess the metabolic profile of whole zebrafish ventricles ex vivo, utilizing the Seahorse assay (extracellular flux assay). This method allows for rapid, real-time assessment of basal and maximal oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) using the XF Mito Stress test on the Seahorse XFe24 analyzer. In this article, we describe a detailed protocol for performing extracellular flux analysis on whole zebrafish ventricles. The ability to quantify the OCR and ECAR in live whole hearts ex vivo will provide the opportunity to elucidate cardiac metabolism at critical timepoints during development, disease progression, and regeneration.

Introduction

In contrast to humans, zebrafish (Danio rerio) maintain the capacity to regenerate lost cardiac tissue throughout their adult life¹. Upon cardiac injury, the cardiomyocytes in the wound border zone de-differentiate and re-enter the cell cycle^2,3 in order to restore cardiomyocyte numbers⁴ and repopulate the wound. Recently, the importance of cardiomyocyte redifferentiation in the regenerative process has been recognised⁵. Both those processes are tightly regulated by metabolic changes in the zebrafish heart, with glycolysis driving proliferation^6,7 and oxidative phosphorylation (OXPHOS) promoting redifferentiation⁸. Although pharmacological and genetic methods to manipulate metabolism are widespread in the field of regeneration^6,7, validation of the metabolic effects of those perturbations remains challenging due to the small size of the zebrafish heart, which requires a complex, customized setup for canulation^9,10,11 or pooling of multiple hearts for mass spectrometry¹² or NMR spectroscopy^8,13.

In order to circumvent these limitations, we recently described the use of the Seahorse XFe24 Analyzer for high-throughput metabolic profiling of whole zebrafish ventricles ex vivo⁸. The extracellular flux assay is designed to measure the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) of cells, pancreatic islets, and spheroids. This is achieved through solid-state sensors that detect changes in oxygen and proton concentration (pH) in the media that result from the respiratory electron transport chain utilization of oxygen and glycolysis-dependent lactate production, respectively (Figure 1A)¹⁴.

The extracellular flux analysis can be performed on a 96-well plate setup (XFe96 and XF Pro), 8-well miniplates, or various iterations of 24-well plates (XFe24). Although the 96-well plates provide higher throughput, the working distance between the sensor and the bottom of the well is not suitable to accommodate a whole zebrafish ventricle. Additionally, on the 96-well plate format, there is no provision for securing the ventricle at the bottom of the well to prevent it from floating in the media and interfering with the sensors. On the other hand, the XFe24 islet capture microplate was originally designed for the metabolic characterization of primary pancreatic islets and, as a result, allows for the deposition of the tissue at the bottom of the well, which is then secured by a screen (grid). Due to the small size of the zebrafish heart, we were able to adapt the islet capture microplate to be used on zebrafish ventricles, which remain secured under the grid for the duration of the assay.

To evaluate mitochondrial respiratory capacity, we utilized the Mito Stress test in which four inhibitors of the electron transport chain and the adenosine triphosphate (ATP) synthase were used (Figure 1A). Firstly, baseline measurements of the OCR are acquired (Figure 1B, magenta). Then, oligomycin, a complex V inhibitor, is injected. Oligomycin inhibits the ATP synthase activity and thus leads to a reduction in OCR, which is attributable to ATP production (Figure 1B, green). The remaining oxygen consumption is a result of uncoupled respiration and non-mitochondrial oxygen consumption (Figure 1B, orange and grey, respectively)¹⁵. Following oligomycin, FCCP is injected, which dissipates the proton gradient across the mitochondrial membrane (Figure 1A). In an attempt to restore the proton gradient, the mitochondria upregulate their oxygen consumption to their maximal capacity (Figure 1B, blue). Finally, a mix of rotenone and antimycin A is used that inhibits complexes I and III, respectively (Figure 1A). This drug combination leads to the complete shutdown of the mitochondrial respiratory electron transport chain activity due to the lack of electron flow from complexes I and III to complex IV¹⁵. This allows for the measurement of non-mitochondrial oxygen consumption (Figure 1B, grey).

Overall, using the islet capture microplate with the Mito Stress test, we were able to rapidly and reproducibly evaluate the oxidative and glycolytic metabolic capacity of whole zebrafish ventricles ex vivo. We describe the detailed protocol for this procedure below. Finally, we show the survival of uninjured, wild-type isolated ventricles ex vivo in the extracellular flux analyzer, and we validate the Mito Stress test on 14 days post-cryoinjury (dpci) transgenic zebrafish ventricles⁸.

Protocol

All procedures described below involving animals were carried out in compliance with the revised Animals (Scientific Procedures) Act 1986 in the United Kingdom and Directive 2010/63/EU in Europe and were approved by Oxford University's central Committee on Animal Care and Ethical Review.

NOTE: This protocol was performed using a Seahorse XFe24 analyzer and the Seahorse DMEM medium without phenol red and bicarbonate.

1. Zebrafish heart cryoinjury (optional)

1. Obtain the appropriate ethical approval from the local institution and the national authorities before starting the experiment.
2. Perform cryoinjury on the zebrafish hearts as previously described^16,17,18.

2. Hydrating the sensor cartridge

1. Open the extracellular flux assay kit and load the 24-well utility plate (Figure 2A) by pipetting 500 µL of the calibrant solution per well.
2. Close the utility plate using the sensor cartridge and hydro booster (Figure 2A).
3. Incubate the cartridge and utility plate in a humidified incubator at 37 °C overnight (18–72 h). Keep the lid on the sensor cartridge throughout the incubation.

3. Preparing the drug solutions and media

1. Make stock solutions of the drugs: dissolve the oligomycin in dimethyl sulfoxide (DMSO) to a concentration of 10 mM, the carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP) in DMSO to a concentration of 5 mM, the rotenone in DMSO to a concentration of 10 mM and the antimycin A in ethanol to a concentration of 10 mM and store in -20 °C if not used immediately.
   NOTE: Preparing fresh oligomycin, rotenone, and antimycin solutions every 2–3 months, and FCCP solution each time before running the assay, is recommended.
2. Prepare working concentration of the drugs (500 µM oligomycin, 300 µM FCCP, 450 µM each rotenone/antimycin A mix) by diluting the stock solution in DMEM media. Ensure the final volume for the working concentrations is 1.5 mL for oligomycin, 1.65 mL for FCCP, and 1.8 mL for rotenone/antimycin A, allowing for 30 ports to be loaded (24 ports per inhibitor on the cartridge plus 6 extra ports per drug to correct for pipetting errors).
   NOTE: Some precipitation will appear in the rotenone/antimycin A, but after mixing well, this will be minimized. Note that the concentrations mentioned in 3.2 are 10x the final concentration of the drugs when they will be injected in the wells.
3. Prepare the DMEM medium (pH 7.4) to contain 10 mM glucose, 1 mM pyruvate, and 2 mM glutamine, and store in -20 °C until use.

4. Turning on the extracellular flux analyzer

1. Turn on the analyzer from the side button overnight before the assay to allow temperature equilibration with the environment.
2. Turn on the computer and open the Wave software.
3. Select the Mito Stress test on the software.
4. Turn off the heater in the bottom left corner of the software so that the temperature in the analyzer drops below 37 °C and reaches 28–29 °C to approach the standard temperature of zebrafish husbandry¹⁹.
5. Set the injection strategy on the Group definitions page (under the injection strategies tab) by adding the compounds that will be deposited in each port from the compound catalogue. For the Mito Stress test, add oligomycin in port A, FCCP in port B, and rotenone/antimycin in port C. There is capacity for the addition of one more drug in port D, which is unused in this protocol.
6. Delete all other pre-designed strategies that exist under the injection strategies tab once the experimental strategy is defined. Specify the experimental groups (e.g., wild-type and mutant) on the Groups side of the page, and specify the injection strategy that they will receive (the one defined in step 4.5). Make sure the injection strategies match between groups and also match the injection strategy designed.
7. On the Plate map page, specify the position of each sample on the 24-well plate by clicking on each group and then clicking on the wells that will contain the samples.
   1. Note that four of the wells are not suitable for sample deposition (Figure 2B) and serve as blanks that will be used for background correction. Ensure that the samples in each group are evenly distributed across the plate to even out potential positional bias in measurements or environmental conditions.
8. On the protocol page, remove the Control or Experimental tab if no additional inhibitor other than those stated above is used, and ensure the software ports match the positions of the drugs on the cartridge.
9. The software automatically sets 3 cycles of mixing, waiting, and measuring per condition. Modify the number of cycles to 10 baseline (untreated), 3 oligomycin, 7 FCCP, and 9 rotenone/antimycin A measurements to allow for stable traces and for the inhibitors to act on the 3D tissue. Ensure each cycle consists of 3 min mixing, 2 min waiting, and 3 min measuring.

5. Analyzer calibration

1. Remove the cartridge and utility plate from the humidified incubator (see steps 2.1–2.3).
2. Remove the hydro booster and the lid (Figure 2A) from the cartridge and place the sensor cartridge on the utility plate containing the calibrant.
3. Gently pipette the inhibitors into their respective ports while touching the side wall of the port with the pipette tip, ensuring a smooth flow of the liquid: 50 µL of oligomycin in port A, 55 µL of FCCP in port B, and 60 µL of rotenone/antimycin A in port C (Figure 2C).
   1. Note that the liquid is held in place only by surface tension at the hole of the port, so avoid abrupt pipetting or taping the cartridge from this step onwards. Ensure that no liquid enters the sensor port (Figure 2C) as this would damage the analyzer.
4. Click start run on the software, name the run, and save it on the computer. At this point, the tray of the analyzer will open to place the cartridge and utility plate.
5. Place the cartridge loaded with the drugs and the utility plate containing the calibrant on the analyzer tray according to the plate map on the tray (A1 well on the top left) and ensure proper fit. Remove the hydrant booster and the lid (Figure 2A) from the plate before loading it on the tray. Failure to do so will damage the analyzer.
6. Once ready, instruct the software to start the assay by pressing run assay. At this point, the analyzer will take in the loaded tray and initiate the calibration.
7. Start the heart isolation (see steps 6.1–6.12). The calibration will last approximately 40 min, but the plate can be retained in the analyzer for as long as it takes to isolate the hearts and place them in the islet capture microplate (Figure 2B,D).

6. Isolating the fish hearts and loading them on the 24-well Islet plate

1. Open the islet capture microplate and load the 24-well islet plate with 500 µL of DMEM (10 mM glucose, 1 mM pyruvate, 2 mM glutamine) at room temperature.
2. Place all the grids in a 35 mm Petri dish containing DMEM media at room temperature, ensuring they are fully immersed and free of large bubbles.
3. Prepare an additional 35 mm Petri dish containing DMEM (10 mM glucose, 1 mM pyruvate, 2 mM glutamine) at room temperature. This will serve to rinse the hearts upon collection and dissect the ventricle from the atria and bulbus arteriosus.
4. Cull one fish in 5 g/L MS-222 and check for the absence of reflexes.
5. Place the fish on a sponge secured on a 92 mm Petri dish, open its cardiac cavity, and dissect the heart as previously described^16,20.
6. Rinse the hearts in the 35 mm Petri dish containing DMEM (10 mM glucose, 1 mM pyruvate, 2 mM glutamine) at room temperature.
7. Remove the atrium and bulbus arteriosus using the forceps in the 35 mm Petri dish.
8. Gently shake and squeeze the ventricle for the blood to drain from its lumen. Do not damage the ventricular wall when washing the heart.
9. Transfer the heart to its designated well on the islet capture microplate and allow it to sink to the center of the appropriate well according to the plate map. Maintain the plate at room temperature. Include all ventricles in the assay, even if they stop beating visibly.
10. Repeat the process for all available fish. If comparing the metabolic activity of different experimental groups, we suggest you isolate one heart per group instead of all hearts from one group before moving on to the next one. Although the hearts can survive for at least 180 min in the analyzer (Figure 3A), we propose the isolation happens one heart per group at a time as best practice to ensure the equal distribution of time in the media between groups.
11. Once all hearts have been isolated (approximately 60 min for a plate of 20 hearts), lay them at the bottom of the islet capture microplate wells (Figure 2B), and position them in the center of the well using thin forceps (Figure 2D,E).
12. Use a forceps to cover them with the grid (approximately 20 min) (Figure 2D). Make sure that the grid is in the correct orientation (mesh facing down towards the heart) and does not contain large bubbles, as this might interfere with the measurements (Figure 2E).
    1. Once the grid is in position, use the forceps to push it in place from the edges until a slight click is heard. Make sure not to crash the heart with the forceps through the grid.

7. Loading the samples on the analyzer and acquiring measurements

1. Once the hearts are enclosed in the islet capture microplate, check on the software that all wells have been successfully calibrated, as noted by green tick marks on the computer screen. Now eject the utility plate by pressing open tray.
   NOTE: The cartridge will not be ejected at this point, and only the utility plate containing the calibrant will appear.
2. Exchange the utility plate for the islet microplate containing the samples on the analyzer tray. Ensure that the islet microplate is positioned correctly (A1 in the top-left corner) and that the lid has been removed. Failure to remove the lid will damage the analyzer.
3. Once the islet plate is positioned, instruct the software to initiate the experiment by pressing load cell plate. During approximately the first 10 min, the sensors will perform an equilibration step, and then the measurements will commence.
4. Once the experiment is completed (approximately 4 h), press eject cartridge for the tray to open and the islet microplate and cartridge to be ejected.
5. Remove the islet microplate and cartridge, and press done for the tray to close. At this point, press view results to see the data collected.
6. Save the measurements, export them from the analyzer computer, and upload them on the analytics interface for further analysis and processing. This will trigger the excision of the islet plate and cartridge from the analyzer.
7. Remove the islet plate and cartridge from the analyzer and close the cartridge using the lid.
8. Freeze the islet plate on dry ice or remove the hearts from the wells (see step 8.2) and store at -20 °C if the protein isolation for normalization (step 8) is not immediately performed.
9. Once the data is saved, turn off the software, the analyzer, and the computer.
10. Upload the data on the analytics interface and visualize it by selecting the OCR graph from the standard views option. On the interface, view the data as OCR and ECAR level or rate, correct for background based on the empty wells, and select wells to visualize by clicking on the plate map to make them appear or disappear.

8. Protein isolation and quantification for normalization

1. Prepare protease inhibitor-containing radioimmunoprecipitation assay (RIPA) buffer by dissolving one protease inhibitor tablet in 10 mL of RIPA buffer and mixing well.
2. If not done already (step 7.8), using the islet capture screen insert tool (Figure 2F), remove the grids from the islet plate to release the hearts. Remove the hearts carefully, as they might be stuck on the grid. If this is the case, use the forceps to release them. If the grid cannot be removed using the screen insert tool, move the plate under the stereoscope and, using microdissection scissors, carefully cut the grid to release the heart.
3. Place each ventricle in a 1.5 mL tube containing 200 µL of RIPA buffer enriched with protease inhibitor on ice.
4. Homogenize the heart using a homogenizer with disposable pestles on ice.
5. Prepare the standards by diluting the provided bovine serum albumin (BSA) in RIPA buffer in the following concentrations: 2000 µg/mL, 1500 µg/mL, 1000 µg/mL, 750 µg/mL, 500 µg/mL, 250 µg/mL, 125 µg/mL, 25 µg/mL, 0 µg/mL.
6. Prepare the Pierce BCA working solution (50:1, reagent A: reagent B, according to the manufacturer's instructions). Calculate the volume of working reagent needed using the following formula:
   (number of standards + number of samples) × 2 × 200 μL
   NOTE: This volume allows the use of 200 µL of working solution per sample and standard, and allows for the acquisition of duplicate measurements per sample and standard.
7. In a 96-well plate, add 25 µL of each standard and each sample per well in duplicates.
8. Add 200 µL of the working solution per well.
9. Incubate the plate at 37 °C for 30 min.
10. Allow the plate to cool down to room temperature for 10 min and seal it using a plate sealer. Make sure there are no bubbles in the wells, as this might obstruct the plate reader. If a limited number of bubbles form, use a tip or the thin forceps to burst them.
11. Measure the absorbance at 562 nm using a plate reader.
12. Generate a standard curve based on the absorbance of the standards and calculate the protein concentration of the samples.
13. Log the protein concentration on the normalization tab (Figure 2G) of the analytics interface for each well.

9. Data analysis

1. After normalizing against the protein content, export the data from the analytics interface as Excel or Prism files for further analysis using the cloud button in the top right corner of the graph.
2. Using the normalized data, calculate the baseline OCR and ECAR as the average of the baseline measurements, excluding the first three acquisitions, as the signal might be unstable at the first measurements.
3. Using the normalized data, to specify the maximal OCR and ECAR, calculate the average measurements of the top three acquisitions of the FCCP treatment.

Results

This protocol enables quantification of the OCR and ECAR of whole zebrafish ventricles. We validated the survivability of the ventricles in an ex vivo setting by incubating uninjured wild-type KCL ventricles in the analyzer in the absence of any drug treatment (Figure 3A). The data indicates that the ventricles have a stable OCR and ECAR, which are significantly elevated from the baseline measurements of the empty wells for at least 180 min (Figure 3A). This data suggests that the ventricles show normal respiratory function and survival in this ex vivo setting.

We performed several OCR measurements using various drug concentrations and media formulations to identify the optimal dosage for the ventricles (Figure 3B). We observed a mild increase in OCR fold change (FCCP/basal) when utilizing the combinations [oligomycin] 10 µM, [FCCP] 8 µM, and [rotenone/antimycin] 25 µM, and a slightly larger increase in [oligomycin] 50 µM, [FCCP] 30 µM, and [rotenone/antimycin] 45 µM (Figure 3B). As a result, we proceeded with the latter concentrations for our experiments. As the ventricle is a 3D structure, it is possible that the drugs strongly act on the external cells of the tissue while not reaching the cells in the luminal side of the ventricle. To address this, we performed a permeabilization step where we treated the hearts with saponin 0.5% for 30 min before incubation in DMEM (10 mM galactose, 1 mM pyruvate, 4 mM glutamine). This resulted in a sustained reduction in OCR and absence of the expected increase in OCR after FCCP injection (Figure 3C), suggesting that the hearts were metabolically failing. Consequently, we did not permeabilize the ventricles for our experiments.

Furthermore, we present here the normalized traces from our original experiment⁸. Harnessing the natural variation in regenerative potential of seven wild-type zebrafish strains, we identified mdh1ab, the cytosolic malate dehydrogenase 1Ab, to have a beneficial effect on regeneration through cardiomyocyte redifferentiation. To validate this finding, we produced transgenic zebrafish overexpressing mdh1ab in a cardiomyocyte-specific manner (mdh1ab c^OE) and control fish overexpressing green fluorescent protein (GFP c^OE). We cryoinjured their ventricles, collected them at 14 dpci, and investigated their metabolism using the Mito Stress test (Figure 3D). The oligomycin treatment did not significantly alter the OCR, although a mild increase in ECAR was observed. Injection of FCCP in the media resulted in the expected increase in OCR and ECAR with the mdh1ab c^OE samples maintaining the highest maximal respiratory capacity (Figure 3D). Finally, rotenone/antimycin led to the complete shutdown of metabolism in both groups. Overall, this method allowed us to identify significant differences in the metabolic capacity of the mdh1ab-overexpressing ventricles and between seven different wild-type zebrafish strains⁸.

Figure 1: Glucose metabolism during a Mito Stress test. (A) Schematic representation of glucose metabolism. Glucose enters the cell through the GLUT transporters and gets converted to pyruvate through glycolysis. Lactate dehydrogenase (LDH) converts pyruvate to lactate, and the resulting protons are secreted from the cell. Pyruvate gets converted to acetyl-CoA and enters the mitochondria and TCA cycle to fuel the respiratory electron transport chain in the inner mitochondrial membrane. The inhibitors of complexes I, III, and V, as well as the uncoupler FCCP, are noted in red. Electrons being transferred through the respiratory electron transport chain are denoted by blue arrows, which eventually lead to the conversion of oxygen and protons to water. (B) Mito Stress test results schematic. Basal respiration is noted in magenta. Upon oligomycin injection, the OCR drops, allowing for the calculation of the ATP-linked respiration (green) as well as the proton leak (orange). FCCP increases the OCR to the maximal respiratory capacity of the cells (blue). Rotenone and antimycin A injection lead to the shutdown of the mitochondrial respiration and only allow non-mitochondrial respiration to take place (grey). Please click here to view a larger version of this figure.

Figure 2: Sensor cartridge and islet capture microplate layout. (A) Image of the lid-sensor cartridge-hydro booster-utility plate complex (top) and the inverted sensor cartridge (bottom). Scalebar: 1cm. (B) Image of an islet capture microplate loaded with zebrafish ventricles. The red circles denote the blank wells that do not contain ventricles and will serve as background correction during the analysis. Scalebar: 1cm. (C) Schematic representation of the port layout in the XFe24 sensor cartridge. Port A contains oligomycin, B contains FCCP, and C contains the rotenone/antimycin mix. Port D can be left empty, or an extra drug can be added. S denotes the sensor port, which should not contain any liquid in order not to damage the analyzer. (D) Representative image of a zebrafish ventricle at the bottom of the well, enclosed by the grid (left) and without a grid (right). Scalebar: 5mm. (E) Schematic cross-section of the islet capture microplate containing a ventricle, the grid (yellow), and the sensor cartridge (green). P denotes the ports and S the sensor. The wavy line represents the medium. (F) The islet capture screen insert tool. Scalebar: 1 cm. (G) Normalization tab icon on the analytics interface. Please click here to view a larger version of this figure.

Figure 3: Representative OCR and ECAR measurements of zebrafish ventricles. (A) Representative OCR and ECAR traces of uninjured wild-type zebrafish ventricles (KCL strain) that have not been treated with the inhibitors used in the assay (n = 2, red line). The traces resulting from empty wells that do not contain any biological material are presented in grey (n = 4); OCR, two-way repeated measures ANOVA, time factor P = 0.0200, time × well interaction P = 0.0196, well factor P < 0.0001, group factor P < 0.0001; ECAR two-way repeated measures ANOVA, time factor P = 0.1323, time × well interaction P = 0.0216, well factor P = 0.0009, group factor P<0.0001; data presented as mean ± SEM. (B) OCR fold change upon various drug concentrations and media used. One-way ANOVA (P = 0.0032) with Tukey's multiple comparisons test. [O] 5 µM, [F] 4 µM, [RA] 2 µM, n = 6; [O] 10 µM, [F] 8 µM, [RA] 25 µM, n = 8; [O] 25 µM, [F] 16 µM, [RA] 45 µM, n = 8; [O] 50 µM, [F] 30 µM, [RA] 45 µM, n = 8; [O] 100 µM, [F] 60 µM, [RA] 90 µM, n = 8. O, oligomycin; F, FCCP; RA, rotenone/antimycin. (C) Trace of saponin-permeabilized ventricles (red, n = 4) compared to untreated ventricles (grey, n = 4); two-way repeated measures ANOVA, time factor P < 0.0001, time × sample interaction P < 0.0001, treatment factor P < 0.0001, sample factor P = 0.824; data presented as mean ± SEM. OCR_max of saponin-treated ventricles compared to untreated ventricles; unpaired student's t-test (P = 0.0850). (D) Representative OCR and ECAR traces from mdh1ab cOE (red, n = 8) and GFP cOE (grey, n = 8) 14dpci ventricles. OCR, two-way repeated measures ANOVA, time factor P < 0.0001, time × sample interaction P < 0.0001, genotype factor P < 0.0001, sample factor P = 0.1098; ECAR, two-way repeated measures ANOVA time factor P < 0.0001, time × sample interaction P = 0.7507, sample factor P = 0.2154, genotype factor P < 0.0001. Data adapted from Lekkos et al., 2025⁸ and presented as mean ± SEM. Please click here to view a larger version of this figure.

Discussion

Here, we describe in detail the protocol for analyzing the OCR and ECAR of ex vivo zebrafish ventricles using the Seahorse XFe24 analyzer and the Mito Stress test. Cardiomyocyte metabolism plays an essential role in heart regeneration^21,22,23,24. However, validation of the metabolic phenotype of the regenerating zebrafish heart has been challenging and is often lacking from publications^6,7. Due to the small size of the zebrafish heart, techniques such as NMR spectroscopy or mass spectrometry require the pooling of several ventricles in order to acquire adequate signal^8,12,13. In contrast, the method we describe here allows for the acquisition of OCR and ECAR measurements from multiple individual ventricles, thus increasing the statistical power.

Classical metabolic profiling methods, such as metabolomics, provide a snapshot of the metabolic profile of the tissue at the moment it was collected. Through the Mito Stress test we were able to collect real-time, live data on how the ventricles are adapting to various metabolic challenges, including their maximal respiratory capacity, which is not evident when using other methodologies. In order to achieve live phenotyping of zebrafish ventricles, perfusion of the zebrafish heart has been attempted^9,10,11. This requires complex customized equipment and highly skilled technicians to achieve canulation while remaining a low-throughput method. The analyzer provides a simpler alternative using a commercially available piece of equipment for high-throughput (20 ventricles at a time) phenotyping of ventricles. To our knowledge, there has been no direct comparison of the method described in this paper with other techniques of live oximetry, such as a Clark-type electrode.

Importantly, the method we describe is adaptable depending on the research question. The assay has been used previously to characterize whole zebrafish embryos^25,26. Additionally to the Mito Stress test described here, the Mito Fuel Flex test has been optimized to evaluate the metabolic fuel utilization of human cardiac slices²⁷. Furthermore, the Cell Energy Phenotype test has been adapted for characterizing the glycolytic capacity of regenerating Astyanax mexicanus hearts²⁸. We anticipate that this protocol will be utilized for other 3D structures such as cardiac organoids, and will be useful for validating genetic perturbation of metabolic pathways in uninjured zebrafish hearts.

We recognize that this technique has certain limitations. In terms of technical challenges, upon isolation, the hearts were handled in DMEM medium that was at room temperature and not maintained at 28 °C, which is the optimal temperature for zebrafish¹⁹. Decreasing the temperature of the water to 18 °C can have significant effects on the function of the zebrafish heart in situ^29,30. However, we do not expect short-term exposure to room temperature to have a significant effect on the physiology of the isolated ventricles, as the ventricles rapidly return to the 28 °C analyzer. In our hands, the majority of the ventricles stopped beating visibly upon dissection of the atrium, pointing to the lack of pacemaker stimulus as an etiology of the non-contraction. Moreover, during the isolation period, the medium was not oxygenated, potentially leading to mild hypoxia, which recovers to normal oxygen levels following mixing in the analyzer before the first measurement.

An additional practical consideration regarding the temperature of the hearts also relates to the heater being turned off in the analyzer, and thus leading to a reduction in temperature from 37 °C to 28–29 °C. The analyzer does not allow for the reduction in heater power to the desired temperature. Nevertheless, when the analyzer is operating, the temperature inside the tray is 6–7 °C higher than the room temperature. This temperature range coincides with the optimal zebrafish husbandry temperature¹⁹.

This assay is performed on bulk zebrafish ventricles. Consequently, it is impossible to distinguish the contributions of individual cell types to the changes observed in the OCR and ECAR. This might prove to be of significant importance when investigating the metabolic rates of early regenerative time points when the high influx of immune cells³¹ might mask the OCR changes of cardiomyocytes. Additionally, residual blood remaining in the ventricular lumen could influence the metabolic readouts. Allowing the hearts to bleed in the presence of heparin⁷ might further reduce the influence of blood cells clotting in the heart. Moreover, in order to enhance the diffusion of the inhibitors used towards the inner cells of the ventricle, we attempted a permeabilization step using saponin. Saponin permeabilization led to a sustained drop in OCR throughout the assay, leading the ventricles to respiratory failure. These limitations could be avoided by the isolation, culture and analysis of isolated cardiomyocytes, although this approach requires the prolonged culture of cells³² which might not be optimal for a regenerative setting.

The injured tissue might influence the normalization method used. In this study, we normalize the OCR by µg protein/mL using the BCA assay. Although the BCA assay does not react strongly with collagen proteins due to their hydroxyproline- and proline-rich composition, which are poorly detected using coper, the presence of other cell-devoid wound components might influence the normalization. Future studies could consider the use of total DNA³³ as an accurate reflection of live cells in the ventricle.

Furthermore, we were unable to reproduce the anticipated reduction in OCR caused by oligomycin-dependent inhibition of complex V¹⁵. Although certain studies show a reduction in OCR by oligomycin treatment of isolated cardiomyocytes^32,34,35, the data here resemble the lack of response reported in a recent study on isolated adult mouse cardiomyocytes³⁶. Since the oligomycin concentration we use is significantly higher than the one reported in the aforementioned studies, it is unclear whether increasing the final concentration to more than 50 μM would have an effect.

Finally, although oxygen consumption primarily depends on respiration, the extracellular acidification is not exclusively driven by the lactate produced by glycolysis, but also through the conversion of carbon dioxide to carbonic acid. It has been reported that in glucose-fed myoblasts, a third of the extracellular acidification results from carbon dioxide in baseline conditions³⁷. Consequently, we advise caution on the interpretation of the ECAR measurements resulting from the Seahorse analyzer if no correction for the proportion of the respiration-driven acidification has been performed³⁸.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1.5ml Eppendorf tubes	Starlab	S1615-5550
96-well plate	ThermoFisher	AB-0800
Antimycin A	ChemCruz	sc-202467A
DMEM medium	Agilent	1035755-100
DMSO	Sigma-Aldrich	D2650-100mL
Dry ice
Ethanol	Supelco	1.00983.2500
Extracellular Flux Assay Kit	Agilent	103518-100	pack contains utility plate, hydro booster, sensor cartridge, lid (comes with the islet plates)
FCCP	Sigma-Aldrich	C2920
Glucose	Agilent	103577-100
Glutamine	Agilent	103579-100
Incubator
Islet capture microplate	Agilent	101122-100	25 grids included in the pack
Islet capture screen inster tool	Agilent	101135-100
microdissection forceps	ideal-tek	3131760	5-S size (2x will be needed to isolate the ventricle from the bulbus arteriosus and atrium)
microdissection scissors	FST	15009-08
MS-222	Sigma-Aldrich	E10521
Oligomycin	MP	151786
Pestle mottor homogeniser	VWR	431-0100
Pestles	VWR	431-0094
petri dish 35mm	Greiner Bio-One	627102
petri dish 92mm	Starstedt	82.1473
Pierce BCA Assay Kit	Thermo Scientific	23227
Pierce Protease and Phosphatase Inhibitor Mini Tablets,EDTA-free	Thermo Scientific	A32961
Pyruvate	Agilent	103578-100
RIPA lysis buffer	santa-crus	sc-364162
Rotenone	MP	02150154-CF
Seahorse Analytics Agilent interphase	Agilent	https://www.agilent.com/en/product/cell-analysis/real-time-cell-metabolic-analysis/xf-software/agilent-seahorse-analytics-787485
Seahorse XFe24 Analyzer	Agilent	https://www.agilent.com/en/product/cell-analysis/real-time-cell-metabolic-analysis/xf-analyzers/seahorse-xfe24-analyzer-740878
Sealer	Thermo Scientific	AB-0558
SPECTROstar Nano	MG LABTECH GmbH	601-0298
Sponge			make an incision long and deep enough to fit the fish ventral side upwards
Wave (version 2.6.1.56)	Agilent	https://www.agilent.com/en/products/cell-analysis/software-download-for-wave-desktop
XF calibrant	Agilent	100840-000