2D and 3D Echocardiography in the Axolotl (Ambystoma Mexicanum)

Summary

Here we present echocardiography protocols for two-dimensional and three-dimensional image acquisition of the beating heart of the axolotl salamander (Ambystoma mexicanum), a model species in heart regeneration. These methods allow for longitudinal evaluation of cardiac function at a high spatiotemporal resolution.

Abstract

Cardiac malfunction as a result of ischemic heart disease is a major challenge, and regenerative therapies to the heart are in high demand. A few model species such as zebrafish and salamanders that are capable of intrinsic heart regeneration hold promise for future regenerative therapies for human patients. To evaluate the outcome of cardioregenerative experiments it is imperative that heart function can be monitored. The axolotl salamander (A. mexicanum) represents a well-established model species in regenerative biology attaining sizes that allows for evaluation of cardiac function. The purpose of this protocol is to establish methods to reproducibly measure cardiac function in the axolotl using echocardiography. The application of different anesthetics (benzocaine, MS-222, and propofol) is demonstrated, and the acquisition of two-dimensional (2D) echocardiographic data in both anesthetized and unanesthetized axolotls is described. 2D echocardiography of the three-dimensional (3D) heart can suffer from imprecision and subjectivity of measurements, and to alleviate this phenomenon a solid method, namely intra/inter-operator/observer analysis, to measure and minimize this bias is demonstrated. Finally, a method to acquire 3D echocardiographic data of the beating axolotl heart at a very high spatiotemporal resolution and with pronounced blood-to-tissue contrast is described. Overall, this protocol should provide the necessary methods to evaluate cardiac function and model anatomy, and flow dynamics in the axolotl using ultrasound imaging with applications in both regenerative biology and general physiological experiments.

Introduction

Ischemic heart disease is a leading cause of death worldwide^1,2. Although many survive a myocardial infarction due to rapid and fine-tuned medical intervention, ischemic incidents in humans often lead to fibrotic scarring associated with hypertrophy, electrical malfunction, and a diminished functional capacity of the heart. This lack of regenerative potential of cardiac tissue is shared among mammals and although controversial claims of mammalian cardiac regeneration have been reported, these have been limited to specific murine strains^3,4 and hypoxia treated mice⁵. Thus, the field of cardiac regenerative medicine and biology is generally limited to non-mammalian animal models to study intrinsic heart regenerative phenomena. The zebrafish (Danio rerio) has in the past decade been established as the most well characterized model for intrinsic heart regeneration^6,7,8,9,10. Due to easy laboratory maintenance, a short generation time and a wide array of molecular tools available, the zebrafish is well adapted as a model for genetic and molecular mechanisms underlying cardiac development and regeneration. However, the minute dimensions of the zebrafish heart make it less suited for functional evaluation, and complicated surgical procedures and the non-tetrapod phylogeny of the zebrafish limits the sensible extrapolation of findings in this species, thus justifying the use of other larger tetrapod models. One of the earliest models of vertebrate heart regeneration was a caudate amphibian, the Eastern newt (Notophthalmus viridescens)¹¹, a species that remains a valuable model¹².

In recent years another caudate amphibian, the Mexican axolotl (A. mexicanum) has entered the scene as a large (up to 100 g of body mass) and highly laboratory adaptable animal model for a wide array of regenerative disciplines spanning limb regeneration, spinal cord injury, and cardiac regeneration^{13,14,15,16,17}. The axolotl is highly amenable to functional measurements on the heart using high frequency echocardiography and the absence of calcified structures on the ventral side of the heart allows for ultrasound imaging with a much lower level of image artifacts (acoustic shadowing and reverberation in particular) than observed in other model animals with calcified sternum and ribs.

The following protocol describes several different methods and preparations (Figure 1, Figure 2) to acquire reproducible echocardiographic measurements on the axolotl heart in both anesthetized (applying three different anesthetics: benzocaine, MS-222, and propofol) and unanesthetized animals in two (Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Supplementary Files 1-12) and three (Figure 8, Figure 9, Supplementary Files 13-14) spatial dimensions. The amphibian heart is three-chambered (two atria and a single ventricle). The atria are supplied by a large sinus venosus and the ventricle empties into the conus arteriosus outflow tract (Figure 2). Since most emphasis is traditionally placed on ventricular regeneration and less on the recovery of atria^{6,7,8,9,10,11,12,14,17}, this protocol mainly focuses on measurements of ventricular function.

Amphibian echocardiography is not well-described in the literature, and the development of the 2D methods described in this paper have been driven by the need to best represent the functionality of the beating axolotl heart at a given time and experimental setting. Thus, the methods described here are applicable in heart regenerative experiments where cardiac function can be repeatedly monitored over the course of a regeneration process. Additionally, the methods can be applied in cardiophysiological experiments on the axolotl in general or modified slightly to span other caudate or anuran amphibian models (e.g.,Xenopus). The axolotl exists in several different strains and color variations (e.g., wildtype, melanoid, white, albino, transgenic white with green fluorescence protein expression), however these characteristics do not affect the compatibility of the axolotl with the described protocol. The method described here to acquire 3D echocardiographic data is a modified version of the spatiotemporal image correlation (STIC) technique developed for clinical ultrasound and the quadratic averaging method described previously in the developing chicken to enhance the signal of blood speckles in soft tissues in species containing nucleated red blood cells^18,19. This method allows for advanced modeling of cardiac contraction and computed fluid dynamics in the axolotl heart.

Protocol

The procedures carried out in this protocol were in accordance with the national Danish legislation for care and use of laboratory animals and the experiments were approved by the Danish National Animal Experiments Inspectorate (protocol# 2015-15-0201-00615).

1. Preparations

1. Prepare axolotl medium.
   1. Apply high quality non-chemically treated tap water as axolotl medium. If this is unavailable, apply 40% Holtfreter's solution.
   2. Prepare 40% (wt/vol) Holtfreter's solution by dissolving 15.84 g NaCl, 0.54 g CaCl₂·2H₂O, 1.11 g MgSO₄·7H₂O, and 0.288 g KCl in filtered and deionized water up to a volume of 1 L.
2. Make immersion anesthetics.
   1. Prepare benzocaine (ethyl 4-aminobenzoate) anesthetic solution by dissolving 200 mg ethyl 4-aminobenzoate in 3 mL acetone and then dissolving this mix in 1 L tap water or 40% Holtfreter's solution.
   2. Prepare MS-222 (ethyl 3-aminobenzoate methanesulfonic acid, also commonly known as tricane) anesthetic solution by dissolving 200 mg ethyl 3-aminobenzoate methanesulfonic acid directly in 1 L tap water or 40% Holtfreter's solution.
   3. Prepare propofol (2,6-diisopropylphenol) anesthetic solution by dissolving 3.3 mg 2,6-diisopropylphenol in in 1 L tap water or 40% Holtfreter's solution. Alternatively, dilute commercially premade solution to 3.3 mg/L.
      CAUTION: Propofol is a powerful human anesthetic (intravenously administered) and should be handled with care, including in the diluted form.
3. Prepare bed and container for echocardiography.
   1. Prepare lip-shaped animal bed for echocardiography by folding a 70 cm x 55 cm piece of soft cloth once and then rolling it into "burrito shape" (Figure 1A). Then bend over the ends until they meet and tape these together (Figure 1B).
   2. Submerge the lip shaped structure in axolotl medium to accommodate the anesthetized axolotl during ultrasound imaging. Secure the animal to the structure and prevent floating using loose rubber bands; position these mid-mandibular and over the sacral region (Figure 1C).
      NOTE: The rubber bands should not squeeze the animal as this will affect hemodynamics.
   3. For 2D echocardiography on unanesthetized axolotls, prepare a hammock by carving out a 16 cm x 8 cm x 5 cm hole in a 33 cm x 27 cm x 5 cm block of polystyrene foam (e.g., a lid from a medium size polystyrene foam container) (Figure 1D).
   4. Push a 33 cm x 27 cm piece of plastic wrap through the hole and secure the edges of the wrap to the top surface of the polystyrene foam block (Figure 1E) to create a hammock. Add axolotl medium to 3 cm of depth in the hammock. The unanesthetized axolotl will sink to the bottom of the hammock allowing for easy ventral access through the plastic wrap (Figure 1F).

2. Anesthetize Axolotls

1. Immerse axolotl in desired anesthetic solution (benzocaine, MS-222, or propofol).
2. Inspect for first signs of sedation, reduced movements and increasing loss of righting reflex, this appears within 10 min in animals < 10 g (< 10 cm) and within 20 min in animals between 10 g and 50 g of body mass (10-22 cm).
3. Inspect for complete lack of body movements, gill ventilation movements, and righting reflex, and make sure the animal is non-responsive to moderate pain stimulation tested by pinching the webbing between digits.
   NOTE: Despite the fact that general anesthesia is accomplished within 30 min in benzocaine anesthetized axolotls, cardiac function is not stabilized until 1 h. This is not the case in MS-222 or propofol anesthetized axolotls (Figure 6A-F).
4. To maintain axolotl under general anesthesia, keep the animal in anesthesia solution or wrapped in wet paper wipes wetted in anesthesia solution.
   NOTE: Anesthesia can be maintained for 7 h without adverse effects on animal's wellbeing given that the skin and especially the gills are kept moist.
5. To reawake axolotl, transfer the animal to anesthesia-free medium.
   NOTE: The first sign of awakening is gill ventilation movements. Animal should be upright and responsive to stimulation within 1 h.

3. 2D Echocardiography on Anesthetized Axolotl

1. Place anesthetized axolotl in a supine position in the lip-shaped animal bed (steps 1.3.1-1.3.2). Secure it from floating using loose rubber bands (Figure 1C). Ensure that the thoracic surface is covered by 3-5 mm of medium.
   NOTE: For a brief acquisition (< 5 min) anesthetic-free medium can be applied. For prolonged acquisition, anesthetic solution should be applied as ultrasound medium to ensure stable cardiac function throughout measurements.
2. Position the transducer over the midline of the animal in the thoracic region parallel to the long axis of the animal (Figure 2A, Figure 3A-B, Supplementary File 2). Use transillumination with a cold light source on white and albino axolotls to ensure the correct placement of the transducer (Figure 2C and Supplementary File 1).
   1. For axolotls weighing < 20 g, use a 50 MHz transducer; for axolotls > 20 g, use a 40 MHz transducer. Ensure positioning of the cranial/anterior direction to the right for standardized image acquisition. If this is not the case rotate the transducer 180 ° or invert the image.
3. Make sure that in the long axis midline view, a small portion of the ventricle (positioned to the right in the thoracic cavity, Figure 2A) appears in the frame at ventricular diastole and a large portion of the left and right atria (positioned at the center/slightly toward the left in the thoracic cavity, Figure 2A) and the sinus venosus are visible both in atrial systole and diastole (Figure 3A, B).
4. Translate the transducer 1-3 mm toward the animal's right to obtain the ventricular long axis view (Figure 2A). The correct position is attained when the cross-sectional area of the end-systole ventricle is at its maximum (Figure 3C-H).
5. In B-mode, acquire ≥ 3 cardiac cycles with > 50 frames/s in either 'general imaging' (high spatial/low temporal resolution) or 'cardiology' (low spatial/high temporal resolution) mode.
   NOTE: This view allows for evaluation of ventricular function. Ventricular function can be evaluated in two dimensions using the ventricular fractional area change (FACv) calculated from the end-diastolic and end-systolic cross sectional area of the ventricle (CSAv) using the equation:
   (1)
   The axolotl's ventricle assumes the shape of a sphere and a geometry based stroke volume [SV(geo)] can be calculated using the equation:
   (2)
6. Translate the transducer along the long axis of the animal until the center of the ventricle is in the middle of the screen. Rotate the transducer 90 ° clockwise to obtain the mid-ventricular short axis view (Figure 5A and B, Supplementary File 10). Evaluate the circular shape of the ventricle by translating the transducer along the long axis of the heart.
7. Return the transducer to the long axis plane and translate it back toward the midline or slightly to the left of the midline to obtain the atrial long axis two chamber view (Figure 2A). Make sure that the correct position is attained by confirming that the cross-sectional areas of the end-systole atria are at their maxima and the two atria combined assume the outline of the number '8' tilted ~ 45 ° toward the left (Figure 4A and B, Supplementary File 6).
8. In B-mode acquire ≥ 3 cardiac cycles with > 50 frames/s in either 'general imaging' or 'cardiology' mode.
   NOTE: The axolotl's atria are irregular in shape and 3D function cannot be directly inferred from 2D data, thus their function must be evaluated as an index measure such as atrial fractional area chance (FACa) based on the combined cross sectional area (CSAa) of both atrial chambers in systole and diastole:
   (3)
9. Translate the transducer toward the right until the outflow tract (conus arteriosus) appears (close to the ventricular long axis view) (Figure 2A).
   NOTE: After leaving the ventricle in an anterior direction, the outflow tract makes a sharp bend and runs at a small angle toward the ventral surface before again assuming an anterior direction and splitting up into gill branches and systemic vessels.
   1. Make sure that the correct outflow tract view is attained by confirming that the diameter of the outflow is at its maximum at ventricular end-systole and two of the three semilunar valves at the entrance of the outflow are visible at mid-ejection (Figure 4E, Supplementary File 8).
      NOTE: The ventral directionality toward the transducer of the intermediate portion of the outflow tract allows for velocity and flow measurements using Doppler imaging.
10. Apply Color Doppler-mode to map blood flow velocities in the outflow tract during cardiac ejection (Figure 4F and Supplementary File 9). Likewise apply Color Doppler and Power Doppler imaging to visualize blood flow in the ventricular and atrial views (Figure 3E-H, Supplementary Files 4-5 and Figure 4C-D, and Supplementary File 7).
11. Apply Pulse Wave (PW) Doppler-mode at the position of maximum blood velocity in the portion of the outflow tract running toward the transducer.
    1. Use 'beam angle' and 'angular correction' up to 45 ° to adjust for the outflow not being completely perpendicular to the face of the transducer (Figure 4G). Make sure that the PW Doppler position is not overlapped by the spiral valve of the outflow during any phase of the cardiac cycle (Figure 4E).
12. In PW Doppler-mode acquire velocity/time data over ≥ 3 cardiac cycles.
13. Return to B-mode and acquire ≥ 3 cardiac cycles at the exact same plane as PWV was acquired.
14. Measure the velocity time integral (VTI) of the blood flow in the outflow tract as the area under the velocity/time curve for one full cardiac cycle (Figure 4G, g1).
    NOTE: From the VTI and the diameter (d) of the outflow tract at end-systole obtained from the B-mode acquisition, a PW Doppler based stroke volume [SV(pw)] can be calculated using the equation:
    (4)
    Heart rate (HR) is measured from the velocity/time curve by measuring the duration of an entire cardiac cycle. Cardiac output [CO(pw)] is calculated using the equation:
    (5)
15. Obtain the oblique paragill view, a view that offers an alternative for the measurement of the blood flow velocity in the outflow tract, by rotating the axolotl 90 ° in the lip shaped bed in such a way that the right part of the animal is facing upward (Figure 2B). Angle and rotate the transducer and position it parallel and just posterior to the protruding gills (Figure 2B). Make sure that the correct position is attained by confirming that the outflow tract is running downward at ~ 45 ° and that the atria appear under the outflow tract during ejection (Figure 5C, Supplementary File 11).
16. Apply PW Doppler-mode at the position of maximum blood velocity in the portion of the outflow tract running away from the transducer (Figure 5D, Supplementary File 12). Use 'beam angle' and 'angular correction' up to 45 ° to adjust for the outflow not being completely perpendicular to the face of the transducer (Figure 5E).
17. In the PW Doppler-mode acquire blood velocity over ≥ 3 cardiac cycles.
18. Return to B-mode and acquire ≥ 3 cardiac cycles at the exact same plane as PWV was acquired.
    NOTE: SV(pw) and CO(pw) are calculated for the oblique paragill view using Equation 4 and Equation 5 as described above for the long axis view.

4. 2D Echocardiography on Unanesthetized Axolotl

1. Place the unanesthetized axolotl in a prone position in the hammock (step 1.3.3).
2. Leave the animal undisturbed for 30-60 min to recover from handling stress.
3. Position the ultrasound transducer with the transducer head facing upward toward the axolotl in the hammock.
4. Apply ultrasound-gel on the transducer.
5. Gently, and without disturbing the animal, position the transducer over the midline of the animal in the thoracic region parallel to the long axis of the animal.
   NOTE: This is the same, but inverted, position as for the anesthetized axolotl (step 3.2).
6. Obtain B-mode, Color Doppler mode, PW mode data in the long axis and short axis view as described in steps 3.2-3.14.
   NOTE: An oblique paragill view is unobtainable in the unanesthetized axolotl. The echocardiographic data in unanesthetized axolotls should be acquired between gill ventilation movements (a 10-20 s period for a resting animal). If the axolotl moves during acquisition, measurements must be repeated.

5. Evaluate 2D Echocardiography Data and Minimize Subjectivity

1. Avoid operator/observer bias in 2D ultrasound imaging and 3D evaluation of cardiac function based on 2D data caused by subjectivity in both the data acquisition and the data analysis phase by performing intra/inter-operator/observer analysis²⁰.
   NOTE: In the startup of studies and when training new personnel this subjectivity must be quantified and minimized using intra/inter-operator/observer analysis.
2. Initiate the intra/inter-operator/observer analysis in a two person setup with operator/observer 1 (less experienced) being tested against operator/observer 2 (more experienced) by performing ≥ 6 consensus measurements together, including both bench work at the ultrasound system (operation) and subsequent analysis of relevant parameters (observation).
3. Reach consensus between operators and observers and operate (operator/observer 1) the ultrasound system to acquire relevant data on ≥ 6 animals (operation 1.1).
4. Directly after, operate (operator/observer 2) the ultrasound system to acquire relevant data on the same animals (operation 2.1).
5. Let animals recover for 3 days. Thereafter, repeat (operator/observer 1) the procedure (operation 1.2).
6. Analyze (operator/observer 1) all measured data (operation 1.1/observation 1.1; operation 2.1/observation 1.1; operation 1.2/observation 1.1) and after 24 h repeat the analysis of operator/observer 2's data (operation 2.1/observation 1.2).
7. Analyze (operator/observer 2) the data acquired by her/himself (operation 2.1/observation 2.1). Note that the values obtained by this analysis are considered closest to the true values.
8. Evaluate variation, tendencies, and bias in comparisons between all acquired parameters using Bland-Atman plotting, QQ plotting, t-test (equal mean), and F-test (equal variance) (Figure 6G).
   1. Note that the operation 1.1/observation 1.1 versus operation 2.1/observation 1.1 comparison demonstrates the inter-operator variation.
   2. Note that the operation 2.1/observation 1.1 versus operation 2.1/observation 2.1 comparison demonstrates the inter-observer variation.
   3. Note that the operation 1.1/observation 1.1 versus operation 1.2/observation 1.1 comparison demonstrates the intra-operator variation.
   4. Note that the operation 2.1/observation 1.1 versus operation 2.1/observation 1.2 comparison demonstrates the intra-observer variation.
9. Make sure that mean and variation of the different measurements are non-significantly different for the four comparisons; the difference between measured values should fall within ± 1.96 standard deviations, and there should appear no tendencies toward less precision of neither small nor large values.

6. 3D Echocardiography on Anesthetized Axolotl

1. 3D acquisition
   1. Place the anesthetized axolotl in a supine position in the lip-shaped animal bed (step 1.3.1). Secure it from floating using loose rubber bands (Figure 1C) and make sure that the thoracic surface is covered by 3-5 mm of medium. 3D acquisition is a lengthy procedure, therefore apply anesthetic solution as ultrasound medium to ensure stable cardiac function throughout measurements.
   2. Position the transducer over the midline of the animal in the thoracic region either parallel to the long axis of the animal (for a sagittal 3D recording) or orthogonal to the long axis (transversal 3D recording).
   3. Translate the transducer in the in-plane dimension (x and y) and the out-of-plane dimension (z or slice) to ensure that the entire cardiac region will be covered in the subsequent 3D capture.
   4. Adjust the frame rate and spatial resolution as desired by selecting either 'general imaging' (high spatial/low temporal resolution) or 'cardiology' (low spatial/high temporal resolution) mode. For 0.33 Hz < HR < 1 Hz use a temporal resolution of 50 frames/s obtained at high spatial resolution ('general imaging') that allows for the cardiac cycle to be reconstructed into 50-150 distinct phases.
   5. Adjust '2D gain' to a level where anatomical structures are barely recognizable in the raw B-mode image (~ 5 dB) to increase signal-to-noise-ratio in the final reconstructions.
   6. For each slice (z step), record ≥ 1,000 frames.
   7. Translate transducer one z step at a time, e.g., 20 µm or 50 µm, and repeat recording until the entire cardiac region is covered.
2. 3D reconstruction (Supplementary Files 13 and 14).
   1. Export acquisitions into Digital Imaging and Communications in Medicine DICOM (little endian).
      NOTE: Each slice containing a set number of frames will compose a single file.
   2. Determine the number of frames in a full cardiac cycle. As HR can vary over time, determine this for both the first and the final slice. Set the highest number of frames per cycle as the initial upper estimation of phase resolution that can later be reduced (step 6.2.8).
   3. Determine crop boundaries and excise irrelevant space surrounding the B-mode window.
      NOTE: These boundaries should be constant throughout slices.
   4. Convert the RGB Color image into 32-bit.
   5. Calculate the correlation value (C) for each frame in the stack and the number of frames included in the first cardiac cycle using the formula:
      (6)
      NOTE: Here  is the signal intensity of the pixel at coordinate (i, j) in the first image  and is the same in the second image, , and , are the mean intensity and standard deviation, respectively, of the first and second image in the comparison, and I and J are the numbers of columns and rows in the image. The resulting array of correlation values will have the size of the product of the number of frames per cardiac cycle and the total number of frames per slice (e.g., 75 × 1,000 = 75,000 in Figure 8) (see exemplary script in Supplementary File 16). The correlation value cannot be calculated if one or both of the images in the comparison have a standard deviation of pixel values of zero, however this is highly unlikely in ultrasonographic images.
   6. Detect local maxima in the array of correlation values (Figure 8A, see exemplary script in Supplementary File 17 to automatically detect local maxima).
   7. Calculate quadratic average Q(AVG) of frames with peak correlation values (i.e., matching cardiac phases) using the formula:
      (7)
      where N is the total number of frames with matching cardiac phases, is the intensity of the pixel at coordinate (i, j) of the n^th image and is the temporal arithmetic mean of of the n^th image (see exemplary script in Supplementary File 18).
   8. Repeat step 6.2.3-6.2.7 for all slices.
   9. Select a slice (reference slice) with easily recognizable anatomical structures (e.g., mid-ventricular) and check if the reconstructed ensemble averaged one cardiac cycle corresponds to exactly one cycle (i.e., if there are additional phases resulting in more than one cardiac cycle). Delete additional phases to yield exactly one cardiac cycle (e.g., going from an overestimated 75 phases/cycle (in reality, 1.07 cycle) in Figure 8 to exactly one cycle containing 70 phases in Figure 8).
   10. On the neighboring slice (test slice), sort the ensemble averaged one cardiac cycle t-stack into matching cardiac phases with the reference slice using the correlation value formula ( Equation 6) (see exemplary scripts in Supplementary File 19 and Supplementary File 20).
       NOTE: Although two non-identical slices will not appear completely similar at any point during the cardiac cycle, adjacent slices with a sufficiently small step size (e.g., 20 µm or 50 µm) will have pronounced similarities resulting in correlation value maxima that can be translated into matching phases.
   11. Repeat step 6.2.9-6.2.10 for all slices.
   12. Collapse the entire 3D reconstruction into a single 3D Tagged Image File format (TIF) containing z slices and t frames or into a stack of DICOM files.
       NOTE: Data can be binned in each dimension to reduce size, increase signal-to-noise ratio and generate isotropic data (in-plane resolution is usually several folds higher than out-of-plane resolution).

Results

Intrapericardial space in the axolotl is dependent on the size of the animal. Smaller animals (2-20 g, 7-15 cm) will have an excess of pericardial fluid (appearing dark in echocardiography) surrounding the cardiac chambers whereas in larger sexually mature animals (> 20 g, > 15 cm) the chambers will occupy most of the intrapericardial space. To provide the best overview for representative results of echocardiographic views of the axolotl heart, a smaller animal (10 g, 10 cm) was app...

Discussion

Echocardiography in the axolotl and other non-mammalian species yields fundamentally different data than mammalian echocardiography because of the nucleated nature of red blood cells in all vertebrates except adult mammals. This results in a pronounced blood signal and less blood-to-tissue contrast in axolotl echocardiographic images compared to e.g., mouse or human echocardiography. This can make image segmentation on unprocessed single frame ultrasound images more difficult as it can be hard to distinguish blo...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Axolotl (Ambystoma mexicanum)	Exoterra GmbH	N/A	All strains (wildtype, melanoid, white, albino, transgenic white with GFP) can be applied for echocardiography
Vevo 2100	Fujifilm, Visualsonics	Vevo 2100	High frequency ultrasound system
MS700	Fujifilm, Visualsonics	MS700	50 MHz center frequency, transducer
MS550s	Fujifilm, Visualsonics	MS550s	40 MHz center frequency, transducer
Micromanipulator	Zeiss	NA
Benzocain	Sigma-Aldrich	94-09-7	ethyl 4-aminobenzoate
MS-222	Sigma-Aldrich	886-86-2	ethyl 3-aminobenzoate methanesulfonic acid
Propofol	B. Braun Medical A/S	NA	2,6-diisopropylphenol
Sodium chloride	Sigma-Aldrich	7647-14-5	NaCl
Calcium chloride dihydrate	Sigma-Aldrich	10035-04-8	CaCl2·2H2O
Magnesium sulfate heptahydrate	Sigma-Aldrich	10034-99-8	MgSO4·7H2O
Potassium chloride	Sigma-Aldrich	7447-40-7	KCl
Acetone	Sigma-Aldrich	67-64-1	Propanone
Soft cloth	N/A	N/A	Any piece of soft cloth measuring appromixately 70 x 55 cm^2 e.g. a dish towel
Styrofoam block	N/A	N/A	Any piece of Styrofoam block measuring approximately 33 x 27 x 5 cm^3 e.g. a medium sized Styrofoam cooler lid
Plastic wrap	N/A	N/A	Any piece of plastic wrap e.g. food wrap
Tape	BSN Medical	72359-02	Leukoplast sleek
Kimwipes	Sigma-Aldrich	Z188956	Kimwipes, disposable wipers
Excel 2010	Microsoft	N/A	Excel 2010 or newer
ImageJ	National Institutes of Health	ImageJ 1.5e or newer. Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, https://imagej.nih.gov/ij/, 1997-2016.