Studying Mechanical Oscillations During Whole-Body Regeneration in Hydra

Abstract

Cells of the freshwater cnidarian Hydra possess an exceptional regeneration ability. In small groups of these cells, organizer centers emerge spontaneously and instruct the patterning of the surrounding population into a new animal. This property makes them an excellent model system to study the general rules of self-organization. A small tissue fragment or a clump of randomly aggregated cells can form a hollow spheroid that is able to establish a body axis de novo. Interestingly, mechanical oscillations (inflation/deflation cycles of the spheroid) driven by osmosis accompany the successful establishment of axial polarity. Here we describe different approaches for generating Hydra tissue spheroids, along with imaging and image analysis techniques to investigate their mechanical behavior.

1 Introduction

Hydra is a simple freshwater animal composed of two epithelial layers, gastrodermis and epidermis, and organized along a single oral/aboral axis. Its regenerative capacities and amenability to experimental manipulation have made it a rich source of insights about regenerating missing body parts already at the dawn of modern experimental biology [1]. Experiments where regeneration has been challenged, as well as transplantation experiments, have substantially shaped the theories of biological pattern formation [2, 3]. Importantly, Hydra does not only offer a platform for manipulating existing patterns but also for observing their emergence de novo. This was shown when cells from dissociated body columns were reaggregated, and they managed to recreate functional animals in a few days [4, 5]. Astonishingly, unlike organoid systems, they are able to do so without the external addition of signaling factors (see ref. 1 for further comparison with organoids).

Initially, the cells in the aggregates sort to re-establish the epidermal and gastrodermal layers, thus creating a symmetric hollow epithelial spheroid composed of cells whose positional identity along the main body axis is to be specified [6]. After approximately 24 h, symmetry is broken and Wnt-expressing organizing centers start to emerge. These centers guide the appearance of head structures at the oral end of the axis [7]. The role of Wnt signaling as a key driver of oral identity is well established both in the homeostatic conditions and in regenerating Hydra [8, 9]. Depending on the size of the aggregates and the axial origin of dissociated tissue, one or more organizing centers can appear [10]. A similar fate awaits spheroids created from small tissue fragments [11]. When a small piece of the body column tissue is excised, it will fold into a hollow spheroid, visually indistinguishable from the one created from reaggregated cells. Since all the cells have a shared identity, body poles need to be defined in this case as well. Wnt signaling centers will emerge and eventually develop into new animal heads.

Interestingly, regenerating spheroids of any origin experience cycles of inflations and deflations on the way to symmetry breaking [12]. These mechanical oscillations are osmotically driven and appear to be important for proper regeneration [13]. Water from the hypotonic medium is entering the cells, which pump it inside the spheroid cavity to maintain their osmotic balance [14]. As a result, the whole spheroid inflates until reaching a threshold of tissue rupture. The accumulated liquid is thus released, the spheroid deflates, and the cycle is repeated. During the spheroid development, the profile of oscillations changes. Initial high-amplitude and low-frequency oscillations (termed Phase I oscillations) eventually transition to faster cycles with lower amplitude (Phase II oscillations). This is a hallmark of symmetry breaking and reflects the emergence of a stable mouth opening that releases the accumulated liquid under lower pressures [15].

Spheroids prepared from small tissue fragments or by single-cell re-aggregation (Fig. 1) are useful to study these mechanical events during regeneration. However, one method might suit specific experimental demands better than the other. Making reaggregates is more laborious, yet offers better control of the spheroid size by using a fixed number of cells. In addition, different cell populations (e.g., expressing different fluorescent markers) can be mixed in one aggregate. Cut tissue pieces, on the other hand, preserve tissue integrity, close faster, and allow better selection of original tissue axial position. Importantly, such spheroids also retain supracellular actin myofibers, which have been recently implicated in mechanically guiding the axis emergence [16]. These structures dissolve upon tissue dissociation and begin to reappear with random orientation in the aggregates. Moreover, they only seem to align and reorient after the symmetry has been broken [17]. This difference thus offers an opportunity to dissect the impact of these actin structures on the tissue mechanical and biological properties.

Fig. 1

Overview of spheroid preparation and development. (a) Using cut pieces as starting material, (b) using cells from dissociated body columns, (c) both methods generate spheroids that will break symmetry and regenerate into full animals

Importantly, Hydra spheroids as an experimental model system do not only offer versatile starting conditions. Additional advantages include short regeneration time, simple culture conditions, amenability to imaging, and experimental manipulations. Perturbing the osmolarity by adding solutes, such as sucrose or sorbitol, to the medium allows slowing down the oscillations in a concentration-dependent manner [13]. Different small molecules (e.g., cytoskeleton-affecting drugs) have also been used to perturb the oscillation dynamics. For example, treatment with the Rho-kinase inhibitor Y-27632 results in a sigmoidal rather than linear inflation behavior [18]. Furthermore, since similar oscillations occur in many other cyst-like structures [19], Hydra spheroids offer a unique system to tackle the biological significance of such phenomena. In the following protocols, we detail techniques for making Hydra spheroids from both cells originating from dissociated animals and cut tissue pieces (Fig. 1). Instructions and tools for live imaging and computational extraction of basic oscillation characteristics from the acquired image data are also provided.

2 Materials

Use distilled water to prepare all the solutions. If not indicated otherwise, solutions can be stored at room temperature.

2.1 Culture Media and Animal Handling

• 1. M-solution: 0.1 M Tris–HCl, 0.1 M NaHCO₃, 0.01 M KCl, 0.01 M MgCl₂, pH 7.4. Filter sterilize.

• 2. Calcium chloride solution: 1 M CaCl₂. Filter sterilize.

• 3. Hydra medium: 2 mL M-solution, 1 mL calcium chloride solution in 1 L water (see Note 1).

• 4. Handling pipettes: Flame the tip of Pasteur pipettes for a few seconds using a Bunsen burner to blunt the edges.

• 5. Agarose gel: 1% (w/v) agarose boiled in Hydra medium. Can be stored for 1–2 weeks.

• 6. Imaging chamber: Multi-chamber glass-bottom imaging slides covered with 2 mm of Agarose gel (see Note 2).

2.2 Cutting and Dissociating Animals

• 1. Microsurgical scalpel (e.g., MICRO FEATHER 15° or 45° ophthalmic incision scalpels, see Note 3).

• 2. Stereomicroscope.

• 3. Dissociation medium: 3.6 mM KCl, 6 mM CaCl₂, 1.2 mM MgSO₄, 6 mM sodium citrate, 6 mM sodium pyruvate, 4 mM glucose, 12.5 mM TES-HCl, pH 6.9. Add antibiotics (0.05 g/L kanamycin, 0.1 g/L streptomycin) and filter sterilize. This solution can be stored at 4 °C for a month.

• 4. Reaggregate medium 1: A 1:3 mixture of Hydra medium and dissociation medium.

• 5. Reaggregate medium 2: A 1:1 mixture of Hydra medium and dissociation medium.

• 6. Reaggregate medium 3: A 3:1 mixture of Hydra medium and dissociation medium.

• 7. Dissociation pipettes: Flame Pasteur pipettes to obtain a narrow opening smaller than 1 mm in diameter. This requires some practice.

• 8. 0.4-mL microcentrifuge tubes (e.g., APEX Scientific mini).

3 Methods

3.1 Spheroid Preparation from Cut Tissue Pieces

• 1. Fill the lid of a 90-mm Petri dish with Hydra medium.

• 2. Transfer a few animals into the dish using a handling pipette.

• 3. Orient the animals with the help of the handling pipette so that they are lying flat and wait until they relax (see Note 4).

• 4. Bisect the body column in 50% of its length with a swift movement of the blade (see Notes 5 and 6).

• 5. Allow the bisected halves to relax again. The tissue immediately next to the cut should appear slightly swollen.

• 6. Make a second cut just below the swelling to obtain a ring of tissue (Fig. 2a). Rings can be taken from both halves of the animal.

• 7. Cut the ring open.

• 8. Perform another cut in the middle of the resulting tissue stripe to create two pieces of equal size (Fig. 2b, see Notes 7 and 8).

• 9. Transfer the pieces into a 35-mm Petri dish with Hydra medium (see Note 9).

• 10. Let them close for 2–2.5 h (see Note 10).

Fig. 2

Critical steps in spheroid preparation protocols. (a) foot half of a bisected animal (note the slight tissue swelling next to the cut side), red line indicates the position of the next cut, (b) ring of tissue, red lines indicate the positions of cuts to prepare fragments of equal size that will give rise to spheres, (c) properly closed spheroids just after closing (left), and after inflation begun (right), (d) improperly closed spheroid releasing cells (arrowheads), (e) tubes with cell suspension positioned in a 50-mL tube and ready for centrifugation, (f) tubes standing in a dish of dissociation medium before the release of aggregates, (g) spheroids mounted in the agarose wells in imaging chamber. All scale bars correspond to 500 μm

3.2 Spheroid Preparation from Dissociated Body Tissue

• 1. Fill the lid of a 90-mm Petri dish with Hydra medium.

• 2. Transfer ~30 Hydra individuals into the dish using a handling pipette.

• 3. Orient the animals with the help of the pipette so that they are lying flat and wait until they relax (see Note 4).

• 4. Cut away the heads of the animals (cut below the tentacles).

• 5. Cut away the feet (cut above the less pigmented zone) of the animals (see Note 11).

• 6. Transfer the resulting body columns into a 15-mL tube with 3 mL of dissociation medium.

• 7. Vortex briefly and wait until the body columns settle at the bottom of the tube.

• 8. Remove as much of the medium as possible.

• 9. Add 3 mL of dissociation medium.

• 10. Slowly pipette the medium in and out of a dissociation pipette (approximately 20 times) to begin dissociating the body column. Avoid introducing any bubbles. The medium should become cloudy.

• 11. Let the suspension sit for about 2 min to allow the sedimentation of bigger tissue pieces.

• 12. Carefully transfer as much of the cell suspension as possible to a new 15 mL tube without disturbing the sediment.

• 13. Repeat steps 9–12 twice with the leftover sediment. Pool all cell suspensions in one tube. You should have collected 9–10 mL of the cell suspension after the third round of dissociation.

• 14. Centrifuge the collected cell suspension for 5 min at 4 °C and 150 rcf.

• 15. Discard most of the supernatant (leave just enough to cover the pellet).

• 16. Resuspend the pellet in 3 mL of dissociation medium (see Note 12).

• 17. Cut the caps of the 0.4-mL microcentrifuge tubes away.

• 18. Fill the tubes with 400 μL of the cell suspension. Pipette the liquid slowly down the side of the tube to avoid creating bubbles. If there is a bubble at the bottom of the tube, tap the tube to release it (see Note 13).

• 19. Position a maximum of four 0.4-mL tubes per 50-mL tubes.

• 20. Centrifuge the 50-mL tubes for 5 min at 4 °C and 150 rcf (Fig. 2e, see Note 14).

• 21. Carefully remove the microcentrifuge tubes from tubes using forceps.

• 22. Add a small amount of dissociation medium to slightly overfill the tube.

• 23. Fill a 90-mm Petri dish with 40 mL of dissociation medium.

• 24. Invert the tubes upside down and place them into the dish with dissociation medium. The tubes should be standing upright with openings completely submerged in the liquid (Fig. 2f, see Note 15).

• 25. Wait for 10–30 min for the aggregates to detach from the tubes and descend into the dish (see Note 16).

• 26. Gently remove the microcentrifuge tubes from the Petri dish.

• 27. Wait for 1 h and carefully transfer the aggregates into the reaggregate medium 3 using a handling pipette.

• 28. Wait for 1 h and transfer the aggregates into the reaggregate medium 2.

• 29. Wait for 1 h and transfer the aggregates into the reaggregate medium 1 (see Notes 17 and 18).

• 30. Culture the aggregates in Hydra medium at 18–23 °C until regeneration is complete.

3.3 Imaging

• 1. Boil the 1% agarose gel (see Note 19).

• 2. Pipette the agarose to the imaging chambers. The agarose layer should cover the flat bottom of the chamber and be approximately 2-mm thick (see Note 20).

• 3. Incubate the chambers at 4 °C until the gel solidifies.

• 4. Create wells in the imaging chamber using a 1000-μL micropipette with the tip attached (see Note 21): Depress the plunger and push the tip through the agarose layer. Rotate the pipette slightly to make sure that the agarose is cut. Release the plunger to suck out the cut piece and take the tip out of the agarose. Create as many wells as necessary for the number of spheroids you intend to image (see Note 22).

• 5. Fill the chambers up with Hydra medium (see Note 23). If air becomes trapped in the wells, release the bubbles by flushing them with a stream of medium from a pipette.

• 6. Select properly formed spheroids (Fig. 2c, d, see Notes 24 and 25).

• 7. Using a handling pipette, carefully transfer spheres to individual wells in agarose (see Note 26). Take care to avoid getting the spheroids in contact with the liquid/air interface, as they will be torn apart by the surface tension. Instead of forcing the spheroids into the wells, hover them over, and wait for them to descend by gravity (Fig. 2g).

• 8. Cover the imaging chamber with a lid and place it on the microscope stage.

• 9. Image in transmitted light with an inverted microscope at temperatures below 23 °C (see Notes 27 and 28). Use magnification that allows fitting the whole agarose well into the field of view. Adjust the light settings so that the spheroid has good contrast against the background—the center of the spheroid will become more transparent as it inflates (Fig. 3a).

Fig. 3

Imaging setup and image analysis. (a) Proper imaging setup, (b) the image from (a) segmented using the described strategy, (c) example of a relative radius trace, shaded area indicates Phase I oscillations, deflation points detected by the “findcollapse” function with a threshold of 1.15 are shown as red dots, (d) plot generated using the “sphereslope” function for the Phase I data in c, black dots indicate the corrected radius (note the absence of deflation), the fitted line is shown in red

3.4 Image Segmentation and Quantification of Oscillation Parameters

• 1. Load the acquired images into ImageJ and apply a coarse median filter (radius = 75 μm, see Note 29).

• 2. Use the Phansalkar segmentation method (radius = 65 μm, parameter 1 and 2 = 0, parameters 1 and 2 correspond to the k and r values in the Phansalkar thresholding method, respectively [20]) in “Auto Local Threshold” function (see Note 30).

• 3. Apply the “Fill holes” function.

• 4. Apply a median filter with a radius of 25 μm.

• 5. Visually inspect the accuracy of the segmentation (Fig. 3b, see Note 31). ImageJ macro for batch segmentation that follows this protocol is provided in Table 1.

• 6. Use the “Analyze Particles” function to measure the size of segmented objects. Include particles bigger than 5000 μm².

• 7. Load the area measurements into Matlab (see Note 32).

• 8. Convert the area measurement into radius according to the formula r = (S/π)^1/2, where S is the measured area and r is the estimated radius.

• 9. Normalize the radius data per sample by dividing them with the respective initial values.

• 10. Use the “findcollapse” function (Table 2) to detect the time and amplitude of spheroid deflations (see Note 33).

• 11. Use the “sphereslope” function (Table 3, see Note 34) to extract the inflation slope.

Table 1 Image segmentation macro that can be used for batch processing in ImageJ. Adjust the parameters to fit the pixel size of your images as described in Note 29

Table 2 The “findcollapse” function. This function identifies deflations and outputs for each instance the time (as frame of the time course) and the amplitude (as the difference of relative radius). Inputs: rad—relative radius data as a column vector, threshold—threshold for detecting the collapse. We recommend a threshold of 1.15 for Phase I oscillations

Table 3 The “sphereslope” function. This function extracts the slope of inflation for a specified period of the time course and plots the fit if requested. Inputs: rad—relative radius data as a column vector, threshold—threshold for detecting the collapse, tinterval—time step of imaging in minutes, frstart—defines the beginning (frame in the time course) of the measured interval, frstop—defines the end of the measured interval, varargin—use either “plot” or “noplot” depending on your preferences. If nothing is specified for varargin, the default option is no plot

4 Notes

• 1. The stock solutions should not be mixed before adding them to the water. This will cause the salts to precipitate.

• 2. We use Lab-Tek (Nunc) or μ-Slide (ibidi) chambers for imaging. Multiwell plates can also be used; however, they are more prone to medium evaporation. Evaporation changes the osmolarity of the medium and impacts on the characteristics of the oscillations.

• 3. Microsurgical blades from different manufacturers are suitable for cutting. The angle of the blade is a matter of experimenters’ preference.

• 4. Use non-budding animals. Be consistent with the feeding status of animals used for experiments since this can influence the results. We typically use animals starved for 24 h.

• 5. If reusing the scalpel, clean the blade with 70% ethanol beforehand. The blade should be sharp and cut through the tissue effortlessly. Use a new scalpel if the blade is damaged. Cut edges that appear squeezed or release threads of cells are an indication of a worn-out blade.

• 6. Body columns can be bisected at different levels, depending on the experimental requirements, for example, to assess the influence of axial position on regeneration.

• 7. It is not recommended to cut the ring in two pieces with a single cut. This usually creates more damage to the resulting pieces than necessary.

• 8. Depending on the proportions of the ring, the animal, and the desired spheroid size, adapt the number of pieces that are cut from one ring. We typically use tissue pieces with the shorter side of ~100–150 μm and the longer side of ~300–500 μm. Animals of some strains (e.g., 105) are thinner, and dividing the ring into two equally sized pieces might result in spheroids that are too small to develop properly. In that case, trim the open ring to the desired size and discard the smaller piece. Similarly, rings with big diameter can be divided into three or more pieces.

• 9. Repeat steps 4–6 in Subheading 3.1 if you want to obtain more rings/spheres from the same animal.

• 10. Tissue pieces from the AEP strain close better than the 105 or Basel strains. To increase the success of closing, dissociation medium can be used instead of the Hydra medium.

• 11. Steps 1–5 in Subheading 3.2 can be omitted but we highly recommend cutting the differentiated parts away before starting the experiments.

• 12. To ensure having a single-cell suspension, the resuspended cells can be passed through a cell strainer (30 μm pore size).

• 13. Less than 400 μL of cell suspension can be used, depending on the desired aggregate size and dissociation efficiency. Always fill up the rest of the tube with dissociation medium.

  For better reproducibility among experiments, cell concentration should be determined using a counting chamber and aggregates prepared with the same number of cells. In our experience, a few thousands of epithelial cells should be used to prepare an aggregate of a final size comparable to a cut spheroid.

• 14. To achieve proper pellet formation, the tubes should be almost vertical during centrifugation. Different rotors might allow for such arrangements even without the use of 50 mL tubes as we describe here. If you want to prepare aggregates with patches of different cells (e.g., expressing two different fluorescent markers), centrifuge first the smaller population of cells, which should form the patch. Then replace the supernatant with the suspension of the other cell population and repeat the centrifugation step.

• 15. The liquid column inside the tubes should be continuous with the medium inside the dish. If you notice any bubbles created at the interface while positioning the tubes, remove them from the dish, add dissociation medium as in step 22 in Subheading 3.2, and place them back into the dish.

• 16. Some aggregates may take longer to detach. These samples often do not develop as well as the faster detaching ones.

• 17. While performing steps 27–29 in Subheading 3.2, you should see the cells sorting and reestablishing the epithelial layers. If you want to image cell sorting, start imaging after the transfer to reaggregate medium 2 and keep the aggregates in this solution.

• 18. The aggregates should stay in the reaggregate medium 1 for 1–2 h before starting the imaging. By this time, the epidermal and gastrodermal layers should be reestablished, which is a sign of successfully completed cell sorting. Depending on the size of the aggregates, cell type ratios, and Hydra strain, the sorting might take longer. In that case, we recommend prolonging the second incubation (reaggregate medium 2). The duration is best determined empirically. Ultimately, the epidermal layer should be continuous, flat, and cells should not be released from spots in the surface.

• 19. Even though the agarose gel can be stored in a closed flask and re-melted when needed, imaging chambers with agarose should always be prepared fresh (e.g., while the spheroids are closing). This prevents the gel in the chamber from drying out and shrinking. Always change tips when pipetting warm agarose to prevent volume changes as the tip heats up.

• 20. It is important to compensate for the presence of agarose in the well when doing chemical treatments. Thermostable compounds (e.g., sucrose) can be added directly to the agarose. Prepare 2% agarose in Hydra medium and mix it, after boiling, with an equal volume of 2× concentrated compound solution. Pipette into the chambers as described. If this approach is not feasible (e.g., for thermolabile molecules), adjust the concentration of the compound, taking the total volume of medium plus agarose into account.

• 21. The wells prevent the spheroids from escaping the field of view during imaging. However, they are spacious enough to allow spheroid expansion without imposing any mechanical constrains.

• 22. You can use a 1000-μL pipette tip or a Pasteur pipette connected to suction to punch the wells and suck out the excess agarose. Alternatively, wells can also be cast using the 200-μL tips or similar custom-made inserts. Position the inserts immediately after pouring liquid agarose into the chambers and make sure they are touching the bottom of the slide. Let the agarose solidify and release the insert afterward by carefully wiggling it out of the gel.

• 23. Various osmolytes (such as sucrose and sorbitol) can be used to osmotically alter the properties of the oscillations. For example, adding 30 mM sucrose approximately doubles the period of Phase I oscillations. If using sucrose or sorbitol, we recommend adding antibiotics to the medium (0.05 g/L kanamycin, 0.1 g/L streptomycin) and filter sterilizing it.

• 24. Properly closed cut spheroids should not be releasing threads of cells and should appear round. The typical size range varies between 300 and 500 μm in diameter. If the spheroids seem to be closing, but the closure point is still visible as a furrow, extend the closing time by 30 min. If you are using Hydra medium for closing the spheroids, some of them may already start inflating after the tissue closes (Fig. 2c). The success rate of closing is strain dependent. We typically observe about 70–95% of successfully closed spheroids in Hydra medium for the AEP strain but a much lower percentage (50–75%) for the 105 strain. The usual causes of unsuccessful closing include cutting with a blunt blade, damage while handling the cut pieces, and poor health of the animals.

• 25. Properly formed spheroids from dissociated cells should be indistinguishable from the spheroids prepared from tissue pieces. Similarly, if the aggregates appear fluffy and/or are releasing threads of cells, they should not be used for further experiments.

• 26. If the imaging medium is not identical with the closing medium, include an additional washing step. Put the spheroids first into a dish with the imaging medium and only then transfer them to the imaging chambers to avoid changing the composition of medium there. Alternatively, the medium in the chamber can also be exchanged after positioning all the spheres in agarose.

• 27. Adjust the parameters of imaging to fit your experimental needs. We routinely image for 60 h with a time step of 10 min.

• 28. This setup is also useful for fluorescence imaging. The bottom half of the sphere can be imaged with good results using a confocal microscope.

• 29. The filter radii are given in μm to be universally applicable. However, the functions in ImageJ require values in pixels. Calculate those according to the pixel size of your image.

• 30. Global segmentation can also be used, but this algorithm tends to outperform it in more challenging situations.

• 31. Common segmentation challenges include cells extruded by the spheroid, agarose pieces in the wells and uneven illumination. Depending on the specific sample, it might be possible to alleviate these issues by adjusting brightness and contrast of the image before segmentation, applying shading and background corrections, or using trainable segmentation algorithms. It is also helpful to only perform the segmentation on the area inside of the agarose well.

• 32. The analysis pipeline can also be implemented in other environments using the information on the algorithm rationale in the following notes and comments within the respective Matlab functions.

• 33. The collapses are detected by dividing the radius values of neighboring time points. For this, the function requires specifying a threshold, which also allows adjusting the sensitivity of detection. We usually use a threshold of 1.15 for Phase I oscillations (Fig. 3c). The amplitude is then calculated as the difference between these points when a collapse is detected.

• 34. This function uses the same deflation detection method as “findcollapse” and uses the amplitude values to correct for the shifts created by spheroid deflations. A straight line is then fitted to the corrected data. This allows measuring the overall slope for long periods, such as the whole Phase I duration (Fig. 3d). To enable slope measurements for different time windows (e.g., one oscillation), the function requires the user to specify an interval for this measurement.