Two-photon manipulation of neuronal activity and behavior in Hydra vulgaris

Summary

The introduction of calcium imaging has rendered cnidarians, such as Hydra vulgaris, valuable model organisms for investigating neuronal activity and behavior. Here, we present a comprehensive protocol to image and manipulate neuronal activity and behavior of Hydra. We describe steps for wide-field imaging and two-photon simulation and ablation of neurons. We then detail imaging behavior and post-ablation analysis. We address challenges that may arise during the preparation and execution of the experiments.

Graphical abstract

Highlights

• •Calcium imaging of neuronal activity in behaving Hydra
• •Selective two-photon activation and ablation of neurons
• •Behavioral effects of neuronal activity manipulation

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

The introduction of calcium imaging has rendered cnidarians, such as Hydra vulgaris, valuable model organisms for investigating neuronal activity and behavior. Here, we present a comprehensive protocol to image and manipulate neuronal activity and behavior of Hydra. We describe steps for wide-field imaging and two-photon simulation and ablation. We then detail imaging behavior and post-ablation analysis. We address challenges that may arise during the preparation and execution of the experiments.

Before you begin

The functional roles of specific neuron types can be studied in Hydra using fluorescent reporters (e.g., GCaMP or GFP). To express the reporter gene in a specific neuron type, we create a plasmid containing promoter and reporter gene sequences. The promoter sequence is taken from the upstream sequence of a gene expressed in the cell of interest.^1,2 The plasmid is then injected into the 1- to 2-cell stage of embryos. It takes about 2 weeks to 1 month for these injected embryos to hatch. After these embryos hatch, hatchlings expressing fluorescent cells are selected and screened for full expression. For background about embryogenesis and transgenesis, please refer to previous papers.^2,3,4 Troubleshooting 1.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Experimental models: Organisms/strains
Hydra vulgaris (interstitial GCaMP6s transgenic)	Hydra vulgaris AEP strain	Dupre and Yuste5
Hydra vulgaris (Hym-355 GFP transgenic)	Hydra vulgaris AEP strain	Yamamoto and Yuste6
Hydra vulgaris (Hym-355 GCaMP6s transgenic)	Hydra vulgaris AEP strain	Yamamoto and Yuste6
Recombinant DNA
pHyVec1	Addgene	Addgene_34789
GCaMP6s Synthetic	GeneArt	Custom
Hym-355 promotor/terminator	GENEWIZ	Custom
Software and algorithms
Codes used for the paper	GitHub	https://github.com/NTCColumbia
Instructions to install DeepLabCut	GitHub	https://deeplabcut.github.io/DeepLabCut/docs/installation.html
Other
Fluorescent stereo microscope	Leica	M165
ORCA-Flash 4.0 LT3 Digital CMOS Camera	Hamamatsu	C11440-42U40
Leica EL6000 External light source	Leica	EL6000
VWR micro cover glass 22 × 22 mm	VMR	48366-067
SecureSeal™ Imaging Spacers	GRACE BIO-LABS	SKU: 654002
Two-photon laser-scanning microscope, Ultima In Vivo	Prairie Technologies/Bruker	N/A
Ti:sapphire laser	Coherent	Chameleon Ultra II
50–160-BK Pockels cell	Conoptics	302RM
Electrically tunable lens	Optotune AG	EL-10-30-C-NIR-LD-MV
Pulse-amplified laser	Spectra-physics	Spirit 1040–8
1147-4-1064 Pockels cell	FastPulse Technology	8025RS-H-2KV
Data acquicition system, Multifunction I/O Device	National Instruments	USB-6008
USB digital microscope	Bysameyee	Microscope-1000×
25× objective	Olympus	XLPLN25XWMP2
Autofluorescent plastic slides	CHROMA	92001
FB650-40 - Ø1″ Bandpass Filter	THORLABS	FB650-40
USB digital microscope	Bysameyee	Microscope-1000×

Materials and equipment

Hydra medium recipe (100 L in deionized water)

Reagent	Final concentration	Amount
Calcium chloride dihydrate	1.0 mM	13.7 g
Magnesium sulfate anhydrous	0.33 mM	3.97 g
Sodium bicarbonate	0.5 mM	4.2 g
Potassium chloride	0.03 mM	0.22 g
diH2O	N/A	100 L

Note: Store at a temperature below 25°C.

Step-by-step method details

Wide-field calcium imaging

Timing: 1 h

Wide-field calcium imaging of the whole body can show correlations between specific neuronal activity and behavior.

Note: Example microscope setup – A fluorescence dissecting microscope (Leica M165) equipped with a long-pass GFP filter set (Leica filter set ET GFP M205FA/M165FC), a 1.6× Plan Apo objective, and an sCMOS camera (Hamamatsu ORCA-Flash 4.0) is used to image Hydra mounted between a coverslip and a glass slide separated by a 150–200 μm spacer. Neuronal activity is imaged at a frame rate of 2 Hz for approximately 1 h.

• 1.
  Mounting Hydra.

  • a.
    Prepare a glass slide and a coverslip: Clean the glass slide and a coverslip with ethanol, rinse with water, and dry.
  • b.
    Attach the spacer: Remove the protective film from one side of the spacer and stick it onto the glass slide.

    • i.
      An optimal preparation will have a firm attachment of the spacer (at a height of 150–200 μm) without any wrinkles.
  • c.
    Mount Hydra: Place transgenic Hydra expressing GCaMP6s in a specific neuron type onto the glass surrounded by a spacer.

    • i.
      An optimal preparation will have Hydra well submerged in Hydra medium and not touching the spacer.
  • d.
    Gently lower a clean glass cover slip onto the top of Hydra without introducing air bubbles (Figure 1).

Note: It is important to use a Hydra of a small size so that it is not overly compressed, as excessive compression can stress the Hydra and alter its neuronal activity and behavior. It may take practice to empirically find the right-sized Hydra for the imaging. Troubleshooting 2.

• 2.
  Imaging.

  • a.
    Prepare the microscope: Turn on the fluorescent microscope and adjust the settings for imaging. Place the glass slide mounted with Hydra onto the microscope stage.

    • i.
      The recommended reference parameters are as follows: Magnification = 3‒5; light intensity = set intensity switch to maximum; exposure time = approximately 200 ms.
  • b.
    Set imaging parameters: Choose imaging parameters that suit experimental needs.
  • c.
    Begin imaging: Start imaging the whole-body Hydra, and continue for the desired length of time. Keep an eye on the Hydra throughout the imaging session to ensure it remains within the field of view.

Note: As an example, the frame rate can be set to greater than 2 Hz to cover faster neuronal activity of Hydra, and exposure/acquisition time can be set to be optimized for fluorescent signal and movement of Hydra.

Note: A dissection microscope can be used to achieve a greater depth of field and a larger field of view, ensuring clear imaging of neuronal activity and body movement during Hydra’s behavior. Alternatively, a compound microscope can be used for the same purpose by using an objective with lower magnification or a larger numerical aperture. The issue of the depth of field with compound microscopes can also be overcome by using techniques such as focus stacking.

CRITICAL: It is important to set the camera frame rate to be over 2 Hz to capture all neuronal spikes, as faster spikes in Hydra can reach up to 1 Hz.⁶ Additionally, it is recommended to image for about 1 h to record less frequent behaviors such as somersaulting, which occurs on average only 1–2 times per h.

Figure 1.

Mounting Hydra for calcium imaging

(A) Hydra cultured in a 10 cm round plastic dish.

(B) Hydra mounted in a confined space: 1. Glass slide, 2. Glass coverslip, 3. Spacer, and 4. Hydra submerged with Hydra media.

Two-photon stimulation

Timing: 1–2 h

The aim of this protocol is to use two-photon laser stimulation to investigate the causal relationship between neuronal activity and behavior using Hydra expressing GCaMP6s in a specific neuron type.

Note: Example microscope setup (used in Yamamoto and Yuste, 2023)⁶ – A modified two-photon laser scanning microscope (Ultima In Vivo, Bruker) was used.⁷ GCaMP6s signals are recorded using a Ti:sapphire laser (Chameleon Ultra II, Coherent) with its wavelength tuned to 940 nm, and photostimulation was performed by using a low repetition rate pulse-amplified laser (Spirit 1040–8, Spectra-physics) with a fixed wavelength of 1040 nm. The power of both lasers is controlled by two independent Pockels cells. The two laser beams on the sample are individually controlled by two independent sets of galvanometric scanning mirrors. Short movies (approximately 270 s) with a sample rate of 3.64 Hz are collected (Imaging laser power <50 mW; dwell time 2 ms/pixel; 256 × 256 pixels in the whole field of view).

CRITICAL: To minimize cell damage, the power of photostimulation is adjusted on each neuron to the lowest value that elicits a response. Single-cell photostimulation is performed with a spiral pattern scanned by a pair of galvanometric mirrors delivered to the center of the cell (3 μm diameter; 3 spiral revolutions) for 30 ms.

Note: An electrically tunable lens (ETL) is used to adjust and keep the same focal plane between imaging and photostimulation. Simultaneous imaging and photostimulation are controlled by PrairieView and custom-made software to control an ETL and data acquisition (DAQ, USB-6008, NI) running in MATLAB.⁸

Note: Optogenetics is not yet feasible in Hydra due to inability to express exogenous opsin proteins using available transgenic techniques. However, two-photon laser can evoke neuronal activity without exogenously expressing channel rhodopsin. This method evokes neuronal activity likely through various mechanisms, including photothermal,⁹ photomechanical,^10,11 and photochemical effects.^12,13,14,15

• 3.
  Optical settings (Figure 2).

  • a.
    Place a stage with a through hole larger in size than the Hydra.
  • b.
    Place a behavior camera beneath the stage.
  • c.
    Install 25× objective.
  • d.
    Place bandpass filters on both the behavior camera and the external illuminator.

Note: It is essential to filter light at the wavelength that is used for imaging and stimulation to not interfere with the recording of Hydra behavior.

• 4.
  Mounting Hydra.

  • a.
    Mount Hydra as described in the previous step (Step 1a).
• 5.Calibration of photostimulation laser.
  As laser systems use various galvanometers, photostimulation (via uncaging galvanometers) needs to be calibrated precisely to target the same x, y coordinates as the imaging laser. This can be done using a built-in function of PrairieView software (Please refer to the manual that is available for download at https://eliceirilab.org/sites/default/files/PrairieViewManual_5_2013_0.pdf). To align the z plane, an ETL is used.

  • a.
    Place an autofluorescent plastic slide.
  • b.
    Start the system, with keeping the light on with the filter.
  • c.
    Open PrairieView software.
  • d.
    Define scan settings (Image Size, Dwell Time, Sampling period, etc.) in the main control window. The same parameter setting work for different experiments.
  • e.
    Start two-photon imaging according to the manual.
  • f.
    Conduct Point Photoactivation/Uncaging to calibrate the location and power of laser stimulation.
  • g.
    Use the electrically tunable lens (ETL) controller program to locate the photostimulation focal plane aligned to the imaging focal plane.

    Note: The alignment of the imaging system and calibration should be done before photostimulation.

    Note: In sub-step d, set the "Image Size" according to the desired field of view so that multiple target neurons can be observed, the "Dwell Time" according to the laser power and desired signal-to-noise ratio (longer dwell times will result in better signal-to-noise ratio but slower imaging), and the "Sampling Period" according to the desired imaging rate (shorter sampling periods will result in faster imaging but lower signal-to-noise ratio). These parameters should be set just enough to see the morphology and the activity of neurons to ensure the well-targeted stimulation. Please refer to the beginning paragraph of this section for approximate parameter values.

• 6.
  Online two-photon stimulation.

  • a.
    Set up the parameters such as time and the number of stimuli (ex. 3 μm diameter; 3 spiral revolutions, and 30 ms).
  • b.
    On the PrairieView software, open the “Mark Points” window and choose “Live/Ablation” to enable laser pulses upon clicking the mouse.
  • c.
    Locate a neuron of interest by moving the manipulator.
  • d.
    Optimize the laser power to determine the lowest power required for efficient stimulation. Start from low power (approximately 5 mW) and increase it by 5 mW steps until the laser reaches neuronal activation.
  • e.
    Start simultaneous recordings of calcium imaging and behavior imaging.
  • f.
    Stimulate the neuron selected during the imaging recordings (online).
  • g.
    Repeat steps from 6c to 6f with different neurons until the Hydra starts to move faster.

Note: These parameters can be determined empirically. The diameter of the spiral should be within the diameter of the cells, which is around 5–10 μm. The number of spiral revolutions and the duration of the stimulation should be set to efficiently stimulate neurons. In our setup, we found the spiral diameter of 3 μm, 3 spiral revolutions, and a duration of 30 ms worked well.

Note: Compressing Hydra under a cover slip limits its movement to a confined space but is necessary for imaging. However, excessive compression can stress Hydra and disturb its neuronal activity and behavior. Therefore, it is crucial to find the right size of Hydra that is compressed enough to limit its movement without being stressed. Troubleshooting 3.

Figure 2.

Two-photon stimulation of Hydra

(A) 25× Objective.

(B) An illuminator with a bandpass filter (650 ± 8 nm) to visualize Hydra behavior.

(C) Hydra mounted as shown in Figure 1, placed above the center of a through hole of the stage for a behavior camera to record the behavior from underneath.

(D) A custom-made stage with through holes at the focus of the imaging/stimulating laser and behavior camera.

(E) A behavior camera with another bandpass filter (650 ± 8 nm) to record Hydra behavior.

Two-photon ablation

Timing: 2–3 h

GFP-expressing neurons are ablated using a two-photon laser to establish the function of specific neuron types for behaviors. The following protocol is based on the same setup as two-photon stimulation with a few additional steps.

• 7.
  Set up lights and filters.

  • a.
    Follow the same setup as two-photon stimulation (Step 3).
• 8.
  Mounting Hydra.

  • a.
    Prepare two clean glass coverslips.
  • b.
    Remove the film from one side of a spacer and stick it onto a glass coverslip to reach a depth of 150–200 μm.
  • c.
    Place a small amount of non-toxic grease on the film of the spacer (non-sticky side), and brush it to form a thin layer.
  • d.
    Mount Hydra expressing GFP in a specific neuron type with Hydra media onto the glass, and place another glass cover slip on top without introducing air bubbles.
  • e.
    Press the glass coverslip to make it stick better.

Note: Neurons on either side of Hydra need to be ablated, so glass coverslips are used on both sides. The grease ensures that there is no leakage of water when handling the sample.

• 9.
  Calibration of laser.

  • a.
    Follow the same step as two-photon stimulation (Step 5).
• 10.
  Two-photon ablation.

  • a.
    Optimize the laser power using the “Live/Ablation” mode of the PrairieView software to determine the lowest power required for efficient ablation.
  • b.
    Using the manipulator to locate each neuron, ablate neurons individually so that all accessible neurons are ablated on one side of the Hydra.
  • c.
    Flip the coverslip to the other side and repeat the ablation process for the remaining neurons. Troubleshooting 4.

Behavioral imaging and analysis

Timing: 3–4 h

After ablation, the freely behaving Hydra is imaged using a bright field microscope. Hydra is mounted in a dish to allow free behavior while limiting movement outside the focal range. The behavior is recorded for 1 h, since some spontaneous behavior may occur only once per hour.

• 11.
  Mounting Hydra (Figure 3A).

  • a.
    Prepare a glass-bottom culture dish and a clean glass cover slip.
  • b.
    Place Hydra on the glass well with media.
  • c.
    Apply a small amount of non-toxic grease around the glass well.
  • d.
    Place the glass coverslip to cover the glass well, and press down to create a tight seal.
  • e.
    Add Hydra media on top of it to prevent evaporation during imaging.

Note: The depth between the coverslip and the glass bottom of the well should be approximately 700 μm for Hydra to behave freely.

• 12.
  Imaging (Figure 3B).

  • a.
    Place the glass-bottom dish on the stage of a microscope equipped with a high-speed camera.
  • b.
    Set the imaging parameter to a frame rate > 1 Hz.
  • c.
    Start recording the behavior of Hydra for 1 h, ensuring that the glass coverslip remains attached during the entire recording. Troubleshooting 5.
• 13.
  Tracking Hydra movements.

  CRITICAL: An open-source software package for markerless pose estimation, DeepLabCut,¹⁶ is used to track the position of Hydra body parts such as mouth, mid-body, foot, and basal disc. This software package has several dependencies, including Python, TensorFlow, OpenCV, Scikit-image, Matplotlib, H5PY, and PyQt5. You must install these dependencies before running DeepLabCut using Python’s package manager, pip. Follow the instructions to ensure that all required packages are installed.

  Note: This part explains how the GUI of DeepLabCut to track the position of Hydra body parts. For specific instructions of how to use the GUI, how to change parameters please refer to Nath et al., 2019.¹⁶

  • a.
    Create a new project: Launch the Anaconda prompt and activate the DeepLabCut environment. Open the DeepLabCut GUI and enter the project name and the video file path.
  • b.
    Assign labels: Edit the config file to assign names for the body parts that are being tracked (e.g., mouth, midbody, foot, basaldisc).
  • c.
    Extract frames: Extract frames to be labeled using the default setting (using kmeans algorithm and openCV).
  • d.
    Label frames: Use the mouse to manually label the Hydra body parts in each frame.
  • e.
    Create training dataset: Create a new training dataset using the default setting (using resnet_50 for the network).
  • f.
    Train network: Select the number of iterations as around 20,000. Train the network.
  • g.
    Create videos: Create a labeled video with the tracked movement. Evaluate the tracking by assessing the labeled video. If the labels do not jump between frames, move to the step 14.
  • h.
    Extract outlier frames: Extract outlier frames that need to be relabeled using the default setting.
  • i.
    Refine labels: Refine the labeling of the outlier frames to improve the accuracy of the network. Use the refined labeling to repeat e-g.

    Note: Move to step 14 when the labeled video has only a few outlier frames (around 10 frames or less), as the values computed during quantification will be smoothed.

    Note: You can save time by refining labels repeatedly, rather than increasing the number of iterations for training the network. Also, by reducing the number of frames in the video by a factor of 10, you can decrease the time for training the network.

• 14.Quantifying Hydra movements.
  We describe a Python code that analyzes the movement of Hydra based on x, y coordinates acquired at Step 13. The code computes parameters including time, body length, body length speed, basal disc distance, basal disc speed, and foot angle. Additionally, as an example of a behavior, the code identifies the number and time locations of somersaults by thresholding the computed parameters.

  Note: To run the code, you need to import the pandas, numpy, and matplotlib packages. Each cell is executed in order. The following sub-steps explain what each cell does when run in Google Colab.

  Our code is available for download at https://github.com/NTCColumbia/Hydra_Somersaulting/blob/main/DLC_Analysis_WY.ipynb.

  • a.
    Upload the file with x, y coordinates in CSV format acquired at the previous step.
  • b.
    Rename the headers of the CSV file to prepare it for analysis.
  • c.
    Convert all the numbers to float.
  • d.
    Compute parameters including body length, body length speed, basal disc distance, basal disc speed, and foot angle.
  • e.
    Throw out noise for the body length using high pass filter.
  • f.
    Plot the filtered body length over time.
  • g.
    Compute the number and time locations of somersaults.

    Note: The threshold for detecting somersaults (substep g) should be determined based on the graph of basal disc speed generated in a previous cell, and the minimum and maximum thresholds should be used to exclude noise.

Figure 3.

Behavioral imaging of Hydra after neuronal ablation

(A) Hydra mounted in a well: 1. Glass coverslip attached to the side of the well with grease, 2. A well depth of 700–750 μm, and 3. Hydra submerged in Hydra media.

(B) A dissection microscope (Leica) attached to a high-speed camera (Hamamatsu) is used for behavior imaging.

Expected outcomes

Wide-field calcium imaging: After making a transgenic animal of which GCaMP6s is expressed in specific neuron types, the neuronal activity can be imaged during behavior. The expected outcome would be that a specific behavior is correlated with specific neuronal activity. For instance, the activity of CB neurons increases when Hydra contracts,^5,17 and the activity of RP1 neurons is increased before Hydra somersaults.⁶

Two-photon stimulation: A small number of specific neurons can be stimulated by a two-photon laser using optimized power. If wide-field calcium imaging shows a correlation between specific neuronal activity and a behavior, the expected outcome is that stimulating those neurons generates the behavior. For instance, contraction is evoked when CB neurons are stimulated, and somersaulting is evoked when RP1 neurons are stimulated (Figure 4; Yamamoto and Yuste).⁶

Figure 4.

Two-photon neuronal stimulation of Hydra

(A) The two-photon imaging shows a stimulated RP1 neuron (arrow).

(B) Another example shows more stimulated RP1 neurons (arrow) after moving the objective to focus on different neurons.

(C) An image from the behavior camera shows that before RP1 neurons are stimulated, the Hydra body is elongated. The arrowhead indicates the location of the basal disc.

(D) Another image shows that after RP1 neurons are stimulated, the Hydra undergoes somersaulting. The arrowhead indicates the location of the basal disc.

Images were taken from the movies that were used for Yamamoto and Yuste (2023).⁶

Two-photon ablation and behavioral imaging: Most neurons of a specific type can be ablated using a high laser power (Figure 5). The expected outcome is the suppression of the particular behavior that correlates with the activity of the set of ablated neurons. For example, when most RP1 neurons are ablated, somersaulting in Hydra is suppressed (Figure 3; Yamamoto and Yuste).⁶

Figure 5.

Two-photon neuronal ablation of Hydra

(Left) Hydra before ablating RP1 neurons. (Right) Hydra after ablating most RP1 neurons.

Limitations

In this two-photon laser method, precise targeting of laser stimulation during Hydra’s constant motion can only be achieved by stimulating neurons one by one after refocusing.

This approach is suited for neurons with strong or bidirectional connections,^18,19 as stimulation of a few neurons spreads over the entire circuit, such as CB neurons or RP1 neurons.²⁰ However, other neuron types such as sensory neurons may not form strong or bidirectional synapses between them.^21,22 Thus, stimulation of a few neurons with the current time scale may not be enough to activate the entire neuronal circuit leading to the behavior. To address this issue, a method to stimulate a population of neurons simultaneously is required, such as optogenetics, chemogenetics, optochemistry or nanoparticles. Alternatively, one could use a Spatial Light Modulator to create a holographic mask of the position of all the targeted neurons, and stimulate them all at once.⁷

Another limitation of our approach is related to the ablation of specific neuron types. Although ablation can reveal the behavioral effect of the loss of specific neuron types, it may not reveal the behavioral effect of inhibiting specific neuron types. To address this issue, a method to inhibit a population of neurons at once is required, such as optogenetics, chemogenetics, or optochemistry.

Troubleshooting

Problem 1

Step: Generating transgenic Hydra.

No fluorescent expression in neurons (before you begin: Step 3).

Potential solution

Hatchlings are healthy but do not have fluorescent cells.

• •Use the promoter of another gene. In Hydra, promoter sequences for each gene have not been systematically identified. Researchers often use the region 1–1.5 kb upstream of the start codon of a gene as a promoter region, and, in most cases, this approach works. However, it is possible that the exact promoter region sits elsewhere. In that case, one can use the region 1–1.5 kb upstream of a different gene that is also expressed in a specific neuron type.
• •Grow Hydra for a longer time until they have buddings. In cases where the expected expression is on the tentacles, it may take longer for a fluorescent marker to appear, as it takes some time for interstitial stem cells to differentiate and migrate to tentacles. In addition, the neurons stay in the tentacles for only a few days before being expelled at the tip of the tentacles. Therefore, timing is important to make sure that neurons with the transgene are captured at the right time.

Hatchlings do not eat.

• •Feed Hydra directly in the mouth. Some hatchlings are not able to catch brine shrimp as they likely don’t make nematocytes in the tentacles. If Hydra does not eat, they shrink and die eventually. One can make non-eating Hydra eat if food is directly presented onto the mouth. In this case, feeding crushed shrimp is helpful to ensure the attachment of the shrimp onto the mouth and its swallowing.
• •Decrease transgene expression. Expressing fluorescent proteins could be toxic to Hydra depending on expression levels and the cell type that they are expressed in. Therefore, it can be helpful sometimes to use promoters with lower expression.

Problem 2

Step: Calcium imaging.

Hydra keeps contracting (Step 1).

Potential solution

• •Wait for 10 min. At the beginning of imaging, Hydra tend to contract more due to the imaging light. Hydra need some time to adapt to the exposure to start behaving more normally.
• •Replace Hydra with a smaller one. Hydra may be too large for the size of the spacer and may keep contracting in response to the mechanical stress of excessive compression. Hydra with a smaller size can be chosen to fit well within the spacer. Due to their elastic body, it is hard to set a specific size that fits the spacer, and the experimenter has to find the optimal size empirically. If the Hydra’s tentacles are relaxed and not contracted, the size is optimal.

Problem 3

Step: Two-photon stimulation

Two-photon stimulation does not lead to any motor behavior (Step 6).

Potential solution

• •Stimulate more neurons in a shorter time (approximately 3 s per neuron). The limiting factor of this method is the time to stimulate each neuron. To minimize the time for stimulation, it is important to move the focus of the objective to a point where more GCaMP6s-labeled neurons are located before starting the stimulation. This process can take 5–10 s.
• •Stimulate neurons during a specific behavioral state. Some neuronal circuits are antagonistic to others. For instance, when a Hydra is contracting, CB neurons are active, which seems to suppress the activity of RP1 neurons. Furthermore, stimulating RP1 neurons during contraction does not generate the required propagation of activity throughout RP1 neurons which results in somersaulting. Thus, it is crucial to stimulate target neurons during different behavioral states.
• •Assess physiological roles rather than motor behaviors. Stimulating neurons may not necessarily result in motor behaviors. Rather, the stimulation of neurons may result in physiological outcomes such as the secretion of growth factors or digestive enzymes. Therefore, it may be worth studying what cells the neurons activate. This could be done by imaging grafted transgenic Hydra that express GCaMP6s in target neurons and in other cell types such as nematocytes, mucous cells, or gland cells.

Problem 4

Step: Two-photon ablation

Two-photon ablation does not suppress behavior (Step 10).

Potential solution

• •Ablate most GFP-labeled target neurons. It is important to ablate most target neurons to see the effect, as any remaining neurons within the neuron type may compensate for the loss by increasing their firing frequency. For instance, decreasing the number of RP1 neurons down to 1/5 is needed to suppress the activity of the remaining RP1 cells.⁶
• •Make sure most target neurons are labeled with GFP. The transgenic Hydra may express GFP in only a portion of target neurons. If there are areas in the body where no GFP-labeled neurons are seen, it is a sign of non-labeled neurons. A further screening process may be necessary to obtain a transgenic line where target neurons are fully labeled with GFP.
• •Assess physiological roles rather than motor behaviors, as mentioned in problem 3.

Problem 5

Step: Behavioral imaging.

Behavior varies in each animal (Step 12).

Potential solution

• •Use the same conditions. To minimize the variation in behavior in each individual, it is crucial to choose animals of similar sizes (without buds, eggs, testes, etc). In addition, Hydra should be cultured in the same conditions, with the same feeding schedule. The temperature of the room should be kept the same for each experiment.
• •Use the same Hydra as a control. Another solution is to use the same Hydra as both a control before ablation, and as a testing group after ablation.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Wataru Yamamoto (wy2303@columbia.edu)

Materials availability

Transgenic Hydra are available upon request.

Data and code availability

This study did not generate/analyze datasets. All code is available on GitHub: https://github.com/NTCColumbia, and Zenodo: https://zenodo.org/record/8011802.