Protocol for recording and analyzing neuronal network activity ex vivo with multi-electrode arrays in mouse brain tissue

Summary

Studying neuronal network activity is essential as brain functions arise from dynamic and coordinated interactions between neuronal populations. Here, we present a protocol for recording and analyzing network bursting activity in mouse hippocampal slices. We describe steps for preparing slices and inducing network activity. We then detail the recording procedure with multi-electrode arrays, followed by analysis. This protocol has potential applications for studying different types of network activity in brain tissues from various areas and species, from mice to humans.

For complete details on the use and execution of this protocol, please refer to Dossi et al.¹

Graphical abstract

Highlights

• •Instructions for preparing acute mouse hippocampal slices
• •Steps for inducing neuronal network bursting activity
• •Guidance on recording bursting activity with multi-electrode arrays
• •Procedures for analyzing network activity with a MATLAB burst detection algorithm

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

Studying neuronal network activity is essential as brain functions arise from dynamic and coordinated interactions between neuronal populations. Here, we present a protocol for recording and analyzing network bursting activity in mouse hippocampal slices. We describe steps for preparing slices and inducing network activity. We then detail the recording procedure with multi-electrode arrays, followed by analysis. This protocol has potential applications for studying different types of network activity in brain tissues from various areas and species, from mice to humans.

Before you begin

The protocol describes the step-by-step procedures to record and analyze hippocampal neuronal network bursting activity ex vivo in mouse brain slices using the multielectrode array (MEA) technique.¹ Nevertheless, this protocol can also be applied to record network activity in different brain regions (i.e. cerebral cortex, cerebellum…) from mice or other animal models and in postoperative human brain tissues, with minor modifications to adapt the storage, incubation, and recording of the slices (i.e. storage and recording temperature, storage conditions and perfusion rate).

Institutional permissions (if applicable)

Experiments were carried out according to the guidelines of the European Community Council Directives of January 1st, 2013 (2010/63/EU). Efforts were made to minimize the number of used animals and their suffering. Experiments were performed in the hippocampus of wild-type (WT) mice and Cx30^−/−Cx43fl/fl hGFAP-Cre (Cx-deficient) mice, provided by Pr. K. Willecke (University of Bonn, Germany).² All mice had a C57BL/6 background and were group-housed on a 12-h light/dark cycle. For all analyses, mice of both genders and littermates were used, and ex vivo slice electrophysiology was performed on P16-P25 mice.³

Researchers utilizing this protocol must obtain approval from their institution’s Animal Ethics Committee.

Preparation of 2.5 M calcium chloride stock solution

Timing: 5 min

• 1.
  Prepare 100 ml of 2.5 M calcium chloride (CaCl₂) stock solution.

  • a.
    Weigh 36.75 g of CaCl₂·2H₂O in a 100 ml beaker.
  • b.
    Add approximately 50 ml of distilled water and stir with a magnetic stirring bar until completely dissolved.
  • c.
    Transfer the solution to a 100 mL volumetric flask.
  • d.
    Rinse the beaker with a small volume (1–2 ml) of distilled water and add it to the flask.
  • e.
    Carefully bring the volume up to 100 mL with distilled water.

Note: Use a Pasteur pipette or a dropper to adjust the volume, avoiding overshooting the calibration mark precisely.

Note: The 2.5 M CaCl₂ stock solution can be prepared in advance and stored at 4°C for a few months.

Preparation of picrotoxin stock solution

Timing: 1 h

• 2.
  Prepare 5 ml of 500 mM picrotoxin (PTX) stock solution (under a chemical hood).

  • a.
    Weigh 1.51 g of PTX in a 10 ml beaker with a precision balance.
  • b.
    Add 5 ml of DMSO.
  • c.
    Place on a stirrer and mix with a magnetic stirring bar until dissolved.

    CRITICAL: Check that the PTX is completely dissolved before proceeding (see troubleshooting 1).

    Note: Protect from light while dissolving, as PTX is light-sensitive.

  • d.
    Prepare 200 μL aliquots and store them at −20°C until use.

    Note: Aliquots can be stored for up to 6 months at −20°C; avoid repeated freeze-thaw cycles.

Preparation of agar block (4%)

Timing: 45 min

• 3.
  Prepare 200 ml for 3 Petri dishes (diameter 10–12 cm).

  • a.
    Weigh 8 g of agar in a 250 mL beaker.
  • b.
    Add 200 mL of ddH₂O.
  • c.
    Place the beaker on a heater/stirrer and heat at 200°C–250°C while agitating with a magnetic stirring bar (800–1000 rpm).
  • d.
    Boil for 1 min after a clear solution.

CRITICAL: Avoid clots and ensure the agar is completely dissolved before proceeding.

• 4.
  Pour the warm agar into the Petri dish.

  • a.
    Remove air bubbles (if present).
  • b.
    Check that the thickness of the agar block is at least 0.8 cm high.

Note: Be careful when handling warm solutions to avoid burns. Always pour heated solutions carefully, and wear appropriate protective equipment, such as heat-resistant gloves if necessary.

• 5.Let the agar cool down and solidify.
• 6.Close the Petri dishes, seal them with parafilm, and store them at 4°C.

Note: 4% agar blocks can be prepared and stored at 4°C for weeks.

Preparation of cutting solution

Timing: 5–10 min

• 7.
  Prepare a 2 L stock of 1× cutting solution (see troubleshooting 2).

  • a.
    Weigh all the required powders in a 1 L beaker.
  • b.
    Add approximately 1 L of distilled water and stir with a magnetic stirring bar until completely dissolved.
  • c.
    Transfer the solution into a 2 L graduated cylinder.
  • d.
    Bring the volume up to 2 L with distilled water.

Note: The 1× cutting solution can be prepared in advance and stored at 4°C for a week.

• 8.
  Complete the solution by adding CaCl₂.

  • a.
    While stirring with a magnetic stirrer (400–500 rpm), add 0.1 ml of 2.5 M CaCl₂ stock solution per 500 ml of cutting solution while oxygenating with carbogen (0.5–1 L/min).
  • b.
    Check osmolarity using an osmometer (target: 320–330 mOsm/L).

Note: Refer to the osmometer manual for detailed instructions.

Preparation of storage and incubation/recording solution

Timing: 30 min

• 9.
  Prepare 2 L of 5× storage solution.

  • a.
    Weigh all the required powders in a 1 L beaker.
  • b.
    Add approximately 1 L of distilled water and stir with a magnetic stirring bar until completely dissolved.
  • c.
    Transfer the solution into a 2 L graduated cylinder.
  • d.
    Bring the volume up to 2 L with distilled water.

Note: The 5× storage solution can be prepared in advance and stored at 4°C for up to two weeks.

• 10.
  Prepare 500 ml of 1× storage solution.

  Note: Prepare the 1× storage solution on the day of the experiment.

  • a.
    Pour approximately 100 ml of 5× storage solution into a 500 ml graduated cylinder.
  • c.
    Bring the volume up to 500 ml with distilled water.
  • d.
    Add 0.5 ml of 2.5 M CaCl₂ stock solution while oxygenating with carbogen.
  • e.
    Check osmolarity using an osmometer (target: 320–330 mOsm/L).

    Note: Refer to the osmometer manual for detailed instructions.

  • f.
    Fill a storage chamber with the prepared storage ACSF and maintain it at 22–24°C under continuous oxygenation.
• 11.
  Prepare 2 L of 5× incubation/recording solution (see troubleshooting 3).

  • a.
    Weigh all required powders in a 1 L beaker.
  • b.
    Add approximately 1 L of distilled water and stir with a magnetic stirring bar until fully dissolved.
  • c.
    Transfer the solution into a 2 L graduated cylinder.
  • d.
    Bring the volume up to 2 L with distilled water.

Note: The 5× incubation/recording solution can be prepared in advance and stored at 4°C for two weeks.

• 12.
  Prepare 1 L of 1× incubation/recording solution.

  • a.
    Pour approximately 200 ml of 5× incubation/recording solution into a 1 L graduated cylinder.
  • b.
    Bring the volume up to 1 L with distilled water.
  • c.
    Add 1 ml of 2.5 M CaCl₂ stock solution while oxygenating with carbogen.
  • d.
    Check osmolarity with an osmometer (target: 320–330 mOsm/L).

    Note: Refer to the osmometer manual for detailed instructions.

  • e.
    Add 200 μL of 500 mM PTX stock solution while stirring (400–500 rpm) and oxygenating continuously (see troubleshooting 1) until clear solution, to avoid local high-concentration spots.

    Note: Protect the PTX-containing solution (storage bottles and incubation chambers) from light throughout the procedure, as PTX is light-sensitive.

    Note: The final DMSO dilution in the incubation/recording solution is 1:5000.

  • f.
    Fill an incubation chamber with the incubation/recording ACSF.
  • g.
    Warm the solution to 32°C–33°C under continuous oxygenation.

    Note: Keep the incubation/recording solution at 22°C–24°C if recordings need to be done at 22°C–24°C.

Preparation of the vibratome and the tools for cutting

Timing: 10–20 min

• 13.
  Prepare the vibratome for cutting.

  • a.
    Turn on the vibratome and carefully insert a blade into the blade holder.
  • b.
    Connect the Vibrocheck device to the vibratome with the appropriate cable.
  • c.
    Perform the calibration procedure to verify proper vibratome function (0.8–1 mm vibration amplitude and 70 Hz vibration frequency).

    Note: Refer to the vibratome user manual for detailed instructions on the use of Vibrocheck.

  • d.
    Once the calibration is complete, remove the Vibrocheck device and install the vibratome buffer tray.
  • e.
    Turn on the cooling system connected to the buffer tray and set the target temperature to 1°C–2°C.

    Note: Refer to the cooling system user manual for setup and operation.

  • f.
    Half-fill the buffer tray with the cutting solution and start continuous oxygenation with carbogen.
  • g.
    Prepare a slush by filling a 250 ml beaker with the cutting solution and placing it at −150°C (freezer) for approximately 15 min.

    Note: Wear protective gloves when placing the cutting solution in the −150°C freezer and when retrieving it to prevent cold-related injuries.

    Note: The cutting solution should be removed from the −150°C freezer once it reaches a semi-frozen state (approximately half ice, half liquid). Thoroughly mix the solution using a spoon or an electric homogenizer to ensure uniform consistency before use.

• 14.
  Prepare the tools for brain extraction and hippocampal dissection.

  • a.
    Fill a container with crushed ice.
  • b.
    Fill a 10 ml beaker and a small plastic Petri dish with cutting solution, place them on ice, and start oxygenation.
  • c.
    Prepare the following dissection tools: large scissors, small scissors, forceps, a razor blade, a small flat spatula, a small spoon, two small, rounded, and bent spatulas, Whatman filter paper, glue, and a small block of 4% agar.

Preparation of the MEA recording system

Timing: 5–10 min

• 15.
  Connect the MEA System components.

  • a.
    Connect the MEA headstage to the interface board via an iX industrial type B cable.
  • b.
    Connect the interface board directly to the computer via USB 3.0 SuperSpeed.
  • c.
    Connect the MEA table to the computer to enable real-time visualization of slice positioning.

Note: Refer to the MEA System user manual for detailed setup instructions.

• 16.Place the MEA Petri dish into the MEA headstage.

Note: Ensure the black arrow on the dish is oriented upward to align the electrodes correctly with the software display.

• 17.
  Configure the perfusion setup.

  • a.
    Insert the heatable inflow and outflow cannulas into two magnetic perfusion holders and secure them with fixation screws.
  • b.
    Place the magnetic holders on the headstage near the MEA Petri dish.
  • c.
    Position the inflow (heatable) cannula on one side of the chamber with the tip touching the inner wall close to the bottom of the chamber.
  • d.
    Connect the inflow cannula to the temperature controller.
  • e.
    Place the outflow cannula on the opposite side, with its tip touching the inner wall.

Note: Avoid positioning the outflow cannula too deep (excessive suction) or too high (solution overflow).

• 18.
  Connect the perfusion tubing.

  • a.
    Connect 2.06/3.76 mm (ID/OD; color: purple-purple) tubing to the inflow cannula.
  • b.
    Connect 2.54/4.24 mm (ID/OD; color: purple-orange) tubing to the outflow cannula.

Note: This tubing combination enables a higher outflow rate, maintaining a stable solution level in the MEA Petri dish. Other combinations may also be used.

• 19.
  Initiate perfusion.

  • a.
    Secure the tubing onto the peristaltic pump.

    Note: Refer to the peristaltic pump user manual for tubing setup details.

  • b.
    Insert the free end of the inflow tubing into the oxygenated incubation/recording solution.
  • c.
    Set the pump rotation speed to achieve a flow rate of 1–2 mL/min.
  • d.
    Turn on the peristaltic pump.
  • e.
    Verify that the solution is flowing through both inflow and outflow lines.
• 20.
  Configure the temperature controller.

  • a.
    Turn on the temperature controller.

    Note: Start perfusion before heating to prevent overheating of air in the cannula.

  • b.
    Activate the channel connected to the inflow cannula.
  • c.
    Set the target temperature at 35°C to achieve ∼32°C in the MEA dish.

    Note: Use an external thermometer to verify the actual temperature and adjust if needed. Refer to the temperature controller user manual for detailed instructions.

• 21.
  Launch the acquisition software.

  • a.
    Open MEA_Monitor to monitor/check the slice position.
  • b.
    Open MC_Rack for data acquisition.
• 22.
  Check for electrical noise and perfusion artifacts (see troubleshooting 4).

  • a.
    Start real-time signal monitoring in MC_Rack during continuous perfusion.
  • b.
    Adjust the position of inflow and outflow cannula tips in the MEA Petri dish to reduce perfusion-related artifacts.

Note: Electrical artifacts may be caused by air bubbles, vibrations, or unstable fluid levels.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
Picrotoxin	Sigma-Aldrich (Merck)	P1675; CAS: 124-87-8
NaCl	Sigma-Aldrich (Merck)	S9888; CAS: 7647-14-5
NaHCO3	Sigma-Aldrich (Merck)	S8875; CAS: 144-55-8
Saccharose	Sigma-Aldrich (Merck)	S0389; CAS: 57-50-1
Glucose	Sigma-Aldrich (Merck)	G8270; CAS: 50-99-7
KCl	Sigma-Aldrich (Merck)	P9541; CAS: 7447-40-7
NaH2PO4∗1H2O	Sigma-Aldrich (Merck)	S9638; CAS: 7782-85-6
MgCl2∗6H2O	Sigma-Aldrich (Merck)	M9272; CAS: 7791-18-6
MgSO4∗7H2O	Sigma-Aldrich (Merck)	M1880; CAS: 10034-99-8
Agar	Sigma-Aldrich (Merck)	A7002; CAS: 9002-18-0
Cyanoacrylate glue	World Precision Instruments	7341; CAS: 7085-85-0
Experimental models: Organisms/strains
Mouse: C57BL/6J	Charles River	RRID: IMSR_JAX:000664
Mouse: Cx30−/−Cx43fl/fl hGFAP-Cre (cross of 3 mouse lines: Cx30−/−,Cx43 fl/fl, and hGFAP-cre)	Pr. K. Willecke (University of Bonn, Germany)	RRID:IMSR_EM:00323 (Cx30−/−) RRID:IMSR_NM-CKO-200180 (Cx43 fl/fl) RRID:IMSR_JAX:004600 (hGFAP-cre)
Software and algorithms
MC Rack 4.5.1	Multi Channel Systems	RRID: SCR_014955
MEA_Monitor	Multi Channel Systems	v.1.0.5
MC_Data Tool	Multi Channel Systems	v.2.6.15
Neuroexplorer 4	Nex Technologies	RRID: SCR_001818
MATLAB	MathWorks	RRID:SCR_001622
Other
Vibratome	Leica Biosystems	VT1200S
Cooler system	Julabo	F250
MEA system	Multi Channel Systems	MEA2100-120
MEA table	Multi Channel Systems	MEA-VMTC-1
MEA dishes	Multi Channel Systems	120MEA200/30iR-Ti
Heatable perfusion cannula	Multi Channel Systems	PH01
2-channel temperature controller	Multi Channel Systems	TC02
Magnetic perfusion holder for perfusion cannula	Multi Channel Systems	MPH
Peristaltic pump	Gilson	MINIPULS, 3
Tubing for peristaltic pump	Tygon LMT-55	070534-21-ND / SC0021T (purple/purple) 070534-23-ND (purple/orange)
Osmometer	ELITechGroup	VAPRO 5600
Slice anchors	Warner Instruments	64-1419 – SHD-41/10

Materials and equipment

Cutting solution

Reagent	Final concentration (mM)	Amount
NaCl	87	5.1 g
NaHCO3	25	2.1 g
Saccharose	75	25.65 g
Glucose	10	1.8 g
KCl	2.5	0.187 g
NaH2PO4∗7H2O	1	0.14 g
MgCl2∗6H2O	7	1.43 g
CaCl2∗2H2O	0.5	0.2 mL/L
ddH2O	N/A	Up to 1 L
Total	N/A	1 L

Store at 4°C for 2 weeks.

Note: Add CaCl₂ 2.5 M before use and oxygenate.

Storage ACSF (5×)

Reagent	Final concentration (mM) in 1× solution	Amount
NaCl	119	69.54 g
KCl	2.5	1.86 g
MgSO4∗7H2O	1.3	3.2 g
NaH2PO4∗7H2O	1	1.38
NaHCO3	26.2	22.01
Glucose	11	19.82
CaCl2∗2H2O	2.5	1 mL/L (1×)
ddH2O	N/A	Up to 2 L
Total	N/A	2 L

Store at 4°C for 2 weeks.

Note: Add CaCl₂ 2.5 M to the 1× solution before use while oxygenating.

Check osmolarity (320–330 mOsm/L).

Incubation/recording ACSF (5×)

Reagent	Final concentration (mM) in 1× solution	Amount
NaCl	119	69.54 g
KCl	2.5	1.86 g
NaH2PO4∗7H2O	1	1.38
NaHCO3	26.2	22.01
Glucose	11	19.82
CaCl2∗2H2O	2.5	1 mL/L (1×)
ddH2O	N/A	Up to 2 L
Total	N/A	2 L

Store at 4°C for 2 weeks.

Note: Add CaCl₂ 2.5 M to the 1× solution before use while oxygenating.

Check osmolarity (320–330 mOsm/L).

Add PTX (0.1 mL of 500 mM PTX/500 mL solution) while agitating with a magnetic stirring bar.

CRITICAL: Picrotoxin: fatal if swallowed (H300). Wash hands thoroughly after use. Wear gloves while manipulating it. Use in a chemical fume hood.

• •MEA dishes: the standard layout (12 × 12 grid, 200 μm inter-electrode distance, 30 μm electrode diameter) has been used for this protocol and in Dossi et al.¹ Other different electrode layouts are available according to the study’s objectives.

Step-by-step method details

Dissection of the hippocampus

Timing: 3–5 min

This section provides step-by-step guidelines for the dissection of the mouse hippocampus for subsequent slicing.

• 1.Sacrifice the mouse by cervical dislocation.
• 2.
  Dissect the brain.

  • a.
    Cut the head off with scissors and place it into a Petri dish covered with Whatman filter paper.
  • b.
    Remove the scalp with a small pair of scissors.
  • c.
    Cut the skull along the midline from the caudal to the rostral part of the brain (Figure 1Ai).
  • d.
    With forceps, remove the skull by pulling the two halves laterally (Figure 1Aii).
  • e.
    Remove the cerebellum with a blade (Figure 1Aiii).
  • f.
    Separate the two hemispheres with a blade (Figure 1Aiv).
  • g.
    Transfer the two hemispheres into a small beaker containing cold oxygenated cutting ACSF using a small spatula (Figure 1Av).

CRITICAL: Step 2 should be completed within 1–2 min to preserve optimal tissue viability.

• 3.
  Dissect the hippocampus.

  • a.
    After 1 min equilibration, place one hemisphere on a dry Whatman filter paper, with the medial surface up (Figure 1Avi-vii).
  • b.
    Remove the diencephalon with two small spoons; the hippocampus is now visible (Figure 1Aviii).
  • c.
    Dissect the hippocampus using a small spoon, beginning at the fimbria, which appears as a white structure (Figure 1Aix-x).
  • d.
    Transfer the hippocampus into a small Petri dish filled with the same cold, oxygenated cutting ACSF.
  • e.
    Do the same for the second hemisphere.

CRITICAL: Step 3 should be completed within 1–2 min to preserve optimal tissue viability.

Figure 1.

Preparation of mouse hippocampal slices

Schematic representation (top row) and corresponding pictures (bottom rows) illustrating the main steps involved in (A) brain extraction and hippocampal dissection and (B) hippocampal slice preparation and MEA recording.

Preparation of hippocampal slices

Timing: ∼15 min

This section provides step-by-step guidelines for cutting transversal slices of the mouse hippocampus for subsequent electrophysiological recordings.

• 4.Cut a small block of agar.
• 5.
  Transfer the hippocampi onto the agar block.

  • a.
    Take one hippocampus with a small spoon and place it on one flat surface of the agar block with the alveus side oriented upward and the ventral side aligned along the edge of the agar block (Figure 1Bi).
  • b.
    Do the same with the second hippocampus (Figure 1Bii).
  • c.
    With Kimtech wipes, gently and carefully dry all excess liquid around the two hippocampi to ensure a secure attachment to the agar.
• 6.
  Cut transverse dorsal hippocampal slices.

  • a.
    Put 2–3 drops of glue on the vibratome tissue support and spread them with the tip.
  • b.
    Glue the agar block with the two hippocampi, with their ventral side toward the glue (Figure 1Biii).
  • c.
    Place the tissue support with the hippocampi/agar into the vibratome buffer tray.
  • d.
    Fill the buffer tray with ice-cold oxygenated cutting ACSF and oxygenate continuously.
  • e.
    Cut 400 μm-thick transverse slices at low speed (0.04–0.05 mm/sec) and 70 Hz vibration frequency.

Storage and incubation of hippocampal slices

Timing: variable (minimum 1 h 45 min)

This section describes how to handle hippocampal slices for recovery after slicing and for incubation to induce neuronal network activity.

• 7.
  Let the slices recover after slicing.

  • a.
    Transfer the slices into the storage chamber with a plastic Pasteur pipette (Figure 1Biv).
  • b.
    Let them rest for 45 min at 22°C–24°C while continuously oxygenating.
• 8.
  Incubate the slices in incubation/recording ACSF to induce population activity.

  • a.
    Transfer the slices into the incubation chamber with a plastic Pasteur pipette.

    Note: Verify the temperature of the incubation chamber before transferring the slices.

  • b.
    Let them rest for 1 h under continuous oxygenation before recording.

    Note: A 1-h incubation is required to ensure stable and reliable neuronal network activity.

Recording of network activity with MEA

Timing: 3–4 h, variable (depending on the number of slices to record)

This section describes how to record neuronal network activity in hippocampal slices using the MEA technique.

• 9.
  Transfer and position a slice into an MEA petri dish (Figure 1Bv).

  • a.
    Take a slice with a plastic Pasteur pipette (tip diameter: 3.4 mm).

    Note: Gently aspirate the hippocampal slice with the plastic pipette to avoid tissue damage. If necessary, for larger slices (i.e., whole brain slices), trim approximately 1.5 cm from the pipette tip to widen the opening and facilitate safe handling.

  • b.
    Transfer it into the MEA petri dish placed in the MEA headstage.

    Note: While no specific orientation of the slice is strictly required, maintaining a consistent orientation across recordings, such as positioning hippocampal slices with CA1 facing upward and the dentate gyrus downward, can facilitate electrode selection for analysis and improve data standardization.

    Note: Coating is not necessary for acute slice recording. According to the manufacturer, MEAs should be stored at 4°C filled with distilled water to maintain the hydrophilicity of the recording surface.

  • c.
    Remove the excess ACSF with the same plastic Pasteur pipette by aspirating along the edge of the MEA Petri dish.

    Note: Remove as much solution as possible, leaving only a thin film to keep the slice moist without submerging it.

    Note: While removing the excess ACSF, make the slice move toward the center of the MEA dish as much as possible to reduce further adjustment of its position.

  • d.
    Put a slice anchor (round anchors with 13.3 mm diameter; Type 316 stainless steel with Lycra threads; 100 μm thick threads at 1 mm spacing) on top of the slice to avoid slice movement during recording.

    Note: Wet the slice anchor before placing it on the slice.

    Note: Place the anchor on the slice very gently to avoid damage.

    Note: Commercial slice anchors are available in various sizes and weights to accommodate different types of slices.

  • e.
    Add some drops of recording ACSF into the MEA dish with a pipette.
  • f.
    Start perfusing the slice with oxygenated recording ACSF at 1–2 mL/min.

    Note: Maintain enough ACSF in the MEA chamber to cover the slice completely.

  • g.
    Adjust the position of the slice into the MEA dish using the MEA_Monitor software.

    Note: Gently reposition the slice anchor with forceps to adjust the slice position.

  • h.
    Take a picture of the slice position.

    Note: This step is essential for hippocampal slices to know which MEA electrodes are located under the different hippocampal regions.

• 10.
  Record neuronal network activity.

  • a.
    Set the recording folder and name.
  • b.
    Set the recording parameters (second-order high-pass filter at 1 Hz to prevent baseline drift with the upper cut-off frequency at 3.3 kHz; 10 000 Hz sampling rate).
  • c.
    Start continuous activity display to check for activity in the hippocampal areas of interest (see troubleshooting 4).

    Note: Refer to MC_Rack manual for more detailed information on how to set the recording protocol.

  • d.
    Start recording.

    Note: For a simple recording of neuronal population activity, a 20-min recording is enough (10 min stabilization + 10 min for analysis). To monitor activity stabilization and baseline noise, enable the long-term display and set the visualization window to 5–10 min. This allows continuous inspection of event frequency and baseline noise or drifts over time.

    Note: Check the recording to confirm population activity and assess the stability and regularity of its frequency.

  • e.
    Stop recording.
  • f.
    Remove the slice anchor with forceps.
  • g.
    Remove the slice with a plastic Pasteur pipette.

    Note: Do not touch the bottom of the MEA Petri dish center with the plastic Pasteur pipette to avoid electrode damage.

  • h.
    Repeat steps 9–10 with other slices.

Data analysis

Timing: ∼2 h (per recording session, 7–8 slices)

This step enables the analysis of population activity recorded in hippocampal slices.

Note: the following analysis starts with “.mcd “ recording files obtained by using MC_Rack recording software (as described in Dossi et al.¹). See troubleshooting 5 when the Multi Channel Experimenter software is used to record and “.msrd” files are obtained.

• 11.
  Selection of the electrodes of interest for the analysis.

  • a.
    Open the MC_Rack software.
  • b.
    Add a “Replayer” by clicking the relative icon (Figure 2A).
  • c.
    Select the file to replay (Figure 2A, red arrow).
  • d.
    Add a Longterm display (Figure 2B, red).
  • e.
    Do a replay to look at the activity and identify the electrodes of interest (for this work, the hippocampal CA1 region).
  • f.
    Add a Butterworth 2^nd order low-pass filter at 499 Hz prior to down-sampling at 1000 Hz (Figure 2B, green).

    Note: This step serves as an anti-aliasing filter to eliminate frequency components above the Nyquist limit (500 Hz) and preserve signal integrity during downsampling.

  • g.
    Replay the file with the recorder function to create a new .mcd file containing the filtered and down-sampled traces of the selected electrodes (Figure 2C, red).
• 12.
  Conversion of the file for MATLAB.

  • a.
    Open the MC_Data Tool software.
  • b.
    Select the newly created .mcd file by clicking the “Open MC_Rack file” icon. All the file information will be displayed (Figure 3A, red).
  • c.
    Convert it into a .txt file by clicking the relative icon (Figure 3A, green); at this point, it is possible to convert all the electrode traces into the file or select a subset of electrodes (Figure3B, green arrow).
  • d.
    Choose the output folder and press save (Figure 3B, red arrows).
• 13.
  Population activity analysis with MATLAB.

  • a.
    Open MATLAB.
  • b.
    Select the folder containing the .txt files to analyze as “Current Folder”.
  • c.
    Load the .txt file previously created and plot all the electrode traces.

    % Load the data: COPY and PASTE the name of the file with .txt

    data = readmatrix('slice_name.txt');

    % Extract time and traces

    time = data(:,1) / 1000; % Convert ms to seconds

    traces = data(:,2:end); % Extract all traces

    num_traces=size(traces,2);

    % Plot each trace in a separate subplot

    figure;

    for i = 1:num_traces

     subplot(num_traces,1,i);

     plot(time, traces(:,i),'b');

     xlabel('Time (s)');

     ylabel('Signal (uV)');

     title(['Trace', num2str(i)]);

     grid on;

    end

    sgtitle('Extracellular Electrode Recordings'); % common title for the figure

    Note: Copy and paste the file name within the parentheses of the “readmatrix” function (in this example, “slice_name.txt”).

    Note: Before plotting the traces, separate time and trace arrays are created; the time array is in sec (the time column of the data array is divided by 1000, as 1000 Hz is the sampling frequency).

    Note: The x-axis, y-axis, and figure are given default titles - Time (s), Signal (μV), and Extracellular electrode recordings- which can be modified as preferred.

  • d.
    Specify the parameters for the detection of bursts and the quantification of burst frequency and duration.

    % Parameters

    threshold_factor = 1.5; %Bursts defined as 1.5× standard deviation

    min_burst_duration = 0.02; %Minimum burst duration in seconds

    sampling_rate = 1000;

    merge_burst_gap = 0.5; %sec; Merge bursts if gap is less than 500ms

    total_recording_time = time(end) - time(1); total_recording_time_minutes = total_recording_time/60;

    % Initialize arrays for burst storage

    burst_data = cell(num_traces, 1);

    Note: The threshold_factor indicates the threshold above baseline to detect a burst and is expressed as a multiple of the standard deviation on the electrode trace. The min_burst_duration (in seconds) specifies the minimum duration a burst must have to be detected as a valid burst. The merge_burst_gap defines the minimum interval (in seconds) required to separate two bursts; bursts closer than this value are combined into a single burst. The total_recording_time and total_recording_time_minutes indicate the recording duration in seconds and minutes, respectively.

    Note: The values of the parameters used in this example have been set based on our experimental dataset. They can be modified to optimize burst detection according to the experimental data by testing the detection on several recordings.

  • e.
    Detect burst phases for each trace.

    %% Step 1: Detect Burst phases for Each Trace

    figure;

    for i = 1:num_traces

     signal = traces(:,i); % Get current trace

     mean_signal = mean(signal);

     std_signal = std(signal);

     % Define positive and negative burst thresholds

     pos_threshold = mean_signal + threshold_factor ∗ std_signal;

     neg_threshold = mean_signal - threshold_factor ∗ std_signal;

     % Detect regions exceeding either threshold

     burst_mask = (signal>pos_threshold)|(signal<neg_threshold);

     % Identify burst phase start and end points

     burst_diff = diff([0; burst_mask; 0]); %Add 0 padding at ends

     burst_starts = find(burst_diff == 1);

     burst_ends = find(burst_diff == -1) - 1;

     % Skip trace if no bursts are detected

     if isempty(burst_starts) || isempty(burst_ends)

     disp(['No bursts detected for trace ', num2str(i)]);

     burst_data{i} = table([], [], 'VariableNames', ... {'Burst_Start', 'Burst_End'});

     continue;

     end

     % Ensure equal number of starts and ends

     min_length = min(length(burst_starts), length(burst_ends));

     burst_starts = burst_starts(1:min_length);

     burst_ends = burst_ends(1:min_length);

     % Convert to column vectors

     burst_starts = burst_starts(:);

     burst_ends = burst_ends(:);

     % Compute burst phase durations

     burst_durations=(burst_ends-burst_starts)/sampling_rate;

     % Filter burst phases by duration

     valid_bursts = burst_durations > min_burst_duration;

     % Convert burst phase indices to time values

     burst_start_times = time(burst_starts(valid_bursts));

     burst_end_times = time(burst_ends(valid_bursts));

     % Store burst phase data in a table for this trace

     burst_data{i} = table(burst_start_times, burst_end_times, ...

     'VariableNames', {'Burst_Start',... 'Burst_End'});

     % Plot original trace using subplot

     subplot(num_traces, 1, i);

     plot(time, signal, 'b');

     hold on;

     % Mark detected burst phases

     valid_burst_starts = burst_starts(valid_bursts);

     valid_burst_ends = burst_ends(valid_bursts);

     for j = 1:length(valid_burst_starts)

     burst_range = valid_burst_starts(j):valid_burst_ends(j);

     plot(time(burst_range), signal(burst_range), 'r',... 'LineWidth', 2);

     end

     % Labels

     xlabel('Time (s)');

     ylabel('Signal (μV)');

     title(['Trace ', num2str(i)]);

     grid on;

    end

    hold off;

    Note: As bursts are complex (biphasic, with positive and negative phases), both a positive and a negative threshold are calculated as multiples of the signal SD. Regions above the positive and below the negative thresholds are detected, and burst starts and ends are identified. Data are stored in the burst_data array.

    Note: To check burst detection, a plot is added where detected bursts are highlighted in red on the traces.

  • f.
    Merge burst phases to get complete biphasic bursts.

    %% Step 2: Merge Burst phases for each trace

    merged_burst_data = cell(num_traces, 1);

    min_merged_burst_duration = 0.400 %in seconds

    burst_frequencies = zeros(num_traces, 1); % Store burst

    % frequency (bursts>min_merged_burst_duration) for each trace

    avg_burst_durations=zeros(num_traces,1); % Store avg burst

    % duration (bursts>min_merged_burst_duration) for each trace

    for i = 1:num_traces

     if isempty(burst_data{i}.Burst_Start)

     disp(['No bursts detected to merge for trace ',... num2str(i)]);

     merged_burst_data{i} = table([], [], 'VariableNames',... {'Merged_Burst_Start', 'Merged_Burst_End'});

     continue;

     end

     % Extract burst start and end times

     merged_starts = burst_data{i}.Burst_Start;

     merged_ends = burst_data{i}.Burst_End;

     % Ensure equal length

     min_length = min(length(merged_starts),length(merged_ends));

     merged_starts = merged_starts(1:min_length);

     merged_ends = merged_ends(1:min_length);

     % Fix merging loop

     j = 1;

     while j < length(merged_starts)

     if j + 1 > length(merged_starts);

     end

     if merged_starts(j+1) - merged_ends(j) < merge_burst_gap

      % Merge bursts

      merged_ends(j) = merged_ends(j+1);

      merged_starts(j+1) = [];

      merged_ends(j+1) = [];

     else

      j = j + 1;

     end

     end

     % Compute merged burst durations

     merged_durations = merged_ends - merged_starts;

     % Store merged burst data

     merged_burst_data{i} = table(merged_starts, merged_ends, merged_durations,...'VariableNames', {'Merged_Burst_Start',... 'Merged_Burst_End', 'Merged_Burst_Duration'});

     % Filter only bursts > min_merged_burst_duration

     long_bursts = merged_durations... (merged_durations>min_merged_burst_duration);

     % Compute burst frequency (bursts/min) for long_bursts

     burst_frequencies(i) = length(long_bursts)/total_recording_time_minutes;

     % Compute average burst duration for long_bursts

     if ∼isempty(long_bursts)

     avg_burst_durations(i) = mean(long_bursts);

     else

    avg_burst_durations(i) = NaN; % No long_bursts

     end

    end

    %% Step 3: Plot Burst phases and Merged Bursts

    figure;

    for i = 1:num_traces

     signal = traces(:,i); % Get the signal trace

     subplot(num_traces, 1, i);

     plot(time, signal, 'b'); % Plot original trace

     hold on;

     % Mark detected bursts in RED

     for j = 1:height(burst_data{i})

     burst_range = find(time >= burst_data{i}.Burst_Start(j) & ... time <= burst_data{i}.Burst_End(j));

     plot(time(burst_range), signal(burst_range), 'r',... 'LineWidth', 2);

     end

     % Mark only merged long_bursts in GREEN

     for j = 1:height(merged_burst_data{i})

     if merged_burst_data{i}.Merged_Burst_Duration(j) > ... min_merged_burst_duration

     merged_range = find(time >=...

     merged_burst_data{i}.Merged_Burst_Start(j) & ...

     time <= merged_burst_data{i}.Merged_Burst_End(j));

     plot(time(merged_range), signal(merged_range), 'g',... 'LineWidth', 2);

      end

     end

      % Labels

     xlabel('Time (s)');

     ylabel('Signal (μV)');

     title(['Trace ', num2str(i)]);

     grid on;

    end

    hold off;

    Note: Positive and negative burst phases detected in the previous step are merged if they are closer than 500 ms (defined by the merge_burst_gap in the parameters section) to reconstruct the complete biphasic bursts.

    Note: To calculate burst frequency (bursts/min) and average burst duration for each trace, only bursts longer than 400 ms (min_merged_burst_duration) are considered (called long_bursts in the script) to exclude small, noisy events that may be detected with a 1.5 SD threshold.

    Note: The merged_burst_data array contains burst start, end, and duration for each burst of each trace. The burst_frequencies and avg_burst_durations tables contain the computed burst frequency and average long burst durations of all traces in the analyzed file.

    Note: A plot is added at the end of this step to display the original trace with detected burst phases in red and merged bursts in green.

    Note: The value of the min_merged_burst_duration used in this example has been set based on our experimental dataset. It can be modified to optimize burst detection according to the experimental data by testing the detection on several recordings.

  • g.
    Save and display the results.

    %% Step 4: Save Burst Statistics to CSV

    burst_stats = table((1:num_traces)', burst_frequencies, ... avg_burst_durations,'VariableNames', {'Trace_Number',... 'Burst_Frequency_per_min','Avg_Burst_Duration_over_400ms_sec'});

    writetable(burst_stats, 'burst_statistics.csv');

    disp('Burst statistics saved to burst_statistics.csv');

    %% Step 5: Display Results

    disp('Burst Frequencies (Bursts per minute) for Each Trace:');

    for i = 1:num_traces

    fprintf('∖nTrace %d: %.2f bursts/min∖n', i, ... burst_frequencies(i));

    end

    disp('Average Burst Durations for Each Trace:');

    for i = 1:num_traces

     fprintf('∖nTrace %d: %.3f sec∖n', i,avg_burst_durations(i));

    end

    Note: A “.csv” file containing the average burst frequency and duration for each electrode is saved.

Figure 2.

Pre-processing of the recordings with MC_Rack

(A) Selection of the recordings to replay using the Replayer function (highlighted in red).

(B) Display of the selected recording using the Longterm Display function (red), and application of a low-pass filter at 499 Hz with downsampling at 1000 Hz via the Filter function (green).

(C) Selection of the electrodes of interest and saving of a new file using the Recorder function (red), containing only the low-pass-filtered and downsampled traces from the selected electrodes.

Figure 3.

File conversion with MC_DataTool

(A) Selection of the file to convert (highlighted in red; prepared with MC_Rack, see Figure 2) and of the desired output format (txt in this case; green).

(B) Selection of electrode traces to convert (from those previously selected with MC_Rack; green) and specification of the output folder (red).

Expected outcomes

This protocol describes the recording and analysis of neuronal network bursting activity in acute mouse hippocampal slices, as previously described in our original publication.¹ Incubation of slices in magnesium-free ACSF containing PTX reliably induces spontaneous and recurrent bursting activity within the hippocampal network. Under optimal slice preparation conditions, ∼80% of the slices develop bursts upon this treatment.

Slice activity recordings are performed using MEAs, which enable the simultaneous recording of activity across different hippocampal subregions (dentate gyrus, CA3, and CA1). Depending on the experimental aim, data can be extracted from all the electrodes covered by the slice or selectively from a subset of them located in specific regions (i.e., CA1 or CA3; Figure 4A).

Figure 4.

Burst detection in mouse hippocampal slices

(A) Left, picture of an acute mouse hippocampal slice placed on a MEA. Electrodes selected for analysis are marked by yellow-filled circles. DG: dentate gyrus. Scale bar: 200 μm. Right, schematic of the analysis pipeline: electrode selection is performed in parallel with low-pass filtering at 499 Hz and downsampling at 1000 Hz. The file is then converted and further analyzed using MATLAB.

(B) Representative traces recorded from the electrodes highlighted in panel A. Left, raw signal traces. Middle, same traces after burst phase detection, with detected phases marked in red. Right, traces after burst merging, with merged bursts indicated in green. Scale bars: 30 s, 200 μV.

(C) Magnified view of the trace segments outlined by the yellow box in panel B. In the middle trace, blue dotted lines represent the positive and negative thresholds used for burst phase detection. Detected burst phases and merged bursts are shown in red and green. Scale bars: 5 s, 200 μV.

(D) Left, representative recordings of bursting activity in WT (top) and Cx-deficient (bottom) mice. Detected bursts are marked in green. Scale bars: 10 s, 100 μV. Right, quantification of burst frequency and duration in WT (white bars) and Cx-deficient (black bars) slices (burst frequency: p = 0.0006, Mann-Whitney test; burst duration: p = 0.0122, unpaired t-test; n = 7 slices from 3 and 5 mice for WT and Cx-deficient). Data are represented as mean ± SEM. Asterisks indicate statistical significance (∗, p < 0.05; ∗∗∗, p < 0.001).

For quantitative analysis, we provide a MATLAB script that allows burst detection and quantification (burst frequency and duration) for each selected electrode (Figures 4B and 4C). In our original manuscript, we applied this protocol to compare bursting activity in WT mice and mice deficient for astroglial connexins 30 and 43 (Cx-deficient mice). We observed that Cx-deficient mice displayed altered bursting properties, characterized by increased burst frequency and decreased burst duration (Figure 4D).

Moreover, this protocol is also applicable for evaluating the effects of pharmacological or molecular manipulations on neuronal network activity. We have indeed applied it to investigate the impact of KCNQ potassium channel modulators, and we have shown that burst dynamics are changed after the inhibition or activation of these channels.¹ Inhibition of KCNQ channels with XE-991 in WT mice mimicked the Cx-deficient phenotype, with more frequent and shorter bursts, while KCNQ channel activation with retigabine in Cx-deficient mice rescued the WT bursting pattern.¹

This protocol can also be adapted to monitor neuronal network activity across distinct brain regions (e.g., cerebral cortex, cerebellum) in mice, other experimental animal models, and resected human brain tissues. In these contexts, only minor modifications are required to optimize parameters for slice maintenance and electrophysiological recording, including adaptation of storage and recording temperatures, incubation conditions, and perfusion flow rates.

Limitations

In the execution of this protocol, it is essential to master the slice preparation. The successful induction of bursting activity indeed relies on optimal tissue viability, which depends on the timing and precision of brain dissection and slice preparation. Deviations in temperature, reduction of oxygenation, or incorrect solution osmolarity during these steps can reduce slice responsiveness to burst induction or induce irregular and unstable bursting patterns.

Under standard conditions, approximately 80% of slices exhibit bursting network activity after incubation in ACSF devoid of magnesium and supplemented with PTX. However, variability across animals or experimental batches may affect reproducibility. Attention must be paid during the preparation of the PTX stock solution, as a high concentration in DMSO is required, which can lead to solubility issues and the risk of precipitation that can affect the final incubation solution’s efficacy.

Furthermore, mechanical artifacts can be introduced by the perfusion system (e.g., air bubbles or unstable fluid levels) and can compromise the quality of recordings. Thus, fine-tuning of the perfusion system is required for proper recordings that can be easily analyzed.

Finally, the MATLAB-based analysis, which allows automated burst detection, depends on preset parameters (e.g., threshold factor, min burst duration, merge burst gap) that require adjustments according to the experimental data to avoid false positives or omissions, especially when applying the protocol to other brain regions or experimental models.

Troubleshooting

Problem 1

Problems of solubility when preparing the 500 mM PTX stock solution in DMSO (related to the “before you begin” section - step 2) and the PTX-containing incubation/recording solution (related to the “before you begin” section - step 12).

PTX is not water-soluble, therefore, the stock solution must be prepared in DMSO. According to the supplier (Tocris), PTX has a maximum solubility of 20 mM in DMSO. However, this concentration is not enough for our experimental needs, as we require a final PTX concentration of 100 μM, and DMSO must be diluted at least 1:1000 to avoid adverse effects on slice viability and neuronal activity. For this reason, we prepare a 500 mM PTX stock solution in DMSO. This high concentration may present challenges both in fully dissolving PTX and when adding it to the incubation/recording solution.

Potential solution

• •To address solubility issues with the 500 mM PTX stock solution, weigh the appropriate amount of PTX in a small beaker, add DMSO, and stir the mixture using a magnetic stirring bar for several minutes. If the compound does not fully dissolve, gently warm the solution while continuing to stir to facilitate dissolution.
• •To prevent precipitation, add PTX to the incubation/recording ACSF gradually while continuously agitating the solution with a magnetic stirring bar.
• •As an alternative, pre-dilute PTX in 1 mL ACSF, then add slowly to 1 L ACSF while stirring.

Problem 2

Alternative solution to prepare hippocampal slices (related to the “before you begin” section - step 7).

This protocol utilized a saccharose-based ACSF for brain dissection and slice preparation. Alternatively, when working with young animals (younger than 2–3 weeks), a standard Ca²⁺- and Mg²⁺-containing storage solution can be used for both slice preparation and storage.

Potential solution

In this case, prepare a larger volume of storage ACSF and chill a portion of it (∼500 mL) for use during brain dissection and slice preparation.

Problem 3

Alternatives to induce neuronal population activity in slices (related to the “before you begin” section - step 11).

This protocol employed a Mg²⁺-free ACSF supplemented with PTX to induce neuronal network activity in mouse hippocampal slices. Alternatively, neuronal activity can be triggered using a Mg²⁺-free ACSF containing elevated KCl (6 mM) instead of PTX, which enhances tissue excitability without directly targeting ion channels or receptors.

Potential solution

In this case, instead of adding PTX in the incubation/recording solution, increase the KCl concentration to 6 mM by adding an appropriate amount of 1 M KCl stock solution.

Problem 4

Electrical artifacts caused by continuous perfusion during the recordings (related to the “before you begin” section - step 22 and step 10).

MEA recordings of acute hippocampal slices are performed under continuous perfusion with incubation/recording ACSF. This is achieved using a perfusion cannula for solution inflow and a needle connected to tubing for outflow, directing the solution through the MEA Petri dish. The inflow and outflow tubing are configured such that the outflow rate slightly exceeds the inflow, maintaining a stable fluid level in the recording chamber. In this configuration, the outflow tubing continuously draws a mixture of liquid and air bubbles. If the bubbles are too large or the tubing alternately draws large bubbles and liquid, this can lead to electrical artifacts during recordings.

Potential solution

• •Ensure that the tubing connected to the peristaltic pump has different internal diameters, with the outflow tubing having a larger diameter than the inflow.
• •In the MEA Petri dish, position the tubing ends so they are in contact with the inner wall of the recording chamber.
• •For the outflow, cut the tip of the tubing at an angle to create a beveled, beak-like shape; this facilitates stable positioning against the inner wall of the MEA dish and helps maintain consistent flow and continuous uptake of small air bubbles mixed with liquid.

Problem 5

Alternatives for the recording system and software (related to steps 10–12).

This protocol used the MEA2100-120 recording system and MC_Rack software by Multichannel Systems (Germany) to record neuronal network activity. The recordings were saved as “.mcd” files and analyzed using the described custom pipeline based on the MC_Data Tool (Multichannel Systems) and MATLAB. Alternatively, the Multi Channel Experimenter software can also be used for activity recording.

Potential solution

In this case, for further processing, use Multi Channel Analyzer to select the electrodes of interest and apply trace filtering, and Multi Channel DataManager to convert “.msrd” files into ASCII format (.csv) for analysis.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Nathalie Rouach (nathalie.rouach@college-de-france.fr).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by the technical contact, Elena Dossi (elena.dossi@college-de-france.fr).

Materials availability

This study did not generate new, unique reagents.

Data and code availability

The original/source data used to write this protocol are included in Dossi et al.¹ The code used to analyze the data is available from the corresponding author/technical contact on request.

Acknowledgments

The authors thank the ‘‘Neuroglial Interactions in Cerebral Physiology and Pathologies’’ laboratory for the discussions on this work. This work was funded by the European Research Council (Consolidator Grant #683154) (N.R.); the European Union’s Horizon 2020 research and innovation programme (Marie Sklodowska-Curie Innovative Training Networks, grant #722053, EU-GliaPhD) (N.R.); Ligue Francaise contre l’Epilepsie (E.D.); and Agence Nationale de la recherche (ANR-22-CE16-0031-01) (E.D.).

Author contributions

E.D. developed the protocol. E.D., R.D., and N.R. wrote the protocol.

Declaration of interests

The authors declare no competing interests.