Surgical Protocol for a Large, Resealable Cranial Window Enabling Longitudinal, Multi-Modal Electrophysiology Recordings of Mouse Default Mode Network

The Default Mode Network or the DMN is a crucial, large-scale network, implicated in a range of cognitive functions and neuropsychiatric disorders such as depression. Studying the DMN's complex dynamics in animal models provides invaluable insights into its function in both healthy and pathological states. However, a significant challenge has been performing stable, long-term, and large-scale electrophysiological recordings from the multiple deep and distributed nodes, the DMN, in awake, behaving mice.

Here, we present a novel two-phase surgical protocol developed in the lab of Eero Castren. This technique creates a large, durable, and resealable cranial window enabling repeated longitudinal recordings from over 1, 000 electrode channels simultaneously. This is achieved by combining surface-level micro-electrocorticography, micro-ECoG, with two Neuropixels probes, allowing unprecedented access to the Default Mode Network.

This video will provide a step-by-step guide to this robust surgical procedure from animal preparation and headplate implantation to the creation of the chronic cranial window, ensuring high-quality, long-term data acquisition for network neuroscience research. All procedures shown were approved by the National Experiment Board in Finland and comply with the European directive for the protection of animals used for scientific purposes. Phase 1, Animal Preparation and Headplate Implantation.

To perform this procedure, begin by preparing all necessary surgical materials and equipment. Anesthetize the mouse with 4%isoflurane and maintain anesthesia at 1.5 to 2.5%with an oxygen flow rate of 0.5 liters per minute. Shave the fur from the top of the head and place the animal on a controlled heating pad equipped with an internal temperature sensor calibrated to hold body temperature at 37 degrees Celsius throughout the procedure.

Apply carbomer eye ointment to prevent corneal drying. Secure the mouse in the stereotactic frame so that the skull lies flat and administer preoperative analgesics and anti-inflammatory agents via subcutaneous injections, Carprofen, Buprenorphine, and Dexamethasone. disinfect the shaved area with povidone-iodine solution and inject lidocaine-epinephrine solution as a local anesthetic under the scalp skin.

Make a small transverse incision at the earline and enlarge the incision progressively to fully expose the top of the skull surface. Carefully clean the exposed skull with acetone until all visible periosteum has been removed to ensure strong adhesion of the implant. Use a blunt, circular scalpel blade to remove any residual periosteal and connective tissue fragments that the acetone wash did not fully resolve.

With the stereotactic device, mark a 4 millimeters by 7.6 millimeters rectangular area on the skull relative to bregma. The rectangle should extend two millimeters laterally from bregma on both sides, three millimeters rostrally, and 4.6 millimeters caudally. Next, mark the two intracranial probe insertion sites on the right hemisphere.

Mark the rostral probe site at 1.66 millimeters anterior and 1.95 millimeters lateral from bregma, and the caudal probe site at 2.2 millimeters posterior and 1.9 millimeters lateral from bregma. Next, engrave a crisscross pattern over all exposed bone surfaces outside the designated window area using a number 11 surgical blade, and produce a shallow but defined grooves approximately 0.2 to 0.4 millimeters deep to form a uniform grid of approximately one by one millimeter squares. Create a fine glue applicator by fitting a sterile 30G needle to the tip of a cotton swab and bend the needle into a V-shape.

Dispense cyanoacrylate glue into a disposable plastic weighing boat and dip the applicator into the glue reservoir. Deposit glue over every exposed and engraved bone surface, applying no more than a single deposit per site so that the coverage is thorough, but never excessive. Allow the glue to dry for seven minutes before proceeding.

For implanting the reference socket, select the drilling site over the left cerebellum position to avoid visible superficial vessels. Drill the pilot hole in five to ten second bursts using a round steel burr at 20, 000 to 25, 000 rounds per minute until the hole becomes more translucent, revealing a light pink tint. Insert the gold-plated reference socket into the pilot hole, secure it with cyanoacrylate glue, and reinforce the junction with UV-curable dental cement.

Cure the dental cement with a LED curing pen light for one minute. Next, place a small quantity of UV-curable dental cement on the apex of the exposed rostral bone and lay the headplate under the skull so that one edge rests on the dental cement scaffolding, cured around the reference socket, and with the opposing edge resting on the fresh rostral cement dab. Make sure the headplate is centered and leveled.

Reinforce the structure by applying dental cement around the base of the head plate and circling the engraved bone and the reference socket until a continuous, sealed enclosure is formed around the perimeter of the future cranial window. Lastly, cured the dental cement with the LED curing pen light for one minute. Once the cement has fully cured, allow the mouse to wake up from the anesthesia.

The first phase is now complete. Allow the mouse to recover for at least 48 hours before proceeding to the next phase. Phase 2, Chronic Cranial Window Creation.

The second phase requires the same tools and medications as the first phase except for the local anesthetic. Additionally, this phase requires a sterile, thin PDMS membrane, cold, artificial cerebrospinal fluid kept on ice, an elastomer silicone sealant, and a custom 3D printed protective cap. At least 48 hours after the first surgery, re-anesthetize the mouse with isoflurane, place it on the heating pad on the stereotactic frame, and administer the preoperative drugs as in the first phase apart from the lidocaine-epinephrine local anesthetic.

Begin thinning the bone with the dental drill along the rectangle outline marked in Phase 1, starting at 25, 000 rounds per minute with the drilling device held in a 90 degree angle relative to the bone surface, and perform one or two slightly deeper passes along the rectangle edges to establish an initial cutting groove. Drill shallow grooves into the bone and dental cement adjacent to the cranial window at the two probe insertion coordinates. These grooves serve as the persistent visual landmarks that remain visible after the bone flap is removed and are used to reposition the intracranial probes reproducibly in subsequent electrophysiological sessions.

Apply ice-cold, sterile artificial cerebrospinal fluid regularly throughout the drilling to prevent thermal damage to the underlying cortex and to minimize bleeding. Reduce the drill speed progressively to 20, 000 rounds per minute as the bone thins. Continue drilling along the rectangle perimeter until the bone within the outline is approximately 90%thinned.

Do not drill completely through the skull. Switch to the fine curved dura hook. Insert the tip under the edge of the thin bone with extreme caution, and slide the hook along the window perimeter, carefully detaching the bone flap from the surrounding skull and underlying tissue.

Lift the successfully detached bone flap carefully with the dura hook in a posterior to anterior direction to approximately 35 degree above the skull surface. Grasp the lifted edge with the forceps and sway gently left and right until the plate releases completely. Gently clear the exposed area of coagulated blood with repeated washes of ice-cold ACSF.

Place a single sterile precast PDMS membrane directly under the dural surface once bleeding has subsided. When correctly sized, it will cover the window precisely and adhere passively to the dura, holding it flat against the brain. Seal the craniotomy by applying silicone sealant.

Fill every crevice between the dental cement enclosure and inner rim of the head plate, fully surrounding and overlapping the edges of the BDMS membrane. Affix a custom 3D printed protective cap onto the headplate with a small quantity of cyanoacrylate glue to protect the window between recording sessions. The mouse is now ready for recovery and should be transferred to its home cage for individual housing to prevent the damage to the implant.

A successful surgery results in a clear, transparent window over the cortex, with visible vasculature and minimal signs of inflammation or infection. This clarity can be maintained for over 21 days, enabling long-term longitudinal studies. Raw electrophysiological signals recorded through the chronic window retained high quality across the longitudinal timeline.

Here, representative broadband micro-ECoG traces sampled from a posterior retrosplenial grid and simultaneous intracranial probe local field potential traces sampled from a superficial cortical channel are shown side by side for the first recording day and 21 days after. Day 21 recordings exhibit comparable signal amplitude, spectral content, and absence of motion and noise artifacts relative to Day 0 recordings from the same animal, confirming that neither the chronic presence of the PDMS membrane and silicone seal nor the repeated dural punctures introduced detectable degradation of either the surface or the intracranial probe signal quality in superficial cortical layers where degradation would be expected to emerge first. The primary validation of this technique is the acquisition of stable, high-quality multimodal electrophysiological data over time.

The resealable window allows for the repeated insertion of probes to record from the same neuronal populations across weeks. Across the alpha-band, phase locking value matrices computed from channels along the caudal high-density intracranial electrode probe show a stable functional architecture between baseline and the Day 21 post-treatment session. This shows reproducible reinsertion to the same cortical location rather than formal tracing of the same individual neurons.

Although minor intercession translation of the shank precludes a same unit claim, the aggregate laminar and regional structure of the alpha-band interactions is preserved, supporting that the chronic window permits longitudinal sampling of the same functional circuit. To directly verify that the chronic window supports reproducible laminar targeting of the same deep structures across sessions, current source density or CSD maps were computed from the probes'LFP channels. Stimulus-evoked CSD profiles show that the expected laminar sink source signatures, including the prelimbic cortex and anterior cingulate area are preserved between sessions.

The laminar power profile overlays between the two sessions are highly similar across sessions with the Pearson correlation being 0.81. The differences observed in the secondary motor cortex are likely due to movement differences between recordings. Because CSD plotting carries anatomical information, it can provide a within-session non-terminal readout of which brain regions each probe is currently sampling independent of postmortem tissue stainings.

Reproducibility of surface micro-ECoG placement across sessions is quantified independently of the intracranial CSD readout. Band-limited spatial power maps computed from the micro-ECoG grid on Day 0 and Day 21 are superimposed, and the pixel-wise correlation of the two maps is computed separately for the rostral and caudal sub-grids. Sub-grid profile correlations remain high and the spatial barcodes of the two sessions visibly co-localized the same cortical power hotspots on both the rostral and the caudal halves of the array.

A sinusoidal alignment queue visible through the translucent grid together with the engraved bone grooves at the probe insertion coordinates therefore supports millimeter-scale reproducibility of micro-ECoG placement over the 21 day longitudinal interval. To further validate the technique, immunohistochemical stainings were performed for GFAP and IBA1 in post-mortem brain slices approximately four weeks after the craniotomy surgery. GFAP are expressed by astrocytes, which are upregulated in reactive gliosis such as brain injury or inflammation.

IBA1 on the other hand is expressed by microglia, which are more active during neuroinflammatory processes. The validation cohort consisted of animals in which the full two-phase surgery was performed and the cranial window was subsequently reopened three times on postoperative day 0, 21 and 22. However, no micro-ECoG placement nor intracranial probe insertions were performed in this cohort.

The window was resealed between sessions as during an electrophysiological recording. This cohort therefore isolates the inflammatory contribution of the chronic window and repeated dural exposure from any probe-induced tissue damage. No significant differences were observed in either of the markers between the no-surgery control group and the vehicle group that underwent surgery.

These representative micrographs show no qualitative evidence of reactive gliosis or microglial activation in the region directly underlying the window. However, a slight increase of microglial and astrocyte activation was observed adjacent to the probe trajectory and in the hemisphere of the probe insertion respectively, which could likely be as a result of too-high insertion speed of the intracranial probe. Successful surgery results in a clear, transparent window over the cortex with minimal inflammation.

After the chronic treatment period, some brain swelling may occur. This can be managed with careful monitoring and if necessary, administration of additional analgesics. For recordings, simply remove the cap and sealant, place the micro-ECoG grids and Neuropixels probes and begin data acquisition from the awake, behaving animal.

In summary, this two-phase surgical protocol provides a reliable and highly effective method for creating a large, chronic cranial window in mice. The procedure's key advantages are its minimal damage to the dura and the large exposure it provides, which is critical for the simultaneous placement of both surface micro-ECoG grids and multiple deep brain Neuropixels probes. This method enables unprecedented longitudinal investigation of network-wide electrophysiological dynamics in the DMN and other large-scale circuits.

It opens the door to deeper investigations into neural underpinnings of complex behaviors and the pathophysiology of brain disorders, ultimately aiding in the development of novel therapeutics.