A Mouse Model of Single and Repetitive Rotational Closed Head Concussive Brain Injury

Summary

A reproducible mouse model of rotational concussive brain injury is described, providing a versatile method to investigate the subtle functional and biological underpinnings - including neurovascular, inflammatory, and behavioral changes - following single or repetitive impact in the absence of overt structural brain damage.

Abstract

Concussive brain injury (CBI), the pathophysiological substrate underlying clinical concussion, is a frequent yet insufficiently understood condition with potential long-term neurological impairment in a subset of patients. Especially repetitive CBI, i.e., in contact sports, has been associated with lasting cognitive deficits and progressive neurodegeneration (e.g., chronic traumatic encephalopathy, CTE). To address this knowledge gap, a reproducible mouse model of closed-head rotational brain injury that recapitulates key biomechanical and pathological features of CBI was established. A stereotactically guided electromagnetic impactor was used to deliver a standardized strike to the intact skull. To reduce focal strain on the skull and adjacent brain tissue, the impactor tip was fitted with a custom-made silicone cap. This configuration reliably induced head rotation with low inter-animal variability while preventing skull fractures or microscopic tissue injury. To preserve physiological neuronal and vascular activity and to avoid potentially neuromodulatory effects of deep anesthesia, brain injury was induced in conscious mice under light sedation using the α2-agonist medetomidine. The induced impacts caused reproducible rotational head motion with only minor variability attributable to head positioning. Structural brain integrity was assessed using in vivo T2-weighted magnetic resonance imaging and confirmed by ex vivo histological analyses, which revealed no evidence of tissue disruption, contusion, or microbleeds but demonstrated a mild, widespread disruption of the microvascular interface. This novel model of rotational closed-head brain injury provides a robust experimental platform for longitudinal investigations of subtle neurovascular, inflammatory, and blood-brain barrier alterations that occur in the absence of overt structural pathology. Its application enables mechanistic insights into the pathophysiology of clinical concussion and potential neurodegenerative consequences of repetitive injury, thereby facilitating the development of urgently needed clinical biomarkers.

Introduction

Traumatic brain injury (TBI) remains a significant cause of long-term disability across all age groups, with adolescents and young adults disproportionately affected^1,2,3. The vast majority of TBIs are classified as mild (mTBI), commonly referred to as a concussion at the clinical level. Although concussion is by definition not associated with overt structural brain damage detectable by conventional neuroimaging, increasing evidence suggests that the underlying injury mechanism—here referred to as concussive brain injury (CBI)—represents a distinct and clinically relevant pathophysiological entity.

In particular, repetitive CBI, especially when occurring in close temporal succession, as frequently encountered in contact sports and military settings, has been linked to persistent cognitive decline, neuropsychiatric symptoms, and the development of chronic traumatic encephalopathy (CTE)^1,2,3. This apparent discrepancy between the mild nature of the initial injury and the potential for long-term neurological sequelae poses a major challenge for both clinical diagnosis and mechanistic research.

Emerging evidence indicates that subtle yet sustained alterations at the neurovascular and neuroglial interface pose a relevant disease mechanism⁴. These changes encompass blood-brain barrier (BBB) integrity, microvascular structure, glial activation, neuronal excitability, and endogenous repair processes, potentially linking cellular-level pathophysiology to the long-term functional and behavioral deficits observed after, especially repetitive, CBI^5,6,7,8. Collectively, such alterations are increasingly recognized as potential contributors to the delayed functional and behavioral deficits following CBI. However, the diffuse and multifaceted nature complicates a mechanistic understanding and has so far impeded the development of targeted therapeutic interventions.

Preclinical models that recapitulate key aspects of the pathophysiological cascade are crucial tools for dissecting these mechanisms and identifying new therapeutic avenues. While a wide range of experimental TBI models exists, most were designed to mimic moderate to severe injuries or rely on focal structural damage^9,10,11. In contrast, there is an astonishing lack of animal models for the characteristic injury mechanism of CBI⁵. Mostly, a cerebral trauma labeled as “mild” is being induced by a focal impact to the rigidly fixated head, or even by direct cortical impact following craniotomy, which fundamentally contradicts the diffuse, acceleration-deceleration-driven injury mechanism that defines concussion in humans^6,7. Although such approaches allow precise control over injury parameters, their limited ability to replicate the dynamics of real-world concussions reduces their translational relevance.

To overcome these limitations, several methods exist to induce a replicable and diffuse closed-brain injury^12,13,14. Building on these concepts and aiming to minimize technical complexity, a controlled rotational injury model in mice using a commercially available electromagnetic impactor often employed in other experimental trauma models (i.e., controlled cortical impact, CCI) was developed. By delivering a reproducible impact to the freely moving head, this approach preserves structural integrity while enabling the longitudinal investigation of neurovascular, neuroglial, and functional alterations following single and repetitive CBI under conditions that closely approximate the biomechanical features of human concussion.

Combined with longitudinal behavioral assessment, in vivo imaging, and ex vivo histology, this model provides a versatile and accessible platform to explore neurovascular, glial, and behavioral alterations following single or repetitive impacts. As such, it is well-suited for studying the subtle secondary disease mechanisms of CBI, identifying translational surrogate parameters, and evaluating potential therapeutic interventions in a clinically relevant experimental setting.

Protocol

All animal procedures complied with the German Animal Welfare Act and were approved by institutional and governmental authorities (LANUV Nordrhein-Westfalen; AZ 81-02.04.2020.A058). Experiments were conducted following ARRIVE guidelines. A schematic workflow is detailed in Figure 1. This protocol reliably induces concussive brain injury, as evidenced by a defined panel of structural, behavioral, and neurovascular validation outcomes (Figure 2). All materials used in this study are detailed in the Table of Materials.

1. General considerations

NOTE: Mice were socially housed under standardized conditions (12 h light/dark cycle) with ad libitum access to food and water. Mice were acclimated to the experimental facility for at least 7 days before surgery. A total of 40 mice were used in this pilot study, with 10 animals allocated to each experimental group (sham as a control group, single CBI group, repetitive CBI in high-frequency, and repetitive CBI in low-frequency group). Set up a high-speed camera system at eye level with the mouse head in a fixed position in front of the mouse (e.g., 10 cm). To ensure consistency across trials, keep the camera position and orientation constant. A GoPro Hero6 was used in this study at 240 frames per second (fps) with a resolution of 1080px. The open-source motion analysis software Kinovea (version 2025.2.0) is used for the kinematic analyses described.

2. Analgesia and anesthesia induction

1. Prepare the setup, including the stereotactic frame and attachment for the electromagnetic impactor.
2. Administer Tramadol (1 mg/mL) via drinking water two days before trauma induction and continue for three days after the (last) surgery.
3. Induce anesthesia using isoflurane in a nitrous oxide/oxygen mixture (O₂ 30%, N₂O 70%), with 4–4.5% isoflurane in an induction chamber.
4. Assess surgical tolerance using the toe-pinch reflex. Proceed once a withdrawal response is absent.
5. Remove the mouse from the induction chamber and place it on a heating pad within the stereotactic frame. Maintain the core temperature at 37 ± 1.0 °C using a fiber-optic rectal probe. Protect the eyes from desiccation by applying ophthalmic ointment.
6. Maintain anesthesia by delivering isoflurane at 2,0–2,5% via a nose cone connected to the stereotactic setup.

3. Preparation of the skull

1. Fix the mouse within the stereotactic frame using specific ear bars.
2. Shave the scalp and thoroughly disinfect the surgical area with an alcohol-based antiseptic.
3. Apply local anesthesia by subcutaneous injection of Bupivacaine (0.1 mg/kg).
4. Reassess the toe-pinch reflex before skin incision. Proceed only in the absence of a withdrawal response to ensure a consistent depth of anesthesia.
5. Perform a midline skin incision of approximately 7 mm to expose the skull.
6. Carefully expose the skull over the right parietal bone between Bregma and Lambda.
7. Ensure that the stereotactic frame is oriented perpendicular to the skull, with the vertical axis positioned at a 90° angle relative to the head surface.
8. Mount a marker pin into the stereotactic holder and center it directly on Bregma.
9. Zero all stereotactic coordinates and move the marker pin to the predefined anteroposterior (AP) and mediolateral (ML) coordinate (in our model, at AP 0.5, ML -2.5). Mark the target point with a water-resistant felt-tip pen.

4. Adjustment of analgesia

1. Release the mouse from the ear bars.
2. Inject medetomidine subcutaneously (0.5 mg/kg) and gradually reduce isoflurane to <0.5 %.
3. Maintain anesthesia at < 0.5% isoflurane while continuously monitoring respiration and body temperature.
4. Allow for equilibration under these conditions for 10 min to reach a steady state of anesthesia while preserving neuronal network activity. Respiratory rate, absence of reflex responses, and stable body temperature (37 ± 1.0 °C) were used as standardized indicators to ensure comparable anesthesia depth across animals.

5. Impact induction

NOTE: Before the experiment, fabricate a head holder for the animal model from commercially available polyethylene (PE) foam pipe insulation, a closed-cell, non-crosslinked polymer foam, that restricts downward acceleration while allowing rotational movement. Additionally, fabricate a custom-made 5 mm spherical impactor tip made from silpuran (ratio 1:1; stiffness 830 kPa).

1. Remove the rectal temperature probe and discontinue isoflurane delivery, then place the mouse in a prone position on the head holder.
2. Attach the custom-made silicone cap to the commercially available impactor tip.
3. Mount the electromagnetic impactor into the stereotactic holder and ensure that the device is oriented at a 10° angle relative to the skull surface.
4. Align the impactor tip above the predefined target region on the skull.
5. Lower the impactor tip in the protracted position until it lightly contacts the skull surface.
6. Retract the impactor tip and advance it to the desired indentation depth (i.e., 3 mm in these experiments) toward the skull surface. For control animals, omit the lowering and raise the tip by the same amount instead.
7. Start video recording. Deliver a single rotational concussive impact by triggering the electromagnetic bolt via the control switch with a velocity of 5 mm/s and a pulse duration of 0.1 ms.
8. Retract the impactor immediately after completion of the impact.
9. Antagonize medetomidine anesthesia using the antagonist atipamazole (0.5 mg/kg, s.c.).

6. Post-impact assessment and wound closure

1. Turn the mouse into a supine position.
2. Assess the righting reflex by recording the time required for the animal to return to a prone position.
3. Re-induce anesthesia with isoflurane in a nitrous oxide/oxygen mixture (O₂ 30%, N₂O 70%) at 4% isoflurane.
4. Remove the mouse from the induction chamber and place it on the heating pad by the stereotactical frame. Maintain the core temperature at 37 ± 1.0 °C using a fiber-optic rectal probe.
5. Fix the mouse in the stereotactic frame using ear bars.
6. Close the skin incision using interrupted sutures with self-absorbing filaments (e.g., Polyglactin 910, 6-0).
7. Transfer the mouse to a warming box and monitor until full recovery of consciousness.
8. Return the animal to its home cage.

7. Repeated impact paradigm

1. Assign mice to sham, single, or repeated CBI groups according to the experimental design.
2. For repeated-injury groups, schedule subsequent CBI sessions at defined intervals (e.g., every 48 h).
3. Before each subsequent CBI, administer preoperative analgesia via tramadol in drinking water as described in section 3.1.
4. Repeat anesthesia, stereotactic marking, and impact following the same procedure outlined in sections 3–6, including local anesthesia with bupivacaine (0.1 mg/kg, s.c.), medetomidine administration (0.5 mg/kg, s.c.) with maintenance of isoflurane at < 0.5%, stereotactic targeting at AP 0.5 and ML −2.5 relative to Bregma, and delivery of a rotational concussive impact using an electromagnetic impactor (indentation depth 3 mm, velocity 5 mm/s, pulse duration 0.1 ms, 10° angle relative to the skull surface).
5. Ensure consistent positioning (prone placement on a polyethylene foam head holder allowing rotational movement), identical impact parameters (angle, depth, velocity, and duration), and standardized peri- and post-procedural conditions, including medetomidine antagonization with atipamazole (0.5 mg/kg, s.c.), re-induction of anesthesia (4% isoflurane in O₂ 30%/N₂O 70%), and temperature control at 37 ± 1.0 °C, across repeated sessions to maintain reproducibility across animals and time points.
6. Ensure consistent positioning, impact parameters, and recovery conditions across repeated sessions to maintain reproducibility across animals and time points.
7. Monitor animals continuously for cumulative behavioral deficits following repeated impacts.

8. Kinematic analysis

1. Export the videos in standard format (e.g., MP4) for analysis and load them into Kinovea software.
2. Navigate to the relevant time window containing pre- and post-impact frames (e.g., -0.025 to 0.3 s from impact). Make sure to set an adequate frame rate (e.g., 240 fps) in the software settings. Define a scaling reference (e.g., inter-eye distance).
3. Use Kinovea’s tracking tool to establish a three-point tracking setup (e.g., right eye, left eye, nose; Figure 3A). Start tracking for each point and manually correct tracking errors, particularly during rapid motion or occlusion.
4. Export tracked coordinates as CSV or Excel files.
   NOTE: Further kinematic analysis was performed using a custom MATLAB script. Time-resolved tracking data exported from Kinovea were imported and preprocessed to remove invalid entries and ensure consistent time scaling. Head translation was defined as the centroid of the three tracked landmarks (right eye, left eye, and nose), from which linear displacement, velocity, and acceleration were derived using numerical differentiation. Head rotation was estimated from the orientation of the eye–nose axis, with angular displacement calculated relative to a pre-impact baseline and angular velocity obtained as its first derivative. To reduce noise amplification, all position and angular signals were smoothed using a Savitzky–Golay filter prior to differentiation. The primary outcome parameter was the maximum angular displacement, while velocity and acceleration profiles were analyzed as relative measures of head motion dynamics across trials.

Results

General considerations:

Application of this protocol reproducibly induces concussive brain injury, as demonstrated by a defined set of structural, behavioral, and neurovascular validation readouts (see Figure 2). Structural integrity was preserved following injury. No overt macroscopic tissue disruption or hemorrhage was detected, indicating that the protocol induces concussive rather than contusive brain injury. Histological assessment confirmed intact tissue architecture without evidence of gross cellular loss or cavitation. In addition, structural imaging and histological analyses consistently failed to detect contusions or microbleeds across animals at the applied impact parameters.

Impact Induction

Head kinematics analysis (Figure 3B) demonstrated consistent head rotation (mean rotation = 18.9°, SD = 6.1°), supporting a predominantly rotational acceleration-deceleration mechanism.

Post-impact assessment

Behavioral readouts revealed transient functional alterations despite preserved structure. Injured animals showed delayed recovery of alertness, increased spontaneous activity, elevated anxiety-like behavior, and mild impairments in motor coordination and spatial cognition. These effects were consistently observed across cohorts, with comparable effect direction and magnitude between independent experimental runs. Specifically, injured animals required significantly longer to regain an upright position following impact and exhibited increased overall activity levels during the first days after injury. In the open field, exploration of the center zone was consistently reduced, indicating anxiety-like behavior, while total locomotor activity remained largely preserved. Spatial cognition, assessed by spontaneous alternation in the Y-maze, was significantly impaired compared with sham controls, and mild motor deficits were observed in the rotating beam test at early time points.

Repeated impact paradigm

Notably, repetitive injury paradigms revealed a frequency-dependent additive effect, with more pronounced and persistent cognitive impairment following high-frequency impacts (e.g., every 48 h) than following lower-frequency impacts (e.g., weekly), consistent with a cumulative injury burden. These behavioral changes were most pronounced at acute time points and resolved during the recovery period, consistent with a clinically reversible, concussive phenotype.

Immunohistological analysis

Neurovascular integrity, however, was selectively compromised. The protocol induced region-specific disruption of BBB integrity, most prominently in the hippocampus, primary motor cortex, and thalamus. Quantitative analysis demonstrated a significant increase in tracer extravasation compared with sham controls, with the strongest effects observed after repetitive high-frequency injury and partial recovery over time. These changes were reproducibly detected across animals and cohorts, with consistent regional patterns and temporal dynamics. The temporal profile of BBB disruption corresponded to cognitive impairments, supporting its role as a mechanistic correlate of injury-induced functional impairments.

Across cohorts, these structural, behavioral, and neurovascular readouts were consistently observed with low inter-animal variability, and frequency-dependent effects were reproducibly detected under standardized conditions. Together, these imaging, histological, behavioral, and neurovascular readouts established successful injury induction and validated the model as a reproducible platform for studying the functional and mechanistic consequences of concussive brain injury. These observations were recently published, providing formal validation of the model and confirming its utility for investigating the neuropsychiatric and cognitive sequelae of (repetitive) CBI⁸.

Figure 1: Experimental workflow for closed-head rotational concussive brain injury in mice. Schematic overview of the protocol illustrating induction of closed-head rotational concussion. Optional validation readouts (BBB integrity, magnetic resonance imaging, histology, or behavior assessment) provide benchmarks for procedural fidelity. The figure was created with BioRender.com. Please click here to view a larger version of this figure.

Figure 2: Validation readouts of concussive brain injury. Thorough evaluation within days after CBI induction reveals preserved structural integrity, as indicated by the absence of overt macroscopic tissue disruption and intact tissue architecture (right; MRI imaging). Despite a lack of structural impairment, animals exhibit transient behavioral alterations (bottom). At the neurovascular level, region-specific disruption of BBB integrity is observed (left; immunoglobulin G extravasation in the hippocampus). The figure was created with BioRender.com. Please click here to view a larger version of this figure.

Figure 3: Kinematic analysis of head motion following impact. (A) Three anatomical landmarks (right eye, left eye, nose) were tracked to define head motion. Rotational displacement (Δθ) was calculated from the eye–nose axis relative to baseline, and translation from the centroid of the three points. (B) Group-averaged kinematic profiles (mean ± SD) aligned to impact (red dashed line, t = 0), showing angular displacement (Δθ), angular velocity (ω), vertical velocity (v_y), and angular acceleration (α). Measures reflect relative head motion dynamics across trials. Please click here to view a larger version of this figure.

Discussion

A novel mouse model of closed-head CBI based on acceleration-deceleration injury is characterized, featuring a mild yet widespread impairment of the BBB and a distinct neuropsychiatric phenotype reminiscent of clinical concussion. By reproducing key hallmarks of CBI in a controlled and consistent manner while preserving structural integrity, the model provides a useful and well-characterized experimental framework for studying the mechanisms underlying CBI and its neuropsychiatric and cognitive sequelae.

Critical steps within the protocol are those that determine the effective mechanical loading of the head. In particular, accurate stereotactic alignment and reproducible head positioning relative to the impactor are required to ensure a consistent rotational injury vector across animals. Standardization of impactor parameters, including position, angle, and settings, as well as routine calibration prior to experimental runs, is essential to minimize variability. Because small deviations in impact location or rotational angle can disproportionately affect injury severity, strict operator consistency and documentation of setup parameters are key to experimental reproducibility. To further enhance methodological rigor and enable troubleshooting of the technique, high-speed video recording of head motion during impact is strongly recommended. Post hoc quantification of linear and rotational head kinematics allows identification of subthreshold impacts, excessive linear acceleration, or off-axis rotations that may otherwise go unnoticed. This approach facilitates differentiation between technical deviations and biological variability and supports iterative refinement of the experimental setup to improve injury consistency.

From a methodological perspective, the protocol integrates multiple complementary outcome measures to verify successful injury induction and overall experimental fidelity. Region-specific BBB disruption provides a sensitive indicator of effective rotational loading, whereas the absence of macroscopic tissue damage, for example, as assessed by longitudinal in vivo magnetic resonance imaging or ex vivo histology, serves as an important control for preserved structural integrity. Together, these readouts enable multidimensional validation of the experimental setup and allow investigators to distinguish successful protocol implementation from suboptimal injury delivery or technical inconsistencies (see Figure 2).

Compared with established experimental models of TBI, the present protocol enables mechanistic isolation that is difficult to achieve with traditional focal-impact paradigms. The model induces diffuse neurovascular alterations while preserving macroscopic tissue integrity, thereby capturing key features of concussive brain injury without confounding effects from focal lesions. Closed-head mTBI models based on weight-drop or related paradigms have demonstrated robust behavioral deficits with minimal structural damage. However, in these approaches, head motion arises from the interaction among impact, body movement, and the supporting surface, making it difficult to control the contribution of rotational versus linear forces and dependent on experimental parameters. In contrast, the present model is specifically designed to generate controlled rotational head motion under standardized conditions. By limiting linear displacement and promoting defined rotational acceleration–deceleration, it enables reproducible and mechanistically focused investigation of diffuse injury processes^12,13,14.

Beyond its use for single-injury paradigms, the technique is readily adaptable to a range of further experimental applications. Based on its standardized implementation and demonstrated consistency across cohorts, the model is well-suited for longitudinal studies of repetitive concussive injury. In addition, the protocol can be combined with advanced behavioral, molecular, and imaging-based readouts, as well as genetic or pharmacological manipulations, to interrogate mechanisms underlying concussion-related dysfunction. These features support its application in preclinical therapeutic screening and in the identification of translational surrogate markers relevant to clinical concussion.

Several limitations of the technique should be acknowledged. While rotational acceleration is a defining biomechanical feature of concussive brain injury, scaling biomechanical forces from mice to humans remains inherently challenging, and absolute injury thresholds cannot be directly translated across species. As with other preclinical TBI models, the present approach therefore prioritizes mechanistic relevance over direct biomechanical equivalence. A further limitation relates to the use of anesthesia, which is required for experimental control and animal welfare and may modulate acute physiological and neurobehavioral responses to injury. Finally, although the model reliably induces functional and neurovascular alterations, it does not recapitulate all aspects of concussive pathology, such as overt axonal disruption or focal tissue damage. The technique should therefore be viewed as complementary to existing traumatic brain injury paradigms rather than as a comprehensive representation of all concussion-related pathologies. In summary, this closed-head rotational injury model provides a technically accessible and biologically relevant platform for studying concussive brain injury in mice.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Medication
Antipamezole	Vetoquinol	VM 06043/4004	Medetomidine antagonist (α2-adrenoceptor antagonist), subcutaneous (0.5 mg/kg)
Bepanthen Ophthalmic Ointment	Bayer	11809917	Eye protection during anesthesia, topical
Bupivacaine	Sigma-Aldrich	B1160000	Local anesthesia, subcutaneous (0.1 mg/kg)
Isoflurane	CP pharma	1214CN	Anesthesia, inhalation (induction and maintenance)
Medotomidine	Vetoquinol	VM 06043/4003	Sedative/anesthetic adjunct (α2-adrenoceptor agonist), subcutaneously (0.5 mg/kg)
Tramadol hydrochloride (100 mg/mL)	Grünenthal	6867645	Analgesic administered via drinking water, 1 mg/mL final concentration for mice, from 2 days pre- to 3 days post-surgery
Equipment
Polyethylene neck cusion	Custom		Support for prone positioning during impact
Sterotaxic Impactor Silupran Cover	Custom		Distributes mechanical load, prevents focal skull damage, attached to impactor tip, 830 kPa
Streotaxic Impactor	Leica Biosystems	39463923/IM10131	Electromagnetic impact device for controlled impact delivery
Tools
GoPro Hero6	GoPro Inc., San Mateo, CA, USA
Kinovea		(version 2025.2.0)
MATLAB	MathWorks, Natick, MA, USA	NA