A mouse model of mechanical stress injury to the basal ganglia using thermosensitive PNIPAM hydrogel for intracerebral hemorrhage research

Abstract

Background

Intracerebral hemorrhage (ICH) is frequently associated with poor clinical outcomes. White matter injury (WMI), particularly to the corticospinal tract (CST), plays a critical role in the development of hemiplegia. However, conventional ICH models tend to induce extensive damage and involve complex blood‐derived components, highlighting the need for a model that can induce direct mechanical stress injury specific to white matter.

Methods

We established a novel mouse model by stereotactically injecting thermosensitive poly(N‐isopropylacrylamide) (PNIPAM) hydrogel into the internal capsule to induce localized mechanical stress on CST. Resulting injury was evaluated by gross pathological examination and transmission electron microscopy. Motor function was assessed using a series of behavioral tests. CST integrity was examined by motor evoked potential (MEP) and nerve tract tracing. The underlying molecular mechanisms were elucidated by RNA sequencing (RNA‐seq) and Western blot.

Results

The model consistently showed the induction of mechanical stress injury in internal capsule, leading to substantial WMI and motor deficits. MEP amplitude was reduced, and nerve tract tracing revealed severe disruption of the CST, which was more pronounced than that caused by the classical blood–injection ICH model. RNA‐seq analysis identified the activation of mechanical stress–related pathways, including tumor necrosis factor (TNF) and fluid shear stress signaling pathways. Western blot assay confirmed the altered expression of WMI markers and upregulation of key molecules involved in these pathways.

Conclusions

This newly established model of mechanical stress injury effectively recapitulates the pathophysiology of CST damage following ICH. It also provides a simple and reproducible tool for conducting preclinical studies on mechanical stress–induced WMI in ICH.

In this study, we established a novel mouse model of intracerebral hemorrhage (ICH) by stereotactically injecting thermosensitive PNIPAM hydrogel into the internal capsule to impose localized mechanical stress on the corticospinal tract, which provides a straightforward and reproducible tool for preclinical studies on focal mechanical stress–induced white matter injury in ICH.

1. INTRODUCTION

Spontaneous intracerebral hemorrhage (ICH) is a devastating stroke subtype associated with high mortality and morbidity rates, with early mortality reaching as high as 30%–40%. ¹ , ² Survivors frequently experience severe and persistent neurological deficits. Only 14%–36% of patients achieve functional independence within 1 year. ³ , ⁴ Motor impairments represent one of the most prevalent and debilitating deficits, significantly diminishing the patient's quality of life. ⁵ A critical pathological contributor to motor dysfunction is injury to white matter tracts, particularly the corticospinal tract (CST), which serves as the principal neural pathway for voluntary somatic motor control. ⁶ , ⁷

Currently, several animal models are employed to investigate the pathophysiology of ICH and evaluate potential therapies. ⁸ The two primary established approaches involve the stereotaxic injection of autologous blood into the caudate nucleus or the administration of bacterial collagenase to induce enzymatic vascular rupture and hemorrhage. ⁹ , ¹⁰ , ¹¹ Both models effectively replicate the secondary injury cascades triggered by blood components, such as thrombin‐mediated toxicity, iron‐induced oxidative stress, and inflammation. However, the precise role of the initial mechanical force in subsequent white matter injury (WMI) and functional impairment remains unclear in these models. Moreover, these models often cause widespread injury within the basal ganglia and are confounded by the concurrent effects of blood‐derived toxins, thereby failing to confine mechanical stress specifically to the CST. These limitations highlight the need for developing a new model capable of isolating the effects of mechanical stress on the CST.

In this context, the present study aimed to establish a novel mouse model of ICH that induces direct mechanical stress injury to the white matter, specifically targeting the CST. This was achieved by stereotactically injecting a thermosensitive poly(N‐isopropylacrylamide) (PNIPAM) hydrogel (PNI gel) into the internal capsule. Upon reaching body temperature, the PNI gel undergoes phase transition from a liquid state to solid state, establishing a defined space‐occupying lesion that mimics the acute mechanical mass effect of a hematoma. Our findings demonstrate that this model reliably reproduces the key features of CST damage. It provides a direct and reproducible tool to investigate mechanical stress–induced WMI. It also facilitates the development of targeted neuroprotective therapies and serves as a complement to existing ICH models.

2. MATERIALS AND METHODS

2.1. Animals

C57BL/6 mice (age: 8–12 weeks, weight: 23–28 g) were obtained from the Experimental Animal Center of the Third Military Medical University (Army Medical University). The animals were housed under a 12‐h light/dark cycle with ad libitum access to food and water. All experimental protocols followed the ARRIVE guidelines (Animal Research: Reporting of In Vivo Experiments) and were approved by the Ethics Committee of Army Military Medical University (approval no.: AMUWEC20232125).

2.2. Experimental design

The study included four parts, as illustrated in Figure 1.

FIGURE 1.

A schematic diagram of the experimental design. CST, corticospinal tract; IF, immunofluorescence staining; MEPs, motor evoked potentials; RNA‐seq, RNA sequencing; TEM, transmission electron microscopy; WB, Western blot assay.

Part A: Pathological changes in the mechanical stress–induced ICH model were assessed in the sham, PNI, PBS, or ICH groups. The animals were euthanized at 0.5 h for fresh brain section analysis and at 24 h for immunofluorescence (IF) assay and transmission electron microscopy (TEM) observation.

Part B: Motor behavior and motor conduction function were assessed before and on days 1, 3, 7, 14, or 21 after modeling using behavioral tests and motor evoked potential (MEP) recordings.

Part C: CST nerve conduction was examined by anterograde and retrograde tracing in the sham, PNI, and ICH groups. Recombinant adeno‐associated virus (rAAV) was injected 3 days postmodeling. The animals were subsequently euthanized for IF analysis 21 days postprocedure.

Part D: At 24 h postmodeling, signaling pathways and protein expression were analyzed using RNA sequencing (RNA‐seq) and Western blot assay of tissue surrounding the lesion (rather than the entire hemisphere). The sampled region is indicated by a green dashed circle.

2.3. Animal surgery

The mechanical stress–induced mouse ICH model was established by stereotaxic injection of PNIPAM hydrogel into the internal capsule (Figure 2). The PNIPAM hydrogel was prepared according to a previously described method. ¹² Mice were deeply anesthetized with 2% isoflurane and secured on a stereotaxic apparatus (RWD, Shenzhen, China). A 1‐mm‐diameter burr hole was drilled at the target coordinates (right: 2.5 mm, posterior: 0.7 mm, depth: 3.2 mm, angle: 16°). A total of 15 μL of PNIPAM hydrogel was then infused into the target site at a rate of 1.5 μL/min. For comparison, an autologous blood‐induced ICH model was also established, as reported previously. ¹³ Briefly, 15 μL of autologous blood was injected into the right caudate nucleus (AP: 0.8 mm, ML: 2.0 mm, DV: 3.0 mm) at the rate of 1.5 μL/min. The needle was left in place for 5 min after injection to prevent backflow. Sham‐operated mice underwent the same surgical procedure including needle insertion but without any injection. In the PBS group, 15 μL of phosphate‐buffered saline (PBS) solution was injected at a rate of 1.5 μL/min at the same coordinate point as the PNIPAM group. After surgery, all mice were placed on a 37°C heating pad until they fully recovered.

FIGURE 2.

Surgical procedure for establishing the PNIPAM‐induced mechanical stress model. (A) After anesthesia induction, the mouse was fixed using ear bars, with the skull aligned to the horizontal plane. (B) After scalp disinfection, a midline incision was made to expose the skull, and a burr hole was drilled at the target coordinates. (C) Magnified view of the exposed skull shows the Bregma point (B′), coronal suture (CS), and sagittal suture (SS); the arrow indicates the drilled hole. (D) A total of 15 μL PNI gel was drawn into a syringe; the inset shows the transition of the PNI gel from the liquid state to the solid state upon warming to body temperature in the experimenter's palm. (E) The right oblique angle of the stereotaxic instrument was set to 16°. (F) The syringe needle was slowly advanced through the burr hole to the target coordinates. (G) PNI gel was infused into the target site at the rate of 1.5 μL/min. (H) Stereotaxic coordinates for model establishment referenced to a coronal section from a mouse brain atlas. Ang, angle relative to the vertical axis; D, depth from the brain surface; P, posterior of B; R, right of B′.

2.4. Immunofluorescence staining

IF staining was performed as described previously. ¹³ Briefly, at 24 h after modeling, mice were deeply anesthetized and perfused with isotonic saline followed by 4% paraformaldehyde. The brains were removed, fixed in 4% paraformaldehyde for 24 h, and subsequently dehydrated with 30% sucrose for 72 h. After embedding, coronal sections of the brains (15 μm) were cut using a cryostat microtome (CM1860UV, Leica, Wetzlar, Germany). The sections were then permeabilized with 0.3% Triton X‐100 for 30 min and blocked with 5% fetal bovine serum for 1 h. Subsequently, the sections were incubated overnight at 4°C with the following primary antibodies: anti‐NF200 (green) (Sigma‐Aldrich; batch number: N4142; St. Louis, MO, USA) and anti‐MBP (red) (Santa Cruz; batch number: sc‐13914; CA, USA). Finally, the samples were incubated with appropriate secondary antibodies at 4°C for 24 h followed by staining with DAPI for 5 min.

2.5. Transmission electron microscopy

At 24 h postmodeling, mice were perfused with isotonic saline, followed by a fixative containing 1.25% glutaraldehyde and 2% paraformaldehyde for transmission electron microscopy (TEM). Approximately 1 mm³ of brain tissue surrounding the lesion site was carefully dissected. The tissue samples were fixed in 1% OsO₄ and dehydrated through a graded series of acetone. After dehydration, the tissues were embedded in Epon resin and 60‐nm‐thick ultrathin sections. The sections were then observed using a TEM system (TECNAI10, Philips, Amsterdam, Netherlands). The G‐ratio (the ratio of the inner axonal diameter to the total outer fiber diameter) was calculated for 120 randomly selected myelinated axons per group from TEM images of the site around the lesioned area. Images were analyzed using ImageJ software by an investigator blinded to the groups. The inner and outer diameters were measured, and the G‐ratio was calculated as the inner axonal diameter divided by the outer diameter according to a previous report. ¹⁴ Myelin injury was assessed using a semiquantitative scoring system: 3 = complete disruption of myelin continuity; 2 = multilamellar vesicular detachment; 1 = separation of a single myelin layer; 0 = normal, intact myelin sheath. A total of 120 myelin sheaths were randomly selected from four mice (30 myelin sheaths per mouse) per group and analyzed for the G‐ratio and myelin injury score.

2.6. Motor behavioral assessment

2.6.1. Open field test

The open field test was conducted in a 50 × 50 cm arena under dim light conditions (50 Lux). Mice were placed in the center of the arena and allowed to explore freely for 10 min. The first 5 min were considered an adaptation period, and data from the last 5 min were analyzed. Total traveled distance (cm) was quantified using the ViewPoint behavioral analysis software.

2.6.2. The forelimb force

Forelimb grip strength was measured separately using a grip strength meter (Columbus Instruments, Columbus, OH, USA). Each mouse was allowed to grasp a metal bar with its forepaws and then gently pulled backward until it released its grip. The peak force was recorded. This test was repeated five times per mouse, and the average value of the three most consistent trials was used for the analysis.

2.6.3. The Basso Mouse Scale (BMS)

BMS scores were assessed by two independent investigators blinded to the experimental groups. Mice were placed in an open field for 4 min, and their hindlimb motor function was scored based on the standard BMS scale. The scores from both investigators were averaged, and inter‐rater reliability was high (Pearson's r > 0.9). The open‐field locomotor rating system uses a 0–9 scale, where a score of 0 indicates no ankle movement and a score of 9 represents frequent or consistent plantar stepping with substantial coordination, parallel paw placement at initial contact and lift‐off, normal trunk stability, and a persistently raised tail.

All motor behavioral tests were performed by observers blinded to the experimental groups. The observers who participated in behavioral evaluation did not participate in random grouping; additionally, the observers involved in data analysis did not participate in random grouping or behavioral evaluation.

2.7. Electrophysiology

MEPs were recorded in mice anesthetized intraperitoneally with 1% sodium pentobarbital (25 mg/kg). Transcranial electrical stimulation was delivered using needle electrodes implanted subcutaneously, with the anode positioned at the midpoint of the binaural line and the cathode placed near the nasal area. Two recording electrodes were placed in the gastrocnemius muscle of the hind limb, and a ground electrode was inserted subcutaneously in the back. ¹⁵ A stimulator (Keypoint, Medtronic, USA) delivered single pulses (7.8 mA, 0.1 ms, 1 Hz) through stimulating needle electrodes. Each mouse received three stimuli at 15‐s intervals. For data analysis, we focused on the peak amplitude, a parameter primarily reflecting the integrity of the primary motor cortex and CST. ¹⁶

2.8. Corticospinal tract nerve conduction

2.8.1. Anterograde tracing

rAAV‐hSyn‐EGFP (BrainVTA Technology, batch number: PT‐1990) was injected into the mouse motor cortex. The procedure was conducted on day 3 postmodeling under anesthesia. Viral injections (200 nL per site) were delivered at three coordinates (relative to bregma: right: +1.0, 0, and −1.0 mm; anterior: 1.0 mm; depth: 0.6 mm). Mice were euthanized on day 21 postinjection, and C3–C7 axial spinal cord tissues were collected for IF analysis of CST axons.

2.8.2. Retrograde tracing

rAAV‐CAG‐mCherry virus (BrainVTA Technology, batch number: PT‐0105) was injected into the left lumbar spinal cord using glass micropipettes. Injections (three sites, 200 nL per site) were performed at depths of 0.6 mm (0.2 mm lateral to the midline) across three segments (L2–L4), with a spacing of 1 mm between adjacent sites. On day 21 postinjection, mice were euthanized, and their brain tissues were collected for IF analysis of corticospinal neurons. ¹⁷

2.9. RNA sequencing

White matter tissues surrounding the lesion area were collected for RNA‐seq. Mice were euthanized at 24 h after modeling. Total RNA was extracted using TRIzol reagent, followed by quantification and purification. The extracted RNA was subsequently reverse transcribed into complementary DNA (cDNA). A U‐labeled second‐strand DNA was synthesized, and after ligation and polymerase chain reaction (PCR) amplification, a cDNA library with an average insert size of 300 bp (±50 bp) was constructed. Sequencing was performed by LC‐biotechnologies (Hangzhou) Co., Ltd.

2.10. Western blot assay

The white matter surrounding the lesion area was dissected and homogenized for protein extraction. Protein samples (50 μg) were separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS‐PAGE) and transferred onto polyvinylidene fluoride membranes. After being blocked with 10% bovine serum albumin for 1 h at room temperature, the membranes were incubated overnight at 4°C with the following primary antibodies: anti‐NF200 (Sigma‐Aldrich; batch number: N4142; Missouri, USA), anti‐MBP (Santa Cruz; batch number: sc‐13914; CA, USA), anti‐TNF‐α (Santa Cruz; batch number: sc‐52746; CA, USA), anti‐NF‐κB (Invitrogen; batch number: PA5‐16545; Carlsbad, CA, USA), anti‐phospho‐NFκB (Invitrogen; batch number: MA5‐15160), and anti‐IL‐1β (Santa Cruz; batch number: sc‐32294; CA, USA). β‐actin (Sangon Biotech; batch number: D191047; Shanghai, China) served as an internal reference.

2.11. Statistical analysis

The normality of the quantitative data and the homogeneity of variances were assessed using Shapiro–Wilk test and Levene's test. All data met the assumptions and are expressed as mean ± standard deviation (SD). A two‐tailed Student's t‐test was used for comparison between two groups. Longitudinal data (behavior, weight, and MEPs) were analyzed using a mixed‐design two‐way analysis of variance (ANOVA). Multigroup data (CST conduction, Western blot assay) were evaluated using one‐way repeated‐measures ANOVA with Scheffé's post‐hoc test. Statistical analyses were performed using GraphPad Prism 8.4.3. Statistical significance was considered at p < 0.05.

3. RESULTS

3.1. Pathological alterations in the white matter in the internal capsule and surrounding regions

In this study, we established a mouse model induced by mechanical stress through stereotaxic injection of 15‐μL thermosensitive PNIPAM hydrogel (referred to as PNI gel) into the internal capsule (Figure 2). IF and fresh brain section analyses revealed consistent space‐occupying lesions involving the CST in the internal capsule of this model (Figure 3A,B), whereas the PBS group showed only minimal tissue disruption along the needle track (Figure S1B).

FIGURE 3.

Pathological changes in the PNIPAM‐induced mechanical stress model. (A) Immunofluorescence (IF) staining showed that PNI gel injection induced a stable mechanical mass effect in the internal capsule. Scale bar: 1 mm. (B) Macroscopic views of mouse brain sections revealed mechanical stress injury in the internal capsule following PNI gel injection. PNI gel transitioned from the liquid state to the solid state after warming to 37°C in vitro. (C) Representative transmission electron microscopy (TEM) images show severe degeneration and disruption of myelin sheaths around the injured brain area. The white arrowhead indicates demyelination. Scale bar: 1 μm. (D) The G‐ratio of myelin indicates thinning and disintegration of myelin sheaths around the injury area (n = 4). (E) Linear regression analysis of the G‐ratio versus axon diameter revealed different slopes and intercepts between the PNI and sham groups (n = 4). (F) Myelin injury scores showed more ruptured myelin around the injury area in the PNI group than in the sham group (n = 4). White and red dashed circles highlight the mechanically injured regions. **p < 0.01.

TEM observation further elucidated the pathological features of myelin degradation, such as splitting or bubbling, in the nerve fibers surrounding the lesion site (Figure 3C). The G‐ratio, an indicator of myelin thickness, was significantly higher in the PNI group than in the sham group (p < 0.01) (Figure 3D). To determine the correlation between myelin sheath thickness and axonal diameter, a coordinate system was constructed using axonal diameter as the independent variable and the G‐ratio as the dependent variable. Linear regression curves were fitted for each group. The collective plotting of the data from both groups showed that the linear function for the sham group (slope = 0.286 ± 0.027, y‐intercept = 0.484 ± 0.017) differed markedly from that of the PNI group (slope = 0.031 ± 0.025, y‐intercept = 0.751 ± 0.020) (Figure 3E). Moreover, more than 87% of myelin sheaths in the PNI group were severely disrupted, with 35% and 52% sheaths receiving myelin injury scores of 3 and 2, respectively. In contrast, fewer than 34% of myelin sheaths in the sham group were severely disrupted, with none assigned a score of 3 (Figure 3F).

In summary, PNI gel injection effectively induced severe WMI in the internal capsule and surrounding regions.

3.2. Changes in motor functions and physiological parameters

Motor function was evaluated with a series of behavioral tests, including the open‐field test, forelimb muscle strength, and BMS assessment (Figure 4A–C). On most days from day 1 to day 21 postmodeling, PNI group mice showed significantly reduced scores in these tests compared to those in the sham group (p < 0.05 or p < 0.01). Moreover, the PNI group showed more motor function deficits than the ICH group during certain time periods.

FIGURE 4.

Motor behavioral and motor conduction function changes in the autologous blood intracerebral hemorrhage (ICH) model and the PNIPAM‐induced mechanical stress model. (A) Left: Representative movement trajectories from the open‐field test in the sham, PNI, and ICH groups. Right: Quantitative results showed that the open‐field test score was significantly reduced in the PNI group up to 7 days postmodeling, whereas in the ICH group, the reduction persisted only up to 3 days. The open‐field test score in the PNI group on day 1 was significantly lower than that in the ICH group (n = 6). (B, C) Motor behavioral scores on the forelimb muscle strength and Basso Mouse Scale (BMS) assessment were decreased in the ICH and PNI group during 1–3 weeks after modeling. The PNI group showed more motor function deficits than the ICH group during certain time periods (n = 6). (D) Left: Representative motor evoked potential (MEP) waveforms. Right: Quantitative analysis shows that the peak amplitude of MEPs in the ICH and PNI groups was significantly lower than that in the sham group until 2 weeks postmodeling. The peak amplitude in the PNI group on day 7 was significantly lower than that in the ICH group (n = 6). *p < 0.05, **p < 0.01 between the sham group and the other groups; ^# p < 0.05, ^## p < 0.01 between the two groups apart from the sham group.

3.3. Changes in motor conduction function and CST nerve transmission

MEPs were used to evaluate motor nerve conduction function in mice (Figure 4D). MEP recording showed that the peak amplitude measured from the left limb of mice in the PNI group was significantly reduced from day 1 to day 14 compared to that in the sham group (p < 0.01). Furthermore, the peak amplitude in the PNI group on day 7 was significantly lower than that in the ICH group.

CST nerve conduction was assessed using both anterograde and retrograde tracing methods (Figure 5A‐F). Compared to the sham group, both the classical autologous blood‐induced ICH model (ICH group) and PNIPAM‐induced mechanical stress model (PNI group) caused significant damage to the CST (p < 0.01). Moreover, CST integrity was more severely impaired in the PNI group than in the ICH group (p < 0.01).

FIGURE 5.

Corticospinal tract (CST) nerve conduction changes in the autologous blood intracerebral hemorrhage (ICH) model and the PNIPAM‐induced mechanical stress model. (A) Schematic of anterograde CST tracing: RAAV‐hSyn‐EGFP was injected into the mouse motor cortex, and C3–C7 axial spinal cord sections were collected for fluorescence intensity analysis 21 days postmodeling. (B) Representative anterograde tracing fluorescence images of the three groups. Scale bar: 50 μm. (C) Anterograde tracing fluorescence density was significantly lower in the PNI and ICH groups than in the sham group and was markedly lower in the PNI group than in the ICH group. (D) Schematic of retrograde tracing: RAAV‐CAG‐mCherry was injected into the mouse L2–L4 spinal cord, and fluorescence‐labeled neurons were quantified in the motor cortex (coronal section). The three injection sites in the spinal cord are also indicated in the images. (E) Representative retrograde tracing fluorescence images for the three groups. Scale bar: 50 μm. (F) The number of retrograde‐traced fluorescently labeled neurons was significantly reduced in the PNI and ICH groups compared to those in the sham group; moreover, the number of these neurons was significantly reduced in the PNI group than in the ICH group. ^## p < 0.01, **p < 0.01.

3.4. Alterations in signaling pathways and protein expression

RNA‐seq was performed to detect gene expression in the PNIPAM‐induced mechanical stress model (Figure 6A). KEGG pathway enrichment analysis revealed significant activation of stress‐related pathways in this novel model, including the tumor necrosis factor (TNF) and fluid shear stress signaling pathways (p < 0.01) (Figure 6B). A volcano plot of differentially expressed genes identified 1622 upregulated and 1316 downregulated genes in the PNI group compared to the sham group, with significant upregulation of TNF, nuclear factor kappa light chain enhancer of activated B cells (NF‐κB), and interleukin‐1 (beta) IL‐1β genes (p < 0.01) (Figure 6C). Gene set enrichment analysis (GSEA) further demonstrated the upregulation of the TNF signaling pathway and fluid shear stress signaling pathway in the PNI group relative to that in the sham group (Figure 6D,E).

FIGURE 6.

Gene expression changes around the lesion at 24 h postmodeling in the PNIPAM‐induced mechanical stress model. (A) At 24 h postmodeling, RNA was extracted from the tissue surrounding the lesion in sham and PNI group mice for RNA‐seq analysis (n = 3). The red dashed circle indicates the lesion area, and the green dashed circle denotes the sampling region. (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed activation of pathways such as tumor necrosis factor (TNF) signaling and fluid shear stress in the PNI group. (C) Volcano plots showed the number of significant differentially expressed genes, with upregulated expression levels of TNF, nuclear factor kappa light chain enhancer of activated B cells (NF‐κB), and interleukin‐1 (IL‐1β) in the PNI group compared to those in the sham group. Differentially expressed genes were identified using thresholds of log2 (fold change) > 1 and adjusted p‐value (adj.p.Val) < 0.05. (D, E) GSEA (Gene Set Enrichment Analysis) showed significant upregulation and activation of the TNF and fluid shear stress signaling pathway in the PNI group compared to those in the sham group.

We also evaluated the expression of these relative cytokines using Western blot (Figure 7A–C). The results confirmed significant upregulation of the protein levels of TNF‐α, p‐NFκB, and IL‐1β in both ICH and PNI groups, with significantly higher levels in the ICH group than in the PNI group (p < 0.05 or p < 0.01). To determine whether the novel model induces alterations in WMI markers, NFH and MBP were analyzed by Western blot assay (Figure 7D–E). NFH and MBP expression levels were significantly downregulated in both ICH and PNI groups compared to those in the Sham group (p < 0.05 or p < 0.01).

FIGURE 7.

Protein expression changes in the autologous blood intracerebral hemorrhage (ICH) model and the PNIPAM‐induced mechanical stress model. (A–C) Top: Representative Western blot assay images of tumor necrosis factor α (TNF‐α), interleukin 1β (IL‐1β), and p‐NF‐κB of the three groups. Bottom: Quantitative analysis showed upregulated expression of these proteins in both ICH and PNI groups, with significantly higher levels in the ICH group than in the PNI group. (D, E) Top: Representative Western blot assay images of NF200 and myelin basic protein (MBP) for the three groups. Bottom: Quantitative analysis of white matter injury (WMI) marker proteins NF200 and MBP indicates significant WMI in both ICH and PNI groups. #p < 0.05, ##p < 0.01; *p < 0.05, **p < 0.01. n.s = no significance.

4. DISCUSSION

ICH is a severe and life‐threatening condition that predominantly manifests within the brain parenchyma. ¹⁸ The basal ganglia region is the most common site of ICH, accounting for 50%–60% of total cases, largely due to the distinctive hemodynamic stresses on the lenticulostriate arteries. ¹ , ¹⁹ A crucial determinant of motor outcome following basal ganglia ICH is whether the hemorrhage extends into the internal capsule—a critical region traversed by the pyramidal tract. ²⁰ This tract comprises the CST, which is responsible for voluntary body movement, and the corticobulbar tract, which controls facial movements. ²¹ Despite its clinical significance, a major gap persists in preclinical research, as there is a lack of animal models that specifically replicate injury to the white matter tracts within the internal capsule.

Several models have been developed to simulate the mass effect of basal ganglia hemorrhage. Early studies, including those by Lopez Valdes et al. ²² and Sinar et al., ²³ used balloon inflation techniques in rats to investigate the implications of acute compression injury. Although these balloon models effectively induce immediate mechanical damage, their clinical relevance is constrained by high cost and the impracticality of deploying the balloon apparatus, which limits their ability to mimic the sustained mass effect of a hematoma. Subsequently, Schlunk et al. ²⁴ used the Onyx liquid polymer model in mice to study secondary bleeding. However, its excessively dense and rigid structure does not fully replicate the biomechanical properties of an actual hematoma. In contrast, the PNIPAM hydrogel rat ICH model developed by Gong et al. ¹² offers a promising alternative because of its biocompatibility and mechanical properties that closely mimic those of an actual hematoma. Although the previous study utilized PNIPAM hydrogel in rats to create a mass effect, our present work advances this concept by specifically targeting the internal capsule in mice. This refined approach enables a more focused investigation of the mechanical stress component associated with ICH. It targets a well‐defined motor pathway—the CST—and facilitates detailed analysis of white matter pathology at the anatomical, functional, and molecular levels. Moreover, a mouse model based on PNIPAM hydrogel has not been reported to date.

PNIPAM is a thermosensitive hydrogel characterized by a distinct transition from a liquid phase to a solid phase upon exceeding its lower critical solution temperature (~32–34°C). ²⁵ This phase transition is driven by a shift in polymer–water interactions from hydrophilic to hydrophobic dominance. PNIPAM is generally considered to possess favorable biocompatibility, underpinning its diverse biomedical applications. ²⁶ These include temperature‐sensitive drug entrapment and controlled release in drug delivery, injectable in situ gelling scaffolds for tissue engineering, and a landmark technique in cell sheet engineering, where intact cell layers can be noninvasively harvested through a simple temperature reduction process, thereby preserving extracellular matrix components. ²⁷

In the present study, we developed a novel mouse model of mechanical stress injury to specifically investigate the impact of mechanical stress on the CST. This was achieved through the stereotaxic injection of 15 μL PNIPAM gel into the internal capsule. Martucci et al. ²⁸ performed in vivo magnetic resonance imaging (MRI) and estimated whole‐brain volume of adult male C57BL/6 mice as 527.2 ± 4.9 mm³. Accordingly, a 15 μL volume corresponds to approximately 37 mL in the human brain (based on the human brain volume of 1294 ± 107 cm³). ²⁹ In current autologous blood ICH models, 25–30 μL of blood is typically infused at the target site in C57BL/6 mice, ⁹ , ³⁰ a volume that may be excessive compared to that observed in clinical practice. The results of the present study also indicate that a 15‐μL blood injection did not produce significant motor deficits in the open‐field test and BMS score at 7 days post‐ICH (Figure 4); this finding markedly differs from the clinical scenario of 30–40 mL ICH in basal ganglia of patients. This discrepancy may be attributed to the anatomical and physiological differences between rodents and humans. Thus, a substantial gap remains between animal models and clinical practice. Moreover, our approach consistently produced a well‐defined, space‐occupying lesion, leading to significant pathological and functional deficits that mirror key features of human ICH involving the internal capsule. Notably, this model isolates the mechanical component of the injury.

Our model effectively recapitulates the critical features of WMI. Pathological analysis revealed the signs of demyelination, including myelin splitting, bubbling, and a significant increase in the G‐ratio. These structural abnormalities were associated with persistent motor deficits, as confirmed by behavioral testing. The strong correlation between severe CST damage and motor impairment supports the premise that direct mechanical disruption of this tract is a primary driver of hemiplegia following ICH. Notably, our model induced more severe CST disruption than the traditional autologous blood–injection model, as confirmed by anterograde and retrograde neural tracing; this finding suggests improved recapitulation of the direct mechanical injury to the CST. This structural damage was functionally confirmed by a concurrent reduction in MEP amplitude.

The results of transcriptomic and protein‐level analyses of the lesion area suggest that mechanical stress may activate an inflammatory cascade. These findings indicate that mechanical stress may promote the production of TNF‐α, a critical pro‐inflammatory cytokine. ³¹ , ³² This activation might upregulate downstream inflammatory mediators such as NF‐κB and IL‐1β, as demonstrated by Western blot assay at the protein level. ³³ , ³⁴ , ³⁵ These cytokines can adversely affect oligodendrocyte viability and impair myelin repair, thereby likely contributing to the progression of WMI. Concurrently, the downregulation of NFH and MBP expression offers a molecular basis for the observed axonal damage and demyelination. Enrichment of the fluid shear stress pathway in RNA‐seq analysis suggests that the vascular system near the lesion site, which includes cerebral blood vessels and the brain lymphatic system, may also respond to mechanical stress and potentially contribute to secondary injury.

Although the present model provides a valuable tool to investigate mechanical stress–induced injury, it has several limitations. First, inherent genetic and anatomical differences between rodents and humans may limit the translational relevance of this research. Second, technical constraints prevent the precise replication of lenticulostriate artery rupture and the resulting hemorrhage in the posterior limb of the internal capsule within the basal ganglia. Additionally, as clinical ICH is more prevalent in elderly populations, establishing this model in aged mice would improve its clinical applicability and translational value. Lastly, although our RNA‐seq analysis revealed the enrichment of fluid shear stress–related pathways, further investigations are required to elucidate the downstream mechanisms.

In summary, we established a reproducible mechanical stress–induced model that consistently produces WMI, motor deficits, and CST disruption. Evidence from this model advances our understanding of direct mechanopathological mechanisms in the white matter and suggests that TNF signaling may represent a key mediator in this process. These findings provide a robust preclinical platform for evaluating potential therapeutic strategies aimed at mitigating WMI resulting from the mechanical stress of ICH.

AUTHOR CONTRIBUTIONS

Mingxi Li: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; visualization; writing – original draft. Yujie Chen: Data curation; formal analysis; investigation; methodology; validation; visualization. Rongsu Huang: Data curation; formal analysis; investigation; methodology. Min Xia: Conceptualization; formal analysis; methodology; resources. Chao Mi: Conceptualization; formal analysis; methodology; resources. Yuan Tian: Data curation; investigation; methodology. Peiwen Guo: Data curation; formal analysis; investigation; methodology. Taotao Jin: Data curation; formal analysis; investigation; methodology. Shilei Hao: Conceptualization; methodology; resources. Yujie Chen: Conceptualization; funding acquisition; methodology; project administration; resources; supervision; writing – review and editing. Zhi Chen: Conceptualization; funding acquisition; methodology; supervision. Hua Feng: Data curation; formal analysis; funding acquisition; investigation; methodology.

FUNDING INFORMATION

This work was supported by the National Natural Science Foundation of China (grant no. 82371361 to Zhi Chen and no. 82371333 to Yujie Chen), the Natural Science Foundation of Chongqing (grant no. CSTB2025NSCQ‐LZX0044 to Yujie Chen), and the Chongqing Municipal Health Commission (grant no. YXGD202451 to Yujie Chen).

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no competing interests.

ETHICS STATEMENT

All experiments were approved by the Laboratory Animal Welfare and Ethics Committee of Third Military Medical University (approval no.: AMUWEC20232125) and reported in accordance with the ARRIVE 2.0 guidelines.

Supporting information

Figure S1. Regions of interest (ROIs) for anterograde and retrograde tracing of CST and histopathology of the PBS and ICH groups. (A) Representative anterograde tracing images of CST nerve fibers (green fluorescence). ROIs are marked with white dashed boxes, and enlarged sample images are provided. Scale bar: 200 μm (Left), 20 μm (Right). (B) Representative gross IF images of the PBS group. White arrowhead indicates that PBS injection produces a minor space‐occupying lesion; white arrow indicates the direction of needle insertion. Scale bar: 1 mm (C) Representative retrograde tracing images of corticospinal neurons (red fluorescence). ROIs are marked with white dashed boxes, and enlarged sample images are provided. Scale bar: 500 μm (Left), 25 μm (Right). (D) Representative gross slice of the ICH group. Black arrowhead indicates blood reflux; black arrow indicates a hematoma.

Li M, Chen Y‐J, Huang R, et al. A mouse model of mechanical stress injury to the basal ganglia using thermosensitive PNIPAM hydrogel for intracerebral hemorrhage research. Anim Models Exp Med. 2026;00:1‐13. doi: 10.1002/ame2.70235

DATA AVAILABILITY STATEMENT

All the data generated or analyzed during this study are included in this published article and supplemental materials. The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.