Material-Driven Therapeutics to Establish a Penetrating Traumatic Brain Injury Rat Model and Implantation of a 3D-Printed Scaffold: Pre-Experimental Pilot Study

Introduction

Traumatic brain injuries (TBIs) are a leading cause of death and disability [1,2]. Annually in the United States, approximately 1.7 million people have a TBI event, resulting in 50,000 deaths [1]. Currently, it is estimated that 3.2 to 5.3 million people are living with a TBI-related disability [1]. Penetrating TBIs are the most lethal form, with a foreign object breaking through the skull, disrupting the blood-brain barrier (BBB), and damaging the brain tissue [3,4]. These are most often caused by accidents, including household firearm accidents, falls, and motor vehicle collisions [2-4]. The severity of the primary injury is worsened by circulating immune cells foreign to the local environment crossing the disrupted BBB, causing persistent inflammation and generation of reactive oxygen species that result in loss of neural tissue and damaged vasculature [1,2,5]. This damage results in severe clinical symptoms, including high cranial pressure and hemorrhage. While no treatments for TBI exist, management strategies include various neurosurgical interventions and pharmacological and nonpharmacological methods [2-4,6-8]. While surgical interventions may improve survival outcomes for severe TBIs, they come with increased risk of infection and longer recovery times and may not result in greater functional outcomes [3]. Currently, researchers are investigating tissue engineering and regenerative medicine strategies to develop biomaterial scaffolds and cellular therapies that may improve upon these poor outcomes [2,6,7,9].

Animals are a necessary preclinical step for translational research in the development of novel therapeutics and can help generate TBI solutions for both animal and human medicine [10]. While large animals such as pigs are important models for TBI, showing more complexity and gray and white matter ratios similar to those of human brains, using rodents such as rats or mice prior to large animal models is a crucial step [11,12]. Outside the advantages of low costs and easy handling, rats offer a high level of standardization and can be assessed for functional outcomes and the pathophysiology of cellular damage in a way that mimics the human brain [11]. Several preclinical models exist to mimic TBIs, including controlled cortical impact injury models, weight drop models, and penetrating ballistic-like models [13]. There are a wide range of TBIs, and while each model excels in representing a specific type of scenario, they do not consistently disrupt the BBB while maintaining controlled, focal damage to the cerebral cortex, as intended in this study [13,14]. Severe mechanical models often impact the brain as a whole, simulating concussion-like injuries alongside the penetrating impact and resulting in high levels of fatality [14]. Meanwhile, the penetrating ballistic-like model, while remaining focal, creates a deep cavity mimicking a combat setting. Adding to the available preclinical models to accurately represent the range of TBIs may alter how clinical care is managed and be useful in evaluating more specific, novel therapeutics [15].

Two oxidized derivatives of graphene nanoparticles, specifically reduced graphene oxide (rGO) and graphene oxide, have been investigated as nanocomposite components for nerve tissue engineering due to their elemental composition and topographic features, which influence their tunable mechanical properties and have a positive effect on cell growth [16,17]. Specifically, rGO can be engineered using a variety of 3D printing techniques to form scaffolds that attach, proliferate, and differentiate viable exogenous or endogenous stem and progenitor cells to aid in nerve repair and regeneration [17-19]. In particular, rGO features the unique property of electrical conductivity to improve functional nerve restoration [16,17,20]. While many studies have focused on graphene’s potential in the peripheral nervous system, graphene and its derivatives have also shown to be a promising component for central nerve injuries [21-23]. Controversy surrounding rGO’s biocompatibility has slowed its translation to clinical use. Numerous variables, such as surface functionalization, particle shape and size, dispersion, concentration, dosage, route of administration, and processing techniques, can all influence how cells respond to rGO [16,20,23-29]. Therefore, we must evaluate whether our specific rGO-containing construct is safe to use in this delicate tissue model.

For 3D printing, graphene-based nanoparticles are always blended with synthetic polymers, including polycaprolactone, polyurethane, polylactic acid, and poly(lactic-co-glycolic) acid (PLGA). PLGA is Food and Drug Administration approved for biomedical devices, primarily because of its tailored biodegradability (ie, its breakdown products, lactic and glycolic acids, are safely metabolized by the body). While PLGA is a Food and Drug Administration–approved and very well–characterized synthetic polymer, graphene’s biocompatibility is highly variable depending on a variety of tunable physicochemical and mechanical properties [30]. Hence, it is important to evaluate the biocompatibility of a PLGA-rGO construct in vivo.

In this pilot study, we performed a craniotomy in rats to manually cut the BBB, managing the disruption to the brain tissue. This allowed us to assess the feasibility of our therapeutic scaffold implantation and the safety of the biomaterial-tissue interaction between the brain tissue and our proposed scaffold containing PLGA and graphene. The use of stereotaxic surgical equipment and cranial landmarks allowed for a standardized injury capable of replication. We used a neurological assessment and immunohistochemistry (IHC) to identify whether disruption of the BBB was sufficient to model a TBI while identifying whether a foreign body reaction occurred with the implanted materials. For this study, we used a previously fabricated scaffold to evaluate the graphene’s biocompatibility with the brain tissue. We hypothesized that the disruption of the BBB would cause behavioral, physical, and histological evidence of a TBI, whereas the implanted material would not show evidence of a foreign body rejection or chronic inflammation 30 days after surgery.

Methods

Biochemicals, Chemicals, and Disposables

All biochemicals, cell culture supplements, and disposable tissue culture supplies were purchased from Thermo Fisher Scientific unless otherwise noted.

Ethical Considerations

Sprague Dawley rats (Charles River Laboratories; N=4; male; aged 6 months) were acclimated prior to beginning procedures. All procedures were conducted in accordance with Public Health Service guidelines for the humane treatment of animals under approved protocols established through the University of Tennessee’s Institutional Animal Care and Use Committee (protocol 2954-0623). The animals were housed in a 12-hour light and dark cycle with ad libitum food and water.

Surgical Procedure

Rats were anesthetized using inhalant isoflurane, and buprenorphine (0.05 mg/kg) was administered subcutaneously prior to surgery. The surgical area was shaved, and eye lubricant was applied. Under aseptic conditions, the rat was mounted onto the stereotaxic apparatus (Kopf Instruments) using a nose cone to maintain anesthesia throughout the procedure [7,31,32]. The dorsal and ventral adaptor was aligned to the −8 setting; the ear bars were aligned to the 7.5 setting on either side. The area was draped, and the scalp was cleaned using iodine prior to cutting a 1- to 2-cm longitudinal incision to expose the skull. The periosteum was gently teased back from the skull, and the area was cleaned using 95% ethanol. A handheld microdrill (Stoelting Co.) with a sterile 5-mm drill bit was used to create a craniotomy to the left of the sagittal suture, midway between the lambda and bregma landmarks. The dura mater was damaged by cutting the layer using forceps and a beaver blade. A 65:35 PLGA (Sigma-Aldrich) and 0.5% rGO (Cheap Tubes) scaffold from a previous study was cut to the 5-mm size and implanted on the cerebral cortex [33]. Briefly, PLGA and rGO were blended with 0.5-mL dimethyl sulfoxide and melted into a homogeneous mixture for extrusion-based 3D printing using the CELLINK BIO X6 printer. The scaffold was printed in 15 filament layers with approximately 80% porosity and in 5-mm (x-axis), 5-mm (y-axis), and 2-mm (z-axis) dimensions. After implantation, the area was covered with bone wax, and the scalp was closed with 4-0 absorbable surgical sutures (Ethicon, Inc). The rats recovered on a heating pad and were individually housed and monitored twice daily for the first 3 days, daily for 1 week, and weekly throughout the remaining 30-day study period. Buprenorphine was administered every 12 hours for 3 days after surgery, and ketoprofen (5 mg/kg) was administered every 24 hours for 3 days after surgery. Baytril (100 mg per 400 mL) mixed into water with flavored Gatorade was provided for 1 week after surgery.

Neurological Function Evaluation

Animal recovery and neurological function were evaluated for each rat on days 1, 2, 3, 7, 14, and 21 after surgery. A modified neurological severity score was used to assess the body weight, general condition, physical appearance, breathing, spontaneous behavior, handling reaction, wound healing, and neurological evaluation of each rat [11,34-36]. These behaviors were graded on a scale from 0 (normal) to 30 and above (severe) to determine any neurological deficits that occurred from the cranial intervention or material implantation.

IHC of Brain Tissue

At the 1-month end-of-study time point, the intact brain was removed as previously described and placed in 10% formalin [37,38]. The cerebellum was removed, and the brain was dissected down the midsagittal plane, dividing the left and right hemispheres of the cerebral cortex. Each half was paraffin embedded and longitudinally sectioned. Sections were deparaffinized, hydrated, unmasked, stained, and mounted as previously described [30]. One histology section from all specimens was stained with hematoxylin and eosin (H&E; Azer Scientific, Inc) for assessment of cellular detail, and one histology section from all specimens was stained with Prussian blue for assessment of hemosiderin deposition using a 30-minute stain with 1:1 hydrochloric acid and potassium ferrocyanide. IHC was performed on the remaining sections using glial fibrillary acidic protein (GFAP; 1:500), neuronal nuclei (NeuN; 1:500; Cell Signaling Technology, Inc), von Willebrand factor (vWF; 1:1000; Abcam), neurofilament light chain (NEFL; 1:20), ionized calcium-binding adapter molecule 1 (Iba1; 1:500), and CD68 (1:500). The secondary antibodies used were goat antirabbit immunoglobulin G horseradish peroxidase (1:500) and rabbit antimouse immunoglobulin G horseradish peroxidase (1:500; Abcam). Briefly, GFAP identifies astrocytes, NeuN identifies neurons, vWF identifies vascularization, NEFL identifies neural cytoskeleton structure, Iba1 identifies microglia, and CD68 identifies phagocytic activity [39-44]. Sections were imaged using a BZ-X Series all-in-one fluorescent microscope (Keyence Corporation) at 5X, 10X, 20X, and 40X magnification. The H&E sections were analyzed by a trained histologist. For each section, multiple images were taken along the length of the section with 20% overlap and stitched together to create a single image of the entire section. Scale bars were added. ImageJ software was used to analyze the stitched images [45]. They were selected for optimal contrast (blue) and threshold (pixel intensity values ranged from 150 to 250). The nerve was isolated from the background and converted to a black-and-white image (black=stained tissue).

Statistical Analysis

For each section, the nerve was measured for percentage of area (percentage of black in the image) [30]. Mean and SD values were obtained. Prism (version 10; GraphPad Software) was used to perform a Shapiro-Wilk test of normality. When all data were determined to be normally distributed, t tests were performed for statistical significance between the control and injured tissues. Statistical significance was determined as a P value of .05 or less.

Results

A TBI is a complex disease in which a primary injury results in an immediate disruption of brain tissue caused by an external source, leading to contusion, hemorrhage, and axonal shearing. The secondary injury evolves over the following hours, days, and months, involving a cascade of degenerative molecular events leading to cell death [13,19]. Patients who experience focal, penetrating TBIs have areas of brain laceration, disruption of the leptomeninges, subdural hemorrhage, and cerebral edema [43]. Our method of injury manually disrupted the dura mater and cerebral cortex with a small blade through the cranial opening (Figure 1A). When the scaffold was placed onto the exposed brain tissue for rats 1, 2, and 4, there was minor bleeding before the scaffold was placed against the brain tissue with mild pressure. With rat 3, there was heavy bleeding after the surgical procedure, and the scaffold was placed onto the brain tissue after ensuring that the bleeding was under control. Evidence of the differences in the surgical implantation was observed 30 days later during the brain dissection. In rats 1, 2, and 4, the scaffold was more strongly incorporated into the surrounding bone of the craniotomy than into the brain tissue (Figure 1B). Some theories as to why this may have occurred include the location of the implanted material, which may not have been pushed sufficiently through the cranial opening; the large z-axis of the implanted scaffold possibly interfering with the implantation process; or excessive cerebral pressure or edema possibly pushing the construct back into the opening due to insufficient closing using the bone wax. Meanwhile, rat 3 showed a greater incorporation into the brain tissue (Figure 1C). This may be due to the more severe damage to the brain, evident in the heavy bleeding and deeper implantation of the scaffold.

‎

During the 30-day study, the neurological assessment evaluated the physical and behavioral changes that may occur with a TBI [11,34-36]. At the 6 time points, all rats showed no changes in any of the areas assessed (Table 1). Rat 3 took slightly longer to come out from anesthesia, which was expected due to the longer procedure time to control the heavy bleeding. However, while rat 3 showed slightly slower spontaneous movement, it still fell within normal parameters. While this indicated no immediate concerns regarding a foreign body rejection of the implanted PLGA and rGO scaffold, it also indicated that the TBI model may not have been severe enough to implicate neurological deficits common with the disease. Other than TBI severity, the points of neurological evaluation may be expanded to include additional behavior tests, identifying more minute behavioral changes in future studies. Monitoring of intracranial pressure throughout a future study may also aid in confirmation of a TBI as it is a signature symptom of the injury.

At the end of the 30-day study, H&E-stained sections were evaluated for local tissue reactions. Similar to the lateral fluid percussion injury model where the TBI is limited to 1 cerebral hemisphere, the contralateral side may be used as a comparison for neural injury [12,13]. As the injury was limited to the left cerebral cortex, the right cerebral cortex was used as uninjured control tissue. The H&E-stained tissue confirmed that no significant pathology was identified in the control tissues for all 4 rats (Figure 2A). For the injured side, rats 1, 2, and 4 showed a slight indentation to the cortex with mildly thickened leptomeninges, reactive small blood vessels, increased stromal cells, and hemosiderin deposition (Figures 2B and 2C). Hemosiderin deposition is a sign of injury, often occurring after hemorrhage, and was confirmed via Prussian blue staining. Together, these findings indicate a mild focal cerebral meningeal fibrosis injury. This type of chronic response may occur after a penetrating TBI [46]. Rat 3 showed a cavitated region of the cerebrum where the meninges was severely expanded by a mass (Figures 2D-2F). The mass was composed of neutrophils and necrotic cell debris, marginated by histiocytic infiltrate, and a periphery of fibrovascular tissue with aggregated plasma cells, lymphocytes, and occasional multinucleate foreign body giant cells. Some of the rGO deposits in the area were associated with distinct vacuoles. Hemosiderin deposition was also evident and confirmed through Prussian blue staining. No evidence of infectious agents was present. Overall, rat 3 had evidence of severe focal chronic pyogranulomatous meningitis and a foreign body reaction to the rGO nanoparticles. It was difficult to decipher whether the injury was more severe in rat 3 or whether the strong inflammatory and immune response was solely due to the scaffold. Future studies will further investigate the separation in these data using a sham surgical group and an untreated TBI group.

‎

The IHC markers that were evaluated had previously shown changes after TBI. After an injury, the glial markers, GFAP and Iba1, increase, whereas the neural marker, NeuN, decreases due to glial scarring and the loss of damaged neural cells in the area, respectively [40,41,43]. vWF increases to identify vascular damage after a severe injury but typically maintains its expression with a mild injury [40]. NEFL increases to identify fragmented neural structures, and CD68 increases to indicate a rise in phagocytic activity [39,43]. Qualitatively, each of these 6 markers showed “normal” anatomical structures (Figures 3A-3F). Quantitatively, GFAP, vWF, NEFL, Iba1, and CD68 showed no significant differences in expression between the injured and control tissues (Figure 3G). However, NeuN expression significantly decreased from 9.65% to 4.56% of area stained in the injured tissue in all rats (P=.02). A significant decrease in NeuN is expected after a TBI, with the associated loss of neural cells due to the sustained damage to the tissue [40,43]. When rat 3 was excluded from the statistical analysis, statistical significance remained unchanged. The remaining markers may not have indicated any changes due to the 30-day time frame. It has been shown that these markers return to preinjury levels by 14 days after a TBI in mild cases [43,44].

Overall, all rats showed evidence of a TBI histologically at 30 days after surgery, although they were asymptomatic according to the neurological assessment. Rats 1, 2, and 4 showed evidence of a mild, focal, penetrating TBI according to the H&E staining and IHC marker (NeuN). Rat 3 showed evidence of foreign body rejection that could not be separated from evidence of a severe TBI.

‎

Discussion

Rat TBI models are diverse and designed to simulate human clinical scenarios and injuries. As a result, the models vary in their severity, location, and ease of animal handling [12]. As our goal in this study was to establish a tissue engineering strategy to address TBI, it required optimization at several steps. For instance, our strategy included selecting the animal in which to create the model; 3D printing the nanocomposite that we wanted to evaluate in vivo; establishing and optimizing procedures for implantation of the biomaterial; and, finally, validating protocols to monitor and evaluate the animals and assess the performance of the biomaterial scaffold. A pilot study of 4 Sprague Dawley rats was established and conducted as presented in this paper. We performed a relatively simple and quick craniotomy in rats to manually cut the BBB. The stereotaxic equipment was set to a consistent setting with respect to the cranial landmarks, allowing us to perform the surgery in a reproducible manner. Using our strategy, and as outlined in the Results section, we were able to prove the feasibility of this study, which allowed for scaffold implantation and in vivo safety evaluation of the biomaterial-tissue interaction. Finally, we were able to establish a scoring system to assess neurological features correlated with the defect. We created a neurological assessment based on published reports. We used histological and immunohistochemical staining to identify whether disruption of the BBB was sufficient to model a TBI, proving that the novel biomaterial was biocompatible, and to evaluate the expression of specific neural target proteins to identify biomarkers for TBI for future studies.

Despite assessing only 4 rats in this study, we were able to validate our system, which could be applied to any future neural tissue engineering project focused on TBI. While this pilot study did not use a surgical control group, the use of a sham surgery in a larger study and with a greater number of animals for statistical analysis can separate the data to determine whether the injury identified through H&E staining was due to a severe TBI or to the implanted scaffold and whether the variability observed in 1 of the 4 rats evaluated was a concern in this model of TBI. Additionally, the surgical procedure was successful for scaffold implantation, although a construct with mechanical properties more aligned with those of the cortex will aid in the successful integration into the appropriate tissue [9,47,48]. A period of 30 days was chosen as this is the estimated time that a scaffold should remain implanted before degradation [49]. This time frame best allows for the alignment of complete degradation with tissue repair and regeneration in the cerebral cortex. The implanted scaffold did not fully degrade at 30 days in vivo and will need to be altered to accelerate degradation within this time frame. Future studies can evaluate the surgical groups at multiple time points, including a shorter time frame such as 24 hours or 3 days, to better confirm the acute injury (0-3 days) using TBI biomarkers such as those evaluated via IHC [19,43,44]. This will also allow for the evaluation of the therapeutic potential of implanted scaffolds and graphene nanoparticles at the acute and chronic TBI phases [9].