Spatial Distribution of Neuropathology and Neuroinflammation Elucidate the Biomechanics of Fluid Percussion Injury

Highlights

Introduction

Clinical and experimental traumatic brain injury (TBI) involve a primary injury and subsequent pathophysiologies that dismantle, repair, and regenerate circuits in the brain.^1–3 In response to the mechanical injury, adaptive repair and regeneration fail to reconstitute the original neuronal circuits, leaving a miswired brain and neurological impairments that decrease quality of life. Clinically, this process is relevant to veterans, athletes, survivors of interpersonal violence, children, and the elderly who experience one or more TBIs.^4,5 We and others apply experimental models of TBI to study the acute and chronic events associated with this disease and use these models to develop therapeutic interventions.

Various models of experimental TBI have been developed, where fluid percussion injury (FPI) is one of the best characterized animal models of TBI.^1,6–9 Specifically, the midline FPI (mFPI) model produces a diffuse, concussive-like TBI in rodents, whereas lateral FPI (lFPI) produces a mixed focal and diffuse injury.^1,8,10–12 To this end, the primary FPI neuropathology is diffuse axonal injury, rather than overt cell death. Clinical relevance of these injuries includes a transient suppression of neurological reflexes and acute motor deficits.^{6,7,11,13–29} In the more chronic phases of injury, evidence of somatic, cognitive, and affective symptoms emerges. Somatic morbidity includes a hypersensitivity to facial whisker stimulation, similar to agitation in people.^30–36 Cognitive performance is degraded in short, long, and working memory, using various cognitive testing modalities.^{11,33,37–47} Affective symptoms are identified by disruption of circulating hormone levels and responses in anxiety tests.^11,48–50 Thus, FPI affords investigations into the development and maintenance of cognitive, somatic, and affective morbidities, which parallel clinical impairments.

Despite the broad implementation of the FPI model in the neurotrauma field, biomechanical models fail to accurately explain the resultant distribution of pathology.^51–58 In the initial implementation of FPI in rodents, Dixon and colleagues describe that this technique does “not attempt to reproduce rapid acceleration-deceleration of the head…. Rather, fluid-percussion brain injury successfully produces graded levels of injury associated with predictable neurologic, physiologic and histologic changes that are comparable to those observed in human brain injury.” ¹⁵ In an approach to observe the mechanical forces of injury on the brain, Dixon and colleagues conducted high-speed cineradiographic studies of the brain inside the skull over the 15 msec of mFPI injury. ¹⁵ In this way, the primary injury was first characterized as “intracranial fluid movement … by rapid radial movement within the epidural space …. suggesting that the image of the indentation acutely may have been caused by lateral fluid displacement following the curvature of the skull.” ¹⁵

Thereafter, subsequent publications comment on the resultant pathology. Hicks and colleagues reported that “it is interesting to note that the primary site of cortical damage is ventrolateral, rather than directly underneath the impact site.” ⁵⁹ These researchers continue to explain this phenomenon as a consequence of biomechanical forces on selectively vulnerable tissue, based on regional or cellular cytoarchitecture. Over the 30+ years of research using FPI in rodents, curious pathology has been reported as focal damage far more lateral than predicted by the location of the applied mechanical forces of injury and remains an enigma. For the midline variant of FPI, the core pathologies are identified millimeters more lateral from the injury site. Further, cortical pathology does not necessarily align with cytoarchitectural landmarks, as posited by Hicks and colleagues. ⁵⁹ The curious pathology occurs far lateral from the injury site in other non-focal TBI models. ⁶⁰

In this communication, we reference the range of pathologies occurring far lateral from the injury site to include, but in no way are limited to, blood–brain barrier disruption, axotomy, plasma membrane permeability, and cell death. Upon re-examination, we recognized that the curious pathology of diffuse TBI was tracked beneath the temporal ridge of the skull. We make a case for differential thickness along the rodent skull as a contributing factor to the direction of biomechanical forces of diffuse brain injury, in addition to inherent properties of the tissue. Thus, we hypothesized that a local thickening of the rodent skull at the temporal ridges serves to focus mechanical forces of brain injury and generate predictable pathology in line with the temporal ridges.

Methods

Results

Rat skull is thicker at the temporal ridge

Naïve rat skulls were cleaned of all tissue using dermestid beetle larvae. Prominence of temporal ridges on the dorsal surface of the skull were evident (Fig. 1). Subsequent in vivo imaging was undertaken to demonstrate the relationship between shapes of the skull and brain. Oblique sections of a 7T MRI in naïve rats were prepared to visualize skull thickness with respect to the brain. Coronal (Fig. 2A,B), horizontal (Fig. 2C,D), and oblique sagittal (Fig. 2E–H) slices were all registered to pass through the temporal ridge of the skull (evident in black). Note that thickness of the temporal ridge extended along the anterior to posterior length of the skull (Fig. 2C). The internal surface of the bone contacting the dura and brain is a smooth surface devoid of macrostructure. To confirm imaging results, coronal sections of rat skulls were taken from rostral to caudal, and contour analysis was performed (Fig. 3A–E). Measurements were taken along the temporal ridge and calvarium (n = 4) of rat skulls (Fig. 3F). The temporal ridge was found to be 75% thicker than the calvarium (t = 4.36; p < 0.01).

OPEN IN VIEWER

FIG. 1.

Rat heads were cleaned by flesh-eating beetle larvae until the skull was clear of all carrion. By gaining an anatomical perspective of the adult rat skull, the temporal ridge clearly protrudes on either side, indicating the increased bone mass in this region.

OPEN IN VIEWER

FIG. 2.

Oblique sections of naïve rat skull using a 7T MRI demonstrated the thickening of the rat skull along the temporal ridges. Coronal schematic (A) and section (B) present the conventional view of the rat brain. Oblique sections (C–H) are identified on the schematic at cross through the space beneath the temporal ridge. Temporal ridges are identified with the solid arrow head. MRI, magnetic resonance imaging; T, Tesla.

OPEN IN VIEWER

FIG. 3.

Local thickening of temporal ridges is shown by contour analysis of coronal sections of rat skull. (A–D) Photographs of rat skull sections are enhanced by black ink on the sectioned edge. Solid arrowheads identify the temporal ridge. (E) Digitized outlines of the sectioned edges show the contour changes of the rat skull from posterior (a) to anterior (d). (F) Measurements were taken from coronal sections of rats to compare thickness of calvarium to the temporal ridge, where the temporal ridge was 75% thicker.

Fluid percussion injury–induced argyrophilic neuropathology under the temporal ridge

Rats were diffuse-brain injured by mFPI and then survived to either 1, 2, 7, or 28 DPI. Brains were then collected, sectioned rostral to caudal, and stained with silver to identify regions of neuropathology that develop after diffuse brain injury. Darker (black) stained regions on sections identify hyperintense deposition of argyrophilic reaction product (Fig. 4). Uninjured sham animals were included in these cohorts and have been published along with more detailed analysis for region-specific neuropathology.^{48–50,64–69}

OPEN IN VIEWER

FIG. 4.

Histological sections of diffuse brain-injured rats are aligned rostral to caudal at 1, 2, 7, and 28 days after midline fluid percussion injury (FPI). Neuropathology was identified by hyperintense deposition of argyrophilic reaction product (amino-cupric silver histochemical technique; black) and occurred primarily along the rostral to caudal extent of sensorimotor cortex. Neuropathology appears to accumulate over 7 days post-injury and mostly subsides by 28 days post-injury. Neuropathology may be diffuse and inconsistent between hemispheres along the rostral-caudal extent of the brain.

For all time points post-injury, argyrophilia was evident lateral to midline, under the temporal ridge, and extended the rostral-caudal length of the brain. This length of pathology lies beneath the temporal ridge, with pathology evident bilaterally over the post-injury course. Argyrophilic staining was predominant in the somatosensory cortex (S1BF), lateral portion of the hippocampus (CA3), and ventral posterior thalamus at 1 and 7 DPI (Fig. 5). This pattern of neuropathology occurred systematically across histological sections, with the deepest penetration of pathology in sections associated with the fluid pulse, but not directly under the site of injury induction (center of the craniectomy). In some sections, areas of increased argyrophilic reaction product were inconsistent across the cortex (Fig. 5), suggesting that a variable other than brain tissue properties may influence the pattern of pathology.

OPEN IN VIEWER

FIG. 5.

Tissue aligned with the temporal ridge (asterisk) showed increased deposition of argyrophilic reaction product (black) over time post-injury. The mechanical forces of diffuse brain injury reflect off the ventral skull into the temporal ridge, thereby inducing pathology in the lateral nuclei of the thalamus and lateral aspects of the hippocampus. Some sections (right side 1 day DPI) showed tissue spared of argyrophilic reaction at the temporal ridge, suggesting that a non-neural variable may influence the pattern of pathology. FPI, fluid percussion injury.

Proposed biomechanical mechanism of rodent traumatic brain injury

Consistent neuropathology occurred beneath and along the temporal ridge after mFPI. We propose a mechanism of injury induced by the mechanical forces of mFPI. Illustrated on a modified coronal MRI section (Fig. 6A), mFPI is initiated (blue arrow) by the fluid pulse and pneumatic forces. This fluid pressure pulse, lasting only milliseconds, then produces mechanical force vectors that propagate throughout the brain (green arcs). Given that the wave propagates through the brain, the force vectors would reflect off the ventral skull, without causing damage. Reflected forces travel dorsal and lateral throughout the cranium, possibly toward the differential thickness of the skull at the temporal ridges (purple arcs). Differential thickness of the skull at the temporal ridges may act as either a pressure sink or a pressure barrier, which ultimately focuses injury-inducing forces back onto the tissue under the temporal ridge. This acts as a “pinch-point” for vulnerable tissue and thereby contributes to the neuropathology observed acutely in the superficial cortical layers (red arrows).

OPEN IN VIEWER

FIG. 6.

Proposed biomechanical mechanism of rodent TBI. (A) Schematic of mechanical forces after midline fluid percussion injury (mFPI) that induce neuropathology. The fluid pulse (blue arrow), generated from the impact of the pendulum on the plunger of a fluid-filled cylinder, lasts only milliseconds and travels through the injury hub into the extradural space, producing mechanical forces that propagate throughout the brain (green arcs). Mechanical forces are then reflected off the ventral portions of the skull and travel throughout the cranium back to the dorsal and lateral portions of the skull (purple arcs). Upon reaching the cortex, increased thickness of the temporal ridge provides a pressure sink or pressure barrier, which ultimately focuses injury-inducing forces on the tissue beneath the temporal ridge, acting as a pinch-point and resulting in observed neuropathology (red arrows). (B) To support this proposed mechanism, temporal ridges were shaved uni- or bilaterally to approximate the thickness of the calvarium before injury. Rats then received mFPI and were prepared for Iba-1 immunohistochemistry at 7 DPI to identify areas of neuroinflammation. Rats who received no shaving to the temporal ridge (top brain slice) show focal increase of microglial activation. However, when the temporal ridge was unilaterally (bottom brain slice) or bilaterally (middle brain slice) shaved, an absence of focal neuroinflammation corresponded with the shaved hemisphere(s). Open arrow heads indicate hemispheres with an intact temporal ridge; solid arrow heads indicate hemispheres with a shaved temporal ridge. DPI, days post-injury; Iba-1, ionized calcium-binding adaptor molecule 1; TBI, traumatic brain injury.

An alternative mechanism of injury would focus extracranial force vectors into the cranium through the temporal ridge, whereby the predicted pathology would initiate at the superficial cortical layers and diffuse ventrally from those points. For either proposed mechanism, the applied forces may remain localized to the originating cerebral hemisphere(s), such that lFPI is lateralized compared to mFPI. We favor the intracranial mechanics proposed mechanism given that pathology shows the largest arc across the cortex at superficial layers.

To support this proposed mechanism, one or two temporal ridge(s) were shaved down before mFPI to approximate the thickness of the calvarium. Rats then received moderate mFPI and were prepared for Iba-1 immunohistochemistry at 7 DPI, the time point with peak neuropathology as identified by de Olmos silver staining. Brains were collected, sectioned, and immunostained with Iba-1 to identify concentrated areas of neuroinflammation, indicative of ongoing neuropathology.^65,70 Representative histological sections show (Fig. 6B) increased focal neuroinflammation (Iba-1⁺ microglial activation) corresponding to hemispheres with an intact temporal ridge. Rats with a single temporal ridge intact showed an absence of focal neuroinflammation in the hemisphere corresponding with the side of the shaved temporal ridge (Fig. 6B, bottom brain section). Rats with both temporal ridges intact showed the predicted concentrated microglial activation in cortical areas beneath the temporal ridge (Fig. 6B, top brain section). Rats with both of the temporal ridges shaved showed diffuse microglial activation (Fig. 6B, middle brain section).

Discussion

Here, we present evidence that the skull's shape and thickness help explain the repeatedly documented pathology that occurs lateral to the injury induction site in experimental TBI. A literature search found dozens of articles with figures that showed aspects of neuropathology preferentially localized to tissue beneath the temporal ridge after mFPI and closed skull TBI models. Changes in skull thickness and skull/brain relationship were confirmed using MRI, skulls, and contour imaging of coronal sections from naïve rat skulls. After mFPI, neuropathology primarily occurs in cortical tissue running the rostral-caudal length of the temporal ridge and includes the S1BF, hippocampus, and ventral thalamus in the dorsal-ventral projections of mechanical force. We thus propose a new biomechanical mechanism of experimental TBI mechanical forces to explain this predictable and consistent neuropathology. Further, the uni- or bilateral surgical thinning of the temporal ridge attenuated focal neuroinflammation under the single shaved temporal ridge after mFPI. Together, these data implicate the shape of the rodent skull, in addition to tissue properties, in defining pathology after experimental brain injury.

Since the implementation of FPI, multiple groups have hypothesized the biomechanics of injury based on injury parameters and extent of tissue pathology, rather than localized tissue pathology and the potential contribution of skull shape.^15,71 As Dixon and colleagues noted, “fluid moves through the epidural space of the brain after FPI.” ⁶ As shown in Figure 3, the rat skull is not uniform in thickness and could influence mechanical force trajectories applied to the head and skull, with the likely consequence of increased strain and stress on neuronal tissue and tissue forced against different parts of the skull. Recent in silico modeling of lFPI included a three-layer hexahedral, element-based skull module with varying Young modulus for each layer. ⁷² The model used a uniformly thick skull and predicted significant strain in regions surrounding the initial injury site that dissipate through the brain, in contrast to our neuropathology results and those of others (Table 1).

An opportunity exists for future finite element models to consider differential skull thickness and shape. Also, although each of these studies describes how the fluid pulse interacts with and thus deforms neuronal tissue, they do not explain the unique pattern of neuropathology observed post-injury, nor the forces reflected off the skull that travel through neuronal tissue. As we characterize here, the rat skull is not a uniform thickness and therefore would not absorb or reflect forces equally. The temporal ridge protrudes externally along the rostral-caudal skull axis, potentially increasing bone rigidity. The calvarium, in contact with the dura and brain, is contoured, smooth, and free of protrusions. One hypothesis to explain the temporal ridge relates to the attachment and early use of the muscles of mastication.^73–75 Suckling and eating after birth can increase strains along the rostral-caudal axis of the rodent skull sufficiently to influence development of the temporal ridge.

In adulthood, the temporal ridge is 75% thicker than the rest of the calvarium, thus providing stiffness along this point of the skull. Undoubtedly, other components of the skull and skeletal development may evolve as the rodent ages. It is not unreasonable to consider that the differences observed in neuropathology between experimental TBI models may be the result of changes to the skull structure as the animal ages, particularly at young ages preceding when the cranial sutures fully fuse. However, the identified literature presents instances of similar pathology across ages and experimental TBI models (Table 1).

Male and female skulls may influence differentially the neuropathology of TBI. Unfortunately, the majority of earlier pre-clinical TBI studies have been performed in male animals, thus limiting the inclusion of sex as a biological variable in the relationship between skull characteristics and neuropathology. Moving forward, inclusion of non-binary sex, non-binary gender, and physiological variables (e.g., hormone cycles) is necessary for a complete understanding of sex-based influence on outcomes post-TBI and may advance personalized medicine. ⁷⁶ With regard to skull development, multiple hormones (e.g., estrogen, progesterone, testosterone, thyroxine, and corticosteroids) interact with bone development and remodeling. ⁷⁷ Fluctuations in hormone levels acutely (e.g., menstrual cycle, estrous cycle) and over time (e.g., menopause) may influence skull structure and therefore how brain injury forces interact with the skull.^76,78 Further, sex differences exist in post-injury neuroinflammation, in terms of Iba-1 staining and microglia morphology, over time and across brain regions after midline FPI and controlled cortical impact (CCI).^79,80 Ultimately, the mechanics of TBI must consider the skull, in addition to the influence of sex, gender, and physiological hormones.

Evidence across the literature indicates that tissue directly underneath the temporal ridge, rather than the impact site, routinely shows TBI-induced neuropathology. We show evidence of this pathology throughout the rostral-caudal extent of coronal sections. This neuropathology, evidenced in multiple histological outcomes by the neurotrauma community, is multi-focal and preferentially in brain regions within proximity to the temporal ridge.^{7,12,19,25,39,42,59,81–119} It is essential to note that the studies listed in Table 1, and Figures 4 and 5, represent decades of multiple neurotrauma laboratories across the world and are the product of numerous surgeons. Thus, the representative neuropathology associated with the temporal ridge is more likely a feature of injury forces in the rodent brain, rather than spurious surgical variation. We contend that the curious neuropathology results from brain tissue forced into the temporal ridge, given that the injury forces reflect off the ventral skull. Transmission through ventral structures and reflection off the ventral skull would predict optic nerve damage, as reported by the Povlishock group, in mFPI.^120,121

In some cases (Fig. 5), neuropathology is not uniform along the cortex near the temporal ridge, suggesting that tissue may warp non-uniformly at the temporal ridge of the skull, thereby sparing some cortical tissue from the full forces of injury. Aspects of mechanical forces that may damage the brainstem, as has been shown for FPI in the cat, may exist and need further investigation.^122,123 Further, gyrencephalic brains may absorb or reflect forces in a manner that minimize the influence of an overall thicker skull. We argue that the relationship of the temporal ridge to injured tissue is not coincidental or attributable to differences in surgical technique.

For brain injury models that intentionally penetrate the dura, neuropathology is expected and localized primarily to the impact site. In these focal injury models, cavitation occurs at the injury site, with a penumbra of tissue damage. ¹²⁴ However, neuropathology after focal CCI spreads into diffuse pathology of the contralateral hemisphere, with accumulation of argyrophilia in cortical areas under the temporal ridge by 7 DPI. ¹²⁴ Alternatively, if the craniotomy for FPI is performed over the temporal ridge, then no overt pathology is observed.^125,126 Finally, cortical neuropathology in murine diffuse TBI models also traversed the temporal ridge, as presented for mFPI (Supplementary Material), further supporting a role for the temporal ridge in the proceeding neuropathology.

Closed TBI in the rodent results in neuropathology, preferentially in the somatosensory cortex (S1BF), hippocampus (CA3), and ventral thalamus. These three regions lie in a wedge or arc from the ventral midline aspect of the skull base toward the temporal ridge. For our biomechanical model of brain injury forces, these regions may not be uniquely vulnerable to TBI pathophysiology, but rather casualties along the force vectors (Fig. 6A). By removing one temporal ridge, neuroinflammation becomes localized to a single hemisphere, whereas removing both temporal ridges may retain bilateral inflammation because of the geometrical shape of the skull, regardless of the temporal ridges. The authors acknowledge that a study limitation exists by excluding uninjured sham animals with and without a thinned temporal ridge. Although the surgical procedure to thin the skull is unlikely to induce neuroinflammation, tissue damage cannot be ruled out. Additionally, the observed neuroinflammation after modifying the dynamics of injury forces only support, and do not confirm, the interpretation that neuropathology (de Olmos silver stain) is similar to neuroinflammation (Iba-1 immunohistochemistry). A concordance supports both neuropathology and neuro-inflammation after experimental TBI as indicators of localized pathological processes.^64,67–69

Further, the neuropathology in these regions would predict neurological and behavioral impairments, including somatosensory and cognitive impairment. With regard to the S1BF and ventral lateral thalamus, whisker hypersensitivity and somatosensory dysfunction have been reported.^{30–36,67,127–129} Cognitive performance involving short, long, and working memory, using multiple established mazes, is also impaired after FPI.^{11,33,37–47} These cognitive impairments may involve CA3 processing to perform object pattern completion, cue retrieval in fear conditioning, episodic memory, and spatial memory.^13,130–137 Thus, the neuropathology associated with the temporal ridge to include the S1BF, CA3, and ventral thalamus manifests with impairments in behavioral performance post-injury.

Conclusions

By analyzing the curious pathology of mFPI and relating the biomechanical model to other closed TBI models, laboratory studies can take advantage of localized pathology, without overt cavitation, to explore post-traumatic reorganization and repair of the cortex, hippocampus, and ventral thalamus.^65,67 By focusing on neuronal circuits that regulate somatosensory and cognitive function, investigators may continue to advance our understanding of the disease process that dismantles, repairs, and regenerates circuits in the brain. The pathology of diffuse TBI is the summation of 1) the mechanical forces of the primary injury, 2) the biomechanics of the impacted substrates, 3) the subsequent signaling cascades, and 4) secondary injuries. Ultimately, the acute events of TBI initiate a disease process that leaves persons with debilitating symptoms, which impair their quality of life. ² Through continued investigation of the consequences of TBI, we strive to improve the quality of life for our patients by advancing diagnostic techniques and therapeutic interventions.