Impaired axoplasmic transport is the dominant injury induced by an impact acceleration injury device: An analysis of traumatic axonal injury in pyramidal tract and corpus callosum of rats

BR A I N R ES E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report

Impaired axoplasmic transport is the dominant injury induced by an impact acceleration injury device: An analysis of traumatic axonal injury in pyramidal tract and corpus callosum of rats Srinivasu Kallakuri, Yan Li, Runzhou Zhou, Sharath Bandaru, Nisrine Zakaria, Liying Zhang⁎, John M. Cavanaugh Department of Biomedical Engineering, Wayne State University, Detroit, MI 48201, USA

A R T I C LE I N FO

AB S T R A C T

Article history:

Traumatic axonal injury (TAI) involves neurofilament compaction (NFC) and impaired

Accepted 25 February 2012

axoplasmic transport (IAT) in distinct populations of axons. Previous quantification

Available online 4 March 2012

studies of TAI focused on limited areas of pyramidal tract (Py) but not its entire length. Quantification of TAI in corpus callosum (CC) and its comparison to that in Py is also lack-

Keywords:

ing. This study assessed and compared the extent of TAI in the entire Py and CC of rats fol-

Traumatic axonal injury

lowing TBI. TBI was induced by a modified Marmarou impact acceleration device in 31 adult

Head impact acceleration model

male Sprague Dawley rats by dropping a 450 gram impactor from either 1.25 m or 2.25 m.

Impaired axoplasmic transport

Twenty-four hours after TBI, TAI was assessed by beta amyloid precursor protein (β-APP-

Neurofilament compaction

IAT) and RMO14 (NFC) immunocytochemistry. TAI density (β-APP and RMO14 axonal swell-

Corpus callosum

ings, retraction balls and axonal profiles) was counted from panoramic images of CC and Py.

Pyramidal tract

Significantly high TAI was observed in 2.25 m impacted rats. β-APP immunoreactive axons were significantly higher in number than RMO14 immunoreactive axons in both the structures. TAI density in Py was significantly higher than in CC. Based on our parallel biomechanical studies, it is inferred that TAI in CC may be related to compressive strains and that in Py may be related to tensile strains. Overall, IAT appears to be the dominant injury type induced by this model and injury in Py predominates that in CC. © 2012 Elsevier B.V. All rights reserved.

1.

Introduction

Traumatic axonal injury (TAI) is a complex pathological process involving disruption of axons of diverse caliber and location following traumatic brain injury. Much of the current understanding on TAI stems from contributions by various investigators with the most notable work coming from Povlishock and colleagues that provided elegant insight into the

course of events leading to axonal pathology from their earlier studies on the fate of reactive swellings in cats subjected to fluid percussion injury (FPI) to their subsequent studies revealing axonal pathology marked by axolemmal changes with neurofilament compaction (NFC) and impaired axoplasmic transport (IAT) in distinct populations of axons (Marmarou et al., 2005; Povlishock and Becker, 1985; Povlishock, 1992, 1993; Povlishock et al., 1997; Stone et al., 2001; Stone et al., 2004).

⁎ Corresponding author at: Department of Biomedical Engineering, Wayne State University, 818 W. Hancock Street, Detroit, MI 48201, USA. Fax: +1 313 5778333. E-mail address: [email protected] (L. Zhang). 0006-8993/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2012.02.065

30

BR A IN RE S E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

The foundation for these two distinct pathologies comes from their previous work showing focal axolemmal permeability and increased neurofilament packing as revealed by reduced inter neurofilament (NF) distance in brains of cats subjected to moderate FPI. Only signs of NF disarray and misalignment were observed in brains with mild FPI (Pettus et al., 1994). Subsequently, in a separate study, proteolysis of NF side arms was reported in giant neurons of lamprey undergoing axotomy (Hall and Lee, 1995) that brought into prominence the use of previously characterized monoclonal RMO antibodies (Lee et al., 1987) targeting specific epitopes exposed on NF-medium chain (NFM) following proteolytic activity (Hall and Lee, 1995). Taking advantage of these antibodies in combination with antibodies for beta amyloid precursor protein (β-APP) as a marker for IAT, it has been shown that NFC occurs independent of IAT following TBI (Stone et al., 2001). This was further supported by a separate investigation that quantified the extent of IAT and NFC in corticospinal tract (CSpT) and medial lemniscus after TBI. IAT in CSpT significantly increased over 24 h with NFC in the same tract showing a significant decline by 24 h compared to their 3 hour count (Marmarou et al., 2005). However, studies on quantification of TAI encompassing the entire length of pyramidal tract are lacking as the previous studies were limited to an area of CSpT just 0.01 mm rostral to the pyramidal decussation (Marmarou et al., 2005) or the pontomedullary junction alone as part of a study to assess the effects of moderate hypothermia (Buki et al., 1999a) or cyclosporine (Buki et al., 1999b) after TBI. Furthermore, studying TAI changes in the entire length of the pyramidal tract rather than in a limited area may offer better insights into the extent of TAI considering the diffuse nature of the pathology itself. This may be of value, for example, in studies that use TAI changes as therapeutic outcomes in TBI treatment. Besides the brainstem white matter tracts, TAI was also documented in corpus callosum and other sub-cortical regions (Foda and Marmarou, 1994; Kallakuri et al., 2003; Li et al., 2011a). However, studies on quantification of both NFC and IAT in corpus callosum of rats following an impact acceleration injury are lacking. Previous studies on quantification of TAI in CC were limited to a semi-quantitative assessment in three sections from immature male and female rats subjected to non contusive controlled cortical impact (DiLeonardi et al., 2009b) with a greater extent of β-APP reactive axons being noted compared to RMO14 reactive axons. However, considering that IAT and NFC occur in different axonal populations (Stone et al., 2001) with axolemmal changes and decreased interfilamental distance occurring in large caliber, heavily myelinated axons of brainstem (Pettus et al., 1994; Stone et al., 2001), understanding these pathologies in mature rat brains encompassing wider number of sections is important. This may be of further significance considering the existence of sex differences in the ratio of unmyelinated to myelinated axons in the genu of corpus callosum of rats with females having a larger population of unmyelinated fibers (Mack et al., 1995). Therefore, considering that there were no previous studies on the expression and comparisons IAT and NFC in the entire pyramidal tract as well as in corpus callosum, the current

study was undertaken to assess the extent of IAT and NFC in representative sections encompassing the entire length of these structures. Also, the effects of varying impact energy on the extent of IAT and NFC distribution in these two regions

Fig. 1 – Anatomic location of pyramidal tract and corpus callosum sections. A): A sagittal view of rat brain showing approximate location of helmet (D = 10 mm) relative to bregma. As seen, parts of CC and brainstem are directly under the site of impact. B) Representative coronal sections relative to bregma selected for observing TAI. The most anterior part of CC (anterior to bregma) is not under the helmet. C) Representative sagittal sections selected for TAI staining. As shown in Fig. 1A, majority of Py is away from the site of impact but demonstrated highest incidence of TAI underscoring the diverse injury mechanisms in different white matter tracts.

BR A I N R ES E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

31

Fig. 2 – Quantification of TAI in Py: A) A panoramic image of pyramidal tract 200 μm away from midline from a representative 2.25 m impacted rat showing β-APP-IR axons across the length of the tract with 200 × 200 μm grid superimposed. B) A sample 200 × 200 μm grid with β-APP-IR axons in the Py (red dots). The number at bottom right indicates total number of quantified injured axon profiles. C) An image of β-APP-IR retraction balls (arrows) and swollen axons (arrow head) in Py. D) A sample image of RMO-IR swellings (arrow head) and retraction balls (arrow) in the most caudal Py.

is also unknown and therefore forms the additional purpose of this study.

2.

Results

As we previously reported, no skull fractures and respiratory depression was observed in rats subjected to TBI from a height of 1.25 m. However, in rats subjected to TBI from 2.25 m a mortality rate of 10.2% and a skull fracture rate of 20.4% were observed (Li et al., 2011a). In both pyramidal tract (Py) (Figs. 2A–B) and corpus callosum (CC), β-APP-IR axons were observed as thin axonal profiles characterized by areas of swollen regions (Fig. 2C). Also observed were many spherical or ovoid retraction balls, with some still having a part of their normal axon attached, giving a tail like appearance. RMO14-IR axons appeared to be thin and occasionally seen as swollen axons or retraction balls and distributed more sparingly (Fig. 2D).

2.1.

of β-APP-IR axons tended to localize caudally in both 2.25 m and 1.25 m groups (Fig. 4A). However, no specific patterns were observed for RMO14-IR axons (Fig. 4B). Furthermore, in Py of rats injured from 2.25 m or 1.25 m no significant differences were observed in the medial–lateral distribution of TAI. In 1.25 m Py, sections from 600 μm region tended to show a high incidence of β-APP-IR axons (Fig. 4C).

2.2.

TAI in corpus callosum

In corpus callosum, quantitatively, the number of β-APP-IR axons were significantly higher in sections from 2.25 m (186 ± 57) compared to sections from 1.25 m impacted animals (20 ± 4; p < 0.05). Furthermore, the number of RMO14-IR axons (105 ± 38) in sections from 2.25 m impacted animals

TAI in pyramidal tract

In pyramidal tract, sagittal sections from both the injured groups showed TAI. The number of β-APP-IR axons was significantly higher in sections from 2.25 m injury group (945.44 ± 218) compared to 1.25 m group (161.76 ± 93; p < 0.05; Fig. 3). Similarly, the number of RMO14-IR injured axons was also high (Fig. 3) in 2.25 m group (116.46 ± 19) compared to the 1.25 m group (14.02 ± 5). In addition, the number of β-APP-IR axons at 2.25 m was significantly higher than RMO14 injured axons at 2.25 m and 1.25 m (p < 0.05). Furthermore, the spatial profile of TAI distribution also revealed that a high number

Fig. 3 – Chart showing average TAI counts in Py at 2.25 m and 1.25 m. TAI counts in 2.25 m were significantly higher than 1.25 m. β-APP-IR axons were significantly higher than RMO14-IR axons at both the heights.

32

BR A IN RE S E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

2.3. Comparison of TAI per square millimeter in pyramidal tract and CC The mean number of β-APP-IR axons per square millimeter in Py at 2.25 m (102.77 ± 19.94) was significantly higher than that in Py at 1.25 m (19.50 ± 10.52) and also from that in CC at both heights (2.25 m:4.77 ± 1.5; 1.25 m:1.06 ± 0.11). No significant differences were observed within groups (Fig. 6A). In the Py the mean number of RMO14-IR axons per square millimeter at 2.25 m (16.69 ± 2.34) was significantly higher than that in Py at 1.25 mm (1.98 ± 0.65) and also from the RMO14 reactive injured axons of CC at both 2.25 m (2.64 ± 1.02) and 1.25 mm (0.58 ± 0.06). No other significant differences were observed (Fig. 6B).

3.

Fig. 4 – TAI distribution charts in Py and CC. A). Chart showing anterior–posterior distribution of IAT in Py that revealed a high incidence of β-APP-IR axons more caudally at both the injured heights. B) Chart showing anterior–posterior distribution of RMO14-IR axons in Py. C) Chart showing medial–lateral distribution of TAI in Py at both the injury heights. β-APP and RMO14-IR axons were highest in 2.25 m impacted rats.

was also significantly different from those in 1.25 m impacted animals (15 ± 3; Fig. 5A). Moreover, the β-APP-IR axons (Fig. 5B) observed in CC from 2.25 m group were also significantly higher than RMO14-IR axons observed in CC of 1.25 m (p < 0.05). Furthermore, in CC, the number of β-APP (Fig. 5C) and RMO14-IR (Fig. 5D) axons were prominent in coronal sections that correspond approximately to regions lying under the helmet (around bregma −0.12 to −2.04) with the most anterior (close to genu) and posterior (close to splenium) sections demonstrating far less axonal injury (Fig. 5B).

Discussion

It has been established that TAI involves two distinct pathologies namely NFC and IAT in distinct axonal populations following TBI (Marmarou et al., 2005; Povlishock and Becker, 1985; Povlishock, 1992; Povlishock, 1993; Povlishock et al., 1997; Stone et al., 2001, 2004). The present study provides a detailed analysis of the extent of NFC and IAT as evidenced by β-APP and RMO14 reactivity in the entire Py and CC of rats subjected to TBI. Quantification of TAI in Py was performed beginning anteriorly at the level of pontomedullary junction that extended caudally to the junction of pyramidal tract and pyramidal decussation and in CC was performed on representative sections encompassing genu, body and splenium. Such comprehensive assessment of TAI following an impact acceleration injury in Py and CC has never been reported. Previously, TAI quantification in Py was restricted to a location 0.01 mm rostral to pyramidal decussation (Marmarou et al., 2005) or to descending corticospinal tract in the pontomedullary junction (Suehiro and Povlishock, 2001; Suehiro et al., 2001). Whereas TAI in CC was assessed in sections from immature rats encompassing three coordinates relative to bregma (DiLeonardi et al., 2009a). However, in this study, TAI in CC was assessed in 13–15 sections encompassing all the regions of CC. Unlike the previous quantitative studies of TAI in Py (Buki et al., 1999b; Marmarou et al., 2005; Suehiro et al., 2001), this study offers details on the extent of TAI in the anterior-posterior direction as well as bilaterally from midline by developing a unique quantification technique which enables localization of quantified TAI relative to a given anatomical location (Fig. 4APy; Fig. 5B-CC). This allows for additional comparisons to finite element (FE) models of rat head–body developed as part of separate ongoing studies (Zhang et al., 2011b). The distribution of axonal changes in relation to localized tissue strain in white matter tracts calculated by the model may provide information related to the biomechanical processes involved in the genesis of TAI. Furthermore, such extensive quantification of TAI in the entire white matter tract can also provide a more precise validation of therapeutic outcomes. In this study, no specific distribution pattern was observed for RMO14-IR axons. However, β-APP-IR axons were prominently found in the entire Py especially at high numbers in the most caudal region which may be related to a convergence of fibers that are

BR A I N R ES E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

33

Fig. 5 – TAI in CC. A) In CC, the TAI was high in animals injured from 2.25 m compared to those from 1.25 m. The incidence of β-APP-IR axons at 2.25 m was significantly higher than RMO14-IR axons at 2.25 m and that of TAI at 1.25 m (p < 0.05). No other significant changes were observed. B) Chart showing that the number of β-APP and RMO14-IR axons of CC was more prominent in coronal sections that correspond approximately to regions lying under the helmet. C) A sample image of IAT in the form of β-APP-IR retraction balls (arrows) and swellings (arrow head) in CC. D) A sample image of NFC in the form of RMO14-IR retraction balls (arrows) and swellings (arrowhead) in lateral CC.

decussating into the pyramidal tract. Furthermore, the fiber bundles are perpendicular to the axis of the sagittal rotation and could have been subjected to high tensile forces induced by rotation (Li et al., 2011b). This was supported by significantly higher IAT and NFC in Py at 2.25 m (both in terms of total number and density) than in Py at 1.25 mm and CC from both heights. The high incidence of TAI in Py was striking considering that CC located deep in the cortex and directly under the influence of impact site is expected to be the logical region for high incidence of TAI. However, CC demonstrated far less TAI compared to Py. TAI can result from angular as well as linear acceleration of head (Gennarelli et al., 2003; King, 2000; McLean, 1995; Zhang et al., 2001) and the precise role of linear (Nishimoto and Murakami, 1998) or angular acceleration (Margulies and Thibault, 1992; Smith, 2003; Zhang et al., 2006) during an impact event has been the subject of several investigations. In the current study, based on analyses of various biomechanical parameters, we suggest that TAI in different brain regions may result from different mechanical loading. Considering that the direction of axons in CC is perpendicular to the impact force vector, TAI in CC may be induced by strain under the impact area mediated by compression of superficial cortical layers (Li et al., 2011b). A change in intracranial pressure (ICP) under the impact site can also contribute to the observed TAI in CC. Increased ICP with increasing depth of depression

in a cortical impact model (Manley et al., 2006) as well as its highest correlation with translational acceleration as assessed by FE simulation of actual recorded sports concussive events (Zhang et al., 2004) also supports the potential role of ICP in inducing TAI. In fact, sudden changes in ICP leading to morphological changes in brain and axonal damage in sub-cortical white matter were also previously reported (Shreiber et al., 1999). On the other hand, TAI in Py may be related to angular head acceleration subjecting axons in brainstem to tensile forces (Li et al., 2011b). An FE analysis of impact acceleration experiments using rat full body FE model indicated maximum principal strains in brainstem especially in Py and medial lemniscus that were higher than in CC. A differential movement between head and body inducing a high brainstem strain at medullary-spinal junction was also illustrated by the model (Zhang et al., 2011b). TAI can also be induced under conditions of ischemia by middle cerebral artery occlusion as evidenced by β-APP-IR swellings and retraction balls (Gresle et al., 2006) similar to those seen in this study. This suggests potential similarities in the genesis of TAI following an ischemic and traumatic insult (Bramlett and Dietrich, 2004). In fact a role for ischemic conditions following a traumatic insult was also supported by elevated vasoconstrictive mediators in animal models of TBI (Armstead, 1996; Armstead, 1999; Dixon et al., 1987; Petrov et al., 2002). Such

34

BR A IN RE S E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

Fig. 6 – Charts showing TAI per square mm in Py and CC. A) Density of β-APP-IR axons in Py at 2.25 m which were significantly higher than β-APP-IR axons of CC at 2.25 m and that of Py and CC at 1.25 mm (p < 0.05). B) Density of RMO14-IR axons in Py at 2.25 m which were significantly higher than CC at 2.25 m and that of both the structures at 1.25 mm (p < 0.05).

elevations of vasoconstrictive mediators were also observed in patients of TBI (Maier et al., 2007; Menon et al., 2002). Furthermore, the density per square millimeter of β-APP-IR axons (102.77 ± 19.94) and that of RMO14-IR axons (16.69 ± 2.34) at 2.25 m from this study was relatively less than those reported previously (APP: 588.38 ± 44.76; RMO14:99.67 ± 46.02); (Marmarou et al., 2005). In the current study, Py quantification was performed in an extensive tract extending from 8.6 mm to 15.8 mm away from bregma. Other studies focused on a localized region at as the pontomedullary junction or a limited area away from pyramidal decussation (Buki et al., 1999a; Marmarou et al., 2005) and where higher densities of TAI were reported. The most anterior regions of Py (Fig. 3A) tend to have less IAT compared to the most posterior region. It was previously reported that thin fibers in human CC were most dense in the genu and decrease in density posteriorly towards the posterior mid body with fibers in splenium increasing again (Aboitiz et al., 1992). This was further supported by others that showed fibers of relatively small diameter were most pronounced in the anterior and posterior third of the CC with larger fibers in the mid body and gradually reducing laterally (Hofer and Frahm, 2006; Witelson, 1989). With the possibility of a similar distribution pattern in rat CC one would expect to see a high incidence of TAI in the most anterior and posterior regions of CC. However, a non-

uniform distribution of TAI longitudinally was observed with a high incidence of TAI in sections directly under the impact site (starting at −0.12 mm to − 2.04 mm away from bregma) which approximately corresponds to the body of the CC, suggesting that the thick fibers are more sensitive to rapid head acceleration compared to thin fibers. It was also suggested that large caliber fibers with their high metabolic demand than small caliber fibers are more susceptible to secondary injury changes following a traumatic event (Shi and Whitebone, 2006). Furthermore, TAI variation in different regions of CC may also be related to the influence of higher compressive forces directly under the helmet than in other parts of CC which are remote from the impacting area. One prominent feature of the present investigation is the dominance of IAT. This is similar to the observations from a previous study that showed a steady increase in the extent of IAT up to 24 h after TBI in Py. However, the extent of NFC increased up to 3 h after TBI that decreased by 24 h in Py (Marmarou et al., 2005). In medial lemniscus, a white matter tract with large caliber axons, the extent of both IAT and NFC was reported to increase up to 24 h. The mechanisms underlying the differences between the Py and medial lemniscus were attributed to inherent differences in cytoskeletal composition. It was suggested that Py has small caliber axons that have proportionately greater number of microtubules per unit area versus large caliber axons in medial lemniscus that have proportionately greater number of neurofilaments per unit area (Malbouisson et al., 1985; Price et al., 1988; Szaro et al., 1990). Comparison of diverse axonal injury patterns becomes even more complicated considering inherent differences within the same fiber systems. The myelinated fibers in medullary pyramids are very small and estimated to have a modal diameter of 0.9 μm. The unmyelinated fibers were found to have a diameter of 0.1–0.3 μm (Harding and Towe, 1985). On the other hand, in a separate study, Leenen et al. (1985) reported that in pyramis medullae unmyelinated fiber diameters ranged from 0.05 to 1.21 μm with a mean of 0.18 μm. The mean diameter of myelinated fibers measured 1.08 μm with a range of 0.25–6.03 μm (Leenen et al., 1985). The predominance of small caliber axons in Py may also support higher incidence of IAT than NFC as in the current study and in a previous guinea optic nerve study (Jafari et al., 1998). It was reported that a significant increase in number of NF in axons with a diameter of 0.51–1.00 μm is associated with a significant reduction in interneurofilament spacing, suggestive of NFC in axons of 0.51–1.00 and 1.01–1.50 μm. They also suggested an inverse relationship to the size of axons with no compaction in axons with diameter of 1.51–2.00 μm (Jafari et al., 1998). Furthermore, Jafari et al. (1998) also reported in axons of > 2 μm diameter, the presence of either NFC or significant increase in spacing between neurofilaments termed as “dispersion.” The increased spacing was postulated to be related to loss of neurofilaments. The presence of a limited incidence of NFC in the Py and CC in the current study and that NFC being more associated with large caliber axons, one may speculate that injury of large caliber of axons in these structures may be related to neurofilaments undergoing dispersion based on findings from guinea pig nerve stretch studies and may warrant further validation.

BR A I N R ES E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

It is of interest to note that Stone et al. (2001) have shown the co-localization of IAT and NFC in fibers of ML, which has a large population of large caliber axons. No such colocalization was reported for fibers in Py (which are of smaller caliber) up to 6 h after injury (Stone et al., 2001). In the current study, RMO14-IR axons in Py were sparse, which was similar to a previous study that showed RMO14 reactive fibers peaking at 3 h, but reduced by 24 h (Marmarou et al., 2005). Furthermore, the preferential IAT in small caliber axons is supported by electrophysiological findings from Reeves et al. (2005) who reported that small caliber unmyelinated fibers in CC continue to show suppressed compound action potential 7 days after injury, unlike large caliber myelinated axons that showed recovery to control levels (Reeves et al., 2005). Taken together, ours and previous studies support the dominance of IAT in white matter regions dominated by small diameter fibers.

4.

Conclusions

TBI induced by an impact acceleration device results in TAI as identified by neurofilament compaction and impaired axoplasmic transport in both Py and CC. TAI in the form of IAT was the more dominant injury type than NFC in both Py and CC. TBI induced by an impact acceleration device results in lower densities of TAI in CC compared to Py.

impactor still weighing 450 g was made up of an aluminum cylinder with enough interior space to house an accelerometer, and a brass impact end, tapered to 19 mm to create an impact interface similar to the original model. The impactor was released from either 2.25 m or 1.25 m height by a custom-made solenoid release device (Li et al., 2011a). 5.3.

Experimental procedures

5.1.

Animal handling and preparation

31 anesthetized male Sprague–Dawley rats (392 ± 13 g) were used to assess the extent of IAT and NFC in pyramidal tract and corpus callosum. TBI of varying severity was induced by dropping a custom-made 450 gram impactor housing an accelerometer from a height of 1.25 m and 2.25 m respectively and the details of this modified method have been described separately (Li et al., 2011a). This device uses a reinforced drop stand and an automatic release device to deliver consistent and repeatable impact energy to rat head. The heights of 1.25 m and 2.25 m, although higher than those used in the original Marmarou model (Marmarou et al., 1994), were chosen to compensate for the loss of velocity caused by the accelerometer cable dragging against the tube (Li et al., 2011a; Zhang et al. submitted for publication). All rats were administered Buprenex (0.3 mg/kg) subcutaneously 20 min prior to impact. Fifteen minutes prior to impact, rats were placed in a sealed acrylic chamber and anesthetized by a mixture of isoflurane (3%) and oxygen (0.6 L/min). The skull was exposed by a midline dorsal incision of the skin and two round stainless steel disks (helmet) each 10 mm in diameter and 3 mm in thickness glued together were positioned midline between bregma and lambda and affixed to the skull vault using cyanoacrylate (Elmer's Products, Columbus, OH).

Induction of traumatic brain injury

The instrumented animals were placed prone on an open-cell flexible polyurethane (PU) foam bed in a Plexiglas box under a 2.5 m long and 57 mm in diameter Plexiglas tube, with the helmet (Fig. 1A) centered directly under the lower end of the tube (Li et al., 2011a; Zhang et al., 2011a). A laser beam was used to guide the positioning of the helmeted head to ensure that the impactor hit the center of the stainless steel disc (helmet). TBI was induced by dropping the impactor from either 1.25 m or 2.25 m. Immediately after the impact, the Plexiglas box was manually removed to avoid a second impact to the rat head and the skull was examined for fractures and then the skin was closed by staples. Rats with skull fractures and those exhibiting severe distress were not included in the study. A new foam (pre-compressed) was used for a series of tests and was replaced by a new foam every three weeks to ensure consistent and repeatable mechanical response to a given impact (Zhang et al., 2011a). 5.4.

5.

35

Termination and fixation

Following TBI, rats were monitored for at least 6 h for signs of distress. Those with skull fracture or signs of severe distress were euthanized. At 24 hour (h) post-TBI, each rat was euthanized with an overdose of sodium pentobarbital (120 mg/kg, intra peritoneal) and transcardially perfused with cold 4% paraformaldehyde in phosphate buffered saline (0.1 M PBS, pH 7.45). The brain was then removed and post fixed (4% paraformaldehyde in 30% sucrose). For analysis of IAT and NFC in corpus callosum, the cerebral hemispheres were coronally cut into 40 μm thick frozen sections (Leica CM 3050, Leica Microsystems Nussloch GmbH, Nussloch, Germany) from the genu of the CC (+2.3 mm anterior to the bregma (0.00 mm)) through the splenium of the CC (−5.2 mm posterior to the bregma; Fig. 1B) (Paxinos and Watson, 2007). The rest of the cerebral hemispheres (−6.0 mm posterior to bregma), with the entire brainstem still attached, were processed for the analysis of TAI in pyramidal tract (Fig. 1C). For this purpose, two bilateral longitudinal cuts were made in the brainstem around 5 mm lateral to midline that ensured inclusion of regions encompassing midbrain, pons and pyramidal tract. Then serial sagittal sections of brainstem (40 μm) were collected. All coronal (Fig. 1B) and sagittal (Fig. 1C) sections were individually placed in 1 × PBS filled multi-well plates for their subsequent selection to match with section co-ordinates in the rat stereotaxic atlas, and stored at 4 °C till further processing.

Impact acceleration device and instrumentation

5.5. β-APP and RMO14 immunostaining for TAI in corpus callosum and brainstem

The diffuse TBI device previously described by Marmarou et al. (1994) was modified (Li et al., 2011a). In this new design, the

β-APP and RMO14 immunostaining was previously used to study distinct populations of axons revealing IAT and NFC

5.2.

36

BR A IN RE S E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

respectively in corpus callosum (DiLeonardi et al., 2009a) and brainstem (DiLeonardi et al., 2009a; Marmarou et al., 2005; Stone et al., 2001). IAT and NFC in corpus callosum were studied separately by single label immunocytochemistry in an average of 14 sections from each 2.25 m (n = 16) and 1.25 m (n = 15) impacted rats. IAT and NFC in pyramidal tract were studied in 7sagittal sections from each 2.25 m (n = 16) and 1.25 m (n = 13) impacted rats. Two sets of successive coronal sections spaced 0.48 mm apart were selected for investigating TAI in along the anteroposterior aspects of corpus callosum. To assess TAI in pyramidal tract, two sets of 7 sagittal sections were selected for β-APP and RMO14 immunocytochemistry respectively. For this purpose, first a section corresponding approximately to midline was chosen and then 6 other sections corresponding approximately to 200 μm, 600 μm and 1000 μm away and from either side of the chosen midline were selected. Briefly, the sections were subjected to antigen retrieval by incubation in a citrate buffer (pH6.0) at 90 °C for 1 h and then allowed to cool to room temperature. Following a wash in 1 × PBS, they were immersed in 0.3% H2O2 for 1 h to quench endogenous peroxidase activity. This was followed by overnight incubation (at room temperature) in a C-terminus specific APP (1 μg/ml; rabbit anti-C-terminus β-APP; catalog #51-2700; Invitrogen, San Francisco, CA) or in mouse RMO14 (1:500; cat #34-1000; Invitrogen, San Francisco, CA) antibody diluted in 2% normal goat serum (Vector Laboratories, Burlingame, CA) and 1% bovine serum albumin. The following day, the sections were incubated either in a goat anti-rabbit or anti mouse IgG (Vector Laboratories, Burlingame, CA) for 1 h. Finally, the sections were visualized via incubation in avidin biotin peroxidase complex (Vectastain ABC Standard Elite Kit, Vector) and were developed by brief incubation in 3,3′-diaminobenzidine (DAB) and hydrogen peroxide. Finally, the sections were washed, dehydrated and coverslipped using Permount. Negative control incubations were performed in the absence of primary antibody. 5.6.

mechanical strain out of corresponding elements of same resolution (200 × 200 μm) as part of a separate study to develop an anatomically detailed finite element (FE) model of the rat head (Zhang et al., 2011b). All β-APP and RMO14-IR axon profiles in each grid were counted using ImageJ software and added to obtain total TAI counts per section. In order to compare the extent of TAI between rats with different number of sections quantified, TAI counts were normalized using the equation: normalized TAI = (Total TAI / number of sections quantified) × normalizing constant. In order to compare TAI density between CC and Py and at different drop heights, average area density was determined. The area density for each section is the total TAI count in given section divided by the area of CC or Py in that section. Then, average area density for all the sections was calculated and expressed as the density of APP-IR or RMO14-IR per square millimeter. Total β-APP or RMO14-IRcount for CC or Py was the sum of β-APP-IR or RMO14-IR counts from panoramic images of all quantified sections. For normalized TAI in CC, a normalizing constant of 14 was used as majority of rats (18 out of 31) had 14 coronal sections. In the case of pyramidal tract a normalizing constant of 6 was used as 16 out of 28 rats had 6 viable quantified sagittal sections. 5.7.

Statistical analysis

Unless otherwise noted, values given are mean ± standard error of the mean (SEM). One-way analysis of variance or ttest was used to assess for significant differences in TAI counts between different structures from the two drop heights. A p value < 0.05 was considered to be statistically significant.

Acknowledgements This research was supported by NIH R01 EB006508.

Quantitative analysis of TAI REFERENCES

Axons undergoing IAT were evidenced by β-APP immunoreactive (β-APP-IR) axonal swellings and retraction balls and RMO14 immunoreactive (RMO14-IR) axons and their profiles were considered for NFC. The total number of β-APP and RMO14 immunoreactive axons in the two sets of coronal and sagittal sections was quantified by a blinded observer. Each section (β-APP-IR or RMO14-IR) was observed under a light microscope (Leica DMLB, Leica Microsystems Ltd, Heerburg, Switzerland) and serial photomicrographs (×10 magnification) encompassing the corpus callosum (2.28 to −4.4 mm away from bregma) or pyramidal tract (8.6 to 15.8 mm away from bregma) were taken with a digital camera system (ProgRes C7, JENOPTIK Laser OptikSysteme, GmbH). The photographs taken at a single focal plane from each rat were then digitally stitched (Photoshop CS2; Adobe Systems Incorporated, CA) to create a panoramic image for corpus callosum and pyramidal tract from all stained sections of both 2.25 m and 1.25 m impacted rats. Then grids measuring 200 × 200 μm were superimposed on each constructed panoramic image (Fig. 2A). This enabled direct correlation of the number of axons showing TAI (Fig. 2B; β-APP or RMO14) to the level of

Aboitiz, F., Scheibel, A.B., Fisher, R.S., Zaidel, E., 1992. Fiber composition of the human corpus callosum. Brain Res. 598, 143–153. Armstead, W.M., 1996. Role of endothelin in pial artery vasoconstriction and altered responses to vasopressin after brain injury. J. Neurosurg. 85, 901–907. Armstead, W.M., 1999. Role of endothelin-1 in age-dependent cerebrovascular hypotensive responses after brain injury. Am. J. Physiol. 277, H1884–H1894. Bramlett, H.M., Dietrich, W.D., 2004. Pathophysiology of cerebral ischemia and brain trauma: similarities and differences. J. Cereb. Blood Flow Metab. 24, 133–150. Buki, A., Koizumi, H., Povlishock, J.T., 1999a. Moderate posttraumatic hypothermia decreases early calpain-mediated proteolysis and concomitant cytoskeletal compromise in traumatic axonal injury. Exp. Neurol. 159, 319–328. Buki, A., Okonkwo, D.O., Povlishock, J.T., 1999b. Postinjury cyclosporin A administration limits axonal damage and disconnection in traumatic brain injury. J. Neurotrauma 16, 511–521. DiLeonardi, A.M., Huh, J.W., Raghupathi, R., 2009a. Impaired axonal transport and neurofilament compaction occur in separate

BR A I N R ES E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

population of injured axons following diffuse brain injury in the immature rat. Brain Res. 1263, 174–182. DiLeonardi, A.M., Huh, J.W., Raghupathi, R., 2009b. Impaired axonal transport and neurofilament compaction occur in separate populations of injured axons following diffuse brain injury in the immature rat. Brain Res. 1263, 174–182. Dixon, C.E., Lyeth, B.G., Povlishock, J.T., Findling, R.L., Hamm, R.J., Marmarou, A., Young, H.F., Hayes, R.L., 1987. A fluid percussion model of experimental brain injury in the rat. J. Neurosurg. 67, 110–119. Foda, M.A., Marmarou, A., 1994. A new model of diffuse brain injury in rats. Part II: morphological characterization. J. Neurosurg. 80, 301–313. Gennarelli, T.A., Pintar, F.A., Yoganandan, N., 2003. Biomechanical tolerances for diffuse brain injury and a hypothesis for genotypic varaiabilityt in response to trauma. In Proceeding of Advanced Autimotive Medicine. Vol. 47, ed.^eds., pp. 62–628. Gresle, M.M., Jarrott, B., Jones, N.M., Callaway, J.K., 2006. Injury to axons and oligodendrocytes following endothelin-1-induced middle cerebral artery occlusion in conscious rats. Brain Res. 1110, 13–22. Hall, G.F., Lee, V.M., 1995. Neurofilament sidearm proteolysis is a prominent early effect of axotomy in lamprey giant central neurons. J. Comp. Neurol. 353, 38–49. Harding, G.W., Towe, A.L., 1985. Fiber analysis of the pyramidal tract of the laboratory rat. Exp. Neurol. 87, 503–518. Hofer, S., Frahm, J., 2006. Topography of the human corpus callosum revisited—comprehensive fiber tractography using diffusion tensor magnetic resonance imaging. Neuroimage 32, 989–994. Jafari, S.S., Nielson, M., Graham, D.I., Maxwell, W.L., 1998. Axonal cytoskeletal changes after nondisruptive axonal injury. II. Intermediate sized axons. J. Neurotrauma 15, 955–966. Kallakuri, S., Cavanaugh, J.M., Ozaktay, A.C., Takebayashi, T., 2003. The effect of varying impact energy on diffuse axonal injury in the rat brain: a preliminary study. Exp. Brain Res. 148, 419–424. King, A.I., 2000. Fundamentals of impact biomechanics: Part I—biomechanics of the head, neck, and thorax. Annu. Rev. Biomed. Eng. 2, 55–81. Lee, V.M., Carden, M.J., Schlaepfer, W.W., Trojanowski, J.Q., 1987. Monoclonal antibodies distinguish several differentially phosphorylated states of the two largest rat neurofilament subunits (NF-H and NF-M) and demonstrate their existence in the normal nervous system of adult rats. J. Neurosci. 7, 3474–3488. Leenen, L.P., Meek, J., Posthuma, P.R., Nieuwenhuys, R., 1985. A detailed morphometrical analysis of the pyramidal tract of the rat. Brain Res. 359, 65–80. Li, Y., Zhang, L., Kallakuri, S., Zhou, R., Cavanaugh, J.M., 2011a. Quantitative relationship between axonal injury and mechanical response in a rodent head impact acceleration model. J. Neurotrauma 28, 1767–1782. Li, Y., Zhang, L., Kallakuri, S., Zhou, R., Cavanaugh, J.M., 2011b. Injury predictors for traumatic axonal injury in a rodent head impact acceleration model. Stapp Car Crash J. 55, 25–47. Mack, C.M., Boehm, G.W., Berrebi, A.S., Denenberg, V.H., 1995. Sex differences in the distribution of axon types within the genu of the rat corpus callosum. Brain Res. 697, 152–160. Maier, B., Lehnert, M., Laurer, H.L., Marzi, I., 2007. Biphasic elevation in cerebrospinal fluid and plasma concentrations of endothelin 1 after traumatic brain injury in human patients. Shock 27, 610–614. Malbouisson, A.M., Ghabriel, M.N., Allt, G., 1985. Axonal microtubules: a computer-linked quantitative analysis. Anat. Embryol. 171, 339–344. Manley, G.T., Rosenthal, G., Lam, M., Morabito, D., Yan, D., Derugin, N., Bollen, A., Knudson, M.M., Panter, S.S., 2006. Controlled cortical impact in swine: pathophysiology and biomechanics. J. Neurotrauma 23, 128–139.

37

Margulies, S.S., Thibault, L.E., 1992. A proposed tolerance criterion for diffuse axonal injury in man. J. Biomech. 25, 917–923. Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J., Kita, H., Demetriadou, K., 1994. A new model of diffuse brain injury in rats. Part I: pathophysiology and biomechanics. J. Neurosurg. 80, 291–300. Marmarou, C.R., Walker, S.A., Davis, C.L., Povlishock, J.T., 2005. Quantitative analysis of the relationship between intra-axonal neurofilament compaction and impaired axonal transport following diffuse traumatic brain injury. J. Neurotrauma 22, 1066–1080. McLean, A.J., 1995. Brain injury without head impact? J. Neurotrauma 12, 621–625. Menon, D.K., Day, D., Kuc, R.E., Downie, A.J., Chatfield, D.A., Davenport, A.P., 2002. Arteriojugular endothelin-1 gradients in aneurysmal subarachnoid haemorrhage. Clin. Sci. (Lond.) 103 (Suppl. 48), 399S–403S. Nishimoto, T., Murakami, S., 1998. Relation between diffuse axonal injury and internal head structures on blunt impact. J. Biomech. Eng. 120, 140–147. Paxinos, G., Watson, C., 2007. The rat brain in stereotaxic coordinates. Vol. Academic Press, New York. Petrov, T., Steiner, J., Braun, B., Rafols, J.A., 2002. Sources of endothelin-1 in hippocampus and cortex following traumatic brain injury. Neuroscience 115, 275–283. Pettus, E.H., Christman, C.W., Giebel, M.L., Povlishock, J.T., 1994. Traumatically induced altered membrane permeability: its relationship to traumatically induced reactive axonal change. J. Neurotrauma 11, 507–522. Povlishock, J.T., 1992. Traumatically induced axonal injury: pathogenesis and pathobiological implications. Brain Pathol. 2, 1–12. Povlishock, J.T., 1993. Pathobiology of traumatically induced axonal injury in animals and man. Ann. Emerg. Med. 22, 980–986. Povlishock, J.T., Becker, D.P., 1985. Fate of reactive axonal swellings induced by head injury. Lab. Invest. 52, 540–552. Povlishock, J.T., Marmarou, A., McIntosh, T., Trojanowski, J.Q., Moroi, J., 1997. Impact acceleration injury in the rat: evidence for focal axolemmal change and related neurofilament sidearm alteration. J. Neuropathol. Exp. Neurol. 56, 347–359. Price, R.L., Paggi, P., Lasek, R.J., Katz, M.J., 1988. Neurofilaments are spaced randomly in the radial dimension of axons. J. Neurocytol. 17, 55–62. Reeves, T.M., Phillips, L.L., Povlishock, J.T., 2005. Myelinated and unmyelinated axons of the corpus callosum differ in vulnerability and functional recovery following traumatic brain injury. Exp. Neurol. 196, 126–137. Shi, R., Whitebone, J., 2006. Conduction deficits and membrane disruption of spinal cord axons as a function of magnitude and rate of strain. J. Neurophysiol. 95, 3384–3390. Shreiber, D.I., Bain, A.C., Ross, D.T., Smith, D.H., Gennarelli, T.A., McIntosh, T.K., Meaney, D.F., 1999. Experimental investigation of cerebral contusion: histopathological and immunohistochemical evaluation of dynamic cortical deformation. J. Neuropathol. Exp. Neurol. 58, 153–164. Smith, M., 2003. Diffuse axonal injury in adults. Trauma 5, 227–234. Stone, J.R., Singleton, R.H., Povlishock, J.T., 2001. Intra-axonal neurofilament compaction does not evoke local axonal swelling in all traumatically injured axons. Exp. Neurol. 172, 320–331. Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu, L.K., Povlishock, J.T., Helm, G.A., 2004. Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp. Neurol. 190, 59–69. Suehiro, E., Povlishock, J.T., 2001. Exacerbation of traumatically induced axonal injury by rapid posthypothermic rewarming

38

BR A IN RE S E A RCH 1 4 52 ( 20 1 2 ) 2 9 –38

and attenuation of axonal change by cyclosporin A. J. Neurosurg. 94, 493–498. Suehiro, E., Singleton, R.H., Stone, J.R., Povlishock, J.T., 2001. The immunophilin ligand FK506 attenuates the axonal damage associated with rapid rewarming following posttraumatic hypothermia. Exp. Neurol. 172, 199–210. Szaro, B.G., Whitnall, M.H., Gainer, H., 1990. Phosphorylation-dependent epitopes on neurofilament proteins and neurofilament densities differ in axons in the corticospinal and primary sensory dorsal column tracts in the rat spinal cord. J. Comp. Neurol. 302, 220–235. Witelson, S.F., 1989. Hand and sex differences in the isthmus and genu of the human corpus callosum. A postmortem morphological study. Brain 112 (Pt 3), 799–835. Zhang, L., Yang, K.H., King, A.I., 2001. Biomechanics of neurotrauma. Neurol. Res. 23, 144–156. Zhang, L., Yang, K.H., King, A.I., 2004. A proposed injury threshold for mild traumatic brain injury. J. Biomech. Eng. 126, 226–236.

Zhang, J., Yoganandan, N., Pintar, F.A., Gennarelli, T.A., 2006. Role of translational and rotational accelerations on brain strain in lateral head impact. Biomed. Sci. Instrum. 42, 501–506. Zhang, L., Zhou, R., Neng, Y., 2010. Biomechanical response of impact acceleration-induced traumatic axonal injury: rat head model development and injury localization. J. Neurotrauma A18, 41. Zhang, L., Gurao, M., Yang, K.H., King, A.I., 2011a. Material characterization and computer model simulation of low density polyurethane foam used in a rodent traumatic brain injury model. J. Neurosci. Methods 198, 93–98. Zhang, L., Zhou, R., Li, Y., Kallakuri, S., Cavanaugh, J.M., 2011b. Local brain strain predicts traumatic axonal pathology in a rodent model of head impact acceleration induced brain injury. J. Neurotrauma 28 (123), p346. Zhang, L., Li, Y., Cavanaugh, J.M., submitted for publication. An improved head impact acceleration device for quantifying head biomechanical responses in a rodent model of traumatic axonal injury.