Accumulation of non-erythroid αII-spectrin and calpain-cleaved αII-spectrin breakdown products in cerebrospinal fluid after traumatic brain injury in rats

Authors


Address correspondence and reprint requests to Dr Brian R. Pike, Department of Neuroscience, University of Florida, 100 S. Newell Dr, Box 100244, Gainesville, FL 32611 USA. E-mail: pike@ufbi.ufl.edu

Abstract

Although a number of increased CSF proteins have been correlated with brain damage and outcome after traumatic brain injury (TBI), a major limitation of currently tested biomarkers is a lack of specificity for defining neuropathological cascades. Identification of surrogate biomarkers that are elevated in CSF in response to brain injury and that offer insight into one or more pathological neurochemical events will provide critical information for appropriate administration of therapeutic compounds for treatment of TBI patients. Non-erythroid αII-spectrin is a cytoskeletal protein that is a substrate of both calpain and caspase-3 cysteine proteases. As we have previously demonstrated, cleavage of αII-spectrin by calpain and caspase-3 results in accumulation of protease-specific spectrin breakdown products (SBDPs) that can be used to monitor the magnitude and temporal duration of protease activation. However, accumulation of αII-spectrin and αII-SBDPs in CSF after TBI has never been examined. Following a moderate level (2.0 mm) of controlled cortical impact TBI in rodents, native αII-spectrin protein was decreased in brain tissue and increased in CSF from 24 h to 72 h after injury. In addition, calpain-specific SBDPs were observed to increase in both brain and CSF after injury. Increases in the calpain-specific 145 kDa SBDP in CSF were 244%, 530% and 665% of sham-injured control animals at 24 h, 48 h and 72 h after TBI, respectively. The caspase-3-specific SBDP was observed to increase in CSF in some animals but to a lesser degree. Importantly, levels of these proteins were undetectable in CSF of uninjured control rats. These results indicate that detection of αII-spectrin and αII-SBDPs is a powerful discriminator of outcome and protease activation after TBI. In accord with our previous studies, results also indicate that calpain may be a more important effector of cell death after moderate TBI than caspase-3.

Abbreviations used
CT

computed tomography

GCS

Glasgow coma scale, GOS, Glasgow outcome scale, PVDF, polyvinylidene fluoride

SBDPs

spectrin breakdown products

TBI

traumatic brain injury.

The incidence of traumatic brain injury (TBI) in the United States of America is conservatively estimated to be more than 2 million persons annually with approximately 500 000 hospitalizations (Goldstein 1990). Of these, about 70 000–90 000 head injury survivors are permanently disabled. The annual economic cost to society for care of head-injured patients is estimated at $25 billion (Goldstein 1990). Thus, accurate and reliable measurement of outcome following head injury is of great interest to both head injury survivors and clinicians. Assessment of pathology and neurological impairment immediately after TBI is crucial for determination of appropriate clinical management and for predicting long-term outcome. The outcome measures most often used in head-injured patients are the Glasgow coma scale (GCS), the Glasgow outcome scale (GOS), and computed tomography (CT) scans to detect intracranial pathology. However, despite dramatically improved emergency triage systems based on these outcome measures, most TBI survivors suffer long-term (for a number of years) impairment, and a large number of TBI survivors are severely affected by TBI despite predictions of ‘good recovery’ on the GOS (Marion 1996). Because of the limitations of current clinical assessments of TBI severity, there has been an increased interest in the development of neurochemical markers for determining injury severity and for clinical evaluation of pathophysiological mechanisms operative in traumatized brain.

For example, TBI results in neuronal tissue death that can cause a variety of neurochemicals such as amino acids, ions and lactate, as well as a number of cellular proteins and enzymes, to be released into the blood and CSF (Goodman and Simpson 1996). Although assessment of cardiac and liver protein levels in the blood has routinely been used in medical practice for years (e.g. creatine kinase MB or troponin-T), assessment of CNS proteins in blood or CSF is far less developed. Thus, recent studies have measured a variety of neurochemical substances in the CSF or blood in attempt to identify specific surrogate markers of cellular damage and outcome after TBI and other CNS disorders (Haber and Grossman 1980; Inao et al. 1988; Robinson et al. 1990; Lyeth et al. 1993; Raabe and Seifert 1999; Raabe et al. 1999; Zemlan et al. 1999; Clark et al. 2000a; Tapiola et al. 2000). For example, creatine kinase BB, lactate dehydrogenase, myelin basic protein, and neuron-specific enolase have been measured in blood or CSF in various CNS disorders including TBI. However, these proteins are non-specific to the brain, offer no insight as to mechanism of injury, and/or prediction of outcome utilizing these proteins has not proven reliable (Goodman and Simpson 1996). Other proteins detected in CSF after brain injury such as S-100B are highly specific to the CNS and have been more robustly correlated with outcome (Raabe and Seifert 1999; Raabe et al. 1999). Although brain-specific surrogate biomarkers like S-100B may be useful indicators of outcome after brain injury, detection of these proteins in blood or CSF offers no insight into neurochemical alterations that mediate brain damage after TBI. Thus, identification of neurochemical markers that are specific to the CNS and that provide information about specific ongoing neurochemical events would prove immensely beneficial for both prediction of outcome and for guidance of targeted therapeutic delivery.

Non-erythroid αII-spectrin is the major structural component of the cortical membrane cytoskeleton, is particularly abundant in axons and presynaptic terminals (Riederer et al. 1986; Goodman et al. 1995), and is a major substrate for both calpain and caspase-3 cysteine proteases (Wang et al. 1998). The calpain-mediated cleavage of αII-spectrin occurs between Tyr1176 and Gly1177 resulting in the formation of calpain-signature spectrin breakdown products (SBDPs) of 150 and 145 kDa (Harris et al. 1988). The caspase-3-mediated cleavages of αII-spectrin occur at Asp1185, Ser1186, Asp1478 and Ser1479 resulting in the formation of caspase-3-signature SBDPs of 150 and 120 kDa, respectively (Wang et al. 1998). Importantly, numerous investigations have documented increased pathological activation of calpain and/or caspase-3 proteases after TBI (Saatman et al. 1996a, 2000; Kampfl et al. 1997; Newcomb et al. 1997; Posmantur et al. 1997; Yakovlev et al. 1997; Pike et al. 1998a; Clark et al. 1999, 2000b; LaPlaca et al. 1999; Okonkwo et al. 1999; Zhang et al. 1999; Beer et al. 2000; Buki et al. 2000). In addition, our laboratory and others have provided extensive evidence that αII-spectrin is processed by calpains and/or caspase-3 to signature cleavage products in vivo after TBI (Beer et al. 2000; Newcomb et al. 1997; Pike et al. 1998a; Buki et al. 2000) and in in vitro models of mechanical stretch injury (Pike et al. 2000b), necrosis (Zhao et al. 1999), apoptosis (Nath et al. 1996a, 1996b; Pike et al. 1998b), glutamate or NMDA excitotoxicity (Nath et al. 2000; Zhao et al. 2000), and oxygen-glucose deprivation (Nath et al. 1998; Newcomb et al. 1998). Moreover, use of selective αII-SBDP antibodies has been used to demonstrate that brain regions with the highest accumulation of SBDPs also have the highest levels of neuronal cell death (Roberts-Lewis et al. 1994; Newcomb et al. 1997). Thus, the ubiquitous distribution of αII-spectrin in the brain coupled with the ability to utilize signature αII-spectrin proteolytic fragments generated by pathological activation of calpain and/or caspase-3 after TBI makes αII-spectrin a potentially important biomarker of brain damage. To test this hypothesis, the present investigation examined alterations in brain levels of αII-spectrin and αII-SBDPs after controlled cortical impact TBI in rodents, and compared these changes to accumulation of αII-spectrin and αII-SBDPs in CSF in the same animals.

Materials and methods

Surgical Preparation and controlled cortical impact traumatic brain injury

As previously described (Dixon et al. 1991; Pike et al. 1998a), a cortical impact injury device was used to produce TBI in rodents. Cortical impact TBI results in cortical deformation within the vicinity of the impactor tip associated with contusion, and neuronal and axonal damage that is constrained in the hemisphere ipsilateral to the site of injury (Gennarelli 1994; Meaney et al. 1994). Adult male (280–300 g) Sprague-Dawley rats (Harlan; Indianapolis, IN, USA) were initially anesthetized with 4% isoflurane in a carrier gas of 1 : 1 O2/N2O (4 min) followed by maintenance anesthesia of 2.5% isoflurane in the same carrier gas. Core body temperature was monitored continuously by a rectal thermistor probe and maintained at 37 ± 1°C by placing an adjustable temperature controlled heating pad beneath the rats. Animals were mounted in a stereotactic frame in a prone position and secured by ear and incisor bars. A midline cranial incision was made, the soft tissues were reflected, and a unilateral (ipsilateral to site of impact) craniotomy (7 mm diameter) was performed adjacent to the central suture, midway between bregma and lambda. The dura mater was kept intact over the cortex. Brain trauma in rats (n = 9) was produced by impacting the right cortex (ipsilateral cortex) with a 5-mm diameter aluminum impactor tip (housed in a pneumatic cylinder) at a velocity of 3.5 m/s with a 2.0-mm compression and 150 ms dwell time (compression duration). Velocity was controlled by adjusting the pressure (compressed N2) supplied to the pneumatic cylinder. Velocity and dwell time were measured by a linear velocity displacment transducer (Lucas Shaevitz™ model 500 HR; Detroit, MI, USA) that produces an analogue signal that was recorded by a storage-trace oscilloscope (BK Precision, model 2522B; Placentia, CA, USA). Sham-injured animals (n = 4) underwent identical surgical procedures but did not receive an impact injury. Appropriate pre- and post-injury management was maintained to insure that all guidelines set forth by the University of Florida Institutional Animal Care and Use Committee and the National Institutes of Health guidelines detailed in the Guide for the Care and use of Laboratory Animals were complied with.

CSF and cortical tissue preparation

The CSF and brain cortices were collected from animals at various intervals after sham-injury or TBI. At the appropriate time-points, TBI or sham-injured animals were anesthetized as described above and secured in a stereotactic frame with the head allowed to move freely along the longitudinal axis. The head was flexed so that the external occipital protuberance in the neck was prominent and a dorsal midline incision was made over the cervical vertebrae and occiput. The atlanto-occipital membrane was exposed by blunt dissection and a 25G needle attached to polyethylene tubing was carefully lowered into the cisterna magna. Approximately 0.1–0.15 mL of CSF was collected from each rat. Following CSF collection, animals were removed from the stereotactic frame and immediately killed by decapitation. Ipsilateral and contralateral (to the impact site) cortices were then rapidly dissected, rinsed in ice cold PBS, and snap frozen in liquid nitrogen. Cortices beneath the craniotomies were excised to the level of the white matter and extended ∼4 mm laterally and ∼7 mm rostrocaudally. The CSF samples were centrifuged at 4000 g for 4 min at 4°C to clear any contaminating erythrocytes. Cleared CSF and frozen tissue samples were stored at − 80°C until ready for use. Cortices were homogenized in a glass tube with a Teflon dounce pestle in 15 volumes of an ice-cold triple detergent lysis buffer (20 mm HEPES, 1 mm EDTA, 2 mm EGTA, 150 mm NaCl, 0.1% SDS, 1.0% IGEPAL 40, 0.5% deoxycholic acid, pH 7.5) containing a broad range protease inhibitor cocktail (cat. #1-836-145Roche Molecular Biochemicals, Indianapolis, IN, USA).

Immunoblot analyses of CSF and cortical tissues

Protein concentrations of tissue homogenates and CSF were determined by bicinchoninic acid microprotein assays (Pierce Inc., Rockford, IL, USA) with albumin standards. Protein balanced samples were prepared for sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) in twofold loading buffer containing 0.25 m Tris (pH 6.8), 0.2 m DTT, 8% SDS, 0.02% bromophenol blue, and 20% glycerol in distilled H2O. Samples were heated for 10 min at 100°C and centrifuged for 1 min at 8160 gin a microcentrifuge at ambient temperature. Forty micrograms of protein per lane was routinely resolved by SDS–PAGE on 6.5% Tris/glycine gels for 1 h at 200 V. Following electrophoresis, separated proteins were laterally transferred to polyvinylidene fluoride (PVDF) membranes in a transfer buffer containing 0.192 m glycine and 0.025 m Tris (pH 8.3) with 10% methanol at a constant voltage of 100 V for 1 h at 4°C. Blots were blocked for 1 h at ambient temperature in 5% non-fat milk in TBS and 0.05% Tween-20. Panceau Red (Sigma, St Louis, MO, USA) was used to stain membranes to confirm successful transfer of protein and to insure that an equal amount of protein was loaded in each lane.

Antibodies and immunolabeling of PVDF membranes

Immunoblots containing brain or CSF protein were probed with an anti-α-spectrin (fodrin) monoclonal antibody (FG 6090 Ab; clone AA6; cat. # FG 6090; Affiniti Research Products Limited, Mamhead Castle, Mamhead, Exeter, UK) that detects intact non-erythroid αII-spectrin (280 kDa) and 150, 145 and 120 kDa cleavage fragments to αII-spectrin. A cleavage product of 150 kDa is initially produced by calpains or caspase-3 proteases (each proteolytic cleavage yields a unique amino-terminal region; Nath et al. 1996b; Wang et al. 1998). The calpain-generated 150 kDa product is further cleaved by calpain to yield a specific calpain signature product of 145 kDa (Harris et al. 1988; Nath et al. 1996a,b) whereas the caspase-3 generated 150 kDa product is further cleaved by caspase-3 to yield a specific caspase-3 signature product of 120 kDa (Nath et al. 1998; Wang et al. 1998). To further confirm the specificity of calpain-cleaved spectrin in CSF after TBI, a second antibody (anti-SBDP150; rabbit polyclonal) that recognizes only the calpain-cleaved N-terminal region (GMMPR) of the 150 kDa αII-spectrin breakdown product (SBDP) was also used (Saido et al. 1993; Nath et al. 1996b). Some immunoblots were immunolabeled with an antibody that recognizes erythroid αI-spectrin (Cat.# BYA10881; Accurate Chemical & Scientific Corp, Westbury, NY, USA). Following an overnight incubation at 4°C with the primary antibodies (FG 6090 Ab, 1 : 4000 for brain tissue and 1: 2000 for CSF; SBDP150 Ab, 1 : 1000; BYA10881, 1 : 400), blots were incubated for 1 h at ambient temperature in 3% non-fat milk that contained a horseradish peroxidase-conjugated goat anti-mouse IgG (1 : 10 000 dilution) or goat-anti-rabbit IgG (1 : 3000). Enhanced chemiluminescence (ECL; Amersham) reagents were used to visualize immunolabeling on Kodak Biomax ML chemiluminescent film.

Statisitical analyses

Semi-quantitative evaluation of protein levels detected by immunblotting was performed by computer-assisted densitometric scanning (AlphaImager 2000; Digital Imaging System, San Leandro, CA, USA). Data were acquired as integrated densitometric values and transformed to percentages of the densitometric levels obtained on scans from sham-injured animals visualized on the same blot. Data was evaluated by least squares linear regression followed by ANOVA. All values are given as mean ± SEM. Differences were considered significant if p < 0.05.

Results

Proteolysis of αII-spectrin in the cortex by calpain, but not caspase-3 after TBI

In the ipsilateral cortex, TBI resulted in decreased protein levels of αII-spectrin (280 kDa) that were associated with concomitant accumulation of calpain-generated 150 and 145 kDa αII-SBDPs (Fig. 1). However, there was little to no detectable increase in the caspase-3-genearated 120 kDa αII-SBDP. These results replicate our previous investigation that reported calpain but not caspase-3 processing of αII-spectrin following a moderate level of lateral controlled cortical impact TBI (Pike et al. 1998a). Decreased αII-spectrin (280 kDa) protein levels were 65%, 48% and 39% of sham-injured protein levels at 24 h, 48 h, and 72 h after TBI, respectively (Fig. 2). Increased 150 kDa αII-SBDP levels were 189%, 157%, and 153% of sham-injured levels at 24 h, 48 h and 72 h after TBI, respectively, while increased 145 kDa αII-SBDP levels were 237%, 203% and 198% of sham-injured levels at 24 h, 48 h and 72 h after TBI, respectively (Fig. 2).

Figure 1.

 Traumatic brain injury (TBI) results in prominent accumulation of αII-spectrin (280 kDa) and calpain-cleaved 150 kDa and 145 kDa αII-SBDPs in CSF (FG 6090 Ab). The caspase-3 generated 120 kDa fragment was also apparent in CSF of some animals. TBI caused proteolysis of constitutively expressed brain αII-spectrin (280 kDa) in ipsilateral but not contralateral cortex. Increases in the caspase-3-mediated 120 kDa αII-SBDP were not as apparent in ipsilateral or contralateral cortex after TBI. Immunolabeling of additional unknown bands at ∼110 kDa and 95 kDa were also detected in CSF at 48 h and 72 h after injury. CSF1 and CSF2 are from two separate series of animals shown to illustrate that there was more variability in CSF levels of SBDPs than in brain levels, which may reflect individual differences in injury severity.

Figure 2.

 Mean arbitrary densitometric units obtained from 280 kDa αII-spectrin and the 150 kDa and 145 kDa αII-SBDPs were converted to percent of sham-injured values. Decreases in 280 kDa αII-spectrin and increases in 150 kDa and 145 kDa αII-SBDPs (ipsilateral cortex) were associated with concomitant increases of these proteins in the CSF. Mean accumulation of the caspase-3 generated 120 kDa fragment in these tissues was relatively flat.

In the contralateral cortex, traumatic brain injury resulted in no apparent alteration in protein levels of αII-spectrin (280 kDa) or in any apparent accumulation of calpain-generated 150 or 145 kDa αII-SBDPs, or in caspase-3 generated 120 kDa αII-SBDP as compared to sham-injured control animals (Fig. 1). These results are also in accord with our previous report that calpain-mediated processing of αII-spectrin is predominately confined to ipsilateral brain regions after moderate lateral controlled cortical impact TBI (Newcomb et al. 1997; Pike et al. 1998a).

Accumulation of calpain-mediated αII-SBDPs in CSF after TBI

Immunoblot analyses of CSF levels of non-erythroid αII-spectrin and αII-SBDPs (FG 6090 Ab) showed no detectable levels of these proteins in CSF of sham-injured control animals (Fig. 1). However, after TBI, accumulation of αII-spectrin (280 kDa) and calpain-generated 150 and 145 kDa SBDPs were markedly increased at 24 h, 48 h and 72 h, after injury (Fig. 1). In addition, there was an increase in the caspase-3-generated 120 kDa fragment in one animal at 48 h after TBI, and in another animal at 72 h after TBI (Fig. 1). Accumulation of αII-spectrin (280 kDa) protein levels was 143%, 212%, and 379% of sham-injured animals at 24 h, 48 h, and 72 h after TBI, respectively (Fig. 2). Similarly, accumulation of 150 kDa αII-SBDP after TBI was 155%, 434%, and 583% of sham-injured levels at 24 h, 48 h, and 72 h, respectively, while accumulation of 145 kDa αII-SBDP after TBI was 244%, 530%, and 665% of sham-injured levels at 24 h, 48 h and 72 h, respectively (Fig. 2). In contrast, although accumulation of the caspase-3 cleaved 120 kDa fragment was detected in two animals, the average response was relatively flat. In addition, several lower molecular weight species of αII-spectrin were detected. The protease(s) responsible for these lower molecular weight fragments are currently unknown. However, future identification of these bands may provide important new information regarding other neurochemical events in the brain after TBI.

To provide further confirmation of calpain-generated αII-SBDP accumulation in CSF after TBI, an additional group of animals (n = 5 per time-point) was injured as described above and immunoblots of CSF samples were probed with anti-SBDP150 Ab. In this experiment, an additional time-point (7 days post-TBI) was also examined. The SBDP150 Ab specifically recognizes only the calpain-cleaved 150 kDa αII-spectrin fragment and does not recognize the intact 280 kDa protein or other proteolytic fragments (Saido et al. 1993; Nath et al. 1996b). Results with the SBDP150 Ab were nearly identical to those obtained with the FG-6090 Ab (Fig. 3). The calpain-cleaved 150 kDa SBDP was nearly undetectable in CSF of sham-injured animals, and a progressive increase in immunoreactivity was observed from 24 h to 72 h after TBI. Importantly, this experiment also demonstrated that levels of calpain-cleaved 150 kDa SBDP were decreased back to sham-injured control levels by seven days after TBI (Fig. 3).

Figure 3.

 N-terminal fragment-specific detection of calpain-generated SBDP150 in CSF after traumatic brain injury (SBDP150 Ab). (a) As with the FG 6090 Ab, the SBDP150 Ab detected a progressive increase in the calpain-cleaved 150 kDaA SBDP from 24 h to 72 h after TBI. Levels of 150 kDa SBDP had resolved to sham-injured control levels by seven days after TBI. (b) Mean arbitrary densitometric units of SBDP150 levels detected with anti-SBDP150 Ab.

Linear regression analyses of Cortical versus CSF levels of αII-spectrin and αII-SBDPs

Least squares linear regression was calculated to determine the relationship between brain and CSF levels of αII-spectrin and αII-SBDPs over days post-injury. The slope of the regression lines for αII-spectrin and αII-SBDPs in brain and CSF were analyzed by ANOVA.

For cortical levels of 280 kDa αII-spectrin protein, the slope of the regression line was relatively steep and negative indicating large decreases over days in cortical levels of native αII-spectrin protein (Fig. 4). In contrast, the slope of the regression line for CSF levels of 280 kDa αII-spectrin was relatively steep and positive indicating large increases over days in CSF levels of αII-spectrin protein after TBI. In addition, ANOVA indicated that there was a significant difference (F = 19.95, p < 0.001) between cortical and CSF slopes for 280 kDa αII-spectrin protein level. This significance indicates that as brain levels of αII-spectrin decrease over days, CSF levels of αII-spectrin increase over days.

Figure 4.

 Cortical versus CSF levels of αII-spectrin (280 kDa) and αII-SBDPs (150, 145, and 120 kDa) over days post-injury. Least squares regression lines of brain and CSF spectrin and SBDP levels were plotted on the same graph. Pearson correlation coefficients for each regression line are indicated. Results indicate that parenchymal decreases in levels of native αII-spectrin (280 kDa) are associated with increases in CSF accumulation while increased parenchymal levels of calpain-mediated αII-SBDPs (150 and 145 kDa) are associated with increased CSF accumulation. On average, there were no changes in parenchymal or CSF levels of the caspase-3-mediated 120 kDa αII-SBDP, although individual rats at different time points showed some increase in CSF levels of the 120 kDa product.

For cortical and CSF levels of 150 kDa αII-SBDP, both slopes of the regressions lines were positive indicating large increases in the calpain-cleaved 150 kDa αII-SBDP in brain and CSF over days (Fig. 4). ANOVA indicated no significant difference (F = 1.86, p = 0. 1873) between slopes indicating that the relative accumulation of 150 kDa αII-SBDP in cortex and CSF were similar. However, the slope for CSF 150 kDa αII-SBDP was relatively steeper than the slope for cortical 150 kDa αII-SBDP. This result reflects the densitometric data (Fig. 2) indicating that, in the cortex, peak levels of the 150 kDa αII-SBDP accumulated rapidly (24 h) and were maintained at 48 h and 72 h post-injury. This result also reflects densitometric data (Fig. 2) indicating that CSF levels of the 150 kDa αII-SBDP accumulated more slowly early after injury (24 h) with a greater rate of further accumulation at 48 h and 72 h post-injury. Observed statistical differences in accumulation rates can be appreciated visually in the immunoblot data (Fig. 1). The stability of αII-spectrin and αII-SBDPs in CSF may be increased due to lack of endogenous proteases. For example, when CSF from TBI animals was stored in individual aliquots at either − 85°C or at ambient laboratory temperature (∼26°C) without protease inhibitors for 48 h, αII-SBDP levels from ambient temperature aliquots were only decreased by 28% compared to aliquots stored at − 85°C (Fig. 5). Importantly, the relative stability of αII-SBDP protein in CSF at ambient temperature further indicates this protein as a useful biomarker after TBI.

Figure 5.

 Stability of αII-SBDP in CSF after prolonged storage in the absence of protease inhibitors at ambient laboratory temperature. The αII-SBDP protein levels only decreased by 28% when stored at ambient temperature (∼26°C) for 48 h compared to identical samples stored at − 85°C for 48 h. These results indicate that αII-SBDPs in CSF are relatively stable at room temperature. This is an important practical consideration for clinical utility.

For cortical and CSF levels of calpain-cleaved 145 kDa αII-SBDP, both slopes of the regression lines were steep and positive indicating large increases in the 145 kDa αII-SBDP in brain and CSF over days (Fig. 4). ANOVA indicated no significant difference (F = 0.69, p = 0.4153) between slopes indicating that the relative accumulation of 145 kDa αII-SBDP in cortex and CSF were similar as compared to the respective controls. Comparison of slopes for 150 kDa and 145 kDa αII-SBDPs in the brain revealed that the slope of the brain 145 kDa αII-SBDP over days was considerably steeper than the slope of the brain 150 kDa αII-SBDP. This result indicates that brain 145 kDa αII-SBDP protein levels accumulate at a greater rate over days than brain 150 kDa αII-SBDP protein levels. This observation is most likely the result of lower basal levels of brain 145 kDa αII-SBDP than brain 150 kDa αII-SBDP in sham-injured animals and of continued calpain digestion of the larger 150 kDa αII-SBDP to the smaller 145 kDa αII-SBDP over time.

For cortical and CSF levels of caspase-3-cleaved 120 kDa αII-SBDP, both slopes were nearly horizontal, indicating no increased accumulation of caspase-3-generated 120 kDa αII-SBDP over days after TBI (Fig. 4). In addition, anova indicated no significant difference between slopes (F = 0.002, p = 0.9621).

Erythroid αI-spectrin versus non-erythroid αII-spectrin

After head injury, the most likely source of CSF contamination will be from blood. Both neurons and blood contain the erythroid form αI-spectrin protein. However, erythrocytes do not contain non-erythroid αII-spectrin protein. To demonstrate that the source of αII-spectrin immunoreactivity in the CSF is not blood borne, we probed immunoblots containing various concentrations of whole blood proteins and brain protein with either an erythroid anti-αI-spectrin antibody or with an anti-αII-spectrin antibody (Fig. 6a,b). As predicted, no immunoreactivity was observed at any concentration of whole blood protein (0–40 µg) while brain spectrin was highly reactive to the non-erythroid anti-αII-spectrin antibody. In contrast, both the brain and blood protein samples were immunoreactive to the erythroid anti-αI-spectrin antibody. This result clearly indicates that use of the non-erythroid, but not the erythroid, anti-spectrin antibody can be used to discriminate non-blood borne spectrin protein in CSF samples.

Figure 6.

 (a) Non-erythroid αII-spectrin protein is not detected in whole blood. After TBI, the most probable source of non-CNS accumulation of proteins in the CSF is from blood. This immunoblot demonstrates that non-erythroid αII-spectrin is detectable in brain protein homogenates but not in blood protein homogenates. (b) In contrast, use of an erythroid αI-spectrin antibody on the same blot that has been stripped and re-probed reveals immunoreactivity for both blood and brain spectrin. These results demonstrate that potential blood contamination of CSF samples does not affect detection of brain-derived αII-spectrin.

Discussion

This paper provides the first evidence for accumulation of non-erythroid αII-spectrin protein and calpain-mediated αII-SBDPs in CSF after TBI. Detection of calpain-specific proteolytic fragments to αII-spectrin were confirmed with two antibodies, one that recognizes both intact αII-spectrin and calpain-specific SBDPs (FG 6090 Ab), and one that recognizes only the N-terminal region of calpain-cleaved 150 kDa SBDP (SBDP150 Ab). Results of this investigation indicate that CSF detection of αII-spectrin and αII-SBDPs can provide both a sensitive surrogate biochemical measure of TBI pathology and provide important information about specific neurochemical events that have occurred in the brain after TBI. To our knowledge, this is the first investigation of any CNS pathology to indicate that identification of accumulated CSF proteins or protein metabolic products can be used to infer specific neurochemical events (i.e. calpain activation) in the brain. Thus, use of αII-SBDPs as surrogate biochemical markers of TBI has important clinical ramifications for assessment of outcome after injury and for determination of specific pathological proteolytic cascades known to occur after TBI. Although other CNS proteins have been detected in CSF after brain injury (e.g. S-100B) and have been correlated with outcome, these proteins offer no insight into pathological mechanisms that have occurred in the brain after TBI. Obviously, identification of metabolic products with known neurochemical etiology will be beneficial for appropriate application of targeted therapeutics (such as calpain inhibitors) after TBI.

Calpain and caspase-3 cysteine proteases are important mediators of cell death and dysfunction in numerous CNS diseases and injuries including TBI. The calpains have historically been associated with necrotic (oncotic) cell death although recent evidence indicates a role in apoptotic cell death as well (Linnik et al. 1996; Nath et al. 1996a,b; Newcomb et al. 1998; Pike et al. 1998b). Numerous investigations have reported calpain activation after TBI (Saatman et al. 1996a, 2000; Kampfl et al. 1997; Newcomb et al. 1997; Posmantur et al. 1997; Pike et al. 1998a) and inhibitors of calpains have been shown to confer neuroprotection after TBI (Posmantur et al. 1997; Saatman et al. 1996b, 2000). Caspase-3 is a critical executioner of apoptosis and caspase-3 activation has been reported in in vitro (Shah et al. 1997; Allen et al. 1999; Pike et al. 2000b) and in vivo (Beer et al. 2000; Yakovlev et al. 1997; Pike et al. 1998a; Clark et al. 2000b) models of TBI. However, it should be noted that at least in our hands, the magnitude of calpain activation after TBI is much greater than that of caspase-3, and that at the moderate level of brain injury employed in the current study, caspase-3 is only transiently elevated in deep, non-cortical brain regions (Pike et al. 1998a). This result most likely accounts for the detection of relatively minimal amounts of the 120 kDa caspase-3-mediated αII-SBDP in CSF after TBI. In contrast to our injury model, Beer et al. (2000) have observed prominent levels of caspase-3 activation in the cortex after cortical impact TBI. However, while our cortical impact model is typically characterized by prominent tissue necrosis and progressive cortical cavitation to the gray–white interface (Kampfl et al. 1996; Newcomb et al. 1997; Dixon et al. 1998; Newcomb et al. 1999; Pike et al. 2000a), the model employed by Beer et al. (2000) was not. Thus, differences in injury magnitude may be important factors affecting calpain and/or caspase-3 activation after TBI, and this hypothesis warrants further investigation. However, it should be pointed out that although caspase-3 activation has not been a prominent feature in our model of cortical impact TBI, we have detected substantial levels of apoptotic cell death in the cortex after TBI (Newcomb et al. 1999). This apparent discrepancy between apoptotic cell death and caspase-3 activation raises the intriguing possibility that apoptosis may occur via a caspase-3-independent pathway after TBI. This observation also warrants further examination.

That different injury magnitudes may result in differential activation of calpain or caspase-3 proteases has important implications for targeted therapeutic intervention after TBI, and importantly, further validates the utility of using surrogate markers of TBI that have known neurochemical etiologies. For example, the current investigation detected CSF accumulation of the calpain-mediated αII-SBDP and not the caspase-3-mediated αII-SBDP. Based on this evidence, administration of calpain but not caspase-3 inhibitors would be predicted to have the most beneficial effect on outcome. However, other injury magnitudes may result in more caspase-3 activation indicating use of caspase-3 inhibitors or a combination of calpain/caspase-3 protease inhibitors. Thus, surrogate measures of TBI will result in selective pharmaceutical therapies based on clinical assessment of neuropathology, and this approach is a superior strategy to promiscuous prophylactic administration of unnecessary and potentially harmful compounds.

The most probable source of peripheral contamination of the CSF after TBI will be blood born. Indeed, we did detect visible red blood cell contamination of CSF after experimental TBI (which was removed by centrifugation). However, our control experiments with brain and whole blood immnuoblots (Fig. 6a,b) clearly demonstrated that the non-erythroid anti-αII-spectrin antibody did not detect any αII-spectrin protein in whole blood samples. Conversly, the erythroid αI-spectrin antibody labeled both brain and blood samples. These results indicate that the major source of potential peripheral CSF contamination after TBI, blood, is not detected by the non-erythroid anti-αII-spectrin antibody. This finding supports the utility of αII-spectrin and αII-SBDPs as surrogate biomarkers of injury after TBI, and importantly, as biomarkers of calpain and/or caspase-3 activation after TBI.

One caveat to the current investigation is the finding that there was more variability in levels of CSF SBDPs than there were in brain levels of SBDPs. This variability is indicated by the larger error bars in Fig. 2 and 4 and can be observed in individual animals in Fig. 1. The reason for the larger variability in CSF protein accumulation is unknown, but may reflect differences in individual animal's CSF circulation after TBI. For example, differences in increased intracranial pressure after TBI may restrict passage of CSF through various foramina that may preclude detection of secreted proteins into the cisterna magna (source of CSF in the present study). Additional studies should examine differences in intraventricular versus intracisternal levels of accumulated SBDPs.

Nonetheless, future studies focused on development of neuron-specific antibodies targeted against calpain-specific and caspase-3-specific αII-SBDPs (such as the SBDP150 Ab) will further strengthen the utility and specificity of αII-SBDPs as surrogate markers of brain injury. In addition, development of enzyme-linked immunosorbent assays (ELISA) will allow greater quantification of calpain and caspase-3 SBDPs and provide a more rapid and practical approach to CSF detection of these proteins.

Acknowledgements

This work was supported by National Institute of Health (NIH) R01 NS39091, NIH R01 NS40182, US Army DAMD17-99-1-9565 to RLH and by NIH award F32-NS10857 and the State of Florida Brain and Spinal Cord Rehabilitation Trust Fund (BSCIRTF) to BRP.

Ancillary