Background
Clinical and experimental studies have implied a pivotal role for the membrane attack complex (MAC, C5b-9) of the terminal complement activation pathway in the pathogenesis of secondary neuronal cell death after traumatic brain injury [
1‐
4]. The complement regulatory molecule CD59 represents the major controller of MAC formation and an essential protector of homologous cell injury after complement activation [
5,
6]. Neurons express CD59 constitutively to protect from autologous "innocent bystander" cell lysis after activation of the complement system in the injured brain [
2,
7]. However, due to low levels of neuronal CD59 expression, the neuronal capacity of controlling complement activation is very limited [
7], which renders neurons susceptible to complement-mediated lysis by the MAC in the setting of intracerebral complement activation [
8‐
10].
One of the putative mechanisms of complement-mediated neuronal death is reflected by the notion that the activation of phosphatidyl-inositol-specific phospholipase C (PI-PLC) after traumatic brain injury [
11] renders neurons vulnerable to MAC-mediated lysis by shedding of the glycosyl-phosphatidyl-inositol (GPI)-anchored glycoprotein CD59 from neuronal membranes [
2]. The intracerebral formation and deposition of MAC on neurons in the contusion area and penumbra zone has been shown to occur after human head injury [
3,
12]. However, the biological significance of CD59 in protecting from complement-mediated neuropathology after traumatic brain injury is far from being fully understood.
The present study was designed to investigate the role of CD59a in a standardized experimental model of closed head injury in mice lacking the gene for
Cd59a (CD59a
-/-). In mice, the
Cd59 gene is duplicated, yielding
Cd59a (widely expressed) and
Cd59b (testis-restricted) [
13,
14]. The CD59a
-/- mice were previously shown to be highly susceptible to complement-mediated demyelination and axonal injury in a model of experimental allergic encephalomyelitis (EAE) [
15], and thus provide an excellent
in vivo model to investigate the role of complement-mediated membrane attack and CD59-dependent neuroprotection in the setting of traumatic brain injury. We hypothesized that CD59a
-/- mice would be more susceptible to complement-mediated secondary brain injury than wild-type littermates in a standardized model of closed head injury.
Materials and methods
Animals
The generation and characterization of CD59a
-/- mice was previously described [
16]. These mice were found to have a spontaneous intravascular hemolysis due to erythrocyte susceptibility to complement-mediated lysis. Despite the chronic hemolysis, the CD59a
-/- mice are healthy and fertile, not anemic, but display elevated reticulocyte counts as a indicator of increased erythrocyte turnover [
16]. The CD59a
-/- mice were generated on a mixed 129/Sv × C57BL/6 genetic background. Wild-type littermates of the 129/Sv × C57BL/6 strain were used as controls. All mice were of age 10–12 weeks, weighing 28–32 g, and of male gender exclusively, in order to avoid a bias regarding gender-related susceptibilty to brain injury. Animals were kept in single cages, bred in a selective pathogen-free (SPF) environment under standardized conditions of temperature (21°C), humidity (60%), light and dark cycles (12:12 h), with food and water provided
ad libitum. A total of
n = 134 animals were used for this study (
n = 67 wild-type;
n = 67 CD59a
-/-). All experimental procedures were performed in compliance with the standards of the
Federation of European Laboratory Animal Science Association (FELASA) and were approved by the institutional animal care committee (
Landesamt für Arbeitsschutz, Gesundheitsschutz und technische Sicherheit Berlin, Berlin, Germany; approval No. G0308/04).
Surgical procedures
Mice were subjected to experimental closed head injury using a standardized weight-drop device, as previously described [
17]. In brief, after induction of isoflurane anesthesia, the skull was exposed by a longitudinal midline scalp incision. The head was fixed and a 250 g weight was dropped on the skull, inducing a focal blunt injury to the left hemisphere from a mean height of 1.7 ± 0.51 cm (mean ± SEM) for CD59a
-/-, and 2.6 ± 0.48 cm for wild-type mice, respectively. The difference in selected weight-drop heights between the knockout mice and wild-type littermates is due to the observed increased susceptibility to head injury in the CD59a
-/- mice, leading to sudden posttraumatic death in initial validation experiments. We therefore titrated the falling height to achieve a similar extent of neurological impairment 1 h after trauma between the groups, based on the baseline posttraumatic NSS (see below). The 1 h NSS reflects the initial extent of head injury. Thus, the reduced weight-drop height in the knockout mice is an indirect sign of increased susceptibility to trauma in the CD59a
-/- group, independent of the additional parameters assessed at later time-points. The stratification of cohorts according to the initial NSS at 1 h has been previously described in this model system as a feasible strategy to attain comparable injury severity between the groups at baseline. This notion is supported by recent studies in the same experimental system which demonstrated a correlation between injury severity determined on MRI with the NSS at 1 hour after trauma [
18,
19].
After trauma, all mice received supporting oxygenation with 100% O2 until fully awake. They were brought back to their cages and monitored at regular time intervals for up to 7 days. Posttraumatic analgesia was provided by injection of 0.1 mg/Kg buprenorphin s.c. immediately prior to the experimental procedure. The posttraumatic neurological impairment was assessed at defined time intervals after trauma (t = 1 h, 4 h, 24 h, and 7 days) using a standardized Neurological Severity Score (NSS), as described below. Sham-operated mice underwent identical procedures with regard to anesthesia, analgesia, and surgical scalp incision, but were not subjected to experimental head trauma. Normal mice were used as an additional internal negative control group and were kept under identical conditions as the trauma and sham-operated mice, but no anesthesia, analgesia, surgical and experimental procedures were performed.
Sample harvesting procedures
Ten mice per group and time-point were euthanized by decapitation under isoflurane anesthesia at t = 4 h, 24 h, and 7 days. Brains were immediately surgically removed, divided into left (injured) and right (contralateral/uninjured) hemispheres, snap-frozen in liquid nitrogen and stored at -80°C until analysis. In addition, serum samples were collected prior to decapitation by intracardiac puncture at identical time-points for determination of neuronal cell death markers by ELISA. Serum samples were collected in sterile tubes, centrifuged at 12,000 rpm at 4°C for 20 min, aliquoted and frozen at -80°C until analyzed.
Neurological severity score (NSS)
A previously characterized 10-parameter score was used for assessment of posttraumatic neurological impairment, as described elsewhere in detail [
20]. The NSS was assessed in a blinded fashion by two different investigators at the time-points
t = 1 h, 4 h, 24 h, and 7 days after trauma. The baseline NSS at 1 hour reflects the severity of the initial injury. The score comprises 10 individual parameters, including tasks on motor function, alertness, and physiological behavior, whereby one point is given for failure of the task, and no point for succeeding. A maximum NSS score of 10 points indicates severe neurological dysfunction, with failure of all tasks. A spontaneous recovery over time, for up to 4 weeks after trauma, is observed in this model system, as previously described [
17,
20,
21].
Quantification of neuron-specific enolase
Serum levels of neuron-specific enolase (NSE), an established marker of neuronal cell death after head injury [
22], were determined by a commercially available ELISA, specific for human NSE (Immuno-Biological Laboratories, Minneapolis, MN). Following confirmation of cross-reactivity of mouse NSE, samples were diluted 1:10 and analyzed by ELISA according to the manufacturer's protocol. Absorbance was read at 450 nm using a "SpectraMax 190" reader (Molecular Devices, Sunnyvale, CA). All samples were analyzed in duplicate and results were calculated from the means of duplicate sample analysis. The sensitivity of the assay was 1 pg/ml, and the standard curve was linear from 1 pg/ml to 140 pg/ml.
Assessment of neuronal cell death
The terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) technique was applied, using the "Fluorescein
In Situ Cell Death Detection Kit" (Roche Diagnostics GmbH, Mannheim, Germany), according to the manufacturer's instructions, to determine the extent of neuronal cell death in tissue sections, as previously described [
23,
24]. In brief, slides were dried for 30 min followed by fixation in 10% formalin solution at RT. After washing in PBS, sections were incubated in ice-cold ethanol-acetic acid solution (3:1), washed in PBS and incubated with 3% Triton X-100 solution for 60 min at RT for permeabilization. Slides were then incubated with the TdT-enzyme in reaction buffer containing fluorescein-dUTP for 90 min at 37°C. Negative control was performed using only the reaction buffer without TdT enzyme. Positive controls were performed by digesting with 500 U/ml DNase grade I solution (Roche). To preserve cells for comparison, slices were covered with Vectashield
® mounting medium containing 4',6'-diamino-2-phenylindole (DAPI; Vector). All samples were evaluated immediately after staining using an "Axioskop 40" fluorescence microscope (Zeiss, Germany) at 460 nm for DAPI and 520 nm for TUNEL fluorescence. Data were analyzed by Alpha digi doc 1201 software (Alpha Innotech, San Leandro, CA).
Real-time RT-PCR
Changes in the mRNA expression profiles of pro-apoptotic (Fas, FasL, Bax) and anti-apoptotic (Bcl-2) mediators were determined by semi-quantitative two-step real-time RT-PCR using commercially available, murine-specific primers (Qiagen, Hilden, Germany). The amplicon lengths were 136 bp for GAPDH (Qiagen No. 241012), 96 bp for Fas (Qiagen No. 241122), 109 bp for FasL (Qiagen No. 241194), 146 bp for Bax (Qiagen No. 241116), and 118 bp for Bcl-2 (Qiagen No. 241118). The detailed technique used in our laboratory was previously described [
24,
25]. In brief, the left brain hemispheres were homogenized as individual hemispheres in Qiazol
® buffer (Qiagen). Total RNA was isolated and further purified using RNeasy
® Mini-kits (Qiagen) and RNA concentrations were measured using a spectrophotometer (Bio-Rad, Munich, Germany). From each brain hemisphere, 2 μg RNA were reversely transcribed using random nonamer and oligo-dT16mer primers (Operon Biotechnologies, Cologne, Germany) with Omniscript
® kits (Qiagen), according to the manufacturer's instructions. Real-time RT-PCR was performed using validated commercially available and custom designed primer-probe
® sets (Qiagen) and optimized protocols on the Opticon
® real-time PCR Detection System (Bio-Rad). For quantification of gene expression levels, GAPDH amplicons were generated and used as a "house-keeping" internal control gene. Relative gene expression levels were calculated in relation to the corresponding GAPDH gene expression levels.
Western blotting
The protein levels of pro- and anti-apoptotic mediators were determined in homogenized mouse brains by Western blot analysis, as previously described [
23,
25]. Briefly, mouse brains were surgically removed under anesthesia, separated into left and right hemispheres, and immediately homogenized in lysis buffer (Sigma) containing 100 mM TRIS-HCl (pH 7.5), 150 mM NaCl, 0.5% sodium dodecyl sulfate (SDS), 0.5% Nonidet P-40, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 5 μg/ml pepstatin, 1 mM phenyl-methyl-sulfonyl fluoride in deionized water, using an Ultra Turrax Homogenizer
® (IKA Werke, Staufen, Germany). After 15 min centrifugation at 13,000 × g, the protein content of the supernatants was determined by commercially available colorimetric protein assay ("BCA Protein Assay", Pierce/Perbio Science, Bonn, Germany). A 60 μg sample of total protein was denatured in loading buffer and separated under reducing conditions on 10% (for Fas, FasL) or 12% (for Bax, Bcl-2) SDS-polyacrylamide gels in parallel with a broad range prestained SDS-PAGE protein standard (Bio-Rad, Munich, Germany). Proteins were then transferred to Protean BA 83 nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany) by electroblotting (Bio-Rad). The blots were blocked overnight and then incubated with either polyclonal rabbit anti-mouse Fas (1:200), rabbit anti-mouse FasL (1:200), rabbit anti-mouse Bax (1:300), or monoclonal anti-mouse Bcl-2 (1:500) antibodies (Santa Cruz Biotechnology, Heidelberg, Germany), and with a monocloncal anti-β-actin antibody (clone AC-15, Sigma) diluted 1:10,000, as internal control for ascertaining equal loading of the bands. After incubation with HRP-labelled secondary antibodies (Dako, Hamburg, Germany, and Santa Cruz Biotechnology, Heidelberg, Germany), diluted 1:5,000, antibody binding was visualized by a non-radioactive chemiluminescence technique using a commercially available ECL
® Western blotting kit (Amersham Pharmacia Biotech, Freiburg, Germany). Equal transfer of proteins to the blotting membrane was confirmed by ponceau red staining (Sigma).
Statistical analysis
Statistical analysis was performed using commercially available software (SPSS 9.0 for Windows™). Differences in intracerebral gene expression levels and NSE serum levels between the groups were determined by the unpaired Student's t-test. The repeated measures analysis of variance (ANOVA) was used for assessing differences in neurological scores (NSS). A P-value < 0.05 was considered statistically significant.
Discussion
The present study was designed to evaluate the role of the complement regulatory molecule CD59a in the posttraumatic neuropathology after experimental closed head injury in mice. We sought to test the hypothesis that mice deficient in the
Cd59a gene (CD59a
-/-) would display a significantly increased susceptibility to complement-mediated secondary brain injury with regard to the extent of posttraumatic neurological impairment and neuronal cell death. CD59a
-/- mice had a significantly impaired neurological outcome 7 days after experimental closed head injury, compared to wild-type controls. This neurological deterioration occurred despite equal titration of initial severity of injury, as determined by a similar baseline NSS in wild-type and CD59a
-/- mice 1 h after trauma (5.7 vs. 6.1 points, mean NSS at 1 h; Figure
1). Strikingly, the titrated weight drop falling height required to induce a similar severity of injury was significantly less in the knockout mice (1.7 cm) than in wild-type mice (2.6 cm), implying an intrinsically increased baseline susceptibility to brain injury in CD59a
-/- mice. Aside from the clinical deterioration within 7 days, CD59a
-/- mice also showed a significant exacerbation of neuronal cell loss in the injured brain, as determined by (a) increased serum levels of NSE – an established marker of neuronal cell death – compared to wild-type mice (Figure
2), and (b) increased neuronal cell death and brain tissue destruction in TUNEL histochemistry experiments (Figure
4).
Since programmed cell death represents an important mechanism of secondary neuronal cell death after head injury, and since complement and the MAC have been associated with neuronal apoptosis [
10,
26,
27], we sought to further investigate the expression of apoptotic mediators in injured brains of wild-type and CD59a
-/- mice. Interestingly, there was no significant difference in intracerebral expression of pro- (Fas, FasL, Bax) or anti-apoptotic mediators (Bcl-2) at the gene and protein level between CD59a
-/- and wild-type mice, at any time-point investigated after trauma (Figures
5 and
6). These findings suggest that the exacerbated induction of neuronal cell death in injured brains of CD59a
-/- mice may reflect direct MAC-mediated cellular lysis in large part, as opposed to indirect cell death mechanisms by differential regulation of apoptotic mediators. However, this notion remains speculative and requires further investigation of the detailed, molecular mechanisms of posttraumatic neuronal cell loss in head-injured CD59a
-/- mice.
The complement system has been implicated for more than a decade in the pathophysiology of traumatic brain injury by contributing to neuroinflammation and secondary neuronal cell death [
2]. However, the exact cellular and molecular mechanisms of complement-mediated neuropathology after head injury remain far from being fully understood. Recent studies have determined that all three activation pathways (i.e. classical, alternative, and lectin) are involved in the pathophysiology of posttraumatic complement activation in the injured brain [
23,
24,
28,
29]. Interestingly, the terminal complement pathway which leads to generation of the MAC (C5b-9), also termed the "killer molecule of complement" [
30], has received less attention in recent research in the field of traumatic brain injury.
We have previously shown that soluble MAC levels are significantly increased in the intrathecal compartment of patients with severe head injuries, and were associated with the extent of posttraumatic blood-brain barrier (BBB) dysfunction [
3]. Other groups have shown that the MAC is a potent inducer of intracerebral neuropathology and neuroinflammation, by mediating the upregulation of adhesion molecules and leukocyte infiltration in the subarachnoid space and cerebral parenchyma within a few hours of intracerebroventricular MAC injection [
31]. In addition, MAC injection into hippocampus was shown to evoke seizures and neurocytotoxicity [
32]. A different study used an
in vitro model of BBB damage and revealed that the co-incubation of normal human cerebrospinal fluid with normal serum from healthy donors resulted in complement activation and soluble MAC (sC5b-9) formation [
33]. Mead
et al. described a crucial role of the MAC in contributing to demyelination and axonal injury in studies of EAE in C6-deficient mice [
9]. In a model of experimental weight-drop head injury in rats, complement C9 deposition was demonstrated around the cerebral contusion site, implying MAC deposition in injured brain tissue [
34].
The complement regulatory molecule CD59 is a GPI-anchored molecule which controls MAC assembly in cellular membranes and thus protects from homologous cell lysis after complement activation [
5]. CD59a has been described as the primary regulator of MAC assembly in the mouse, since mice have a testis-restricted CD59b expression [
13]. We have recently described a significant induction of
Cd59a gene upregulation in the injured mouse brains, using the same experimental closed head injury model as in the present study [
25]. It appears that the upregulation of CD59a in injured brains represents a feedback mechanism aimed at protecting neurons from accidental homologous cell lysis related to posttraumatic complement activation [
2]. However, in the complex setting of head injury-induced neuroinflammation, the shedding of the GPI-anchored glycoprotein CD59 from cell surfaces by activation of phospholipases, such as PI-PLC, may render neurons vulnerable to complement mediated attack, independent of upregulation of the
Cd59 gene, as briefly described in the introduction [
2,
5,
11,
25].
CD59a
-/- mice which lack the
Cd59a gene [
16] provide an excellent
in vivo model to test the relevance of the terminal complement pathway in contributing to secondary neuropathology after traumatic brain injury. Recent experimental studies revealed that CD59a
-/- mice are more susceptible to ischemia/reperfusion injuries [
35] and to disease severity, myelin loss and axonal damage after EAE [
15]. The latter study showed that areas of myelin loss and axonal damage in spinal cords of CD59a
-/- mice were associated with MAC deposition, implicating the complement MAC as a crucial mediator of neuropathology of this autoimmune disease [
15]. The findings of exacerbated neurological impairment and increased neuronal cell loss in our current study on experimental closed head injury concur with the previous findings in the EAE model [
15] and further underline the importance of the complement MAC in the pathophysiology of inflammatory central nervous system (CNS) disorders.
Traumatic brain injury is currently still lacking a specific pharmacological therapy designed to avoid induction of secondary brain injuries and delayed neuronal cell death [
36]. However, in the field of therapeutic complement inhibitor development, there have been significant advances in recent years [
37‐
40]. While some of these inhibitors have been successfully tested in experimental head injury models [
24,
25,
41], the "bench-to-bedside" extrapolation to clinical applications in head injured patients has yet to be accomplished [
36]. In this regard, the identification of human CD59 complement binding interfaces, as well as the recent development of human soluble mutant CD59-based compounds, which have been shown to exert an up to 3-fold increased complement inhibitory activity [
42], may represent a promising future strategy for attenuating the terminal complement pathway-mediated neuropathology and the extent of secondary brain damage in head-injured patients.
Competing interests
PFS is the co-inventor of a US patent filed (No. 11,441,828) entitled: "Inhibition of the alternative complement pathway for treatment of traumatic brain injury, spinal cord injury, and related conditions." The authors declare no other potential conflict of interest related to this project.
Authors' contributions
PFS and BPM designed the study. BPM provided the CD59a-/- mice and wild-type littermates. PFS and IL performed and supervised the animal experiments. MAF and BMT performed the ELISA experiments. IP, CS, CC, and IL performed the PCR and Western blot experiments. BPM, WRS, KB, OIS, and WE contributed to analysis and interpretation of the data and corrections of the manuscript. PFS and MAF wrote the first draft of the paper. All authors read and approved the final version of this manuscript.