The neuroprotective effects of z-DEVD.fmk, a caspase-3 inhibitor, on traumatic spinal cord injury in rats
Xeref Barut, MDa, Yusuf Atilla U¨ nlq, MDa, Alper Karaog˘lan, MDa, Matem Tunc¸demir, MScb, Fatma Kaya Dag˘istanli, MScb, Melek O¨ ztqrk, PhDb,
Ahmet C¸ olak, MDa,T
aNeurosurgery Clinic, Taksim Education and Research Hospital, Istanbul, Turkey 34144
bDepartment of Medical Biology, CerrahpaYa School of Medicine, Istanbul University, Istanbul, Turkey 34144
Received 31 March 2004; accepted 21 March 2005

Abstract Background: Apoptosis is one of the most important forms of cell death seen in a variety of physiological and pathological conditions, including traumatic injuries. This type of cell death occurs via mediators known as caspases. Previous studies have investigated the roles that apoptosis and different caspases play in the pathogenesis of secondary damage after spinal cord injury (SCI). The
aim of this research was to assess the neuroprotective effect of z-DEVD.fmk, a caspase-3 inhibitor, in a rat model of SCI.
Methods: Forty-five Wistar albino rats were studied in 3 groups of 15 animals: sham-operated control animals (group 1); trauma-only control animals (group 2); and rats subjected to trauma +
z-DEVD.fmk treatment (group 3). Spinal cord injury was produced at the thoracic level using the weight-drop technique. Responses to injury and the efficacy of z-DEVD.fmk were assessed by light microscopy and terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling staining in cord tissues collected at 4 and 24 hours posttrauma. Five rats from each group were used to assess functional recovery at 7 days after SCI. The functional evaluations were done using the inclined-plane technique and a modified Tarlov motor grading scale.
Results: At 4 hours postinjury, the mean apoptotic index in groups 1, 2, and 3 was 0, 33.01 F 6.62, and 16.40 F 4.91, respectively. The group 3 count was significantly lower than the group 2 count
( P b .01). At 24 hours postinjury, light microscopic examination of group 2 tissues showed widespread hemorrhage, necrosis, polymorphonuclear leukocyte infiltration, and vascular thrombi. The group 3 tissues showed similar features. The prominent findings in group 2 were hemorrhage and necrosis, whereas the prominent findings in group 3 were focal hemorrhage and leukocyte infiltration. The mean inclined-plane angles in groups 1, 2, and 3 were 64.58 F 1.08, 41.58 F 1.38,
and 478 F 2.08, respectively. Motor scale results in all groups showed a similar trend.
Conclusion: Local application of z-DEVD.fmk after SCI in rats reduces secondary tissue injury and
helps preserve motor function. These effects can be explained by inhibition of apoptotic death in all cell types in the spinal cord.
D 2005 Elsevier Inc. All rights reserved.
Keywords: Apoptotic cell death; Caspase 3; Programmed cell death; Spinal cord injury; z-DEVD.fmk

Abbreviations: AC, apoptotic cell number; AI, apoptotic index; CNS, central nervous system; DMSO, dimethylsulfoxide; IC, intact cell number; SCI, spinal cord injury; TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling.
T Corresponding author. Kartaltepe Mahallesi, Terakki Caddesi, No. 47/7, BakVrkfy, Istanbul, Turkey. Tel.: +90 212 543 55 30; fax: +90 212 252 63 00.
E-mail address: [email protected] (A. C¸ olak).

0090-3019/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.surneu.2005.03.042

1.Introduction

It is well known that cell death occurs via 2 morpholog- ically distinct pathways: necrosis and apoptosis [6,17-22,26]. The latter is considered as physiological or programmed cell death and may be induced by external or internal stimuli. Initially, it was thought that apoptosis in the CNS occurred only during development, specifically in size matching of cell populations and in the formation of normal synaptic connections [30]. Inappropriate apoptosis has been impli- cated in many CNS diseases, including stroke, degenerative conditions, and traumatic injury [8-10,21,25,27,35]. Recent work have shown that apoptosis plays a significant role in the pathological loss of CNS cells such as neurons, astrocytes, and oligodendrocytes [4,20,21,39].
Liu et al [20] investigated neuronal and glial apoptosis after SCI in a rat model. They observed marked neuronal and glial apoptosis in the gray and white matter zones of the cord at the lesion site within the first 24 hours after injury, and noted oligodendrocyte apoptosis in distant white matter after several days. Lou et al [21] demonstrated apoptosis at the damaged site 4 hours after SCI in a rat model. Emery et al [7] studied the spinal cords of 15 patients who died between 3 hours and 2 months after SCI and reported apoptotic cell death in these tissues as well.
Apoptosis is mediated through activation of members of the caspase family, the human homologues of enzymes that were first discovered in Caenorhabditis elegans [30]. To date, more than 15 different caspases have been identified [5,24,37]. Studies have shown that these aspartate-specific cysteine proteases play a central role in the apoptosis induced by various insults to CNS tissue [21,24,31]. Caspases are classified in 3 groups according to their functional properties: cytokine activators, initiators of apoptosis, and effectors of apoptosis. Previously, the function of these enzymes has been investigated in experimental and clinical studies focused on the pathogen- esis of secondary damage after SCI [3,4,7,13-15,17, 19 -22,36,39 - 41]. Experiments have also revealed that caspase activity in the spinal cord can be impaired in time- dependent and region-specific ways [13,15,19,20,32]. Cas- pase activation and activity are critical for appropriate and timely execution of apoptosis [5,30]. Specifically, these substances participate in the transmission of proteolytic signals that trigger apoptotic cell death [5,24]. Caspase 3 is known to be an important effector of apoptosis, and previous work have implicated activation of this enzyme in the cell damage that occurs after cerebral ischemia and SCI [7,19,23,32,38]. A study by Casha et al [3] on SCI in rats revealed that caspase 3 is activated between days 1 and 7 postinjury and that tissue levels of this enzyme return to normal within 14 days. In other work on rats, Springer et al
[32] demonstrated activation of caspase 3 and caspase 9 after SCI, implicating both these enzymes in apoptosis. Emery et al [7] documented apoptosis with colocalization of caspase 3 in damaged human spinal cord tissues.

Based on the known functions of caspases, it is assumed that caspase inhibitors block apoptotic cell death and inhibit cytokine production [6,18,19,21,29,36]. Liu et al [20] showed that intraperitoneal injection of cycloheximide, a well-known antiapoptotic agent, can improve outcome after SCI in rats. The molecule z-DEVD.fmk is a selective, irreversible caspase-3 inhibitor that has anti-inflammatory properties and protects against ischemia [18]. Previous investigators have assessed the effects of z-DEVD.fmk in the settings of cerebral ischemia and traumatic brain injury [12,18,38]. The present study is the first evaluation of z-DEVD.fmk in a rat model of traumatic cord damage. Our aim was to investigate the neuroprotective effects of this caspase-3 inhibitor in relation to secondary damage after experimental SCI.

2.Material and methods
Forty-five adult male Wistar albino rats that weighed 250 to 350 g were divided into 3 groups. The groups were subjected to different experimental conditions (detailed below), and spinal cord tissues and functional recovery were compared at different stages postinjury. The animals were fed a normal diet throughout the study period. All 45 rats underwent multilevel laminectomy. Before the operation, each animal was anesthetized with an intramuscular injection of 9 mg/kg of xylazine (Rompun, Bayer, Istanbul, Turkey) and 50 mg/kg of ketamine (Ketalar, Parke-Davis, EczacVbaYV,
Istanbul, Turkey). The posterior region of each animal was
shaved and cleansed with povidone-iodine solution. Body temperature was monitored with a rectal thermometer and maintained at 378C throughout the procedure using a heating lamp and a heating pad. Arterial pressure and heart rate were also continuously monitored. To expose the cord, 3-level
laminectomies (T6-8) were performed with the aid of a surgical microscope under ×10 magnification. Care was taken to avoid damaging the dura mater.
2.1.Traumatic injury model
The SCI model used in this study was the previously described weight-drop technique [1]. The force applied in the trauma was 40 g-cm. Briefly, a 5-mm-diameter cylindrical glass tube was positioned at a 908 angle on the surface of the exposed dura mater, and a 4-g cylindrical
constant weight was dropped from a 10-cm height through the tube onto the spinal cord.
2.2.Experimental protocol
The animals were randomly divided into 3 groups of 15 animals.
2.2.1.Group 1: Sham-operated control animals
In each animal, a skin incision was made, the para- vertebral muscles were dissected, and the laminae were exposed. Laminectomies were carried out at T6-8 as noted above, but the cord was not damaged or manipulated in any

way. Once this was complete, the muscles and the skin were closed with no. 3.0 silk sutures. At 4 hours after the operation, 5 animals were given lethal intraperitoneal injections of pentobarbital. Five other rats were killed the same way at 24 hours postsurgery. In each of these 10 rats, the area where the laminectomies had been performed was exposed and approximately 2 cm of spinal cord was removed under the microscope. The dura was dissected from the cord. The cord tissues were fixed in 10% neutral- buffered formalin solution and then prepared for light microscopic examination and TUNEL assay (detailed below). The remaining 5 animals in this group were used for functional recovery experiments. These rats were neurologically examined at 7 days postinjury using the inclined-plane technique and a modified version of the Tarlov motor grading scale [28,33].
2.2.2.Group 2: Trauma-only control animals
Laminectomies were performed and SCI was induced as described above. In each rat, 5-lL DMSO was applied locally at the trauma site. Five animals were killed at each of the time points noted above (4 and 24 hours postinjury) and spinal cord samples were collected and processed as described for group 1. The remaining 5 rats were used to assess functional recovery at 7 days postinjury.
2.2.3.Group 3: Trauma + treatment animals
Each rat in this group underwent the same laminectomy and trauma procedures detailed for group 2. Dry-form z-DEVD.fmk (Sigma, Kimeks, Istanbul, Turkey) was dis- solved in DMSO. Immediately after weight-drop trauma, 320 ng of z-DEVD.fmk in 5 lL DMSO was applied locally at
the trauma site using an automatic micropipette. As in the
other groups, 5 rats were killed at 4 and 24 hours posttrauma and spinal cord specimens were removed and prepared as detailed above. The other 5 animals were used for functional recovery evaluation.
2.3.Morphological methods
2.3.1.Light microscopy
The cord specimens obtained at 4 and 24 hours postinjury were prepared for histological study. Each approximately 2-cm cord segment centered at the injury site was fixed in 10% neutral-buffered formalin solution and then embedded in paraffin. Five-micrometer-thick coronal sections of dam- aged cord were cut and stained with hematoxylin-eosin and Kresylecht Violett. Slides were examined under the light
microscope at ×10 and ×40 magnification.
2.3.2.The TUNEL method and quantification of apoptotic cells
Apoptosis was assessed using an in situ nick-end labeling (TUNEL) staining technique [11]. Five-microm- eter-thick coronal sections were cut from the abovemen- tioned paraffin blocks of cord tissues collected at 4 and 24 hours postinjury, and TUNEL staining was performed using an apoptosis detection kit (Intergen, Inc, Purchase,

NY). Deparaffinized tissue sections were incubated with Proteinase K (20 lg/mL). Tissue sections subjected to 3% H2O2 for endogenous peroxide inhibition were incubated with 1× equilibration buffer at room temperature for
30 minutes. The digoxigenin-labeled deoxynucleoside
triphosphate tail was incubated with terminal deoxynu- cleotidyl transferase for 1 hour at 378C and the slides were washed in stop/wash buffer for 10 minutes at room temperature. Tissue sections incubated with anti– digoxigenin-peroxidase antibody at room temperature for 30 minutes were stained with diaminobenzidine and color reaction was evaluated in the microscope. Cells with clear brown nuclear labeling were defined as TUNEL positive.
For staining specificity controls, dexametazon (5 mg/kg) was applied to rat thymus tissue sections, which were used as positive controls. As for negative controls, distilled water instead of terminal deoxynucleotidyl transferase enzyme was used.
Morphometric analysis of the positive cells in tissue stained by TUNEL method was performed under high-power magnification (×400) in a blinded fashion. On each slide, 15 fields were randomly selected. To quantitate the extent of
apoptosis, we recorded numbers of apoptotic cells (TUNEL- positive cells) in cord sections from the 3 groups. The researcher totaled all the TUNEL-positive cells and intact cells in those fields and then calculated the AI a mean count

Fig. 1. Photomicrographs of spinal cord tissues from sham-operated control animals (group 1). A: A view of gray and white matter zones (hematoxylin- eosin, original magnification ×20). B: Cord tissue stained with TUNEL
shows no TUNEL-positive cells (TUNEL, original magnification ×40).

per slide. Apoptotic index was calculated according to the formula AI = (AC/AC + IC) × 100, where AC is the apoptotic cell number and IC is the intact cell number. Finally, an overall mean count for each set of specimens (4 and 24 hours) in each group was calculated and the group mean
values at each time point were compared.
2.4.Functional recovery assessment
Functional recovery was assessed with objective and subjective testing (inclined-plane technique and modified Tarlov motor grading scale, respectively) at 7 days postinjury [28,33]. For motor function, animals were

classified using a 5-point scale: grade 5 = able to walk normally; grade 4 = able to walk with mild spasticity or incoordination of the hind limbs; grade 3 = able to stand but unable to walk; grade 2 = minimal voluntary hind limb movements but unable to stand; and grade 1 = no voluntary hind limb movement.
2.5.Statistical analysis
Data are expressed as mean F SD where appropriate. Mann-Whitney U test analysis was used to assess differ- ences among the 3 groups. P values less than .05 were considered to indicate statistical significance.

Fig. 2. Photomicrographs of cord tissues from trauma-only control animals (group 2). A: A gray matter region at 4 hours postinjury shows advanced motor neuron degeneration (arrow) and diffuse hemorrhage (H) (hematoxylin-eosin, original magnification ×20). B: A gray matter region at 4 hours postinjury shows
3 apoptotic cells (arrows). Note the identifying features of condensed chromatin in the nucleus and decreased cell size (hematoxylin-eosin, original magnification ×100). C: Cord tissue from gray matter at 24 hours postinjury shows necrotic cells (arrowheads) and normal cells (arrow) (Kresylecht Violett, original magnification ×100). D: Numerous apoptotic cells (arrows) in spinal cord tissue from gray matter at 24 hours (TUNEL, original magnification ×40). E: Two shrunken neurons (arrows) within the lesion area that were labeled by TUNEL at 24 hours. (TUNEL, original magnification ×40).

3.Results
3.1.Light microscopy findings
As expected, light microscopic examination of the cord samples from group 1 (sham-operated animals) showed normal findings (Fig. 1A).
In the group 2 (trauma-only animals) tissue samples, we observed diffuse hemorrhage and congestion in the gray matter at 4 hours postinjury (Fig. 2A). At 24 hours, there

Fig. 3. Photomicrographs of cord specimens from the z-DEVD.fmk–treated animals (group 3). A: A gray matter region 4 hours postinjury shows large numbers of healthy cells occupying considerable space and focal areas of
hemorrhage (H) (Kresylecht Violett, original magnification ×20). B: A view
of gray matter at 24 hours postinjury reveals both healthy cells (arrowheads) and necrotic motor neurons (arrows) (Kresylecht Violett, original magni- fication ×40). C: TUNEL-stained spinal cord specimens collected 24 hours postinjury. Labeled apoptotic cells were significantly decreased in
z-DEVD.fmk–treated rats. (TUNEL, original magnification ×100).

Table 1
Inclined-plane technique results and motor scores for the 3 study groups at 7 days postinjury (mean F SD)

Group Inclined-plane angle (8) Motor scores

1 64.5 F 1.0 5 F 0
2 41.5 F 1.3 1 F 0
3 47 F 2.0 2.2 F 0.42

Note the marked effect of local administration of z-DEVD.fmk on functional recovery after traumatic SCI.

were marked hemorrhagic necrosis, vascular thrombosis, and widespread edema in both white and gray matter zones. At this stage, the traumatized segment also showed infiltration with polymorphonuclear leukocytes, erythro- cytes, and macrophages, as well as cystic vacuolar degeneration (Fig. 2C).
In group 3 (trauma + treatment animals), the cord tissues obtained at 4 hours postinjury showed similar features with group 2 specimens (Fig. 3A). In marked contrast to group 2, the tissues collected at 24 hours showed mild edema and infiltration with small numbers of polymorphonuclear leukocytes and macrophages in both white and gray matter zones (Fig. 3B).
3.2.Findings from TUNEL staining
There were no apoptotic cells present in the cord sections of the sham-operated control group (Fig. 1B).
In the group 2 tissues, we observed typical TUNEL- positive cell nuclei (Fig. 2D and E). At 4 hours after injury, apoptotic cells were scattered throughout the gray and white matter zones of the cord. The most intense staining was in motor neurons at the periphery of the damaged site, where injured tissues met surrounding healthy tissues. We also observed apoptosis in glial cells. Labeled apoptotic cells were observed to be significantly decreased in the trauma + treatment group compared with the trauma-only group both on 4th and 24th hour ( P b .01 and P b .05, respectively; Table 2; Fig. 3C).
3.3.Inclined-plane results
The inclined-plane results are summarized in Table 1. In group 2, the inclined-plane angle at 7 days after trauma ranged from 408 to 42.58 (mean, 41.58 F 1.38). In group 3,
the range was from 458 to 508 (mean, 47.08 F 2.08). These
mean values were significantly different ( P = .008).
3.4.Motor function findings
At 7 days after trauma, the respective mean motor scores in groups 1, 2, and 3 were 5.0 F 0.0, 1.0 F 0.0, and 2.2 F
0.42 (Table 1). The difference between groups 2 and 3 was statistically significant ( P = .008).

4.Discussion
The results of this study demonstrate that z-DEVD.fmk reduces apoptosis in different cell types in the rat spinal

cord. Recent experiments have suggested that both global and focal ischemic spinal cord cell loss may, in part, be caused by programmed cell death [19-22]. Research have also shown that apoptosis occurs after traumatic insults to the cord and in the setting of neurodegenerative diseases [16,20,25,26,35]. In animals, traumatic injuries that involve spinal cord contusion produce lesions similar to those seen in human SCI. The weight-drop method used in this study causes this type of damage. Young et al [39] and Li et al
[19] demonstrated apoptotic cell death in 4 cell populations in the spinal cord, namely, neurons, astrocytes, oligoden- drocytes, and microglia. Emery et al [7] observed marked apoptosis in oligodendrocytes in white matter tracts at lesion sites and in descending tracts below the level of injury in the human spinal cord. In our study, we found the greatest extent of apoptosis at the periphery of the lesion, where damaged tissues met healthy tissues. This pattern of cell death radiating from the lesion center might explain how lesions in traumatized spinal cord expand, as other authors have reported [7,14,20].
Previous investigations have shown that there is neuronal necrosis in the dorsal horn of the spinal cord within 6 hours of ischemic injury, that this remains evident at 7 days, and that apoptosis is not apparent at 6 hours but begins later in the first day postinjury [19,20,39]. Liu et al [20] assessed neuronal and glial apoptosis after traumatic SCI in rats. They detected apoptosis in neurons within the lesion area by 4 hours and found that this neuronal death peaked at 8 hours. The same authors also noted apoptosis in glial cells within the lesion area between 4 hours and 14 days and recorded a peak at 24 hours [20]. Lou et al [21] studied apoptosis after acute SCI in rats. These investigators found marked apoptosis in both neuronal and glial cells at 4 hours after injury but observed no discernable apoptotic activity in the cord at 24 hours. Our findings are somewhat similar to those of Liu et al but conflict with those of Lou et al. This may be because of the different severities and types of trauma used. In contrast to the work of Lou et al, our study involved a less severe contusive injury to the cord. In our experiments, we feel that mitochondrial function in the cord tissue may have been relatively preserved to a greater degree, thus allowing neuronal cell death by apoptosis.
In a research on a mouse model of cord compression, Li et al [19] showed that apoptosis was distributed over several spinal cord segments between 4 and 9 days postinjury. They found that most apoptotic cells were in the longitudinal tracts of the white matter and considered these to be oligodendrocytes. Liu et al [20] reported detection of a second wave of apoptotic cells (which they also identified as oligodendrocytes) throughout the white matter at 7 days and

population, especially oligodendrocytes, and less damage to axonal myelin. In our opinion, all these findings support the proposals of Li et al [17] and Bresnahan et al [2] that traumatic insults to the spinal cord result in delayed oligodendrocyte apoptosis.
The TUNEL technique involves in situ labeling of degraded internucleosomal DNA, which is the hallmark of apoptotic cell death [11]. We used this method because it is a highly sensitive and specific means of identifying DNA fragmentation [11]. As noted, we assessed brown staining of cell nuclei in the cord specimens as confirmation of apoptotic cell death. One potential confounding issue with this technique is that internucleosomal DNA fragmentation may also occur in necrotic cell death, which means that necrotic neurons can be TUNEL positive. We found that the mean AIs (4 and 24 hours) in the z-DEVD.fmk–treated group were both significantly lower than the corresponding findings in the trauma-only group ( P b .01 and P b .05; Table 2; Fig. 3C). In line with this, the animals treated with z-DEVD.fmk had better inclined-plane results and motor scores than those in the trauma-only group (Table 1). Inhibition of caspase-3 activity by z-DEVD.fmk resulted in less apoptosis, and this lower-level tissue damage allowed for better functional recovery. This agent also had an inhibitory effect on cytokine production; therefore, less edema and infiltration with a small number of polymorpho- nuclear leukocytes were observed in treated animals than in untreated ones (Fig. 3C).
Caspases are synthesized as inactive proenzymes that require proteolytic activation and participate in a proteolytic signaling cascade [5,24]. Activation of one caspase can lead to cleavage and activation of additional molecules of the same protease or other proteases, leading to an amplified protease cascade [34]. As supported by our findings, inhibition of caspase activity by specific inhibitors can prevent apoptosis and help maintain cell viability. Yakovlev et al [38] investigated the role of caspase 3 in traumatic brain injury in a rat model and found that administration of a caspase-3 inhibitor reduced the rise in caspase-3 activity that usually occurs under these conditions. Hara et al [12] studied the activity of caspase 3 in mice subjected to focal cerebral ischemia and demonstrated that a caspase-3 inhibitor reduced infarct volume and neurological deficits. As noted above, work by Casha et al [3] revealed that

Table 2
Mean AIs for trauma-only and trauma + treatment groups at 4 and 24 hours Group 4-h AI 24-h AI
Trauma only (n = 5) 33.01 F 6.62 [32.5] 34.69 F 10.28 [37.8]

noted that this was associated with degenerative axons. Casha et al [3] suggested that wallerian degeneration of

Trauma + treatment
(n = 5)

16.40 F 4.91T [15.2] 17.57 F 3.24TT [16.3]

axons might initiate secondary apoptosis of some oligoden- drocytes. In our study, the rats treated with z-DEVD.fmk showed better functional recovery (Table 1). We think that this finding may be related with preservation of the glial cell

At both time points, there were statistically significant differences between
the counts in the trauma-only and the z-DEVD.fmk–treated rats. Values are expressed as mean F SD [median].
T P b .01.
TT P b .05.

caspase 3 is activated between days 1 and 7 after SCI in rats and that levels normalize by 14 days. To the best of our knowledge, the present study is the first to have researched the neuroprotective effects of the caspase-3 inhibitor z-DEVD.fmk in rat SCI. The z-DEVD.fmk is one of the potent inhibitors of caspase 3 and can be applied locally. The findings suggest that local administration of a caspase-3 inhibitor after acute surgical decompression and stabiliza- tion might be a valuable component of SCI treatment in human beings. The localized nature of SCI makes these lesions particularly well suited for local application of a therapeutic agent, as demonstrated in our study.

5.Conclusion
Use of the caspase-3 inhibitor z-DEVD.fmk as an antiapoptotic agent in this rat model of SCI helped limit secondary damage and provided significant neuroprotection. The latter was reflected in less extensive TUNEL staining and superior functional recovery postinjury. Our results indicate that agents that block apoptotic pathways may be of value for treating SCI in the future.

References
[1]Allen AR. Remarks on the histopathological changes in the spinal cord due to impact. An experimental study. J Nerv Ment Dis 1914;41:141 – 7.
[2]Bresnahan JC, Schuman SL, Beattie MS. Evidence for apoptosis of oligodendroglia in long tracts undergoing wallerian degeneration after spinal cord injury (SCI) in monkeys. Soc Neurosci 1996;22:1185 [abstr].
[3]Casha S, Yu WR, Fehlings MG. Oligodendroglial apoptosis occurs along degenerating axons and is associated with fas and p75 expression following spinal cord injury in the rat. Neuroscience 2003;103:203 – 18.
[4]Crove MJ, Bresnahan JC, Shuman SL, et al. Apoptosis and delayed degeneration after spinal cord injury in rats and monkeys. Nat Med 1997;3:73 – 6.
[5]Deveraux QL, Stennicke HR, Salvesen GS, et al. Endogenous inhibitors of caspases. J Clin Immunol 1999;19:388 – 98.
[6]Ekert PG, Silke J, Vaux DL. Caspase inhibitors. Cell Death Differ 1999;6:1081 – 6.
[7]Emery E, Aldana P, Bunge MB, et al. Apoptosis after traumatic human spinal cord injury. J Neurosurg 1998;89:911 – 20.
[8]Fink KB, Andrews LJ, Butler WE, et al. Reduction of post-traumatic brain injury and free radical production by inhibition of the caspase-1 cascade. Neuroscience 1999;94:1213 – 8.
[9]Friedlander RM, Brown RH, Gagliardini V, et al. Inhibition of ICE slows ALS in mice. Nature 1997;388:31.
[10]Friedlander RM, Yuan J. ICE, neuronal apoptosis and neurodegenera- tion. Cell Death Differ 1998;5:823 – 31.
[11]Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 1992;119:493 – 501.
[12]Hara H, Friedlander RM, Gagliardini V. Inhibition of interleukin 1b; converting enzyme family proteases reduced ischemic and excitotoxic neuronal damage. Proc Natl Acad Sci U S A 1997;94:2007 – 12.
[13]Katoh K, Ikata T, Katof S, et al. Induction and its spread of apoptosis in rat spinal cord after mechanical trauma. Neurosci Lett 1996;216:9 – 12.
[14]Kato H, Kanellopoulos GK, Matsuo S, et al. Neuronal apoptosis and necrosis following spinal cord ischemia in the rat. Exp Neurol 1997;148:464 – 74.

[15]Keane RW, Kraydich S, Latocki G, et al. Apoptotic and anti-apoptotic mechanisms following spinal cord injury. J Neuropathol Exp Neurol 2001;60:422 – 9.
[16]Kerr JFR, Wyllie AH, Currie AR. Apoptosis. A basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 1972;26:239 – 57.
[17]Li GL, Brodin G, Farooque M, et al. Apoptosis and expression of Bcl-2 after compression trauma to the rat spinal cord. J Neuropathol Exp Neurol 1996;55:280 – 9.
[18]Li H, Colbourne F, Sun P, et al. Caspase inhibitors reduce neuronal injury after focal but not global cerebral ischemia in rats. Stroke 2000;31:176 – 82.
[19]Li M, Ona VO, Chen M, et al. Functional role and therapeutic implications of neuronal caspase-1 and 3 in a mouse model of traumatic spinal cord injury. Neuroscience 2000;99:333 – 42.
[20]Liu XZ, Xu XM, Hu R, et al. Neuronal and glial apoptosis after spinal cord injury. J Neurosci 1997;17:5395 – 406.
[21]Lou J, Lenke LG, Ludwig FJ, et al. Apoptosis as a mechanism of neuronal cell death following acute experimental spinal cord injury. Spinal Cord 1998;36:683 – 90.
[22]Lu J, Aswell KWS, Waite P. Advances in secondary spinal cord injury: role of apoptosis. Spine 2000;25:1859 – 66.
[23]Namura S, Zhu J, Fink K, et al. Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia. J Neurosci 1998;18:3659 – 68.
[24]Nicholson DW, Thornberry NA. Caspases: killer proteases. Trends Biochem Sci 1997;22:299 – 306.
[25]Portera-Cailliau C, Hedran JC, Price DL, et al. Evidence for apoptotic cell death in Huntington disease and excitotoxic animal models. J Neurosci 1987;15:3775 – 87.
[26]Raff M. Cell suicide for beginners. Nature 1998;396:119 – 22.
[27]Rink A, Fung KM, Trajanowski JQ, et al. Evidence of apoptotic cell death after experimental traumatic brain injury in the rat. Am J Pathol 1995;147:1575 – 83.
[28]Rivlin AS, Tator CH. Objective clinical assessment of motor function after experimental spinal cord injury in the rat. J Neurosurg 1977; 47:577 – 81.
[29]Rudel T. Caspase inhibitors in prevention of apoptosis. Herz 1999;24:236 – 41.
[30]Samali A, Zhivotovsky B, Jones D, et al. Apoptosis: cell death defined by caspase activation. Cell Death Differ 1999;6:495 – 6.
[31]Shulz JB, Weller M, Moscowitz MA. Caspases as treatment targets in stroke and neurodegenerative diseases. Ann Neurol 1999;45: 421 – 9.
[32]Springer JE, Azbill RD, Knapp PE. Activation of the caspase-3 apoptotic cascade in traumatic spinal cord injury. Nat Med 1999; 5:943 – 6.
[33]Tarlov IM. Spinal cord compression: mechanism of paralysis and treatment. Spingfield (Ill)7 Charles C Thomas; 1957.
[34]Tewari M, Quan LT, Q’Rowke K, et al. Yama/CPP 32 beta, a mammalian homologue of CED-3, is a CrmA-inhibitable protease that cleaves the death substrate poly (ADP-ribose) polymerase. Cell 1995;81:801 – 9.
[35]Thompson CB. Apoptosis in the pathogenesis and treatment of disease. Science 1995;267:1456 – 62.
[36]Wada S, Yone K, Ishidou Y, et al. Apoptosis following spinal injury in rats and preventative effect of N-methyl-d-aspartate receptor antago- nist. J Neurosurg (Spine) 1999;91:98 – 104.
[37]Van de Craen M, Van Loo G, Pype S, et al. Identification of a new caspase homologue: caspase-14. Cell Death Differ 1998;5: 838 – 46.
[38]Yakovlev AG, Knoblach SM, Fan L, et al. Activation of CPP32-like caspases contributes to neuronal apoptosis and neuronal dysfunction after traumatic brain injury. J Neurosci 1997;17:7415 – 24.
[39]Young C, Arnold PM, Zoubine M, et al. Apoptosis in cellular compartments of rat spinal cord after severe contusion injury. J Neurotrauma 1998;15:459 – 72.

[40]Zhang Z, Krebs CJ, Guth L. Experimental analysis of progressive necrosis after spinal cord trauma in the rat: etiological role of the inflammatory response. Exp Neurol 1997;143:141 – 52.
[41]Zurita M, Vaquero J, Zurita I. Presence and significance of CD-95 (Fas/Apo1) expression after spinal cord injury. J Neurosurg (Spine) 2001;94:257 – 64.Z-DEVD-FMK

Commentary

This is an interesting study using adequate controls and a reasonable number of animals to demonstrate a neuro- protective effect of a caspase inhibitor in a rat model of SCI. The simultaneous demonstration of neuropathological pro-

tection and a behavioral benefit is encouraging. The key question is—what comes next? A variety of treatments have succeeded in rats, yet we have no proven and effective neuroprotection for SCI. Even a high dose of methylpred- nisolone has received mixed reviews.
It is possible that only a combination therapy works in human beings or that physiological differences make rodent models relatively limited.

Ben Roitberg, MD Department of Neurosurgery University of Illinois at Chicago Chicago, IL 60612, USA