Cellular localization and enzymatic activity of cathepsin B after spinal

cord injury in the rat

Rebecca C. Ellisb,f, Wilbur A. O’Steenb,f, Ronald L. Hayesb,d,e,f, Harry S. Nickb,f,

Kevin K.W. Wangb,c,e,f, Douglas K. Andersona,b,d,f,TaMalcom Randall VAMC, University of Florida, PO Box 100244, Gainesville, FL 32610, USA

bDepartment of Neuroscience, University of Florida, PO Box 100244, Gainesville, FL 32610, USAcDepartment of Psychiatry, University of Florida, PO Box 100244, Gainesville, FL 32610, USA

dDepartment of Neurosurgery, University of Florida, PO Box 100244, Gainesville, FL 32610, USAeCenter for Traumatic Brain Injury Studies, University of Florida, PO Box 100244, Gainesville, FL 32610, USA

fEvelyn F. and William L. McKnight Brain Institute of the University of Florida, PO Box 100244, Gainesville, FL 32610, USA

Received 10 September 2004; revised 19 November 2004; accepted 30 November 2004

Abstract

Mechanical spinal cord injury (SCI) initiates a cascade of pathochemical and pathophysiological events, collectively known as the

secondary injury. There has been a long-standing interest in understanding the activation and involvement of proteases in this secondary

injury process. Several proteases including the calpains, caspases and matrix metalloproteinases are activated by perturbations to the spinal

cord and have been linked to cell death following SCI and in other models of CNS disease and insult. Cathepsin B (Cath B), a potent

lysosomal protease, has also been implicated in the pathology of CNS diseases including brain tumors, Alzheimer’s disease, amyotrophic

lateral sclerosis and stroke. Previously, we reported significant increases in Cath B mRNA and protein expression following contusion-SCI.

This characterization of Cath B continues with the experiments reported herein, which were designed to examine Cath B enzymatic activity

and cellular localization following contusion-SCI in the rat. Cath B enzymatic activity was significantly increased in the injury epicenter at 5

and 7 days post-injury and was highly correlated with increases in the active forms of the Cath B protein reported earlier. Furthermore, the

immunohistochemical analyses revealed that the post-injury increases in expression and enzymatic activity at the injury epicenter were due to

the presence of a large and diverse population of inflammatory cells. However, in areas adjacent to the injury epicenter, it appears that

parenchymal neurons may also contribute to these increases. Our findings coupled with the documented role of Cath B in other CNS

pathologies make this potent protease an attractive candidate for involvement in the tissue destruction associated with the secondary injury

cascade following SCI.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Spinal cord injury; Protease; Cathepsin B; Rat; CA-074; Enzymatic activity

Introduction

There are approximately 243,000 people currently living

in the United States (1,000,000 worldwide) with spinal cord

injury (SCI) and roughly 11,000 new injuries are sustained

each year. Spinal cord injury is extremely complex and is

comprised of a mechanical primary insult followed by a

delayed secondary injury cascade. The primary injury can

be contusive or concussive in nature and can involve

shearing or stretching of the spinal cord. The secondary

injury cascade is made up of a series of uncontrolled

endogenous biochemical reactions that collectively extend

the tissue damage to segments rostral and caudal of the

original injury site. Ultimately, significant cell death occurs,

leading to loss of motor and sensory function below the

0014-4886/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.expneurol.2004.11.034

T Corresponding author. Department of Neuroscience, University of

Florida, PO Box 100244, Gainesville, FL 32610, USA. Fax: +1 352 846

0250.

E-mail address: anderson@mbi.ufl.edu (D.K. Anderson).

Experimental Neurology 193 (2005) 19–28

www.elsevier.com/locate/yexnr

injury site. At the present time, the prognosis for substantial

functional recovery in the vast majority of SCI individuals is

limited.

Among the many pathophysiological processes that

have been identified following SCI is the activation and

release of lytic enzymes such as proteases (Banik et al.,

1986; Iizuka et al., 1986; Iwasaki et al., 1987; Taoka et al.,

1998; Yashon et al., 1975). Proteases have the potential to

cause significant tissue damage due to the hydrolysis of a

wide variety of intracellular and extracellular substrates.

Since many of these proteases, including the calpains,

caspases, matrix metalloproteinases and cathepsins, are

widely expressed and extremely potent, elaborate cellular

mechanisms have evolved to regulate their expression and

activity. However, following SCI, it is likely that these

regulatory controls are overwhelmed, thereby leading to

the over-expression, over-activation and possibly the

inappropriate intracellular and extracellular release of a

number of proteases from a variety of cell types.

Uncontrolled release of proteases can exacerbate the

ongoing tissue damage initiated by the primary mechanical

injury. Thus, many of these proteases have properties that

make them strong candidates for involvement in the

secondary injury cascade.

To date, the role of cathepsins in SCI pathophysiology

has been understudied and as a result, minimal information

is available concerning the expression and activity of this

important class of proteases in the injured spinal cord.

Microarray analyses have shown increases in the transcript

expression of many cathepsins after insults to the CNS.

Following spinal root avulsion, Hu et al. (2002) reported an

increase in the expression of cathepsins B, C, D, H, I, L and

S. Furthermore, while only cathepsin B (Cath B) was

upregulated following hemisection of the spinal cord (Fan et

al., 2001), the expression of Cath B and cathepsin D were

elevated after peripheral sciatic nerve crush (Fan et al.,

2001). These microarray studies suggest that while different

types of CNS injuries can increase the mRNA expression of

several cathepsins, Cath B appears to be particularly

sensitive to perturbations as it was upregulated in all three

injury models.

We recently demonstrated that both Cath B mRNA and

protein expression were increased following contusion-SCI

(Ellis et al., 2004). Thus, the aims of the present study are

to extend the post-SCI characterization of Cath B by

determining if an increase in enzymatic activity occurs as a

result of the induction of Cath B mRNA and protein, and

then, to identify the source of these increases in Cath B

expression. Our data indicate that following SCI, Cath B

enzymatic activity is significantly increased at the injury

epicenter and that the primary source of Cath B at the

injury site is derived from a diverse population of

inflammatory cells. In areas adjacent to the injury epicenter,

Cath B immunoreactivity appears elevated in neurons and

most Cath B+ inflammatory cells are restricted to the dorsal

columns.

Materials and methods

All experimental procedures were conducted in accor-

dance with the guidelines set forth by the University of

Florida’s Institute Animal Care and Use Committee

(IACUC).

Laminectomy and injury

Adult female Long-Evans rats weighing approximately

230–300 g (Harlan, Indianapolis, IN) were used in this

study. All surgical procedures were conducted under sterile

conditions with supplemental heat. Intraperitoneal admi-

nistration of Nembutal (sodium pentobarbital—50–60 mg/

kg) was used to induce anesthesia. Following a T12

laminectomy (intact dura mater), injury to the spinal cord

was produced with the NYU impactor device (i.e., 10 g

dropped 25 mm onto the exposed dura). The sham-injury

animals received a laminectomy and were placed in the

injury apparatus, but were not injured. The incision was

closed in layers (i.e., muscle then skin). Animals were

recovered in a heated incubator with food and water

available ad libitum. Bladders were manually expressed

and fluids were administered as frequently as required.

Analysis of Cath B enzymatic activity

Spinal cord tissue was collected after extending the

laminectomy to allow three segments of tissue (each

segment is approximately 6 mm in length) to be removed

(i.e., tissue from the injury site and tissue from the regions

immediately rostral and caudal to the injury site). The fresh

tissue was rinsed with cold PBS and flash frozen with liquid

nitrogen. Frozen spinal cords were crushed with a chilled

mortar and pestle. Triton extraction buffer (20 mM Tris

(pH 7.4), 150 mM NaCl, 5 mM EGTA, 1.0% Triton X-100,

0.2 mM DTT) was added to the crushed tissue, which was

then placed on ice for 60 min. Samples were vortexed every

15 min for the next hour. At the end of the second hour,

samples were centrifuged at 15,000 rpm for 10 min at 48 C.The supernatant was removed and placed into a clean

Eppendorf. Glycerol was added to stabilize the lysates.

Protein concentrations of the tissue lysates were determined

by bichinchoninic acid (BCA) assay (Pierce Inc., Rockford,

IL). All samples were equalized to a common protein

concentration for ease of handling.

Cath B enzymatic activity was assessed at 1, 2, 3, 5 and 7

days post-injury in both sham- and contusion-injured rats (n =

3–4 rats/group/time point). Enzymatic activity of Cath B was

also measured in the spinal cord of 5 normal rats. Enzymatic

activity assays were carried out in 96 round-bottom well

plates (Costar, Inc., Coring, NY). Lysates from spinal cord

tissue were incubated with 20 mM DTT, 100 mM MES (pH

5.5), H2O and the Cath B specific substrate Z-Arg-Arg-7-

amido-4-methylcoumarin hydrochloride (200 AM; Sigma-

Aldrich, St. Louis, MO). The level of fluorescence, generated

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–2820

through the hydrolysis of Z-Arg-Arg-7-amido-4-methylcour-

marin hydrochloride by Cath B only, was measured

(excitation at 355 nm, emission at 460 nm) using the Spectra

Gemini XS (Molecular Devices, Sunnydale, CA).

CA-074 was chosen as the inhibitor of Cath B in these

experiments. Previously, CA-074 had been shown to be a

specific Cath B inhibitor in vitro (Murata et al., 1991) and in

vivo (Towatari et al., 1991). Furthermore, Buttle et al., 1992

reported that CA-074 was a rapid inactivator of Cath B

(second order rate constant 112,000 M�1d s�1) with barely

detectable inhibitory action on Cath H (b10), L (20F 2) and

S (b10) and m-calpain (b10).To test the efficacy of CA-074 [N-(l-3-trans-propylcar-

bamoyloxirane-2-carbonyl)-l-isoleucyl-l-proline] (Sigma-

Aldrich, St. Louis, MO) as a specific and irreversible

inhibitor of Cath B in this assay, purified Cath B from the

bovine spleen (Sigma-Aldrich, St. Louis, MO) was incu-

bated with various concentrations of CA-074 (0 AM–

5000 AM). Enzymatic activity levels were then determined

as previously described. Lastly, to confirm that the increases

in fluorescence were due to an increase in Cath B hydrolysis

rather than autofluorescence or substrate hydrolysis by a

related protease, CA-074 (25 AM) was added to the reaction

vessel containing lysates of injured spinal cord tissue. Cath

B enzymatic activity was then determined as previously

described.

Immunohistochemical localization of Cath B

Immunohistochemistry was performed on tissue sections

from normal (n = 2) and 7 days contusion-injured rats (n =

2). Animals were perfused intracardially initially with 0.9%

saline and then 4.0% paraformaldehyde (pH 7.4). The

perfused animal was stored overnight at 48C, after which thecord was removed. After the dura was carefully dissected

away, the spinal cord was cryoprotected in 30% sucrose. A

segment of tissue containing the injury site was sectioned

(14 Am) and mounted onto slides (Fisher Scientific,

Pittsburgh, PA) using the Frigocut 2800 (Reichert-Jung).

Sections were fluorescent immunolabeled with two

primary antibodies (AB) in the following experiments:

polyclonal AB against Cath B (1:1000; Upstate Biotechno-

logy, Inc., New York) and monoclonal AB against (1) glial

acidic fibrillary protein or GFAP for astrocytes (1:1000;

Sternberger Monoclonals, Lutherville, MD), (2) neuron-

specific nuclei marker or NeuN for mature neurons (1:1000;

Chemicon, Temecula, CA), (3) lysosomal-associated mem-

brane protein or LAMP for lysosomal membranes (1:1000,

Stressgen, British Columbia, Canada), (4) galactocerebro-

side or GalC (1000; Sigma, Saint Louis, MO), 2V, 3V-cyclicnucleotide 3V-phosphodiesterase or CNPase (1:1000; Sigma,

St. Louis, MO) and myelin basic protein or MBP (1:1000;

Chemicon, Temecula, CA) for oligodendrocytes, (5) OX-42

for resting/activated microglia and macrophages (1:1000;

Serotec Inc., Raleigh, NC), (6) OX-8 (1:1000; Serotec Inc.,

Raleigh, NC) for cytotoxic CD8+ cells (and a subset of

activated macrophages) and ED-2 (1:1000, Serotec Inc.,

Raleigh, NC) for perivascular macrophages. The nuclear

dye DAPI (in Vectashield, Vector Laboratories, Burlingame,

CA) was used to label the nuclei. The first primary antibody

was incubated at room temperature (RT) with a 10% goat

serum–10% horse serum–0.2% Triton-X 100 in 0.1 M PBS

(block) solution followed by overnight incubation with the

second primary antibody also in block at RT. The tissue

sections were then incubated in fluorescent-tagged secon-

dary antibody (1:1000; Molecular Probes, Eugene, OR) and

cover-slipped. Fluorescent images were viewed and digi-

tally captured with a Zeiss Axioplan 2 microscope equipped

with a SPOT Real Time Slider high-resolution color CCD

digital camera (Diagnostic Instruments, Inc., Sterling

Heights, MI). Confocal images with z-plane sectioning

were collected with the 1024 ES Confocal Microscope laser

scanning system with LaserSharp software (Bio-Rad,

Hercules, CA). This system has true 24-bit confocal

imaging and allows simultaneous three-channel acquisition

and display.

Statistical analysis

The average level of Cath B enzymatic activity for the

group of normal animals was determined. The drawT activitylevels of individual sham- and contusion-injured animals

were then normalized to this averaged value. All normalized

levels within the sham- and contusion-injury groups were

then averaged to generate a Fold Increase vs. Normal value

(FSEM) for that group. Following a two-way ANOVA (P b0.05), statistically significant differences between the two

groups were detected with Tukey’s post hoc test (SigmaStat

Statistical Software, SPSS Inc., Chicago, IL).

Results

Contusion-spinal cord injury increases Cath B enzymatic

activity

Cath B enzymatic activity levels were determined by

the production of a fluorescent product generated speci-

fically through Cath B-mediated hydrolysis. Levels of Cath

B enzymatic activity following both sham-(gray bars) and

contusion-injury (black bars) were measured from tissue

lysates of the segment containing the injury epicenter and

those segments immediately rostral and caudal to it at five

post-injury time points (1, 2, 3, 5 and 7 days). Cath B

enzymatic activity in sham-injured animals is minimally

elevated at every time point in all three tissue segments

(Figs. 1A–C). Following contusion-SCI, the level of Cath

B enzymatic activity at the injury epicenter begins to

increase at 3 days post-injury and reaches values of 5.3-

and 6.6-fold above normal at 5 and 7 days post-injury,

respectively (Fig. 1A). The contusion-injury-induced

increases at these latter two time points are significantly

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–28 21

greater than sham-injury levels (**P b 0.01). Furthermore,

regression analysis reveals the levels of the 30 kDa and 25

kDa forms of Cath B (Ellis et al., 2004) are positively

correlated with the levels of enzymatic activity (r2 =

0.9891 and r2 = 0.9828, respectively) at the injury

epicenter. Cath B enzymatic activity was not elevated in

segments rostral (Fig. 1B) and caudal (Fig. 1C) to the

injury epicenter.

Inhibition of Cath B enzymatic activity in spinal cord lysates

To confirm the usefulness of CA-074 as an inhibitor of

Cath B activity in this experimental assay, CA-074 was

added to a reaction vessel containing purified bovine Cath

B. With increasing concentrations of CA-074 (0–5 mM),

the level of Cath B enzymatic activity dramatically

diminishes (Fig. 1D), thereby confirming its efficacy as

an inhibitor of Cath B activity. Furthermore, the addition of

CA-074 (25 AM) to lysates of the tissue segment containing

the injury epicenter (Fig. 1E) essentially eliminates the

increases in Cath B enzymatic activity induced by

contusion-SCI (shown previously in Fig. 1A), thereby

demonstrating that the fluorescent signal produced in this

assay is a product of Cath B hydrolysis. Despite the lack of

Cath B enzymatic activity in either the sham- or contusion-

injury groups following the addition of CA-074, the

difference between them is statistically significantly (P b0.01) at 5 days post-injury (Fig. 1E).

Cath B immunoreactivity appears restricted to neurons in

the normal spinal cord

To investigate the cellular source of Cath B, frozen

sections of adult spinal cord were double immunostained

both with antibodies to Cath B and a variety of cell-specific

markers. In the normal spinal cord, Cath B immunoreacti-

vity (green) is most prominent in neuronal cell bodies (Fig.

2A, dorsal horn). These neurons are identified by their

morphology and by staining (red) for the neuron-specific

nuclear marker known as NeuN (Fig. 2B). The co-loca-

lization of staining for both Cath B and NeuN with the

nuclear stain DAPI (blue) is presented in Fig. 2C.

Cath B immunoreactivity (green) in normal spinal cord

tissue, however, does not coincide with staining for the

astrocytic marker glial acidic fibrillary protein or GFAP

(red, Fig. 3A) or for OX-42, a marker for resting microglia

(red, Fig. 3B). Cath B immunoreactivity is also below

detection in oligodendroglia of the normal spinal cord as

indicated by the absence of co-staining for myelin basic

protein or MBP (red, Fig. 3C). This absence of Cath B

immunoreactivity is further confirmed by the lack of Cath B

co-localization with other oligodendroglia markers includ-

ing those for 2V, 3V-cyclic nucleotide 3V-phosphodiesterase orCNPase (red, Fig. 3D) and galactocerebroside or GalC (red,

Fig. 3E). Fiber bundles of myelinated tracts (red) among the

Cath B+ neurons (green) in the dorsal horns are clearly

evident (Figs. 3C–E).

Fig. 1. Cath B enzymatic activity levels increases following contusion-injury. Cath B enzymatic activity levels (reported as Fold Increase vs. Normal) were

recorded from tissue lysates of the injury site (A) and the segments immediately rostral (B) and caudal (C) to it at five post-injury time points. Following

contusion-injury (black bars), Cath B enzymatic activity reaches 5.3- and 6.6-fold above normal at 5 and 7 days post-injury, respectively, both of which are

significantly higher than sham-injury levels (gray bars). CA-074 is an appropriate Cath B inhibitor. Cath B activity levels were assessed through the generation

of a Cath B specific fluorescent cleavage product. The addition of the specific and irreversible Cath B inhibitor CA-074 in increasing concentrations to purified

bovine Cath B nearly eliminates the fluorescent signal (D). Furthermore, the addition of CA-074 (25 AM) to the injury site tissue lysates (E) suppresses the

contusion-injury induced increases in Cath B enzymatic activity previously presented in (A).

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–2822

Cath B immunoreactivity (green) in the neurons of the

non-injured spinal cord is characterized by a punctate

granular quality (Fig. 4A), which is characteristic of a

lysosomal localization for this protease (Berquin and Sloane,

1996; Demchik et al., 1999). Indeed, the lysosomal marker,

lysosomal associated membrane protein, (LAMP, red, Fig.

4B) co-localizes with Cath B (Fig. 4C; DAPI (blue). This

same neuron was viewed with confocal microscopy with

z-plane sectioning (Figs. 4D (Cath B); E (LAMP) and F

(merged)). An additional set of confocal image of neurons is

shown in Figs. 4G–I, in support of the lysosomal localization

of Cath B in neurons of a normal spinal cord.

Contusion-spinal cord injury increases Cath B

immunoreactivity and alters Cath B localization

Although we previously detected increases in Cath B

protein levels following SCI (Ellis et al., 2004), the

cellular source of these increases was not identified. For

the present study, we used immunohistochemical techni-

ques to identify the potential source(s) of the elevation in

Cath B at 7 days post-injury. The normal spinal cord is

shown in Fig. 5A for reference. At the injury epicenter,

Cath B immunoreactivity is evident in non-neuronal cells

that essentially encompass the entire gray matter and much

of the white matter (Fig. 5B).

The Cath B-immunopositive cells at the injury epicenter

(green, Fig. 6A) co-localize with a diverse population of

leukocytes. OX-42 (red), a marker of activated microglia

and macrophages, is shown in Fig. 6B and as a merged

image with the nuclear stain DAPI (blue) in Fig. 6C

(magnified in the inset). After injury, staining for Cath B

also co-localizes with OX-8, a marker for cytotoxic T-cell

and a subset of activated macrophages (Fig. 6D). As with

the normal spinal cord (Figs. 3A, E), Cath B immuno-

reactivity is not seen in cells that are GFAP (Fig. 6E and

inset) and GalC (Fig. 6 F and inset) immunopositive.

In areas adjacent (rostral) to the injury epicenter, the

number of Cath B+ inflammatory cells is reduced and

appears to be generally located in the dorsal columns

(Fig. 7A). Unlike the injury epicenter at 7 days post-

injury (Fig. 5B), Cath B immunoreactive neurons are still

evident in areas adjacent to the injury epicenter, parti-

cularly in the ventral horns (Fig. 7B). When compared to

the punctate character of Cath B staining in the neurons

of the normal spinal cord (Fig. 7C), neurons in the 7 days

Fig. 2. Cath B staining is localized to gray matter neurons of the normal spinal cord. Cath B immunoreactivity (green) is localized morphologically to neurons

in the dorsal horn of the normal spinal cord (A). Staining for the neuronal marker NeuN (red, B) co-localizes with Cath B staining (green) in these neurons and

is presented with the nuclear stain DAPI (blue) in (C).

Fig. 3. Cath B staining is absent from other cell types in the normal spinal cord. Cath B (green) immunoreactivity does not co-localize with staining for the

astrocytic marker GFAP (red, A), the resting microglial marker OX-42 (red, B) or the oligodendroglial markers MBP (red, C), CNPase (red, D) or GalC (red, E)

in tissue sections of the normal spinal cord. Panels (A) and (B) were obtained from the ventral horn whereas (C), (D) and (E) were taken from the dorsal horn.

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–28 23

post-injury spinal cord appear more intensely immunor-

eactive (Figs. 7D–F).

Discussion

It is well known that mechanical trauma to the spinal

cord initiates a complex cascade of biochemical processes

that collectively contribute to neuronal and glial cell death,

tissue cavitation and sensory and motor deficits. Currently,

there is limited opportunity to improve and/or restore

function to the individual suffering a SCI. While a number

of the pathophysiological events contributing to the

secondary injury have been identified, Nixon and Cataldo

(1993) suggested that lysosomal leakage or rupture with the

subsequent release of proteases represents the greatest threat

to neuronal survival. Since this report, however, research on

the role of lysosomal proteases in the etiology of secondary

SCI has lagged. As indicated previously, we recently

reported that contusion-SCI significantly increased Cath B

Fig. 5. Cath B immunoreactivity is increased and altered following contusion-SCI. Cath B staining (green) is primarily restricted to gray matter neurons in the

normal spinal cord (A). In a comparable section from a 7-day post-injury spinal cord, the injury site is characterized by the robust presence of rounded Cath B-

immunopositive cells in both the gray and white matter (B). Cath B+ neurons are not evident 7 days post-injury at the injury epicenter. Both images were

obtained under identical settings (i.e., magnification, exposure time, gain).

Fig. 4. Cath B is staining is distinguished by its lysosomal localization. Using fluorescent microscopy, Cath B immunoreactivity (green) is characterized by its

punctate granular appearance (A) in neurons of the normal spinal cord. Similarly, the staining pattern of the lysosomal marker LAMP (red) also appears

punctate (B). A merged image of these two panels with the nuclear stain DAPI is shown in (C). The insets of (A) and (B) are the negative primary antibody

controls (absence of 18 antibodies). The same neuron presented in panels A–C was also examined by confocal microscopy with z-plane sectioning and is

presented in panels D (Cath B), E (LAMP) and F (merged). An additional set of confocal panels is seen in G (Cath B), H (LAMP) and I (merged).

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–2824

mRNA and protein expression at the injury site, and to a

lesser degree, in segments rostral and caudal to the injury

site (Ellis et al., 2004). The experiments described herein

were designed to determine: (1) if an increase in enzymatic

activity is associated with the previously described SCI-

induced increase in Cath B protein and (2) the post-SCI

cellular source of Cath B. Our results indicate that Cath B

enzymatic activity significantly increases at the injury

epicenter and that the primary source of this increase is

the presence of a large, diverse population of inflammatory

cells.

Although a trend towards an increase in Cath B

enzymatic activity was initially seen at 3 days post-injury,

activity in the injured tissue was not significantly higher

than sham-injury levels until post-SCI days 5 and 7. The

increase in Cath B activity at the injury site was highly

Fig. 6. Cath B-immunopositive cells in the injury epicenter are inflammatory in origin. Cath B+ cells (green) in the injury epicenter (A) are immunoreactive for

the activated microglia/macrophage marker OX-42 (red, B). These panels are combined with the nuclear stain DAPI (blue) in (C) and magnified in the inset.

Cath B+ cells also stain for OX-8, a marker of CD8+ cytotoxic T cells and a subset of activated macrophages (red, D). Staining for the astrocytic marker GFAP

(red, E, magnified in inset) and the oligodendroglial/myelin marker GalC (red, F and inset) is absent from Cath B immunoreactive cells in the 7-day post-injury

spinal cord.

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–28 25

correlated with the elevated levels of the active forms of

Cath B protein. Interestingly, post-injury increases in Cath B

enzymatic activity were not observed in the rostral or caudal

segments; a finding that most likely reflects both the small

increase in Cath B protein (generally b2-fold) and the

diminished population of inflammatory cells (restricted to

portions of the dorsal columns) observed in these adjacent

areas.

Immunohistochemical analysis showed that in the normal

spinal cord, Cath B staining was localized primarily to

lysosomes of neuronal cell bodies. Cath B staining in

astrocytes, microglia, or oligodendroglia was absent. Fol-

lowing contusion-injury, however, intense Cath B immu-

noreactivity was seen in large numbers of inflammatory

cells (e.g., activated microglia, macrophages, cytotoxic

CD8+ T cells) that essentially fill the injury epicenter.

Indeed, by 7 days post-injury, Cath B+ inflammatory cells

were so uniformly distributed through the gray and white

matter at the injury epicenter that Cath B+ neurons could not

be identified. In the areas adjacent to the injury epicenter,

Fig. 7. Cath B immunoreactivity increases in both neuronal and non-neuronal cells in areas adjacent to the injury epicenter. Cath B+ inflammatory cells (green)

are restricted to portions of the dorsal columns in areas rostral to the injury epicenter at 7 days post-injury (A). Ventral horns neurons are also Cath B

immunoreactive (green) in the injured spinal cord (B). Cath B staining in neurons of the normal spinal cord (C) is punctate, defined and not as robust as that

seen in the injured spinal cord (B). Multiple examples of increased Cath B immunoreactivity following injury are presented panels D, E and F. The images in

panels C–F were obtained under identical settings (i.e., magnification, exposure time, gain).

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–2826

however, smaller numbers of Cath B+ inflammatory cells

were evident and appeared to be limited to the dorsal

columns. Many of the neurons identified in these rostral and

caudal areas seemed to be more intensely immunoreactive

than those seen in the uninjured spinal cord. Collectively,

these findings suggest that the contusion-injury induced

increases in Cath B expression and enzymatic activity at the

injury epicenter are primarily due to the presence of large

numbers of Cath B+ inflammatory cells. However, this does

not exclude the possibility that parenchymal neurons

contribute to the smaller increases in Cath B mRNA and

protein expression that were measured shortly after injury.

Furthermore, changes in character and intensity of Cath B

immunoreactivity may precipitate and/or signal the death of

these neurons, resulting in the release of Cath B into the

cytosol, an event that has been documented in other models

of CNS trauma (Kohda et al., 1996; Yamashima et al.,

1998).

The presence of a large population of Cath B+ in-

flammatory cells in the injured spinal cord is consistent with

the suggestion that Cath B may be a contributor to the

secondary injury cascade. This proposal is supported by

previous studies that show Cath B expression is signifi-

cantly elevated in activated (but not resting) macrophages

(Kominami et al., 1988; Lah et al., 1995) and that Cath B is

produced by and released from activated microglia (Reddy

et al., 1995; Ryan et al., 1995). Furthermore, activated

microglia produce and secrete various molecules including

free radicals and cytokines that are also likely to stimulate

production and secretion of potentially damaging proteases

like Cath B (Giulian and Corpuz, 1993; Giulian and Vaca,

1993; Kim and Ko, 1998; Rothwell et al., 1996). Enzymes

produced by activated microglia have been shown to play a

direct role in neuronophagia (Banati et al., 1993; Thanos,

1991), in degradation of components of the extracellular

matrix (ECM) (Gottschall et al., 1995; Maeda and Sobel,

1996; Nakanishi, 2003) and in neuronal cell death (Gan

et al., 2004; Nitatori et al., 1996). Previous reports by

Popovich et al. (2003, 1997) have extensively characterized

the inflammatory response following SCI including the

identification and emergence of activated microglia, macro-

phages and cytotoxic T lymphocytes at the injury epicenter.

The results of this study now localize Cath B to these same

cell types in the injured spinal cord. Considering these

reports and the data contained herein, it is not entirely

surprising that reducing or blocking the influx of neutrophils

and macrophages into the injury site following SCI

decreases the extent of tissue damage (Blight, 1994; Giulian

and Robertson, 1997; Hamada et al., 1996; Mabon et al.,

2000; Popovich et al., 1999). Since our data confirm the

presence of a large and diverse population of Cath B+

inflammatory cells in the spinal cord after injury, this potent

protease is in a location where its release could certainly

exacerbate the secondary injury cascade.

Although the current study does not directly implicate

Cath B in the secondary injury cascade after SCI, the

documented involvement of this protease in a variety of

other CNS pathologies including amyotrophic lateral

sclerosis (Kikuchi et al., 2003), Alzheimer’s disease (Gan

et al., 2004, Mackay et al., 1997), multiple sclerosis (Bever

and Garver, 1995), tumor progression (Demchik et al.,

1999; Mikkelsen et al., 1995; Rempel et al., 1994;

Sivaparvathi et al., 1995; Strojnik et al., 2001) and cerebral

ischemia (Kohda et al., 1996; Seyfried et al., 1997;

Yamashima et al., 1996) demonstrate the capacity of Cath

B to participate in cell death and tissue loss processes in the

CNS. Thus, Cath B remains a strong candidate for

involvement in the tissue destruction that occurs following

SCI, a determination, however, that requires interventional

inhibitory studies.

Acknowledgments

This work was supported by (i) the State of Florida’s

Brain and Spinal Cord Injury Rehabilitation Trust Fund,

(ii) the C.M and K.E. Overstreet Family Chair in Spinal

Cord Regeneration and (iii) the Department of Veteran

Affairs. The authors thank Dr. Lucia Notterpek, Dr. Gerry

Shaw, Dr. Jake Streit and Mr. Erik Johnson for their

technical expertise.

References

Banati, R.B., Rothe, G., Valet, G., Kreutzberg, G.W., 1993. Detection of

lysosomal cysteine proteinases in microglia: flow cytometric measure-

ment and histochemical localization of cathepsin B and L. Glia 7,

183–191.

Banik, N.L., Hogan, E.L., Powers, J.M., Smith, K.P., 1986. Proteolytic

enzymes in experimental spinal cord injury. J. Neurol. Sci. 73,

245–256.

Berquin, I.M., Sloane, B., 1996. Cathepsin B expression in human tumors.

Adv. Exp. Med. Biol. 389, 281–294.

Bever Jr., C.T., Garver, D.W., 1995. Increased cathepsin B activity in

multiple sclerosis brain. J. Neurol. Sci. 131, 71–73.

Blight, A.R., 1994. Effects of silica on the outcome from experimental

spinal cord injury: implication of macrophages in secondary tissue

damage. Neuroscience 60, 263–273.

Buttle, D.J., Murata, M., Knight, C.G., Barrett, A.J., 1992. CA074 methyl

ester: a proinhibitor for intracellular cathepsin B. Arch. Biochem.

Biophys. 299, 377–380.

Demchik, L.L., Sameni, M., Nelson, K., Mikkelsen, T., Sloane, B.F., 1999.

Cathepsin B and glioma invasion. Int. J. Dev. Biol. 17, 483–494.

Ellis, R.C., Earnhardt, J.N., Hayes, R.L., Wang, K.K., Anderson, D.K.,

2004. Cathepsin B mRNA and protein expression following contusion

spinal cord injury in rats. J. Neurochem. 88, 689–697.

Fan, M., Mi, R., Yew, D.T., Chan, W.Y., 2001. Analysis of gene expression

following sciatic nerve crush and spinal cord hemisection in the mouse

by microarray expression profiling. Cell. Mol. Neurobiol. 21, 497–508.

Gan, L., Ye, S., Chu, A., Anton, K., Yi, S., Vincent, V.A., von Schack, D.,

Chin, D., Murray, J., Lohr, S., Patthy, L., Gonzalez-Zulueta, M.,

Nikolich, K., Urfer, R., 2004. Identification of cathepsin B as a

mediator of neuronal death induced by Abeta-activated microglial cells

using a functional genomics approach. J. Biol. Chem. 279, 5565–5572.

Giulian, D., Corpuz, M., 1993. Microglial secretion products and their

impact on the nervous system. Adv. Neurol. 59, 315–320.

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–28 27

Giulian, D., Robertson, C., 1997. Inhibition of mononuclear pha-

gocytes reduces ischemic injury in the spinal cord. Ann. Neurol.

27, 33–42.

Giulian, D., Vaca, K., 1993. Inflammatory glia mediate delayed neuronal

damage after ischemia in the central nervous system. Stroke 24,

I84– I90.

Gottschall, P.E., Yu, X., Bing, B.J., 1995. Increased production of

gelatinase B (matrix metalloproteinase-9) and interleukin-6 by activated

rat microglia in culture. Neurosci. Res. 42, 335–342.

Hamada, Y., Ikata, T., Katoh, S., Nakauchi, K., Niwa, M., Kawai, Y.,

Fukuzawa, K., 1996. Involvement of an intercellular adhesion molecule

1-dependent pathway in the pathogenesis of secondary changes after

spinal cord injury in rats. J. Neurochem. 66, 1525–1531.

Hu, J., Fink, D., Mata, M., 2002. Microarray analysis suggests the

involvement of proteasomes, lysosomes, and matrix metalloproteinases

in the response of motor neurons to root avulsion. Eur. J. Neurosci. 16,

1409–1416.

Iizuka, H., Iwasaki, Y., Yamamoto, T., Kadoya, S., 1986. Morpho-

metric assessment of drug effects in experimental spinal cord injury.

J. Neurosurg. 65, 92–98.

Iwasaki, Y., Yamamoto, H., Iizuka, H., Yamamoto, T., Konno, H., 1987.

Suppression of neurofilament degradation by protease inhibitors in

experimental spinal cord injury. Brain Res. 406, 99–104.

Kikuchi, H., Yamada, T., Furuya, H., Doh-ura, K., Ohyagi, Y., Iwaki, T.,

Kira, J., 2003. Involvement of cathepsin B in the motor neuron

degeneration of amyotrophic lateral sclerosis. Acta Neuropathol. (Berl.)

105, 462–468.

Kim, W.K., Ko, K.H., 1998. Potentiation of N-methyl-d-aspartate-mediated

neurotoxicity by immunostimulated murine microglia. J. Neurosci. Res.

54, 17–26.

Kohda, Y., Yamashima, T., Sakuda, K., Yamashita, J., Ueno, T., Kominami,

E., Yoshioka, T., 1996. Dynamic changes of cathepsins B and L

expression in monkey hippocampus after transient ischemia. Biochem.

Biophys. Res. Commun. 228, 616–622.

Kominami, E., Tsukahara, T., Hara, K., Katunuma, N., 1988. Biosynthesis

and processing of lysosomal cysteine proteinases in rat macrophages.

FEBS Lett. 231, 225–258.

Lah, T.T., Hawley, M., Rock, K.L., Goldberg, A.L., 1995. Gamma-

interferon causes a selective induction of the lysosomal proteases,

cathepsins B and L, in macrophages. FEBS Lett. 363, 85–89.

Mabon, P.J., Weaver, L.C., Dekaban, G.A., 2000. Inhibition of monocyte/

macrophage migration to a spinal cord injury site by an antibody to the

integrin alphaD: a potential new anti-inflammatory treatment. Exp.

Neurol. 166, 52–64.

Mackay, E.A., Ehrhard, A., Moniatte, M., Guenet, C., Tardif, C., Tarnus,

C., Sorokine, O., Heintzelmann, B., Nay, C., Remy, J.M., Higaki, J.,

Van Dorsselaer, A., Wagner, J., Danzin, C., Mamont, P., 1997. A

possible role for cathepsins D, E and B in the processing of beta-

amyloid precursor protein in Alzheimer’s disease. Eur. J. Biochem. 244,

414–425.

Maeda, A., Sobel, R.A., 1996. Matrix metalloproteinases in the normal

human central nervous system, microglial nodules, and multiple

sclerosis lesions. J. Neuropathol. Exp. Neurol. 55, 300–309.

Mikkelsen, T., Yan, P.S., Ho, K.L., Sameni, M., Sloane, B.F., Rosenblum,

M.L., 1995. Immunolocalization of cathepsin B in human glioma:

implications for tumor invasion and angiogenesis. J. Neurosurg. 83,

285–290.

Murata, M., Miyashita, S., Yokoo, C., Tamai, M., Hanada, K., Hatayama,

K., Towatari, T., Nikawa, T., Katunuma, N., 1991. Novel epoxysuccinyl

peptides. Selective inhibitors of cathepsin B, in vitro. FEBS Lett. 280,

307–310.

Nakanishi, H., 2003. Microglial functions and proteases. Mol. Neurobiol.

27, 163–176.

Nitatori, T., Sato, N., Kominami, E., Uchiyama, Y., 1996. Participation of

cathepsins B, H, and L in perikaryal condensation of CA1 pyramidal

neurons undergoing apoptosis after brief ischemia. Adv. Exp. Med.

Biol. 389, 177–185.

Nixon, P.A., Cataldo, A.M., 1993. The lysosomal system in neuronal cell

death: a review. Ann. N. Y. Acad. Sci. 679, 87–109.

Popovich, P.G., Wei, P., Stokes, B.T., 1997. Cellular inflammatory response

after spinal cord injury in Sprague–Dawley and Lewis rats. J. Comp.

Neurol. 377, 443–464.

Popovich, P.G., Guan, Z., Wei, P., Huitinga, I., van Rooijen, N., Stokes,

B.T., 1999. Depletion of hematogenous macrophages promotes partial

hindlimb recovery and neuroanatomical repair after experimental spinal

cord injury. Exp. Neurol. 158, 351–365.

Popovich, P.G., van Rooijen, N., Hickey, W.F., Preidis, G., McGaughy, V.,

2003. Hematogenous macrophages express CD8 and distribute to

regions of lesion cavitation after spinal cord injury. Exp. Neurol. 182,

275–287.

Reddy, V.Y., Zhang, Q.Y., Weiss, S.J., 1995. Pericellular mobilization of the

tissue-destructive cysteine proteinases, cathepsins B, cathepsin L, and

cathepsin S, by human monocyte-derived macrophages. Proc. Natl.

Acad. Sci. U. S. A. 92, 3849–3853.

Rempel, S.A., Rosenblum, M.L., Mikkelsen, T., Yan, P.S., Ellis, K.D.,

Golembieski, W.A., Sameni, M., Rozhin, J., Ziegler, G., Sloane, B.F.,

1994. Cathepsin B expression and localization in glioma progression

and invasion. Cancer Res. 54, 6027–6031.

Rothwell, N.J., Luheshi, G., Toulmond, S., 1996. Cytokines and their

receptors in the central nervous system: physiology, pharmacology, and

pathology. Pharmacol. Ther. 69, 85–95.

Ryan, R.E., Sloane, B.F., Sameni, M., Wood, P.L., 1995. Microglial

cathepsin B: an immunological examination of cellular and secreted

species. J. Neurochem. 65, 1035–1045.

Seyfried, D., Han, Y., Zheng, Z., Day, N., Moin, K., Rempel, S., Sloane, B.,

Chopp, M., 1997. Cathepsin B and middle cerebral artery occlusion in

the rat. J. Neurosurg. 87, 716–723.

Sivaparvathi, M., Sawaya, R., Wang, S.W., Rayford, A., Yamamoto, M.,

Liotta, L.A., Nicolson, G.L., Rao, J.S., 1995. Overexpression and

localization of cathepsin B during the progression of human gliomas.

Clin. Exp. Metastasis 13, 49–56.

Strojnik, T., Zidanik, B., Kos, J., Lah, T.T., 2001. Cathepsins B and L are

markers for clinically invasive types of meningiomas. Neurosurgery 48,

598–605.

Taoka, Y., Okajima, K., Murakami, K., Johno, M., Naruo, M., 1998. Role

of neutrophil elastase in compression-induced spinal cord injury in rats.

Brain Res. 799, 264–269.

Thanos, S., 1991. The relationship of microglial cells to dying neurons

during natural neuronal cell death and axotomy-induced degeneration of

the rat retina. Eur. J. Neurosci. 3, 1189–1207.

Towatari, T., Nikawa, T., Murata, M., Yokoo, C., Tamai, M., Hanada, K.,

Katunuma, N., 1991. Novel epoxysuccinyl peptides. A selective

inhibitor of cathepsin B, in vivo. FEBS Lett. 280, 311–315.

Yamashima, T., Saido, T.C., Takita, M., Miyazawa, A., Yamano, J.,

Miyakawa, A., Nishijyo, H., Yamashita, J., Kawashima, S., Ono, T.,

Yoshioka, T., 1996. Transient brain ischaemia provokes Ca+2, PIP2 and

calpain responses prior to delayed neuronal death in monkeys. Eur. J.

Neurosci. 9, 1932–1944.

Yamashima, T., Kohda, Y., Tsuchiya, K., Ueno, T., Yamashita, J., Yoshioka,

T., Kominami, E., 1998. Inhibition of ischaemic hippocampal neuronal

death in primates with cathepsin B inhibitor CA-074: a novel strategy

for neuroprotection based on dcalpain–cathepsin hypothesisT. Eur. J.Neurosci. 10, 1723–1733.

Yashon, D., Bingham, W.G., Friedman, S.J., Faddoul, E.M., 1975.

Intracellular enzyme liberation in primate spinal cord injury. Surg.

Neurol. 4, 43–51.

R.C. Ellis et al. / Experimental Neurology 193 (2005) 19–2828