F

Neurobiology of Aging xxx (2005) xxx–xxx

Elevation of cytoskeletal protein breakdown in aged Wistar rat brain3

Eric Bernathb, Nancy Kupinab, Ming Cheng Liub,c, Ronald L. Hayesa,b,c,Colleen Meeganb,c, Kevin K.W. Wanga,b,c,d,∗

4

5

a Department of Psychiatry, McKnight Brian Institute, L4-100, P.O. Box 100256, University of Florida, Gainesville, FL 32610, USA6b Department of Neuroscience, McKnight Brian Institute, L4-100, P.O. Box 100256, University of Florida, Gainesville, FL 32610, USA7

c Center for Traumatic Brain Injury Studies, McKnight Brian Institute, L4-100, P.O. Box 100256, University of Florida, Gainesville, FL 32610, USA8d Center for Neuroproteomics and Biomarkers Studies, McKnight Brian Institute, L4-100, P.O. Box 100256, University of Florida,

Gainesville, FL 32610, USA9

10

Received 27 April 2004; received in revised form 4 February 2005; accepted 25 February 2005

11

Abstract12

Previous studies indicated there is an overall increase of proteolysis in aging rat brains. We monitored the potential degradation of cytoskeletalproteins in neuronal tissue taken from cerebral cortex and cerebellum of young (3 month) and aging (17, 21 and 23.5 month) Wistar rats.W thec DP150a ggesting thei o calpaind ather, therei usceptibilityo

13

14

15

16

17

18

19

20

21

©22

K23

24

125

26

(27

p28

[29

p30

A31

o32

f33

h34

c35

d36

i37

t38

on39

- 40

sing41

a 42

ellu-43

lar44

by45

46

pro-47

re of-48

, cer-49

od-50

ro-51

an de-52

opto-53

rote-54

dis-55

56

1 02 d

C

OR

RE

CTE

D P

RO

Oe found significant age-dependent proteolysis of cytoskeletal proteins (�II-spectrin and microtubule-associated protein MAP-2A/B) inerebral cortex and the cerebellum. The pattern of�II-spectrin breakdown shows a marked increase in 150- and 145-kDa fragments (SBnd SBDP145, respectively), but we did not detect the caspase-3-mediated 120-kDa fragment (SBDP120) in aged rat brains, su

nvolvement of the calpain proteases. The pattern of MAP-2A/B breakdown in aged rat brains mirrors that produced by in vitrigestion of 3-month control rat brain MAP-2A/B. In aged rat brains, there is no significant increase in pro-caspase-3 processing; r

s a moderate reduction in pro-caspase-3 protein and caspase-3 hydrolytic activity in the cortex. These results point to selective sf cytoskeletal proteins to calpain-mediated degradation, but not caspase-3 in aging rat brains.2005 Published by Elsevier Inc.

eywords:Aging; Alzheimer’s disease; Calpain; Caspase; MAP-2A/B;�II-spectrin

. Introduction

Dysfunction in the homeostatic regulation of calciumCa2+) has been one of the many notable theories pro-osed to account for age-related neuronal degradation

7,12,13,16,30,56,59]and (of more immediate therapeuticotential) neurodegenerative diseases such as stroke andlzheimer’s disease[6,49]. Our study was designed to focusn experimental evidence supporting the calcium hypothesis

or aging[29,54]. Concerning this theory, past investigatorsave focused on studying the regulation of Ca2+ in intra-ellular and extracellular areas and its dysfunctional link toegenerative alterations in the physiology of neuronal activ-

ties. Of specific interest are the effects of Ca2+ on synapticransmissions[36,53] and hypoxic-ischemic cellular death

∗ Corresponding author. Tel.: +1 352 294 0031; fax: +1 352 392 2579.E-mail address:kwang@psychiatry.ufl.edu (K.K.W. Wang).

[9,24]. Many laboratories, including ours, have focusedthe secondary dysfunctional role of Ca2+ on cellular physiology and the potential role of calpain-overactivation in cauneurodegenerative responses[53,60]. It has been shown innumber of studies that a sustained elevation in intraclar Ca2+ triggers calpain activation, which leads to celludamage[5,21,24,51]. Activated calpain exerts its damagedegrading various cellular substrates such as�II-spectrin,calmodulin binding proteins, microtubule-associatedteins (MAPs) and neurofilaments. These processes aten associated with oncotic necrosis (oncosis). Howevertain pathophysiological conditions (such as hypoxia or merate excitotoxicity) can also trigger activation of the papoptotic protease caspase-3. Like calpain, caspase-3 cgrade a large number of neuronal proteins, leading to apsis[57]. In fact, both caspase-3 and calpain-mediated polytic events have been implicated in neurodegenerativeeases such as Alzheimer’s disease[1,4,25,26,44,45,48,50].

197-4580/$ – see front matter © 2005 Published by Elsevier Inc.oi:10.1016/j.neurobiolaging.2005.02.013

N NBA 6255 1–9

U

DO

F

2 E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx

Therefore, it is conceivable that calpain and caspase-3 pro-57

teases might also be activated in the normal aging process58

[10,61]. In fact, there have been some sporadic reports show-59

ing increased activities of calpains[3,5,21,52]and caspases60

[28,62]in aging brain and/or other organs. Hinman et al.[18]61

also recently showed that there is a significant increased of62

activated calpain-1 immunoreactivity in antigen-DR (HLA-63

DR)-positive microglia in the white matter of the aged rhesus64

monkey.65

In the present study, we evaluated the hypothesis that pro-66

teolysis plays a role in cellular damage due to aging by exam-67

ining calpain- and caspase-mediated proteolysis in neuronal68

tissue during the process of normal aging in the Wistar rat.69

2. Materials and methods70

2.1. Tissue preparation71

All procedures were carried out in strict compliance with72

the Institutional Animal Care and Use Committee of Parke-73

Davis and Pfizer Global Research and Development and the74

Society of Neuroscience. Male Wistar rats (Charles River75

Laboratories), age’s 3, 12, 7, 21 and 23.5 months were anes-76

thetized with 4% isoflurane before being sacrificed via car-77

diac puncture. Animals were first perfused with saline, then78

t sent-79

i ther80

t d fe-81

m on,82

e over83

d84

f ain-85

i ains86

w pes-87

t88

289

using90

a me91

o92

E d93

p to the94

f 2 h95

w ared96

w ere97

d the98

s99

100

t ide,101

N V102

f ia a103

B 2 h.104

T Tris105

b and106

then incubated in primary antibody [beta-actin (1:1000,107

Sigma Chemical), MAP-2A/B (1:1000, Sigma Chemical),108

caspase-3 (1:500, Santa Cruz),�II-spectrin (1:1000, Affin- 109

ity Co., UK) and SBDP150i-specific antibody (Cell Sig-110

naling Technology, #2121)] overnight at 4◦C. Secondary 111

biotinylated antibodies [anti-mouse IgG or anti-rabbit IgG112

(1:500, Caliches)] were exchanged for 1 h, followed by a ter-113

tiary streptavidin–alkaline phosphatase conjugate (1:3000,114

Caliches) for 30 min. The blots were developed in a115

linear range with substrate 5-bromo-4-chloro-3-indolyl-116

phosphate and nitroblue tetrazolium (BCIP/NBT, Kirkegaard117

& Perry). Densitometric analysis of the protein bands118

was quantified using NIH Image software on a Macintosh119

computer. 120

2.3. Assay for caspase-3 activity in neuronal tissue 121

Sample lysate was taken immediately following prepa-122

ration and mixed with a buffer containing 100 mM Hepes,123

10% glycerol, 10 mM DT, 0.5 mM EGTA and 40 mM acetyl-124

Asp-Glu-Val-Asp-7-amido-4-methylcoumarin (Ac-DEVD-125

MCA, Peptide International). After 2 h, the assay was mea-126

sured on a Millipore CytoFluor 2300 plate-reader[31] us- 127

ing a fluorescence of excitation 380± 15 nm and emission 128

460± 15 nm. 129

2 130

sec-131

t es132

w laced133

o e 134

w aline135

( the136

c oge-137

n rum138

( en139

i , 140

S vel-141

o ces142

w ride143

( nd144

E h a145

p 146

3 147

3 148

a 149

time150

c l pro-151

t ic-152

t Gel153

l ata154

UN

CO

RR

EC

TE

he brains were extracted and divided into regions repreng the cerebral cortex, midbrain and cerebellum, while oissues were removed (hearts, liver, kidneys, muscle anur) for inclusion in a parallel study. Following dissectiach brain region was either flash frozen in isopentanery ice (−70◦C) or stored in 4% paraformaldehyde at 4◦C

or 12 h before processing for immunohistochemical stng. Immediately preceding protein extraction, frozen brere crushed into powder using a pre-cooled mortar and

le over dry ice.

.2. Protein extraction and sample analysis

Protein samples were extracted from neuronal tissuemodified Triton extraction method. A proportionate voluf 20 mM Tris–HCl buffer (pH 7.4), 5 mM EDTA, 5 mMGTA, 150 mM NaCl, 1 mM DTT, 1% Triton X-100 anrotease inhibitor (Roche #11 836 153 001) was added

rozen (−70◦C) powdered tissue and stored in wet ice forith intermittent agitation. The homogenate was then cleith centrifugation, protein concentrations of the lysate wetermined using a modified Lowry assay (BioRad) andamples were stored in 50% glycerol at−70◦C.

Each sample was normalized to contain 25�g of pro-ein and run on SDS-PAGE [4–20% (w/v) polyacrylamovex] with a Tris/glycine running buffer solution at 200

or 2 h. Gels were transferred to a PVDF membrane vioRad semidry electrotransferring unit at 20 mA forhe blots were then placed in Blotto (Tween–TBS:ase, Tris–HCl, Tween-20 and 10% non-fat dry milk),

PR

O

NBA 6255 1–9

.4. Immunohistochemistry in neuronal tissue

Paraformaldehyde-stored brains were transverselyioned at 100�m, using a chilled tissue slicer. The slicere placed in a sucrose solution (20%) before being pn glass microscope slides and stored at 4◦C. Samples werashed in a series, which included phosphate buffered s

PBS), 80% formic acid, 0.5% Triton buffer (to permeateell membrane), 1% hydrogen peroxide (to quench endous peroxidase activity) and blocked in 2% horse seprovided by Vectastain Elite ABC kit). Slides were thncubated over night at 4◦C with anti-MAP-2A/B (1:500igma Chemical) and then washed in PBS. Following depment using the Vectastain Elite ABC Kit, the brain sliere visualized with a diaminobenzidine tetrahydrochlo

DAB) solution and counter-stained with Hematoxylin aosin (H&E) Staining. All sections were examined withase contrast microscope at 400× resolution.

. Results

.1. Immunoblot analysis ofαII-spectrin in young andged rat brains

Western immunoblots were prepared to evaluate aourse of degradation of two protease-sensitive neuronaeins (�II-spectrin and MAP-2A/B) to gauge an overall pure of biochemical functions during the aging process.oading controls with beta-actin were performed for all d

OR

RE

CTE

D P

RO

OF

E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx 3

Fig. 1. Age-related increase of�II-spectrin breakdown in rat cortex. (A) Two sets of representative immunoblots of�II-spectrin breakdown during timecourse are shown (totaln= 5–8). Spectrin intact bands (280 kDa). SBDP150 and SBDP145 are indicated by arrows. (B) Time course of protein extracts ofcerebral cortex tissue at 3, 17, 21 and 23.5 months were assayed in immunoblots against�II-spectrin antibody (Affiniti) at 1:1000 concentration. Dark barsindicate 150-kDa spectrin breakdown products (SBDP150) while shaded bars indicate intact SBDP145. Data are shown as SBDP/intact spectrin ratio. Dataare means± S.E.M. (n= 5 for control and 23.5 month,n= 6 for 21 month,n= 8 for 17 month). Values showing significance different from 3-month controlanimal are indicated by∗p≤ 0.05 or** p< 0.005, with Student’s unpairedt-test. (C) One set of representative immunoblots of caspase-produced SBDP150i(Cell Signaling Technology) was shown (samples as in top panel of A). The far right lane (+) is SDBP150i positive control (lysate from human SH-SY5Y cellssubjected to staurosporin treatment)[33]. (D) For comparison,�II-spectrin was degraded in rat cortex lysate digested with calpain and caspase-3, producingSBDP150/SBDP145 and SBDP150i/SBDP120, respectively. Analysis of variance (ANOVA) was also performed on SBDP levels with respect to age.

(Figs. 1 and 2). Since the loading was relatively even, we155

did not further normalize�II-spectrin and MAP-2A/B data156

to actin. Control Wistar rats (3 months,n= 5) were compared157

with rats of ages 17 months (n= 8), 21 months (n= 6) and158

23.5 months (n= 5). Immunoblot samples were separated by159

brain section, with testing done on the cerebral cortex and160

cerebellum.161

Analysis with�II-spectrin antibody revealed significant162

breakdown of this protein in the cortex and cerebellum in163

aged rats (Figs. 1 and 2). Fig. 1A shows cortical samples sub-164

jected to�II-spectrin immunoblot analysis. The intact�II-165

spectrin appears in all lanes as a band of 280 kDa. However,166

levels of�II-spectrin breakdown products 150 and 145 kDa167

(SBDP150 and SBDP145) increased significantly during the168

time course of aging. To standardize comparison across dif-169

ferent blots, we expressed SBDP level as a ratio of the intact170

�II-spectrin level in the same samples. In the cortex, the high-171

est increase of SBDP150 was observed at 17 months, but172

SBDP150 levels were also significantly elevated at 20 and173

23.5 months, based upon Student’st-test analysis (Fig. 1A174

and B). Similar to the SBDP150 counterpart, SBDP145 is175

also significantly elevated in the cortex of 17-, 20- and 23-176

month-old rats (Fig. 1). On the other hand, in the cerebellum,177

SBDP150 and SBDP145 were also significantly and progres-178

sively elevated in 17, 20 and 23 months, when compared179

to young rat controls (3 months) (Fig. 2A and B). Further, 180

ANOVA (analysis of variance) showed that SBDP145 sig-181

nificantly increases in both cortex and cerebellum with re-182

spect to animal age, while age-related SBDP150 increasing183

did not reach statistical significance. In addition, evenness184

of gel loading was confirmed with beta-actin as the standard185

(Figs. 1A and 2A). 186

To compare the SBDPs observed in the aged rat brains187

to those generated by these proteases (calpain and caspase-188

3), we subjected naıve 3-month rat cortex lysate to in vitro 189

calpain and caspase-3 digestion, respectively (Fig. 1C). Both 190

SBDP150 and SBDP145 were prominently generated by cal-191

pain proteolysis[43,47,57]. In contrast, in vitro caspase-3 di-192

gestion generated and a fragment just below 150-kDa band193

(termed “SBDP150i”) and a characteristic 120-kDa spectrin194

breakdown product (SBDP120)[33,58]. Both SBDP150 and 195

SBDP145 were prominently generated by calpain proteolysis196

[47,57]. In comparison, while both SBDP150 and SBDP145197

were prominently observed in aged rat cortex and cerebel-198

lum, the SBDP120 was noticeably absent (Figs. 1A and 2A). 199

We further examined, if there was also a lack of casapse-200

3-produced SBDP150i. Indeed, using an anti-SBDP150i-201

specific antibody (cell signaling), we confirmed that caspase-202

3-generated SBDP150i was also absent in cortex and cere-203

bellum from all age groups (Fig. 1C).

UN

C

NBA 6255 1–9

CO

RR

EC

TED

PR

OO

F

4 E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx

Fig. 2. Age-related�II-spectrin breakdown in rat cerebellum. (A) Two sets of representative immunoblots of�II-spectrin breakdown during time course areshown (totaln= 5–8). Spectrin intact bands (280 kDa), SBDP150 and SBDP145 are indicated by arrows. (B) Time course of protein extracts of cerebral cortextissue at 3, 17, 21 and 23.5 months were assayed in immunoblots against�II-spectrin antibody (Affiniti) at 1:1000 concentration. Dark bars indicated 150-kDaspectrin breakdown products (SBDP150) while shaded bars indicate SBDP145. Data are shown as SBDP/intact spectrin ratio. Data are means± S.E.M. (n= 5for control and 23.5 month,n= 6 for 21 month,n= 8 for 17 month). Values showing significance different from 3-month control animal are indicated by∗p≤ 0.02 or** p< 0.05, with Student’s unpairedt-test. In the�II-spectrin intact band (280 kDa), arrows indicate SBDP150 and SBDP145. (C) One set ofrepresentative immunoblots of caspase-produced SBDP150i was shown (samples as in top panel of A). The far right lane (+) is SDBP150i positive controlasin Fig. 1[33]. ANOVA was also performed on SBDP levels with respect to age.

3.2. Immunoblot analysis of MAP-2A/B in young and204

aged rat brains205

Young and aged rat cortical and cerebellar samples were206

also subjected to analysis with Western immunoblotting us-207

ing a microtubule associated protein 2A and 2B (MAP-2A/B)208

antibody. In this case, the intact MAP-2A/B appeared as a209

doublet of about 300 kDa in all lanes (Fig. 3A (cortex) and 4A210

(cerebellum)). A dramatic and decreased intact MAP-2A/B211

density over the time course is seen in the cortex (Fig. 3B)212

and in the cerebellum (Fig. 4B) (based upon Student’st-test213

analysis). The reduction of MAP-2A/B protein levels could214

not be due to uneven gel loading, since equal amounts (20�g)215

were loaded to each lane. Furthermore, results with beta-actin216

showed constant levels between control and aged samples in217

both cortex and cerebellum, indicating sample loadings were218

even (see “actin” panels inFig. 7). In parallel with the reduc-219

tion of intact MAP-2A/B in aged rat brains, multiple break-220

down products of MAP-2A/B were also apparent in cortex221

and cerebellum from aged rats (Figs. 3A and 4A, respec-222

tively). Similar to SBDP analysis, evenness of gel loading223

was confirmed with beta-actin as standard (Figs. 3A and 4A).224

Additional ANOVA post hoc analysis also showed that in-225

tact MAP-2A/B levels significantly decreased with respect226

to animal age (p< 0.05) in both cortex and cerebellum.227

Since MAP-2A/B was reported to be extremely sensitive228

to calpain-mediated proteolysis[15,19,20,51]and no reports 229

have been published on its sensitivity to caspase-3, we sub-230

jected naıve cortex (3 month) lysate to calpain digestion and231

found that the in vitro calpain-proteolyzed MAP-2A/B also232

shows multiple lower molecular weight fragments (Fig. 3C), 233

as was the case observed in aged rat cortex and cerebellum234

(Figs. 3A and 4A). 235

We also performed immunohistochemical analysis with236

MAP-A/B antibody with cortex and cerebellum sam-237

ples from young (3 month) and aged (23.5 month) rats.238

Paraformaldehyde-fixed brains were transversely sectioned239

at 100�m and slices were incubated over night at 4◦C 240

with MAP-2A/B (1:500, Sigma Chemical) and visualized241

with a diaminobenzidine tetrahydrochloride (DAB) (brown)242

(Fig. 5). Slices were also counter-stained with Hematoxylin243

and Eosin Staining (blue/red).Fig. 5 showed representa- 244

tive slices from young cortex have more MAP-2 reactiv-245

ity when compared to aged counterpart (compare panel246

A to B, Fig. 5). Consistent with the Western blot re-247

sults, the decline of MAP-2A/B staining in aged cere-248

bellum appeared more severe (compared panel D to C,249

Fig. 5).

UN

NBA 6255 1–9

CO

RR

EC

TED

PR

OO

F

E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx 5

Fig. 3. Alteration of MAP-2A/B levels in rat cortex during aging. (A) Two representative immunoblots (totaln= 4) of MAP-2A/B breakdown during timecourse are shown. Indicated is the intact MAP-2A/B (280 kDa). Arrows indicate multiple breakdown products (BDPs). (B) Time course of intact MAP-2A/Blevels of cerebellum tissue at 3, 17, 21 and 23.5 months were assayed in immunoblots. Data are ratios of intact MAP-2A/B at time X/MAP-2A/B of 3-monthcontrol animals. Data are means± S.E.M. (n= 4). Values showing significance different from 3-month control animal are indicated by∗p≤ 0.02 or** p< 0.001with Student’s unpairedt-test. (C) For comparison, MAP-2A/B was degraded into multiple BDPs in rat cortex lysate digested with calpain. ANOVA was alsoperformed on MAP-2A/B levels with respect to age.

Fig. 4. Alteration of MAP-2 levels in rat cerebellum during aging. (A) Two representative immunoblots (totaln= 4) of MAP-2 breakdown during time courseare shown. Indicated is the intact MAP-2 (300 kDa). Arrows indicate multiple breakdown products (BDP). (B) Time course of intact MAP-2A/B levels ofcerebellum tissue at 3, 17, 21 and 23.5 months were assayed in immunoblots. Data are ratio of intact MAP-2A/B at time X/MAP-2A/B of 3-month controlanimals. Data are means± S.E.M. (n= 4). Values showing significance different from 3-month control animal are indicated by∗p≤ 0.05 with Student’s unpairedt-test. ANOVA was also performed on MAP-2A/B levels with respect to age.

UN

NBA 6255 1–9

RR

EC

TED

PR

OO

F

6 E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx

Fig. 5. Immunohistochemical staining of MAP2A/B of young and agedrat brains. Paraformaldehyde-stored brains were transversely sectioned at100�m. Cortex (A and B) and cerebellum (C and D) sections from young(3 months, panels A and C) or old (23.5 months, panels B and D) ratswere treated and incubated with anti-MAP-2A/B (1:500, Sigma Chemical),washed and visualized with a diaminobenzidine tetrahydrochloride (DAB)solution (brown). The immunoreactive product appears brown in color overa background of Hematoxylin and Eosin (H&E) Staining (blue/red). All sec-tions were examined with a phase contrast microscope at 400× resolution.Shown here are representative sections. Scale bar represents 100�m. (Forinterpretation of the references to color in this figure legend, the reader isreferred to the web version of the article.)

3.3. Analysis of caspase-3 processing and activity in250

aged rat brains251

The lack of increase of SBDP120 in aged rat brains252

prompted us to further investigate the status of caspase-3253

Fig. 6. Time course of caspase-3 activity in rat cortex and cerebellum. TritonX-100 extracts were incubated in reaction mixtures the caspase-3 preferredAc-DEVD-AMC peptide substrate. Solid bars indicate cortex levels andshaded bars indicate cerebellum levels. All data are measured as fluorescenceunits divided by sample protein level (�g). Data are means± S.E.M. (n= 5for control and 23.5 month,n= 6 for 21 month,n= 12 for 17 month). Valuesshowing significance different from 3-month control animal are indicatedby ∗p< 0.05 or** p< 0.005, with Student’s unpairedt-test.

processing and activity during aging in neuronal tissue. Im-254

mediately following protein extraction, a time course of fresh255

cortical and cerebellar samples were assayed against a fluo-256

rogenic caspase-3 substrate acetyl-DEVD-AMC to examine257

if aging produced an elevated caspase-3 hydrolytic activity.258

Fig. 5 shows that the opposite trend was seen: a significant259

decrease in caspase-3 activity (rather than increase) in both260

cortex (28% decrease) and cerebellum (25% decrease) was261

seen with aged (23.5 month) rats compared to the 3-month262

control animals (Fig. 6). ANOVA also showed that caspase263

activity declines significantly with age. 264

F llum. I d( se-3 (S of caspase.N ed by a g aging wd formed s( ed rats kss indicat

C

O

ig. 7. Lack of caspase-3 processing in aged rat cortex and cerebe23.5 month;n= 5) rat cortex (A) and cerebellum (B) using anti-caspao apparent processing of the 32 kDa pro-form of caspase-3 (indicatetected. As controls, Immunoblot analysis of beta-actin was also peras pro-caspase-3/beta-actin ratio) is calculated for the young and aghowing significance different from 3 month (white bars) samples are

UN

mmunoblot analysis of pro-caspase-3 levels in 3-month control (n= 4) and ageanta Cruz, H-277, 1:1000), which detects pro- and activated formsrrow) to the 17 kDa activated form (indicated by open triangle) durinasusing anti-beta-actin antibody (Sigma). (C) Quantification of pro-caspase-3 level’ cortex and cerebellum, respectively. In the cortex, 23.5 month (darbars) valueed by∗p< 0.05 with Student’s unpairedt-test.

NBA 6255 1–9

ED

OF

E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx 7

Parallel immunoblotting analysis of the same samples265

were then performed and probed with an anti-caspase-3 anti-266

body, which detects both pro- and activated forms of caspase-267

3 (Fig. 7A). The 32 kDa pro-caspase-3 protein was observed268

in the cortex and cerebellum of both young and aged rats,269

while the fragment for the 17 kDa activated form was not270

observed, either in young or aged rat brains. Western im-271

munoblots were also performed with anti-beta-actin anti-272

body, which showed relatively constant levels between young273

and aged cortex and cerebellum sample group (Fig. 7A and274

B, respectively). In fact, densitometric analysis showed that,275

with respect to beta-actin levels (as controls), the pro-caspase-276

3 level was actually lower in cortex but remained constant277

in the cerebellum in aged rats with comparison to young rats278

(Fig. 7C). Taken together, these data would suggest that there279

might be an age-related reduction of pro-caspase-3 expres-280

sion and no evidence of increased activation of caspase-3.281

4. Discussion282

The results of this study present evidence of elevated283

proteolysis of�II-spectrin (Figs. 1 and 2) and MAP-2A/B284

(Figs. 3 and 4) in a time course of aging in Wistar rats. Our285

results provide evidence that cytoskeletal protein degrada-286

tion is a common occurrence in aged rat brains. Both of these287

p ele-288

t t of289

t xons290

a291

p hich292

p traf-293

fi294

w eu-295

r egra-296

d se-297

v298

eath299

a have300

s -301

o rone302

t e303

c in the304

p down305

a ellum306

( that307

c gion308

[ rs to309

c urons310

i g311

t any312

s n the313

i in314

n ion315

c ping316

b ins317

had on degeneration. They showed conclusively that MAP-2318

was specifically at risk for being degraded due to the sensi-319

tivity as a calpain substrate. Others highlighted the changes320

in phosphorylation that occurs with the MAP-2 protein[20], 321

which, due to age-related changes, will increase its suscepti-322

bility with time. This happens when the regulation of binding323

changes in the microtubule network, making it more open to324

calpain hydrolysis. It is possible that changes due to spectrin325

breakdown could contribute to this. With the loss of MAP-326

2A/B, dendrites appear to collapse. One of the first studies to327

examine the effects of MAPs in an aged brain[8] saw almost 328

total MAP-2A/B loss in the cortex and hippocampus with329

associated loss in synaptic plasticity. 330

One of the major hypotheses that promoted this study was331

that caspase- and calpain-mediated proteolysis could also be332

identified as a contributor to the protein degeneration seen in333

the aging process. Many studies have identified its increase in334

ischemic events[31,33,35,40]and in head trauma[23,41,46]. 335

The caspase family of proteases has been shown to mediate336

cell suicide[34] in cells either damaged by trauma or having337

undergone irreversible degeneration to the point where pro-338

grammed death is required. Our experiments focused on the339

evaluation of caspase- and calpain-mediated proteolysis in340

aged rat brains. These studies, utilizing anti-caspase-3 West-341

ern immunoblots and an active caspase-3 assay (ac-DEVD-342

AMC) indicated no increase in activated forms of the 32 kDa343

p ase-3344

( s 345

s is to346

S ibed347

c oduct348

( 349

n about350

1 i- 351

b 150352

w bility,353

w here354

w g or355

a 356

w t sig-357

n 358

s nted.359

B ain-360

m , but361

n and362

c DP363

w am-364

p (13365

m lpain-366

m ing in367

b e 368

o use369

v 370

ele-371

t ed372

r age-373

UN

CO

RR

EC

T

roteins play important roles in the stability of the cytoskal system.�II-spectrin is the major structural componenhe cortical membrane cytoskeleton and is abundant in and presynaptic terminals[16,42]. MAP-2A/B is a dendriticrotein that interacts with and stabilizes microtubules, wrovide both structural elements and tracks for organellec [11,27]. The degradation of�II-spectrin and MAP-2A/Bould seriously compromise the normal functions of n

onal tissue. Our results also indicate that age-related dation of both�II-spectrin and MAP-2A/B appears lessere in the cortex than in the cerebellum.

Prior studies have looked at various aspects of cell dnd calcium-related changes in the brain. Experimentshown region-specific changes in calcium[17] and that homestasis in the hippocampus and cortex are especially p

o dysfunction in aged rats, while the Ca2+ changes in therebellum and striatum are less severe. Interestingly,resent study, aging-induced cytoskeleton protein breakppears to affect both the cerebral cortex and the cerebFigs. 1–4). Studies of the human brain have implicatedalpain levels are among the highest in the cerebellar re3]. Also, evaluation of the rat brain has led researcheonclude that aging produces a more drastic loss of nen the cerebellum than in the cortex[37]. When examininhe effects of calpain-mediated spectrin breakdown, mtudies have highlighted the role that proteases play ontegrity of the neuronal structure[24] and the increaseeuronal death[46]. Early studies focused on the relatalpain-mediated proteolysis had on the young develorain [14] and highlighted the susceptibility these prote

PR

O

NBA 6255 1–9

ro-caspase fragment, or any increase in active caspFigs. 5 and 6). Furthermore, anti-�II-spectrin immunoblothowed that there is significant age-related proteolysBDP150 and SBDP145, while the previously descraspase-3 generated 120-kDa spectrin breakdown prSBDP120)[57] was not observed (Figs. 1 and 2). It is worthoting that caspase-3 can also produce a fragment of50 kDa (termed SBDP150i, see[33,58]). It is, thus, possle that SBDP150i can in part contribute to the SBDPe observed in aged rat brain tissues. To test this possie used a SBDP150i-specific antibody and found that tere no detectable signals of SBDP150i in either younged cortical or cerebellar samples (Figs. 1C and 2C). Thus,e conclude that caspase-mediated proteolysis is noificantly elevated in aged rat brains. Calpain-mediated�II-pectrin proteolysis in aging has in fact been documeahr et al.[2] showed a rather selective increase of a calpediated SBDP in both the cortex and telencephalonone in the cerebellum of aged mouse brains. Vicenteolleagues[54] also showed that a calpain-mediated SBas significantly elevated in the neocortex and hippocus but not in thalamus or cerebellum in mid-aged miceonths). These data contrast our observations that caediated SBDP150 and SBDP145 are elevated with agoth cortex and cerebellum (Figs. 1 and 2, respectively). Onf the contributing factors could be species variation (moersus rat).

Our analysis of the extensive proteolysis of the cytoskal protein MAP-2A/B in the cortex and cerebellum of agats supports the importance of calpain activation in

DO

F

8 E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx

related degeneration, since MAP-2A/B is known to be more374

susceptible to calpain- and not caspase-3 activation. MAP-375

2A/B (Figs. 3 and 4). It is worth noting that MAP-2A/B break-376

down products in the cortex are in fact more intense than their377

counterparts in the cerebellum (Fig. 3A versusFig. 4A). Yet,378

the levels of remaining intact MAP-2A/B in aged cerebel-379

lum (23.5 month) were decreased to about 55% of young380

cortex controls, while the MAP-2A/B levels in aged cerebel-381

lum (23.5 months) were further decreased to 25% of young382

cerebellum controls (Fig. 3B versusFig. 4B). We hypothe-383

sized that proteolysis in the cerebellum was even more ac-384

tive, thus, not only reducing the level of intact MAP-1A/B,385

but more efficient in further degrading the fragments into386

smaller peptides, as is the case with digestion with calpain387

in vitro (Fig. 3C). In addition, immunohistochemical analy-388

sis, while not quantitative, showed a more dramatic reduction389

of MAP-2A/B levels in the cerebellum than in cerebral cor-390

tex in the aged (23.5 month) rats, when compared to their391

younger counterpart (3 month) (Fig. 5). Hinman et al.[18]392

recently showed that there is also a significant increase in ac-393

tivated calpain-1 immunoreactivity in microglia and possibly394

myelin-bearing oligodendrocytes in the white matter of the395

aged rhesus monkey. It is, therefore, possible that there is a396

concerted increase of calpain activity in multiple cell-types397

within the aged brain.398

Chronic neurological diseases, such as Parkinson’s and399

A d by400

a in ac-401

t n402

o n403

t ents404

[ lpain405

a rther406

d rod-407

u the408

c d tha409

m ues410

a tion.411

W con-412

c here413

c brain414

a ative415

d416

U417

418

A419

stions420

a nnis421

a ven422

D nts;423

Dr. Stephen Larner (University of Florida) for his statistical424

expertise and Dr. Brian Pike (NIH) concerning aging research425

was greatly appreciated. 426

References 427

[1] AyalaGrosso C, Ng G, Roy S, Robertson GS. Caspase-cleaved428

amyloid precursor protein in Alzheimer’s disease. Brain Pathol429

2002;(12):430–41. 430

[2] Bahr BA, Vanderklish PW, Ha LT, Tin MT, Lynch G. Spectrin break-431

down products increase with age in telencephalon of mouse brain.432

Neurosci Lett 1991;131(2):237–40. 433

[3] Banay-Schwartz M, DeGuzman T, Palkovits M, Lajtha A. Calpain434

activity in adult and aged human brain regions. Neurochem Res435

1994;(19):563–7. 436

[4] Bancher C, Lassmann H, Breitschopf H, Jellinger KA. Mech-437

anisms of cell death in Alzheimer’s disease. J Neural Transm438

1997;(50):141–52. 439

[5] Benuck M, Banay-Schwartz M, DeGuzman TT, Lajtha A. Changes440

in brain protease activity in aging. J Neurochem 1996;(67):2019–29.441

[6] Brinton DR, Yamazaki RS. Advances and challenges in the442

prevention and treatment of Alzheimer’s disease. Pharm Res443

1998;(15):386–98. 444

[7] Carlson JC, Riley JC. A consideration of some notable aging theo-445

ries. Exp Gerontol 1998;(33):127–34. 446

[8] Chauhan N, Siegal G. Age-dependent organotypic expression of447

microtubule-associated proteins (MAP1, MAP2, and MAP5) in rat448

brain. Neurochem Res 1997;(22):713–9. 449

[9] Choi DW. Calcium-mediated neurotoxicity: relationship to spe-450

rosci451

452

[ dur-453

ndent454

. 455

[ lles456

iol457

458

[ ms459

460

[ urol461

462

[ pro-463

2 in464

465

[ ad-466

(De-467

468

[ JE,469

ull470

471

[ ific472

rain.473

474

[ Ac-475

of476

477

[ ule-478

rative479

480

[ tein 2481

rosci482

483

[ in II484

nth-485

486

UN

CO

RR

EC

TE

lzheimer’s disease have both been found to be affecteltered calcium homeostasis resulting in increased calpa

ivity [48], �II-spectrin breakdown[38,43] and degradatiof MAP-2A/B in the cytoskeleton[20]. It has been show

hat neurofibrillary tangles contain broken MAP-2 segm22]. Although these results associate the affects of cas having key roles in the aging process, studies of fuetail are necessary to pinpoint all related breakdown pcts, their roles in the changing physiology of the cell, andascade effects on the system as a whole. It is understooany other effects of aging, from DNA damage to plaqnd toxins, play fundamental parts in neural degeneraith increased examination of these factors, working in

ert with calpain-mediated proteolysis, it is hoped that tan be further understanding of the aging of the humannd its relationship to the pathobiology of neurodegeneriseases[7].

ncited references

[32,39,55].

cknowledgments

The authors sincerely appreciate the support, suggend technical assistance of Rathna Nath, Dr. Kim McGind Michelle Field. Also the help of Drs. Feng Bian, Steuddy and Larry Walker in the execution of experime

PR

O

NBA 6255 1–9

t

cific channel types and role in ischemic damage. Trends Neu1988;(11):465–9.

10] Chondrogianni N, Fragoulis EG, Gonos ES. Protein degradationing aging: the lysosome-, the calpain- and the proteasome-depecellular proteolytic systems. Biogerontology 2002;3(1–2):121–3

11] Cole NB, Lippincott-Schwartz J. Organization of organeand membrane traffic by microtubules. Curr Opin Cell B1995;7(February (1)):55–64.

12] Cuervo AM, Dice FJ. How do intracellular proteolytic systechange with age? Front Biosci 1997;(3):25–43.

13] Drachman DA. Aging and the brain: a new frontier. Ann Ne1997;(42):819–28.

14] Fischer I, Romano-Clarke G, Grynspan F. Calpain-mediatedteolysis of microtubule associated proteins MAP1B and MAPdeveloping brain. Neurochem Res 1991;(16):891–8.

15] Friedrich P, Aszodi A. MAP2: a sensitive cross-linker andjustable spacer in dendritic architecture. FEBS Lett 1991;295cember (1–3)):5–9.

16] Goodman SR, Zimmer WE, Clark MB, Zagon IS, BarkerBloom ML. Brain spectrin: of mice and men. Brain Res B1995;36(6):593–606.

17] Hartmann H, Velbinger K, Eckert A, Muller W. Region-specdownregulation of free intracellular calcium in the aged rat bNeurobiol Aging 1995;(17):557–63.

18] Hinman JD, Duce JA, Siman RA, Hollander W, Abraham CR.tivation of calpain-1 in myelin and microglia in the white matterthe aged rhesus monkey. J Neurochem 2004;289(2):430–41.

19] Johnson GV, Litersky JM, Jope RS. Degradation of microtubassociated protein 2 and brain spectrin by calpain: a compastudy. J Neurochem 1991;56(May (5)):1630–8.

20] Johnson GVW, Jope RS. The role of microtubule-associated pro(MAP-2) in neuronal growth, plasticity, and degeneration. J NeuRes 1992;(333):505–12.

21] Kenessey A, Banay-Schwartz M, DeGuzman T, Lajtha A. Calpaactivity and calpastatin content in brain regions of 3- and 24-moold rats. Neurochem Res 1990;(15):243–9.

ED

OF

E. Bernath et al. / Neurobiology of Aging xxx (2005) xxx–xxx 9

[22] Kosik KS, Orecchio LD, Bakalis S, Duffy L, Neve RL. Partial se-487

quence of MAP-2 in the region of a shared epitope with Alzheimer488

neurofibrillary tangles. J Neurochem 1988;(51):587–98.489

[23] Kupina NC, Detloff MR, Bobrowski WF, Snyder BJ, Hall Ed. Cy-490

toskeletal protein degradation and neurodegradation evolve differ-491

ently in males and females following experimental head injury. Exp492

Neurobiol 2003;180(March (1)):55–73.493

[24] Lee K, Frank S, Vanderklish P, Arai A, Lynch G. Inhibition of494

proteolysis protects hippocampal neurons from ischemia. Proc Natl495

Acad Sci USA 1991;(88):7233–7.496

[25] Lee MS, Kwon YT, Li M, Peng J, Friedlander RM, Tsai LH.497

Neurotoxicity induces cleavage of p35 to p25 by calpain. Nature498

2000;(405):360–4.499

[26] Lee MS, Tsai LH. Cdk5: one of the links between senile plaques500

and neurofibrillary tangles? J Alzheimers Dis 2003;(5):127–37.501

[27] Lippincott-Schwartz J, Cole NB. Roles for microtubules and ki-502

nesin in membrane traffic between the endoplasmic reticulum and503

the Golgi complex. Biochem Soc Trans 1995;(23):544–8.504

[28] Lynch AM, Lynch MA. The age-related increase in IL-1 type I505

receptor in rat hippocampus is coupled with an increase in caspase-506

3 activation. Eur J Neurosci 2002;(15):1779–88.507

[29] Lynch MS. Analysis of the mechanisms underlying the age-related508

impairment in long-term potentiation in the rat. Rev Neurosci509

1998;(July–December (9)):169–201.510

[30] Morrison JH, Hof PR. Life and death of neurons in the aging brain.511

Science 1997;(278):412–24.512

[31] Namura S, Zhu J, Fink K, Endres M, Srinivasan A, Tomaselli KJ,513

et al. Activation and cleavage of caspase-3 in apoptosis induced by514

experimental cerebral ischemia. J Neurosci 1998;(18):3659–68.515

[32] Nath R, Probert A, McGinnis KM, Wang KKW. Evidence516

for activation of caspase-3-like protease in excitotoxin-517

hem518

519

[ ng520

onal521

522

[ IBS523

524

[ is of525

526

[ la-527

aged528

stasis529

530

[ rain531

532

[ . In-533

imer534

535

[ , et536

loss537

nce538

539

[ ang540

ing541

chem542

543

[ yes544

, cal-545

enta-546

i Res547

[42] Riederer BM, Zagon IS, Goodman SR. Brain spectrin (240/235)548

and brain spectrin (240/235E): two distinct subtypes with different549

locations within mammalian neural cells. J Cell Biol 1986;102(June550

(6)):2088–97. 551

[43] Roberts-Lewis JM, Savage MJ, Marcy VR, Pinsker LR, Siman552

R. Immunolocalization of calpain I-mediated spectrin degradation553

to vulnerable neurons in the ischemic gerbil brain. J Neurosci554

1994;14(June (6)):3934–44. 555

[44] Rohn TT, Head E, Nesse WH, Cotman CW, Cribbs DH. Activa-556

tion of caspase-8 in the Alzheimer’s disease brain. Neurobiol Dis557

2001;(8):1006–16. 558

[45] Rohn TT, Rissman RA, Davis MC, Kim YE, Cotman CW, Head E.559

Caspase-9 activation and caspase cleavage of tau in the Alzheimer’s560

disease brain. Neurobiol Dis 2002;(11):341–54. 561

[46] Saatman KE, Bozyczko-Coyne D, Marcy V, Siman R, McIntosh TK.562

Prolonged Calpain-mediated spectrin breakdown occurs regionally563

following experimental brain injury in the rat. J Neuropathol Exp564

Neurol 1996;(55):852–62. 565

[47] Saido T, Yokota M, Nagao S, Yamaura I, Tani E, Tsuchiya T, et al.566

Spacial resolution of fodrin proteolysis in postischemic brain. J Bio567

Chem 1993;(268):25239–43. 568

[48] Saito K, Elce JS, Hamos JE, Nixon RA. Widespread activation569

of calcium-activated neutral proteinase (calpain) in the brain of570

Alzheimer’s disease: a potential molecular basis for neuronal de-571

generation. Proc Natl Acad Sci USA 1993;(90):2628–32. 572

[49] Sandson FO. The neurobiology and pharmacotherapy of Alzheimer’s573

disease. J Neuropsychiatry Clin Neurosci 1999;(11):19–31. 574

[50] Selznick LA, Holtzman DM, Han BH, Gokden M, Srinivasan575

AN, Johnson Jr EM, et al. In situ immunodetection of neuronal576

caspase-3 activation in Alzheimer disease. J Neuropathol Exp Neu-577

rol 1999;(58):1020–6. 578

[ lpain579

ron580

581

[ CR.582

ndro-583

the584

585

[ lls.586

587

[ INS588

589

[ ctrin590

Mol591

592

[ ing593

594

[ rends595

596

[ nia597

in by598

ugust599

600

[ m-601

602

[ cal-603

604

[ Mol605

606

[ e ac-607

ntol608

609

UN

CO

RR

EC

T

and hypoxia/hypoglycemia-injured neurons. J Neuroc1988;(71):186–95.

33] Nath R, Raser KJ, McGinnis K, Nadimpalli R, Stafford D, WaKKW. Effects of ICE-like protease and calpain inhibitors on neurapoptosis. Neuroreport 1996;(8):249–55.

34] Nicholson DW, Thornberry NA. Caspases: killer proteases. T1997;(22):299–306.

35] Nixon RA. A “protease activation cascade” in the pathogenesAlzheimer’s disease. Ann N Y Acad Sci 2000;(924):117–31.

36] Ouanounou A, Zhang L, Charlton M, Carlin P. Differential modution of synaptic transmission by calcium chelators in young andhippocampal CA1 neurons: evidence for altered calcium homeoin aging. J Neurosci 1999;19(February (3)):906–15.

37] Peng MT, Lee LR. Regional differences of neuron loss of rat bin old age. Gerontology 1979;(25):205–11.

38] Peterson C, Vanderklish P, Schubert P, Cotman C, Lynch Gcreased spectrin proteolysis in fibroblasts from aged and Alzhedonors. Neurosci Lett 1991;(121):239–43.

39] Posmantur R, Kampfl A, Siman R, Liu SJ, Zhao X, Clifton GLal. A calpain inhibitor attenuates cortical cytoskeletal proteinafter experimental traumatic brain injury in the rat. Neuroscie1997;(77):875–88.

40] Posmantur R, McGinnis K, Nadimpalli R, Gilbertsen RB, WKKW. Characterization of CPP32-like protease activity followapoptotic challenge in SH-SY5Y neuroblastoma cells. J Neuro1997;(68):2328–37.

41] Pike BR, Zhao X, Newcomb JK, Wang KK, Posmantur RM, HaRL. Temporal relationships between de novo protein synthesispain and caspase 3-like protease activation,and DNA fragmtion during apoptosis in septo-hippocampal cultures. J Neurosc1998;52(June (5)):505–20.

PR

O

NBA 6255 1–9

51] Siman R, Noszek JC. Excitatory amino acids activate caI and induce structural protein breakdown in vivo. Neu1988;(1):279–87.

52] Sloane JA, Hinman JD, Lubonia M, Hollander W, AbrahamAge-dependent myelin degeneration and proteolysis of oligodecyte proteins is associated with the activation of calpain-1 inrhesus monkey. J Neurochem 2003;84(January (1)):157–68.

53] Verkhratsky A, Kettenmann H. Calcium signaling in glial ceTrends Neurosci 1996;19(August (8)):346–52.

54] Verkhratsky A, Toescu EC. Calcium and neuronal ageing. T1998;(21):2–7.

55] Vicente JS, Munirathinam S, Bahr BA. Calpain-mediated spebreakdown identifies vulnerable brain regions at middle age. JBiol Biotech 2002;3:1–22.

56] Walker LC, Cork LC. Alzheimer’s disease: the neurobiology of agin nonhuman primates 1999; 233–243 [chapter 15].

57] Wang KKW. Calpain and caspase: can you tell the difference? TNeurosci 2000;(23):20–6.

58] Wang KKW, Posmantur R, Nath R, McGinnis K, Whitton M, TalaRV, et al. Simultaneous degradation of alphaII- and betaII-spectrcaspase 3 (CPP32) in apoptotic cells. J Biol Chem 1998;273(A(35)):22490–7.

59] Wang KKW. Principles of medical biology intracellular calciubinding proteins 1996;4:255–74.

60] Wang KKW, Yuen P. Development and therapeutic potential ofpain inhibitors. Adv Pharmacol 1997;(37):117–51.

61] Ward WF. Protein degradation in the aging organism. ProgSubcell Biol 2002;(29):35–42.

62] Zhang Y, Chong E, Herman B. Age-associated increases in thtivity of multiple caspases in Fisher 344 rat organs. Exp Gero2000;(37):777–89.