Apoptotic Changes in the Aged Brain Are Triggered by Interleukin-1 (cid:1) -induced Activation of p38 and Reversed by Treatment with Eicosapentaenoic Acid*

Among the several changes that occur in the aged brain is an increase in the concentration of the proinflammatory cytokine interleukin-1 (cid:1) that is coupled with a deterioration in cell function. This study investigated the possibility that treatment with the polyunsaturated fatty acid eicosapentaenoic acid might prevent interleu-kin-1 (cid:1) -induced deterioration in neuronal function. Assessment of four markers of apoptotic cell death, cytochrome c translocation, caspase-3 activation, poly-(ADP-ribose) polymerase cleavage, and terminal dUTP nick-end staining, revealed an age-related increase in each of these measures, and the evidence presented indicates that treatment of aged rats with eicosapentaenoate reversed these changes as well as the accompanying increases in interleukin-1 (cid:1) concentration and p38 activation. The data are consistent with the idea that activation of p38 plays a significant role in inducing the changes described since interleukin-1 (cid:1) -induced activation of cytochrome c translocation and caspase-3 activation in cortical tissue in vitro were reversed by the p38 inhibitor SB203580. The age-related increases in inter-leukin-1 (cid:1) 4- 22-month-old

Increased expression of the proinflammatory cytokine interleukin-1␤ (IL-1␤) 1 has been linked with neurodegenerative disorders like Down's syndrome, Alzheimer's disease, and Parkinson's disease (1,2). Consistent with the view that IL-1␤ plays a role in deterioration of cell function are the findings that IL-1␤ expression is increased, in parallel with cell damage, in experimental models of ischemia (3), excitotoxicity (4), and traumatic lesions (5). Indeed, IL-1␤ has been shown to trigger cell death in primary cultures of human fetal neurons (6) and inhibition of caspase-1, which leads to formation of active IL-1␤, and blocks lipopolysaccharide-induced changes in cell morphology, which are consistent with cell death (7). IL-1␤ has been shown to stimulate the mitogen-activated protein kinases p38 and c-Jun NH 2 -terminal kinase (8,9), and activation of both c-Jun NH 2 -terminal kinase (10,11) and p38 (10,(12)(13)(14)(15)(16) has been closely linked with apoptotic cell death. Significantly, an increase in p38 activity has been coupled with apoptotic changes in Alzheimer's disease (17,18). Concomitant increases in IL-1␤ concentration and p38 activity have been reported in the aged rat brain (19 -21); in hippocampus these changes are correlated with compromised synaptic function and with an age-related impairment in long term potentiation (LTP) (19 -22), while consistent with the high expression of IL-1␤ and IL-1RI in hippocampus is the finding that the cytokine depresses LTP in dentate gyrus (8,19,20,23,24). Significantly, we have recently reported that the age-related increases in IL-1␤ concentration and c-Jun NH 2 -terminal kinase activity, as well as the decrease in LTP, are reversed by treatment with the n-3 polyunsaturated fatty acid docosahexaenoic acid (22).
In this study we have attempted to identify the downstream consequences of the coupled age-related increases in IL-1␤ concentration and p38 activation in neuronal tissue. In particular, we have focused on assessing whether these changes might trigger apoptotic changes in neuronal tissue as it does in other tissues and have analyzed the effect of the ethyl ester of the -3 fatty acid eicosapentaenoic acid (EPA) on age-related changes in cortex and hippocampus. The data indicate that dietary manipulation reversed several changes in the aged cortex that are indicative of apoptotic cell death as well as age-related changes in IL-1␤ concentration, p38 activation, and LTP in hippocampus.

EXPERIMENTAL PROCEDURES
Animals-Groups of young and aged male Wistar rats (300 -350 g), maintained at an ambient temperature of 22-23°C under a 12-h lightdark schedule, were subdivided into those that were fed on a diet enriched in eicosapentaenoic acid (ethyl eicosapentaenoate, 10 mg/rat/ day for 3 weeks and 20 mg/rat/day for 5 weeks; Laxdale Research Ltd.) or standard laboratory chow for 8 weeks. Daily food intake was assessed for 2 weeks prior to commencement of the treatment: mean values (ϮS.E.) were 21.25 Ϯ 1.4 and 18.55 Ϯ 0.6 g/day for 4-and 22-month-old rats, respectively. At this time the mean body weights of young rats assigned to control and experimental groups were 265.6 Ϯ 9.1 and 250.2 Ϯ 11.7 g, respectively; corresponding values in aged rats were 483.6 Ϯ 9.8 and 481.2 Ϯ 7.9 g, respectively. Diet was prepared fresh each day, and rats were offered 100% of their daily intake. Mean daily food intake in all groups remained unchanged throughout the 8-week treatment period, and at the end of this time mean body weights of young rats assigned to control and experimental groups were 366.4 Ϯ 13.4 and 354.5 Ϯ 19.4 g, respectively; corresponding values in aged rats were 473.7 Ϯ 11.4 and 462.9 Ϯ 6.3 g, respectively. At this time rats were 4 and 22 months old. Rats were maintained under veterinary supervision for the duration of this experiment.
Induction of LTP in Perforant Path-Granule Cell Synapses in Vivo-At the end of the 8-week treatment, LTP was induced as described previously (19). Rats were anesthetized by intraperitoneal injection of urethane (1.5g/kg), recording and stimulating electrodes were placed in the molecular layer of the dentate gyrus (2.5 mm lateral and 3.9 mm posterior to bregma) and perforant path, respectively (angular bundle, 4.4 mm lateral to lambda), and stable baseline recordings were made before electrophysiological recording at test shock frequency (1/30 s) commenced for 10 min before and 40 min after tetanic stimulation (three trains of stimuli; 250 Hz for 200 ms; intertrain interval, 30 s). Rats were then killed by decapitation, and the hippocampus and cortex were removed, cross-chopped into slices (350 ϫ 350 m), and frozen separately in 1 ml of Krebs' solution (136 mM NaCl, 2.54 mM KCl, 1.18 mM KH 2 PO 4 , 1.18 mM MgSO 4 ⅐7H 2 O, 16 mM NaHCO 3 , 10 mM glucose, 1.13 mM CaCl 2 ) containing 10% dimethyl sulfoxide (22). For analysis, thawed slices were rinsed three times in fresh buffer and used as described below.
Analysis of IL-1␤ Concentration-IL-1␤ concentration was analyzed in homogenate prepared from cortex and hippocampus by enzymelinked immunosorbent assay (R&D Systems). Antibody-coated (100 l; 1.0 g/ml final concentration diluted in phosphate-buffered saline (PBS), pH 7.3; goat anti-rat IL-1␤ antibody) 96-well plates were incubated overnight at room temperature, washed several times with PBS containing 0.05% Tween 20, blocked for 1 h at room temperature with 300 l of blocking buffer (PBS, pH 7.3 containing 5% sucrose, 1% bovine serum albumin, and 0.05% NaN 3 ), and washed. IL-1␤ standards (100 l; 0 -1,000 pg/ml in PBS containing 1% bovine serum albumin) or samples (homogenized in Krebs' solution containing 2 mM CaCl 2 ) were added, and incubation proceeded for 2 h at room temperature. Secondary antibody (100 l; final concentration, 350 ng/ml in PBS containing 1% bovine serum albumin and 2% normal goat serum; biotinylated goat anti-rat IL-1␤ antibody) was added and incubated for 2 h at room temperature. Wells were washed, detection agent (100 l; horseradish peroxidase-conjugated streptavidin; 1:200 dilution in PBS containing 1% bovine serum albumin) was added, and incubation continued for 20 min at room temperature. Substrate solution (100 l; 1:1 mixture of H 2 O 2 and tetramethylbenzidine) was added and incubated at room temperature in the dark for 1 h after which time the reaction was stopped using 50 l of 1 M H 2 SO 4 . Absorbance was read at 450 nm, and values were corrected for protein (25) and expressed as pg of IL-1␤/mg of protein.
Analysis of p38 Phosphorylation, Cytochrome c Translocation, and PARP Cleavage-p38 phosphorylation was analyzed in samples of homogenate prepared from hippocampus and cortex; cytosolic cytochrome c and expression of 116-kDa PARP were analyzed in cortical tissue. p38 activity was also assessed in freshly prepared hippocampal and cortical tissue that was incubated for 20 min in the absence or presence of IL-1␤ (3.5 ng/ml), while cytochrome c translocation was assessed in vitro following incubation of slices of cortex with IL-1␤ (3.5 ng/ml) in the presence or absence of IL-1ra (350 ng/ml) or SB203580 (50 M). For analysis of p38 in all experiments and also for PARP, homogenate was diluted to equalize for protein concentration, and aliquots (10 l, 1 mg/ml) were added to 10 l of sample buffer (0.5 mM Tris-HCl, pH 6.8, 10% glycerol, 10% SDS, 5% ␤-mercaptoethanol, 0.05% (w/v) bromphenol blue), boiled for 5 min, and loaded onto 10% SDS gels. In the case of cytochrome c, cytosolic fractions were prepared by homogenizing slices of cortex in lysis buffer (20 mM HEPES, pH 7.4, 10 mM KCl, 1.5 mM MgCl 2 , 1 mM EGTA, 1 mM EDTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 5 g/ml pepstatin A, 2 g/ml leupeptin, 2 g/ml aprotinin), incubating for 20 min on ice, and centrifuging (15,000 ϫ g for 10 min at 4°C). The supernatant (i.e. cytosolic fraction) was suspended in sample buffer (150 mM Tris-HCl, pH 6.8, 10% (v/v) glycerol, 4% (w/v) SDS, 5% (v/v) ␤-mercaptoethanol, 0.002% (w/v) bromphenol blue) to a final concentration of 300 g/ml, boiled for 3 min, and loaded (6 g/lane) onto 12% gels. In all cases proteins were separated by application of 30 mA constant current for 25-30 min, transferred onto nitrocellulose strips (225 mA for 75 min), and immunoblotted with the appropriate primary and secondary antibodies. In the case of p38, anti-phospho-p38 (Santa Cruz Biotechnology; 1:500 in phosphate-buffered saline-Tween (0.1% Tween 20) containing 2% nonfat dried milk) and peroxidaselinked anti-mouse IgM (1:1,000; Amersham Biosciences) were used. In the case of PARP, we immunoblotted with an antibody (1:2,000) raised against the epitope corresponding to amino acids 764 -1014 of poly-(ADP-ribose) polymerase of human origin (Santa Cruz Biotechnology), and immunoreactive bands were detected using peroxidase-conjugated anti-rabbit IgG (Sigma) and ECL (Amersham Biosciences). To assess cytochrome c, a rabbit polyclonal antibody raised against recombinant protein corresponding to amino acids 1-104 of cytochrome c (Santa Cruz Biotechnology) was used. In addition to loading equal amounts of protein, some blots were reprobed for analysis of total (rather than phosphorylated) p38, and in other cases blots were probed with an anti-actin FIG. 1. The age-related increases in IL-1␤ concentration and p38 activity in cortex are abolished by EPA. a, IL-1␤ concentration was significantly enhanced in cortical tissue prepared from aged rats fed on the control diet (n ϭ 10) compared with young rats fed on either diet (*, p Ͻ 0.05, ANOVA; n ϭ 6), but this change was not evident in tissue prepared from aged rats fed on the EPA-enriched diet (n ϭ 10). b, p38 activity was significantly enhanced in cortical tissue prepared from aged rats fed on the control diet (lane 3) compared with young rats fed on either control (lane 1) or EPA (lane 2) diet (*, p Ͻ 0.05, ANOVA), but this change was not evident in tissue prepared from aged rats fed on the EPA-enriched diet (lane 4). c, IL-1␤ (lane 2) significantly enhanced p38 activity in cortical tissue in vitro (*, p Ͻ 0.05, Student's t test for paired values; n ϭ 6). Con, control.
antibody to confirm equal loading. Data from these experiments showed no differences between treatment groups. In all experiments, immunoreactive bands were detected using peroxidase-conjugated anti-rabbit antibody (Sigma) and ECL (Amersham Biosciences). Quantification of protein bands was achieved by densitometric analysis using two software packages, Grab It (Grab It Annotating Grabber, Version 2.04.7, Synotics, UVP Ltd.) and Gelworks (Gelworks ID, Version 2.51, UVP Ltd.) for photography and densitometry, respectively.
Analysis of Caspase-3 Activity-Slices of cortical tissue prepared from young and aged rats were washed, homogenized in lysis buffer (400 l; 25 mM HEPES, 5 mM MgCl 2 , 5 mM dithiothreitol, 5 mM EDTA, 2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 10 g/ml pepstatin, pH 7.4), incubated on ice for 20 min, analyzed for protein concentration, and diluted to equalize for protein concentration. In some experiments, slices prepared from control young rats were incubated for 60 min at 37°C in the presence or absence of IL-1␤ (3.5 ng/ml) to which IL-1ra (350 ng/ml) or SB203580 (50 M) was added; these samples were treated as described above. All samples (98 l) were added to 2 l of caspase-3 substrate (Ac-DEVD-AFC peptide, Alexis Corp.; 5 M), transferred to a 96-well plate, and incubated for 1 h at 37°C. Fluorescence was assessed (excitation, 390 nm; emission, 510 nm), and enzyme activity was calculated with reference to a standard curve of AFC (0 -10 M) concentration versus absorbance.
Analysis of Caspase-3 mRNA-Total RNA was extracted from cortical neurons (26) using TRI reagent (Sigma). cDNA synthesis was performed on 1 g of total RNA using oligo(dT) primer (Superscript reverse transcriptase, Invitrogen). Equal amounts of cDNA were used for PCR amplification for a total of 30 cycles at 94°C for 1 min and 58°C for 2 min. A final extension step was carried out at 70°C for 10 min. Multiplex PCR was performed using the Quantitative PCR Cytopress Detection kit (Rat Apoptosis Set 2; BioSource International, Camarillo, CA) generating caspase-3 PCR products of 320 bp and glyceraldehyde-3phosphate dehydrogenase PCR products of 532 bp. The PCR products were analyzed by electrophoresis on 2% agarose gels, photographed, and quantified using densitometry. Expression of glyceraldehyde-3phosphate dehydrogenase mRNA was used as a standard to quantify the relative expression of caspase-3 mRNA.
Preparation of Dissociated Cells and Colocalization of p38 and Caspase-3-Cortical slices (350 ϫ 350 m) were incubated at 37°C for 30 min in HEPES-buffered Krebs' solution (145 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 2 mM CaCl 2 , 1 mM Mg 2 SO 4 , 1 mM KH 2 PO 4 , 10 mM glucose, 30 mM HEPES at pH 7.4) containing trypsin (1 mg/ml), DNase (1,600 kilounits/liter), protease X (1 mg/ml), and protease XIV (1 mg/ml). Slices were washed, triturated, and passed through a nylon mesh filter. Cells were centrifuged (1,000 rpm for 1 min), resuspended in HEPES-buffered Krebs' solution, plated onto coverslips, fixed in 4% paraformaldehyde in PBS (w/v) for 30 min, and permeabilized in 0.1% Triton in PBS (v/v) for 15 min. Cells were incubated in normal goat serum in PBS (v/v) to block nonspecific binding, treated with anti-phosphospecific p38 antibody (1:100; Santa Cruz Biotechnology) or anti-caspase-3 (1:500; Bio-Source), and incubated overnight at 4°C. Cells were washed and incubated in the dark for 2 h in either fluorescein isothiocyanate-labeled goat anti-mouse IgG and IgM (1:100; BioSource) or R-phycoerythrinlabeled goat anti-rabbit IgG (1:100; BioSource) to visualize labeling with p38 and caspase-3, respectively. Following a further wash, slides were mounted using 2 mg/ml p-phenylenediamine in 50% glycerol in PBS (v/v) and sealed. Fluorescence was analyzed using the Bio-Rad MRC-1024 laser scanning confocal imaging system in which the fluorochromes were excited by laser light emitted at 565 and 494 nm and detected at 578 and 520 nm, which measured bound R-phycoerythrin and fluorescein isothiocyanate, respectively. Cells were analyzed at ϫ63 magnification under oil immersion with the laser at 100% power. The images were analyzed using the Bio-Rad software, and the Kalman filter was used to decrease background. In this system R-phycoerythrinlabeled cells are stained red, and fluorescein isothiocyanate-labeled cells are stained green. In a separate series of experiments, groups of aged and young rats were killed and brains were rapidly removed, coated in OCT compound, immersed in an isopentane bath over liquid nitrogen, and used to prepare sections for analysis of phosphorylated p38 as described above.
TUNEL Staining-Apoptotic cell death was assessed using the DeadEnd colorimetric apoptosis detection system (Promega). Cells were permeabilized and fixed in paraformaldehyde as described above. In separate experiments cultured cortical neurons were prepared from neonatal rats as described previously (26) and maintained in neurobasal medium for 12 days before incubating in the absence or presence of IL-1␤ (5 ng/ml) for 72 h. Biotinylated nucleotide was incorporated at 3Ј-OH DNA ends by incubating cells with terminal deoxynucleotidyltransferase for 30 min at 37°C. Washed cells were incubated in horseradish peroxidase-labeled streptavidin and then incubated in 3,3Ј-diaminobenzidine chromogen solution, and TUNEL-positive cells were calculated as a proportion of the total cell number.

RESULTS
IL-1␤ concentration and p38 activity were both significantly increased in cortical tissue prepared from aged rats fed on the control diet compared with young rats (p Ͻ 0.05, ANOVA; Fig. 1,  a and b), but EPA suppressed these age-related changes so that the values in tissue prepared from EPA-treated rats were not significantly different from control values. In vitro analysis revealed that IL-1␤ significantly enhanced p38 activity in cortical tissue (Fig. 1c). In parallel with this observation, we found that cytochrome c translocation was significantly increased in cortical tissue prepared from aged rats fed on the control diet compared with tissue prepared from either group of young rats (p Ͻ 0.05, ANOVA; Fig. 2a). A causal relationship between the age-related changes in cytochrome c translocation and IL-1␤ is suggested by the finding that cytochrome c translocation was significantly enhanced by IL-1␤ (p Ͻ 0.05, ANOVA; Fig. 2b) and that this action relied on IL-1RI activation since the IL-1␤-induced change was inhibited by IL-1ra. Fig. 2b also demonstrates that the IL-1␤-induced change was inhibited by SB203580 suggesting that the effect was mediated by activation of p38. One downstream consequence of cytochrome c translocation is activation of caspase-3, therefore we analyzed enzyme activity in tissue prepared from aged and young rats fed on either the control or experimental diet. Fig. 3a demonstrates that there was a significant age-related increase in caspase-3 activity; thus the mean value in cortical tissue prepared from aged rats fed on the control diet was significantly increased compared with the value in tissue prepared from young rats (p Ͻ 0.05, ANOVA), but EPA suppressed this change. In an effort to establish whether the change in caspase-3 activation was coupled with the increases in IL-1␤ concentration and p38 activation, a series of in vitro experiments were undertaken that revealed that IL-1␤ significantly enhanced caspase-3 activity in cortical tissue (p Ͻ 0.05, ANOVA; Fig. 3b); this effect was blocked by IL-1ra, suggesting that the action of IL-1␤ was dependent on receptor interaction, and by SB203580, indicating that the action of IL-1␤ also required activation of p38. Cells prepared from cortex of young and aged rats fed on both diets were stained for phosphorylated p38 and caspase-3, and staining was assessed using confocal microscopy. We found no cell in preparations obtained from young rats in which there was evidence of colocalization of activated p38 and caspase-3. In contrast, several cells prepared from aged rats fed on the control diet stained positively for both; an example is shown in Fig. 3d. However, while some cells prepared from aged EPAtreated rats stained positively for p38, there was little staining for caspase-3, and we found no evidence of colocalization. These findings support the idea that caspase-3 activation is closely coupled with p38 activation. In addition to the stimulatory effect of IL-1␤ on caspase-3 activity, we observed that IL-1␤ significantly increased caspase-3 mRNA in cultured cortical cells (Fig. 3c).
In an effort to consolidate these findings, which suggested

FIG. 3. The age-related increase in caspase-3 activity is mimicked by IL-1␤
: evidence for a role for p38 activation. a, caspase-3 activity was significantly enhanced in cortical tissue prepared from aged rats fed on the control diet compared with young rats fed on either diet (*, p Ͻ 0.05, ANOVA), but this change was not evident in tissue prepared from aged rats fed on the EPA-enriched diet. b, in vitro analysis indicated that IL-1␤ significantly increased enzyme activity (*, p Ͻ 0.05, ANOVA; n ϭ 6) but that the IL-1␤-induced effect was inhibited by IL-1ra and also by SB203580. c, incubation of cultured cortical neurons in IL-1␤ (lane 2) for 72 h significantly increased caspase-3 mRNA (*, p Ͻ 0.05, Student's t test for paired means); values were normalized with reference to expression of glyceraldehyde-3-phosphate dehydrogenase (lower bands). d, colocalization of activated p38 and caspase-3 was observed in several cells obtained from aged rats that were fed on the control diet but in none of the cells obtained from young rats fed on either diet. There was some evidence of p38 staining in aged rats fed on the experimental diet but no evidence of colocalization with caspase-3. Con, control; SB, SB203580. that there was significant evidence of cell death in the aged cortex, we investigated cleavage of PARP by assessing expression of the 116-kDa form of the enzyme. Fig. 4 indicates, in a sample immunoblot and by analysis of the mean data obtained from densitometric analysis, that there was a significant decrease in expression of 116-kDa PARP in cortical tissue prepared from aged rats fed on the control diet compared with young rats (p Ͻ 0.05, ANOVA) but that this effect was reversed in tissue prepared from aged rats treated with EPA. These data were paralleled by changes in TUNEL staining; thus a significantly greater number of cells stained positively for TUNEL in preparations obtained from aged rats fed on the control diet compared with young rats fed on either diet (p Ͻ 0.05, ANOVA; Fig. 5a). The number of TUNEL-positive cells in preparations obtained from aged rats fed on the EPA-enriched diet was similar to that in tissue prepared from young rats. Fig. 5b shows that exposure of cultured cortical neurons to IL-1␤ for 72 h also significantly increased the number of TUNEL-positive cells (p Ͻ 0.05, Student's t test for independent means).
In an effort to explore the synaptic changes that might occur as a consequence of IL-1␤-induced cell death, we turned to analysis of changes in hippocampus and first assessed age-and diet-related changes in IL-1␤ concentration and p38 activation. Fig. 6a shows that, in parallel with the age-related findings in cortical tissue, IL-1␤ concentration was significantly increased in hippocampal tissue prepared from aged rats fed on the control diet compared with young rats fed on either diet (p Ͻ 0.05, ANOVA). There was no evidence of a similar age-related increase in aged rats fed on the EPA-enriched diet. Analysis of p38 activation revealed a similar pattern; thus there was a significant age-related increase in p38 activity (p Ͻ 0.05, ANOVA, aged rats fed on the control diet versus young rats; Fig. 6b), which was not observed in tissue prepared from aged rats fed on the EPA-enriched diet. A likely causal relationship between IL-1␤ concentration and p38 activation is suggested by the finding that IL-1␤ significantly increased p38 activity in hippocampus in vitro (Fig. 6c). In a separate series of experiments, in which no dietary manipulation was made, cryostat sections of tissue were prepared from young and aged rats. We observed that there was a marked increase in p38 staining in hippocampus of aged compared with young rats; sample sections are demonstrated in Fig. 6d.
Analysis of LTP in dentate gyrus was undertaken in the same rats in which biochemical analyses were performed. Fig.  7 demonstrates that LTP was successfully induced in both groups of young rats (Fig. 7a) but was impaired in aged rats fed on the control diet (Fig. 7b); in the latter group of rats the mean percent changes in population excitatory postsynaptic potential slope in the 2 min immediately following tetanic stimulation (Fig. 7c) and in the last 5 min of the experiment (Fig. 7d; compared with the mean value in the 5 min prior to the tetani) were 136.5 Ϯ 3.92 and 111.8 Ϯ 0.86, respectively. In contrast, the corresponding values in aged rats fed on the EPA-enriched diet were 181.3 Ϯ 3.12 and 149.1 Ϯ 0.82, respectively. These latter values were similar to those observed in young rats fed on the control (190.5 Ϯ 4.83 and 147.48 Ϯ 1.31, respectively) and EPA-enriched (184.0 Ϯ 3.24 and 154.8 Ϯ 0.92, respectively) diets, indicating that EPA treatment restored the ability of aged rats to respond to tetanic stimulation. DISCUSSION We present evidence demonstrating that treatment with EPA prevents neuronal cell death in aged rats and show that suppression of the age-related coupled increases in IL-1␤ concentration and p38 activity is the key to inhibiting the cascade of cellular events that leads to apoptosis. The age-related increase in IL-1␤ concentration in cortical and hippocampal tissue, which confirms previous findings (19 -21), is coupled with increased TUNEL staining in vivo. The implicit suggestion that endogenous IL-1␤ induces cell death is supported by the finding that IL-1␤ also enhances TUNEL staining in cultured cortical cells, although a comparison of data from two such different experimental conditions must be made with caution. Significantly, treatment with EPA, which has been shown to have anti-inflammatory properties (27,28), prevented the age-related increases in IL-1␤ concentration and TUNEL staining in cortex. These data support the previous finding that fish oils, which contain EPA, reduce production of proinflammatory cytokines in circulating cells (29 -32) and chondrocytes (33) and suppress the lipopolysaccharide-induced increase in circulating IL-1␤ (34). EPA and mimicked by IL-1␤. a, the number of cells that stained positively for TUNEL was significantly enhanced in cortex of aged rats fed on the control diet (e.g. left-hand picture) compared with that in young rats fed on either diet (*, p Ͻ 0.05, ANOVA); dietary manipulation with EPA reversed this effect. b, incubation of cultured cortical neurons in the presence of IL-1␤ significantly increased the number of TUNEL-positive cells (*, p Ͻ 0.05, Student's t test for independent means; n ϭ 5). Con, control; ϩve, positive.

FIG. 5. The increase in TUNEL staining with age is blocked by
Increased activation of p38 accompanied the age-related increase in IL-1␤ concentration consistent with previous observations in hippocampal cells (8, 19 -21) and in other cell types (35)(36)(37)(38)(39). The evidence presented here pinpoints IL-1␤-induced increased activation of p38 as a pivotal event in triggering changes that are characteristic of apoptotic cell death, for example cytochrome c translocation and caspase-3 activation. This suggestion concurs with previous findings. Thus inhibition of p38 by SB203580 has been shown to prevent singlet oxygen-induced apoptosis in HL-60 cells (40), while another p38 inhibitor, SB2390663, prevents caspase cleavage in thioloxidant-induced apoptosis in forebrain neuronal-enriched cell cultures (41). That inhibition of p38 proffers protection is further supported by the finding that its activation spared dopa-minergic neurons deprived of serum (42) and markedly reduced infarct size induced by ischemia (16). In parallel with the effect of dietary manipulation on IL-1␤ concentration, the data indicate that EPA treatment blocked the age-related increase in p38 activation in hippocampus and cortex.
The data from in vitro analysis suggested that IL-1␤, through an action on IL-1RI and mediated through activation of p38, stimulates cytochrome c translocation in cortical tissue. Since IL-1␤ concentration and p38 activation were enhanced in cortical tissue prepared from aged rats, it was predicted that translocation of cytochrome c might also be a feature of the aged brain; the present data indicate that there was an agerelated increase in cytochrome c translocation that paralleled the increases in IL-1␤ concentration and p38 activation. Sig-FIG. 6. The age-related increases in IL-1␤ concentration and p38 activity in hippocampus are abolished by EPA. a, IL-1␤ concentration was significantly enhanced in hippocampal tissue prepared from aged rats fed on the control diet (n ϭ 10) compared with young rats fed on either diet (*, p Ͻ 0.05, ANOVA; n ϭ 6), but this change was not evident in tissue prepared from aged rats fed on the EPA-enriched diet (n ϭ 10). b, p38 activity was significantly enhanced in hippocampal tissue prepared from aged rats fed on the control diet compared with young rats fed on either control or EPA diet (*, p Ͻ 0.05, ANOVA), but this change was not evident in tissue prepared from aged rats fed on the EPA-enriched diet (lane 4). c, IL-1␤ significantly enhanced p38 activity in hippocampal tissue in vitro (*, p Ͻ 0.05, Student's t test for paired values; n ϭ 6). d, expression of phosphorylated p38 is markedly increased in cryostat sections prepared from aged compared with young rats. Con, control. nificantly, this increase was absent in cortical tissue prepared from EPA-treated aged rats. Disruption of the mitochondrial transmembrane potential, which is accompanied by cytochrome c translocation from the mitochondria to the cytosol (43), is a relatively early event in apoptotic cell death. It has been shown that among the stimuli that lead to these events is oxidative stress (43); indeed it is possible that enhanced reactive oxygen species accumulation, which we have shown to accompany increased IL-1␤ concentration in other studies (19)(20)(21) and in the present one (data not shown), is responsible for the observed age-related increase in cytochrome c translocation.
Cytochrome c translocation triggers a cascade of reactions initiated by interaction with apoptosis protease-activating factor-1 and propagated by activation of caspase-9 and subse-quently other caspases that eventually culminates in cell death (44). The importance of the role of cytochrome c translocation in this cascade was underscored by the demonstration that injection of active cytochrome c into cells led to apoptosis (45). The present findings, which show parallel changes in cytochrome c translocation and caspase-3, are consistent with the view that caspase-3 activation is triggered by cytosolic cytochrome c (46,47). Our data also indicate that caspase-3 mRNA, as well as enzyme activity, was increased by IL-1␤ in vitro and, furthermore, that caspase-3 activation was dependent on interaction of IL-1␤ with IL-1RI and mediated through the subsequent activation of p38. The observation that caspase-3 colocalized with activated p38 was a further indication of a coupling between these parameters. This finding is strikingly similar to data reported in a recent study that showed that SB203580 diminishes caspase activity and protects SH-Sy5Y neuroblastoma cells and cultured cortical neurons from NO-induced cell death (48). EPA reversed the age-related increases in p38 activation, cytochrome c translocation, and caspase-3 activity, and we view this concurrence as strong evidence of a causal relationship between the parameters.
One downstream substrate for caspase-3 is the DNA repair enzyme PARP, and the present findings indicate that, in parallel with the effects of age and diet on caspase-3 activation, PARP cleavage was increased in aged rats fed on the control, but not the EPA-enriched, diet. We have previously reported that cleavage of PARP was increased in entorhinal cortex of lipopolysaccharide-treated rats, and this change was accompanied by changes in morphology of cells that were indicative of cell death (20). Indeed cleavage of PARP has been considered to be a reliable marked of apoptosis (49,50). Significantly, we observed that these lipopolysaccharide-induced changes were blocked by caspase-1 inhibition pinpointing a role for IL-1␤ in the cascade of events leading to cell death (8). In the present study the age-related enhancement in PARP cleavage was coupled with evidence of reduced cell viability, but evidence of both changes was absent in tissue prepared from EPA-treated aged rats.
Our data indicate that the age-related increases in IL-1␤ concentration and p38 activation observed in cortical tissue were also observed in the hippocampus; these findings support our previous observations (19,20). However, we report that no such changes were observed in hippocampus of aged rats that were fed on the EPA-enriched diet. LTP in perforant pathgranule cell synapses was markedly depressed with age, confirming data from several previous studies (19 -22, 51); significantly, this age-related impairment in LTP was completely absent in aged rats fed on the EPA-enriched diet. The negative correlation between the IL-1␤ concentration and p38 activation and the expression of LTP together with the observation that the IL-1␤-induced inhibition of LTP was suppressed by p38 inhibition provides strong evidence of a causal relationship between these measures. The question of how polyunsaturated fatty acid, specifically EPA, uptake into neuronal tissue is achieved should be considered. It has been shown that circulating fatty acids cross the blood-brain barrier, and the rate of incorporation is proportional to plasma concentration; evidence indicates that transport is mainly by diffusion, although a facilitated process may also contribute (52). The question of the underlying cause of the decrease in polyunsaturated fatty acids in aged rats remains to be fully resolved, but it appear that fatty acid uptake into brain tissue is not altered with age (53).
Our working hypothesis is that aging is coupled with a significant increase in IL-1␤ concentration in neuronal tissue that is likely to exert multiple effects including activation of p38. We propose that increased p38 causes mitochondrial mem-FIG. 7. The age-related impairment in LTP in dentate gyrus is suppressed by EPA. a and b, tetanic stimulation (time 0) resulted in an immediate and sustained increase in the mean population excitatory postsynaptic potential (Epsp) slope in young rats (a) fed on either diet. A similar change was observed in aged rats (b) fed on the EPA-enriched diet, but aged rats that were fed on the control diet exhibited a marked attenuation in response to tetanic stimulation. c and d, analysis of the mean changes in the 2 min immediately following tetanic stimulation and in the last 5 min of the experiment (compared with the mean excitatory postsynaptic potential slope in the 5 min preceding the tetanus) revealed a significant decrease in excitatory postsynaptic potential slope in aged rats fed on the control diet compared with the other three groups (*, p Ͻ 0.01, ANOVA; n ϭ 8 in the case of young and n ϭ 12 in the case of aged rats). Con, control. brane perturbation leading to translocation of cytochrome c. One consequence of these changes is an increase in caspase-3 activation; caspase-3 acts on its substrate, PARP, resulting in its cleavage. Apoptotic changes in cells occur since DNA repair is compromised as a result of this action. The evidence presented points to a pivotal role for IL-1␤ in triggering the cellular events that lead to an increase in cell death in the aged brain and identify activation of p38 as a key mediator. It is proposed that, although EPA abolished several age-related changes, the primary action of EPA may be to block the agerelated increase in IL-1␤ concentration.