Neurotoxic Mechanisms Caused by the Alzheimer’s Disease-linked Swedish Amyloid Precursor Protein Mutation OXIDATIVE STRESS, CASPASES, AND THE JNK PATHWAY*

Autosomal dominant forms of familial Alzheimer’s disease (FAD) are caused by mutations of the amyloid precursor protein (APP) gene and by mutations of the genes encoding for presenilin 1 or presenilin 2. Simulta-neously, evidence is provided that increased oxidative stress might play a crucial role in the rapid progression of the Swedish FAD. Here we investigated the effect of the Swedish double mutation (K670M/N671L) in the (cid:1) -amyloid precursor protein on oxidative stress-in-duced cell death mechanisms in PC12 cells. Western blot analysis and cleavage studies of caspase substrates revealed an elevated activity of the executor caspase 3 after treatment with hydrogen peroxide in cells containing the Swedish APP mutation. This elevated activity is the result of the enhanced activation of both intrinsic and extrinsic apoptosis pathways, including activation of caspase 2 and caspase 8. Furthermore, we observed an enhanced activation of JNK pathway and an attenua-tion of apoptosis by SP600125, a JNK inhibitor, through protection of mitochondrial dysfunction and reduction of caspase 9 activity. Our findings provide evidence that the massive neurodegeneration in early age of FAD patients could be a result of an increased vulnerability of neurons through activation of different apoptotic pathways Transfection—

Alzheimer's disease is the most frequent neurodegenerative disorder in the aged population and is characterized by the presence of senile plaques and neurofibrillary tangles in the brain. Senile plaques are deposits of amyloid-␤-protein (A␤), 1 which arises by proteolysis from the amyloid precursor protein (APP) through an initial ␤-secretase cleavage by the aspartyl protease beta-site APP-cleaving enzyme (1, 2) followed by an intramembranous cut of ␥-secretase, which resides in a high molecular weight complex composed of presenilin (3), nicastrin (4 -6), PEN-2 (5), and probably Aph-1 (6).
Despite all of these findings, the underlying mechanisms responsible for the massive neurodegeneration in early age of FAD patients are still not completely understood. Persuasive evidence indicates that oxidative stress plays an important role in the neuropathological process in AD (24 -27). AD-related mutations probably enhance oxidative stress. Increased oxidative stress levels have been found in the temporal inferior cortex from Swedish FAD patients (28), and brains of mice transgenic for human presenilin 1 show reduced antioxidative enzyme activity (29).
Moreover, convincing evidence indicates that A␤ is neurotoxic, probably via an apoptotic pathway (12, 30 -34). A␤ is believed to play a major role in promoting neuronal degeneration and death by rendering neurons vulnerable to age-related increases in levels of oxidative stress and impairments in cellular energy metabolism (32,35). New evidence indicates that reactive oxygen species-induced cellular events implicate the activation of mitogen-activated protein kinases (36,37). Oxidative stress may cause the activation of c-Jun N-terminal kinase (JNK; also known as stress-activated protein kinase) in degenerating neurons in AD (38). An involvement of JNK pathway by the induction of Fas ligand in A␤-induced neuronal apoptosis was described by Morishima et al. (39). Activation of JNK and p38 associated with amyloid deposition was observed in mice transgenic for the human APPsw mutation (40). The inhibition of the JNK pathway was proposed as a potential therapeutic target in AD (41).
Caspases can be divided into initiator caspases and effector caspases based on the presence of a large prodomain at their amino-terminal region. Initiator caspases containing a long prodomain, like caspase 2, caspase 8, caspase 9, and caspase 10 generally act in early stages of a proteolytic cascade, whereas effector caspases, like caspase 3, caspase 6, and caspase 7 act downstream and are involved in the cleavage of specific cellular proteins (for a review, see Refs. 42 and 43). Among the identified caspases, caspase 3 is of particular interest, since it appears to be very important in the progression of AD (12,44,45). In addition to caspase 3, other caspases have been associated with AD (46 -49).
Here we report on the effects of FAD-related APP mutation, using stably transfected PC12 (rat pheochromocytoma) cells expressing the Swedish mutant APP, wild-type APP, or empty vector. We examined the effect of hydrogen peroxide on the activation of different caspases and the JNK pathway to elucidate the relationship between mutant APP expression, increased A␤ production, and neuronal cell death. Furthermore, we investigated the protective effects of caspase inhibitors as well as JNK inhibitor in preventing oxidative stress-mediated cell death.

EXPERIMENTAL PROCEDURES
Materials-Phospho-JNK (Thr 183 /Tyr 185 ) and phospho-c-Jun (Ser 63 ) antibodies were purchased from Cell Signaling Technology, and cytochrome c antibody was from BD Biosciences.
Cell Culture and Transfection-The generation and characterization of PC12 cell lines expressing human APPwt-, mutant APP (APPsw)-, and vector-transfected control clones have been described previously (12). In brief, PC12 cells were transfected with DNA constructs harboring human mutant (APPsw; K670M/N671L) or wild-type APP (APPwt) gene, inserted downstream of a cytomegalovirus promotor, using the FUGENE technique (Roche Applied Science). The transfected cell lines APPwt PC12, APPsw PC12, and vector-transfected PC12 were cultured in Dulbecco's modified Eagle's medium supplemented with 10% heatinactivated fetal calf serum, 5% heat-inactivated horse serum, 50 units/ml penicillin, 50 g/ml streptomycin, and 400 g/ml G418 at 37°C in a humidified incubator containing 5% CO 2 . The stably transfected clones (APPwt M5, N10, and U7 and APPsw Q8 and Q9) were selected for the present study based on their similar expression of APP695 and the 5-fold increased secretion of A␤-  in the APPsw clones ( Fig. 1). To rule out potential influences on the APP expression levels, APPsw and APPwt clones were routinely screened in parallel to the experiments presented in this work. No alterations in the expression of APP were found.
Detection of A␤-For the detection of secreted A␤-(1-40), we used a specific sandwich enzyme-linked immunosorbent assay employing monoclonal antibodies as described previously (50).
Measurement of Caspase Activity-PC12 cells were plated the day before at a density of 5 ϫ 10 6 cells/culture dish (10 cm diameter). After treatment with H 2 O 2 , the cells were harvested, and after a centrifugation step, the culture medium was aspirated. The cell-containing pellet was washed with PBS and lysed in 100 l of lysis buffer (10 mM HEPES, 0.1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 0.1 mM EDTA, 1 mM dithiothreitol, 1 g/ml pepstatin A, 1 g/ml leupeptin, 0.1% CHAPS, pH 7.4). Lysates were centrifuged at 20,000 ϫ g for 10 min at 4°C, and the supernatant was used for caspase assay. The approximate protein amount of the probes was confirmed by a Lowry protein test. Caspase activity was measured by cleavage of colorimetric substrates (Ac-DEVD-pNA, Ac-VEID-pNA, Ac-VDVAD-pNA, Ac-IETD-pNA, and Ac-LEHD-pNA for caspase 3, caspase 6, caspase 2, caspase 8, and caspase 9, respectively). The production of pNA was monitored over 30 min in a photometer ( ϭ 405 nm). The caspase activity is expressed as change in absorption units. One unit was defined as the amount of enzyme required to cleave 1 pmol of NA per min of incubation per 5 ϫ 10 6 cells.
Quantification of Apoptosis by Flow Cytometry-Apoptosis was determined by propidium iodide (PI) staining and fluorescence-activated cell sorting analysis as described previously (12). Briefly, H 2 O 2 -treated PC12 cells were lysed in buffer (0.1% sodium citrate and 0.1% Triton X-100) containing 50 g/ml propidium iodide. Samples were analyzed by flow cytometry (FACSCalibur) using Cell Quest software (Becton Dickinson). Cells with a lower DNA content showing less propidium iodide staining than G 1 have been defined as apoptotic cells (sub-G 1 peak).
Phospho-JNK Pathway Western Blot Analysis-PC12 cells were plated at a density of 5 ϫ 10 6 cells/culture dish (10 cm diameter) and treated with H 2 O 2 . Cells were harvested and lysed in SDS sample buffer: Tris-HCl (62.5 mM, pH 6.8), 2% SDS, 10% glycerol, 50 mM dithiothreitol, 0.01% bromphenol blue. After sonicating and boiling, proteins were separated by electrophoresis on a 16% polyacrylamide gel. The separated proteins were transferred to a polyvinylidene difluoride membrane (ImmobilonTM-P; Millipore). These membranes were stained with Ponceau S red (reversible stain) to visualize the proteins. Nonspecific sites on the membranes were blocked with 5% fat-free milk in Tris-buffered saline, 0.1% Tween 20 (TBST). The phosphorylated JNK was visualized with the appropriate primary phosphoantibody, phospho-stress-activated protein kinase/JNK (Thr 183 /Tyr 185 ), which detects stress-activated protein kinase/JNK only dually phosphorylated at Thr 183 /Tyr 185 . The phosphorylated c-Jun was detected by a phosphoc-Jun (Ser 63 ) antibody. After thorough washing with TBST, membranes were covered with ECL TM detection reagents and quickly exposed to an autoradiography film. Membranes were routinely stripped for actin control.
Caspase Immunoblot Analysis-PC12 cells were plated at a density of 5 ϫ 10 6 cells/culture dish (10 cm diameter) and treated with H 2 O 2 . After the indicated periods of time, the medium was aspirated, and cells were processed for Western blotting with rat-specific cleaved caspase 3, caspase 8, and caspase 9 antibodies. Blots were scanned (Umax Astra 4000 U), and the band intensity was determined after background subtraction using densitometric analysis.
Determination of Cytochrome c Release-The amount of cytochrome c released from the mitochondrial intermembrane space into the cytosol was determined by digitonin permeabilization (51). Briefly, 5 ϫ 10 6 cells were exposed to oxidative stress for different periods of time. After washing with ice-cold phosphate-buffered saline, cells were resuspended in permeabilization buffer containing 75 mM NaCl, 1 mM NaH 2 PO 4 , 8 mM Na 2 PO 4 , 250 mM sucrose, 1 mM phenylmethylsulfonyl fluoride, additional protease inhibitors, and 0.05% digitonin. Following a centrifugation step at 800 ϫ g at 4°C for 10 min, the supernatant was separated from the pellet consisting of mitochondria and cellular debris. The supernatant containing cytoplasmic proteins was purified by centrifugation at 13,000 ϫ g at 4°C for 10 min. Equal amounts of protein (10 g) were loaded on an 18% acrylamide gel and separated by SDS-PAGE. After immunoblotting with a monoclonal cytochrome c antibody, polyvinylidene difluoride membranes were stripped and reprobed either for COX4 or actin, ensuring equal protein loading and the absence of mitochondrial contamination.
Determination of Mitochondrial Membrane Potential-PC12 cells were plated the day before at a density of 2 ϫ 10 5 cells/well in a 24-well plate. The cells were pretreated for 1 h with the JNK inhibitor SP600125, and H 2 O 2 was added for 6 h. The membrane potential of the inner mitochondrial membrane was measured using the dye rhodamine 123. The dye was added to the cell culture medium at a concentration of 0.4 M for 15 min. The cells were washed twice with Hanks' balanced salt solution, and the fluorescence was determined with a fluorescence reader (Victor Multilabel counter; PerkinElmer Life Sciences). Transmembrane distribution of the dye depends on the mitochondrial membrane potential (⌬⌿ m ).
Statistical Analysis-Data are given as mean Ϯ S.E. For statistical comparison, paired t test, Student's t test, or one-way ANOVA followed by Tukey's post hoc test or two-way ANOVA were used. p values less than 0.05 were considered statistically significant.

RESULTS
Generation and Characterization of PC12 Cell Lines Overexpressing APPwt and APPsw-To investigate the effects of APP overexpression in oxidative stress-induced apoptosis, we used several clones of PC12 cells stably transfected with APP cDNAs. Similar expression of human APP in the investigated cell lines was confirmed by Western blotting. As expected, the human-specific antibody does not recognize any APP species in vector-transfected PC12 cells (Fig. 1A), whereas equal expression levels of human APP were detected in clones of APPwt cells and in clones of cells containing the Swedish mutation. Culture supernatants of the APP-transfected cells were also analyzed for A␤-(1-40) production revealing 4 -5-fold elevated A␤-(1-40) levels in clones of APPsw cells compared with clones of APPwt cells (Fig. 1B).
APPsw Mutation Leads to an Enhanced Vulnerability to Oxidative Stress-induced Apoptosis-We induced apoptosis in transfected PC12 cells using the oxidative stressor hydrogen peroxide. Nuclear DNA fragmentation was quantitatively detected by propidium iodide staining and flow cytometry. Very importantly, clones bearing the APPsw mutation are sensitized to an exposure to oxidative stress for 24 h in a concentrationdependent manner with a maximum increase of apoptotic cells at a concentration of 500 M H 2 O 2 (12). Treatment with H 2 O 2 led to a significantly enhanced maximum increase of apoptotic cells in APPsw-bearing PC12 cells (51.4 Ϯ 5.3%) compared with APPwt-bearing cells (28.4 Ϯ 4.7%) and vector control cells (21.3% Ϯ 4.3%). Different clones with the same transfectant behave similarly in their sensitivity to H 2 O 2 -induced cell death (Fig. 1C).
Oxidative Stress Induces Activation of Caspase 2 in PC12 Cells-Oxidative stress in the human brain has been implicated as one major cause of neuronal cell loss in AD patients. However, the exact mechanism still remains unknown. A␤ appears to bind redox-active metals like zinc, copper, or iron with high affinity, resulting in production of hydrogen peroxide and autoxidation of the metallopeptide complex (52,53). The signaling cascades activated following the oxidative stress have not been widely studied. Therefore, we studied the implication of caspase 2 by measuring the activity and by Western blotting. Cleavage of the photometric substrate Ac-VDVAD-p-nitroanilide by cytosolic protein extracts indicates the presence of caspase 2 protease activity in the cultures after exposure to hydrogen peroxide. The activity was already induced early in APPsw cells within 2 h of induction ( Fig. 2A). Caspase 2 activity was continuously elevated over time compared with APPwt and vector-transfected cells. Maximum caspase 2 activity was measured after 2 h, with a 4-fold higher activity in APPsw cells than in APPwt and vector PC12 cells, respectively (Fig. 2B).
The mechanism of activation of caspases occurs by a sequential cleavage of the zymogen to release the large and the small cleavage products. In accordance with our findings from the determination of activity, Western blotting confirmed cleavage of caspase 2 within 2 h of hydrogen peroxide induction (Fig.  2C). Again, the activation of caspase 2 was noticeably higher in APPsw cells compared with vector-transfected controls.
Involvement of Caspase 8 -Apoptosis associated with the Fas/tumor necrosis factor receptor family of death receptors requires caspase 8 activity and adaptor proteins such as FADD. The involvement of caspase 8 in processing of APP during apoptosis by caspase 8 has been reported recently (46). The blocking of neuronal death by caspase 8 inhibitor IETD-fmk in A␤-induced cell death was also demonstrated (54). Hence, we examined the activation of caspase 8 in our transfected cell lines overexpressing intracellular high levels of A␤ in order to ascertain the involvement of this initiator caspase in oxidative stress. We utilized Ac-IETD-pNA to measure caspase 8 activity in lysates of treated cells. In parallel to the early activation of caspase 2, caspase 8 activity increased about 3-fold after 2 h of stress induction in PC12 cells expressing the Swedish double mutation and was significantly enhanced in a time-dependent manner compared with vector transfected controls (Fig. 3, A  and B). Activation of caspase 8 involves a two-step proteolysis: the cleavage of the procaspase 8 to generate a 43-kDa fragment and a 12-kDa fragment that is further processed to 10 kDa. The large fragment is then cleaved to yield p26. Here, we used an antibody recognizing the 26-kDa large fragment of caspase 8. Western blot analysis showed a notable expression of active caspase 8 even in the untreated APPsw PC12 cells. Over the whole time period, the protein expression in APPsw cells was continuously higher compared with vector control (Fig. 3C).
In addition, we analyzed the activities of caspase 1 and caspase 6 using AC-YVAD-pNA and AC-VEID-pNA. Interestingly, no changes in the activities of caspase 1 and caspase 6 were found in oxidative stress-induced cell death between mutant APP-bearing PC12 cells and controls (data not shown).
Increased Caspase 3 Activity of APPsw PC12 Cells in Response to Oxidative Stress-Furthermore, we prepared extracts from H 2 O 2 -treated PC12 cells and measured their ability to cleave the colorimetric substrate to Ac-DEVD-pNA, which is specifically cleaved by caspase 3. Cellular extracts from untreated control cultures showed very low Ac-DEVD-pNA cleav- age in all cell lines (Fig. 4). Under these base-line conditions, no significant differences could be detected between vector-, APPwt-, and APPsw-transfected PC12 cells. After exposure to oxidative stress, caspase 3 activity increased gradually and reached a maximum activity after 6 h of incubation, demonstrating a delayed increase as compared with the initiator caspases 2 and 8. APPsw and APPwt PC12 cells showed a significantly enhanced caspase 3 activity compared with vector-transfected clones with the strongest increase in the APPsw cells (Fig. 4A). By Western blotting, we also found a significantly increased protein expression of the active caspase 3 in APPsw PC12 cells compared with control cells (Fig. 4B). The highest protein expression was evidenced after 6 h of H 2 O 2 exposure by densitometric immunoblot analysis (Fig. 4C).

Activation of Caspase 9 in Response to Oxidative Stress-
Mitochondria play a prominent role in the intrinsic apoptosis pathway. During mitochondrial dysfunction, several essential players of apoptosis, including procaspases, cytochrome c, and apoptosis-inducing factor, are released into the cytosol. The multimeric complex formation of cytochrome c, apoptotic protease-activating factor-1, and caspase 9 activates downstream caspases, leading to apoptotic cell death (55). Therefore, we examined whether oxidative stress in PC12 cells activates caspase 9 by using an polyclonal antibody specific for the inactive full-length procaspase 9 and for the cleaved fragments of the protein. We detected a time-dependent increase in cleaved fragments of caspase 9 after treatment of the cells with hydrogen peroxide (Fig. 5A). APPsw PC12 cells showed an elevated expression of the active cleaved fragments of caspase 9 compared with vector-transfected cells. A time-dependent cytochrome c increase in the cytosol was also detected by Western blot using a specific antibody (Fig. 5D). The highest cytochrome c amount in the cytosolic fraction was detected after a 4-h oxidative stress induction. In addition, caspase 9 activities were measured colorimetrically using Ac-LEHD-pNA as substrate in lysates of hydrogen peroxide-treated cells (Fig. 5B). We ascertained an early induced activation of caspase 9 after 2 h, whereas the activity was increased in the APP-transfected cells compared with vector-transfected controls but was not different between APPsw and APPwt cells (Fig. 5C).
Enhanced Activation of JNK Pathway in APPsw PC12 Cells-The mobilization of the stress-activated protein kinase (JNK) plays an important role in apoptosis induced by several types of environmental stress, like UV, ␥-radiation, inflammatory cytokines, and DNA damage (56 -58). In oxidative stressinduced cell death, the activation of mitochondrial apoptosis machinery by JNK in adult cardiac myocytes has been observed (37). Hence, we addressed the question whether oxidative stress-mediated activation of the JNK pathway is affected by overexpression of APP. We performed a Western blotting of immunocomplexes to detect activation of JNK and c-Jun. H 2 O 2 treatment of APPsw PC12 cells induced JNK phosphorylation within 2 h (Fig. 6A). The highest expression was detected after 4 h with an elevated expression in APPsw cells compared with APPwt cells and vector control. When activated, JNK translocates to the nucleus to regulate transcription through its effects on c-Jun and other transcription factors. As for JNK, we observed a strong impact of the Swedish APP transfection on the activation of c-Jun (Fig. 6B). In our experiments, the kinetics of activation of c-Jun correlated with the phosphorylation of JNK with the highest detection of phosphorylated c-Jun after 6 h of H 2 O 2 treatment. The expression of phosphorylated c-Jun was markedly elevated in APPsw compared with treated APPwt or vector-transfected controls after 4 and 6 h of hydrogen peroxide treatment.
Prevention of Apoptosis by Caspase Inhibitors-Due to the enhanced activation of several caspases in response to oxidative stress, we examined the potential against several caspase inhibitors in protecting PC12 cells of oxidative stress-induced apoptosis. Caspase inhibitors were added to cultures of PC12 cells in a time-and concentration-dependent manner (data not shown). Best protection was provided by adding caspase inhibitors at a concentration of 10 M 3 h prior to H 2 O 2 exposure. As shown in Fig. 7, only the inhibitor of the effector caspase 3, AC-DEVD-CMK, was able to rescue APPsw cells from apoptosis below the level of that from vector-transfected control cells. The pretreatment of the APPsw cells with inhibitors of initiator caspase 8 and caspase 9 revealed similar levels than did vector cells.
Inhibiton of JNK by SP600125-As detected by Western blotting, we could show that the transfection of the Swedish FAD mutation in PC 12 cells leads to an enhanced activation of the JNK pathway under oxidative stress. Therefore, we investigated the effect of SP600125, a selective anthrapyrazolone inhibitor of JNK (59), in our cell model. We measured the caspase activities of different caspases and the mitochondrial membrane potential after pretreatment with JNK inhibitor to attribute the JNK involvement in oxidative stress-induced apoptosis. We found a highly significant reduction of caspase 9 activity (two-way ANOVA, p Ͻ 0.001) and caspase 3 activity (two-way ANOVA, p ϭ 0.0384) in cell lysates of APPsw cells pretreated with 200 nM JNK inhibitor, whereas the activities of caspase 2 and caspase 8 were not significantly altered (Fig. 8).
Corresponding to the reduction of caspase 9 activity, we were able to detect a protection against a decrease in the mitochondrial membrane potential by SP600125 (Fig. 9), indicating an involvement of JNK upstream of mitochondria in oxidative stress signaling. Whereas the treatment with H 2 O 2 (250 M) led to a significant increase of apoptotic cells in APPsw-bearing PC12 cells (42.21 Ϯ 3.71%) compared with vector cell line (30.87 Ϯ 5.97%), APPsw cells pretreated with the JNK inhibitor (33.77 Ϯ 4.06%) exhibited similar apoptotic cell levels as the corresponding vector cells (Fig. 7). DISCUSSION In this study, we provide evidence for the enhanced cell death vulnerability caused by the Swedish APP mutation and elucidate cell death mechanism probably initiated by intracellularly produced A␤. Although A␤-induced cell death has been a topic of research in many studies, the relevance of these in vitro findings with regard to in vivo processes is not clear. In most studies, synthetic A␤ peptides have been applied at micromolar levels that are in contrast to the low nanomolar levels of natural A␤ found in the brain and cerebrospinal fluid of AD patients.
Therefore, studies of A␤-induced neurotoxic effects under more likely physiological conditions mimicking the situation in AD brains are needed (60). Moreover, A␤ is produced intracellularly and can accumulate within cells (21,61,62) before its deposition in senile plaques. Intracellular accumulation of A␤ might impair cellular functions and may represent the primary event within the neurotoxic A␤ cascade. In order to mimic the situation in vivo, we decided to transfect PC12 cells with human APP and the FAD-related Swedish APP mutation. APPsw and APPwt PC12 cells showed a moderate expression of human APP, leading to low levels of A␤-(1-40) production (APPsw, 90 pg/ml; APPwt, 20 pg/ml) close to the situation in vivo. These two cell lines with different production levels of A␤ may addi- tionally allow us to study dose-dependent effects of A␤. Expression of APPsw rendered PC12 cells more vulnerable to the induction of cell death after exposure to oxidative stress. It seems likely that increased production of A␤ at physiological levels primes APPsw cells to undergo cell death only after additional stress, a scenario that has been suggested to occur in AD brain. Moreover, we were able to show that caspase 3 activity was significantly enhanced in APPsw compared with APPwt and vector-transfected cells. Interestingly, APPwt also showed an increased activity of caspase 3 compared with empty vector control cells, although to a lesser extent than APPsw cells. Our data suggest that already very low A␤ levels in APPwt cells are probably sufficient to prime cells to undergo apoptosis. This is confirmed by our findings that caspase 9 activity is increased and at the same time mitochondrial membrane potential is lowered in APPwt as well as in APPsw cells compared with vector-transfected control cells after the induction of oxidative stress. In addition, we examined the implication of two other initiator caspases, which has been described primarily in the extrinsic apoptotic pathway. We demonstrated that caspase 8, involved in death receptor signaling, is activated in oxidative stress signaling in PC12 cells. Importantly, this activation is markedly elevated in APPsw cells compared with APPwt cells. Whether this increase is due to effects of extracellularly secreted A␤ on cell death receptors or is mediated by intracellularly accumulated A␤ is not clear and still under investigation. In parallel, we observed in PC12 cells an activation of caspase 2 in oxidative stress signaling. In APPsw cells, the maximal activity was 4-fold higher compared with APPwt and vector PC12 cells, respectively. This finding is of particular interest, since recent data describe caspase 2 to be localized in several intracellular compartments, including the mitochondria, Golgi, cytosol, and nucleus (63)(64)(65). Cytotoxic stress causes activation of caspase 2, which is required for the permeabilization of mitochondria (66) and induces the release of cytochrome C (64,67,68). It is assumable that the enhanced A␤ production in the APPsw-transfected cells occurs through an increased expression and activity of beta-site APP-cleaving enzyme (26). The elevated intracellular levels of A␤ in the mutant cells may therefore lead to an enhanced activation of caspase 2, which in turn implements the activation of the intrinsic apoptotic pathway by permeabilizing mitochondria (69). As a consequence, cytochrome c is released to the cytosol, resulting in an activation of the apoptotic protease-activating factor-1/caspase 9 apoptosome complex. Simultaneously, other effects of caspase 2 need to be considered, since we determined no differences in cytochrome C release and caspase 9 activation between APPsw and APPwt cells. Recent data indicate that A␤ induces the release of other proapoptotic factors from the mitochondria (e.g. Diablo/Smac (second mitochondria-derived activator of caspase)) (70). Thus, Smac or other unknown factors may be involved in our cell model. Since caspase 2 activation occurs early in our cell model and neurons from caspase 2 null mice are resistant to ␤-amyloid-induced death (47), this caspase seems to play a critical role.
Given the important role that mitogen-activated protein kinase pathways play in cellular stress signaling, we investigated the relationship between activated JNK, apoptosis, and the Swedish APP mutation. We demonstrated that oxidative stress leads to an activation of JNK. JNK phosphorylates specific sites in the N-terminal region of c-Jun, thereby enhancing its transcriptional activity. Very importantly, we were able to demonstrate that in PC12 cells bearing mutant APPsw, the activation of JNK and c-Jun was significantly enhanced compared with APPwt cells. At the same time, activation of JNK and c-Jun is increased in APPwt compared with vector cells. This suggests that the different A␤ levels induce a dose-dependent activation of JNK pathway. Interestingly, we could observe a reduction of caspase 9 activity using the selective JNK inhibitor SP600125, whereas activities of caspase 2 and caspase 8 were unaltered. Furthermore, the inhibitor compensated the mitochondrial abnormality when oxidative stress was induced. Due to the fact that members of the antiapoptotic Bcl-2 family proteins are inactivated through phosphorylation by JNK (71,72), our results support evidence that JNK alters mitochondrial function. Recently, it was reported by different groups that JNK-interacting proteins JIP1b and JIP2 bind the cytoplasmic domain of APP (73)(74)(75)(76). JIP family proteins are scaffold proteins that organize specific members of the JNK/ mitogen-activated protein kinase cascade to facilitate signaling. It was proposed that these scaffold proteins could serve as cargo for the microtubule motor kinesin to mediate the transportation of several transmembrane proteins (77), in which APP is supposed to be a cargo receptor (78). Our data indicate that the altered processing of APPsw might play a crucial role in activation of the stress kinase pathway. Whether this occurs by a redistribution of these scaffold proteins has to be clarified in the future.
In addition, we investigated the protective effects of caspase inhibitors as well as JNK inhibitor in preventing oxidative stress-mediated cell death. Only the caspase 3 inhibitor was able to reduce oxidative stress-induced apoptosis significantly in APPsw cells below the level of control cells. The inhibition of caspase 8 or caspase 9 was less effective in preventing apoptosis. The lack of efficacy of the initial caspase inhibitors to abolish completely apoptotic cell death could be attributed to the compensatory pathway. Although the treatment of PC12 cells with hydrogen peroxide leads to activation of these initiator caspases, their inhibition is not enough to avoid apoptosis, because different caspase pathways co-exist (Fig. 10). Such a compensation of caspase pathway was ascertained in sympathetic neurons from caspase 2 null animals. A lack in caspase 2 expression was compensated by an increased expression of both mRNA and protein for caspase 9 and DIABLO/Smac in trophic factor deprivation-mediated death (79).
In summary, we propose a hypothetical sequence of events linking FAD, A␤ production, JNK activation, and mitochondrial dysfunction with caspase pathway and neuronal loss for our cell model (Fig. 10). The brain has a high metabolic rate and is exposed to gradually rising levels of oxidative stress during life. In Swedish FAD patients, the levels of oxidative stress are increased in the temporal inferior cortex (28). Our study using a cell model mimicking the in vivo situation in AD brains indicates that probably both increased A␤ production and the gradual rise of oxidative stress throughout life lead to an increased vulnerability to apoptotic cell death in neurons from FAD patients.