PS2 mutation increases neuronal cell vulnerability to neurotoxicants through activation of caspase‐3 by enhancing of ryanodine receptor‐mediated calcium release

To determine neurotoxic mechanisms of PS2 mutation, disruption of cellular calcium homeostasis through ryanodine receptor pathway, interaction between PS2 and ryanodine receptor, and sensitization of neurons to neurotoxicants by activation of caspase-3 activity were investigated in the stably transfected PC12 cells and cortical neurons from transgenic mice expressing wild or mutant PS2 (N141I). PC12 cells expressing mutant PS2 and cortical neurons from mutant PS2 transgenic mice were significantly more vulnerable against L-glutamate and Aβ treatment compared with PC12 cells expressing wild-type (WT) PS2, and cortical neurons from WT PS2 mice, as determined by MTT reduction assay in the PC12 cells, and by Trypan blue staining in cortical neuron or DAPI staining (data not shown). To determine whether PS2 mutations increase calcium release and thereby contribute to an increased vulnerability to L-glutamate and Aβ, we measured the intracellular calcium levels of PC12 cells, and compared with that in PC12/neo, PC12/PS2wt, and PC12/PSmt cells. The intracellular calcium level was elevated in PC12 cells expressing WT PS2 or mutant PS2 by L-glutamate and Aβ in a dose-dependent manner (data not shown). However, the magnitude of the calcium release was much greater in the PC12 cells expressing mutant PS2. Calcium release from cortical neurons after L-glutamate (30 mM), Aβ25–35 (50 fM) and Aβ1– 42 (10 fM) treatment was also examined and found to be significantly greater (2-to 3-fold higher in increase) in cortical neurons from mutant PS transgenic mice compared with WT or nontransgenic (NT) mice (data not shown). PS is an integral protein in the endoplasmic reticulum (ER), and PS may be important for regulating ER calcium signals, which have been implicated in neuronal cell death by mutation of presenilin. We therefore determined whether PS2 is coexpressed or colocalizes with neuronal type RyR (type 3) in PC12 cells and cortical neurons. Double-labeling immunostaining confocal analysis showed that the PS2-reactive compartment in the ER of PC12 cells was also RyR-positive (Fig. 1A). Colocalization of PS2 and RyR was observed in the cortex and hippocampus of transgenic mice expressing PS2 in the ER (data not shown), and the expression of RyR and PS2 (full length, N-and C-terminal PS2) was greater in the cortex and hippocampus of PS2 mutant transgenic mice than in WT PS2 transgenic or age-matched NT mice. Similarly, Western blot analysis data showed that immunoprecipitates of lysates with the N-PS2 antibody immunoreacted with the RyR antibody, and much greater coexpression was found in PC12/PS2mt cells (Fig. 1B, C) and in the brain of transgenic mice expressing WT and mutant PS2 (data not shown). Expression of RyR protein and colocalization with PS2 in PC12 cells expressing PS2. A) Confocal laser scanning micrograph showing a merged image of PC12 cells double-labeled with PS2 and RyR antibodies. Red, N-terminal PS2 immunoreactivity; green, RyR immunoreactivity; yellow, sites of colocalization. B) Expression of N-terminal PS2 and RyR in lysates of each type of PC12 cells after with or without treatment of L-glutamate and Aβ. C) Coimmunoprecipitated (IP) proteins from lysates of each type of PC12 cells. Proteins immunoprecipitated with N-terminal PS2 (left) and RyR (right) antibodies or IgG antibody (negative control) were detected on Western blots with antibodies to RyR (left) or N-terminal PS2 (right). Similar results were obtained in three different experiments. N-PS2, N-terminal PS2 fragment; control, negative control. Caspase-3 activity measured by the biochemical assay was much greater in PC12/PS2mt cells exposed to L-glutamate and Aβ25–35 than in other cell types (Fig. 2A, B). Much greater increase in expression and caspase-3 reactive cell number were observed in PC12/ PS2mt cells than in PC12, PC12/neo, or PC12/PS2wt cells after treatment with L-glutamate, Aβ25–35, and Aβ1– 42 (Fig. 2C–E). Like PC12 cells, cortical neuron cells from mutant PS2 transgenic mice exhibited much higher caspase-3 activity compared with cortical neurons from WT transgenic mice after L-glutamate (30 mM), Aβ25–35 (50 fM) and Aβ1– 42 (10 fM) treatment (Fig. 2F). To further confirm the significance of intracellular calcium changes by RyR in the cell death and activation of caspase-3, cell viability and apoptotic cell death as well as caspase-3 activity were determined by the pretreatment of dantrolene. Pretreatment with dantrolene 30 min before treatment with L-glutamate (30 mM) and Aβ25–35 (50 fM) caused a reduction in cell viability in a dose-dependent manner (data not shown) and diminished the induction of apoptotic cell death in PC12/PS2mt cells (data not shown) as well as cortical neuronal cells expressing WT and mutant PS2 (data not shown). Pretreatment with dantrolene prevented the L-glutamate-and Aβ25–35-orAβ1– 42-induced increase caspase-3 activity in PC12 cells and in cortical neurons expressing mutant PS2 (data not shown). Caspase-3 activity, expression of active caspase-3 and caspase-3 immunoreactive cell number in cultured PC12 cells and cortical neurons expressing WT or mutant PS2. Caspase-3 activity was determined in cytosolic protein extracts from PC12 cells treated for 24 h (A, B). Data are means ± se from 4 to 5 independent experiments with triplicates. *P < 0.05 compared with corresponding values for PC12/PS2wt cells. Expression of active caspase-3 was determined by Western blot analysis of cytosolic protein extracts from PC12 cells treated for 24 h (C). Similar patterns of expression of active caspase-3 were found in 3 individual experiments. Caspase-3 immunoreactive cells were quantified 24 h after treatment by immunocytochemistry using anti-active caspase-3 (D), and the number of immunoreactive cells was counted under the microscope (X200) (E). The frequency of immunoreactive cells is expressed as the average number of immunoreactive cells per high-power field (visible immunoreactive cells/HPF) from four experiments with triplicates. *P < 0.05 compared with corresponding values for PC12/neo cells. Caspase-3 activity was determined in cytosolic protein extracts of cortical neurons from NT or transgenic mice treated for 24 h (F). Data are means ±se for 4 mice. *P < 0.05 compared with corresponding value for NT mice. #P < 0.05 compared with corresponding value for PS2 WT transgenic mice. PC12 cells and cortical neurons expressing mutant PS2 (N141I) were more vulnerable to L-glutamate and Aβ25–35 or Aβ1–42 compared with those expressing WT PS2. The increased vulnerability conferred by mutant PS2 is associated with an increased intracellular calcium level in both cell types. We found a much higher level of type 3 RyR in the ER of PC12 cells and in the brain of transgenic mice overexpressing mutant PS2. Moreover, treatment of PC12 cells with L-glutamate, Aβ25–35 and Aβ1–42 resulted in a strongly enhanced expression of RyR. However, treatment with dantrolene, which blocks RyR, protected cells from death induced by L-glutamate and Aβ. Increased calcium release mediated by RyR has received attention as a critical factor in the pathogenic mechanism of PS mutation-induced neuronal cell death. This interpretation is further supported by our observations: first, the protein level of RyR was much greater in PC12 cells and cortical neurons expressing mutant PS2 than in those expressing vector alone or WT PS2, and the level of RyR in PC12 cells expressing mutant PS2 was greatly enhanced by treatment with L-glutamate and Aβ. Second, we found that the RyR inhibitor dantrolene attenuated L-glutamate-and Aβinduced cell death in PC12 cells and in cortical neurons expressing mutant PS2. We found that PS2 directly interacts with RyR. Double-label confocal analysis and coimmunoprecipitation studies using N-terminal PS2 and RyR antibodies showed that PS2 and RyR colocalize in the ER. An interaction between PS2 and RyR was found in the brain of transgenic mice; this interaction was stronger in mice expressing mutant PS2. These data suggest that changes in RyR expression in the ER may be associated with the increased vulnerability of cells expressing mutant PS2 through the perturbation of calcium homeostasis. The molecular mechanism by which mutant PS2 induces increases in RyR-mediated cell death is not clear. Here we show that the level of expression of an active form of caspase-3, and the level of caspase-3 activity, was much higher in PC12 cells and cortical neuronal cells expressing mutant PS2. We found a direct linkage between the activation of caspase-3 and PS2 mutation-induced calcium release, as evidenced by the finding that blockage of changes in the intracellular calcium level by pretreatment with dantrolene inhibited caspase-3 activity and reversed neuronal cell death induced by L-glutamate and Aβ. Collectively, mutant PS2 increases the vulnerability of PC12 cells and cortical neuronal cells, and this effect may be associated with the increased release of intracellular calcium level through the ryanodine receptor-sensitive calcium pool, and activation of caspase-3 may be a significant signal in the pathway of mutant PS2-induced neuronal cell death. Overall possible mechanism was elucidated in Fig. 3. Schematic illustration of the possible molecular mechanism of presenilin 2 mutation-induced increase of neural cell vulnerability. Presenilin 2 mutation significantly increases intracellular calcium level through activation of ryanodine receptor, thus activating caspase-3-dependent cell death signal resulting in increase of vulnerability of the neuronal cells in the brain of Alzheimer's disease.