Our study demonstrated that the acetylcholinesterase inhibitor donepezil antagonizes the ability of Aβ25-35 to decrease cell viability. This protective effect of donepezil was partially offset by the PKC inhibitor GF109203X, indicating that the PKC pathway participates in the cellular protective effect of donepezil. Donepezil increased the expression of P-PKC and P-MARCKS and altered the distribution of PKCα and PKCε, suggesting that donepezil has an activating effect on PKC, which may be the mechanism by which donepezil protects against Aβ25-35.
Twenty-four hours after treatment with 5–50 μmol/L Aβ25-35, the viability of PC 12 cells was decreased in a dose-dependent manner; however, donepezil at the same concentration did not significantly affect cell viability. In contrast, 1–50 μmol/L donepezil antagonized the neurotoxicity of 20 μmol Aβ25-35 at. Compared with the control, treatment with Aβ25-35 led to marked changes in PC 12 cells, including cell shrinkage, diminished cell bodies, reduced protrusions and decreased adherence ability. These cell changes were protective responses that resulted from exposure to foreign toxic substances, and pretreatment with donepezil significantly antagonized the toxic effects. However, the specific proteins that participate in the morphological changes of cells and the detailed mechanism remain to be explored.
PKC plays a very important regulatory role in the initiation and development of AD, and PKCα and PKCε are thought to be closely related to the nosogenesis of AD. Normally, PKC is located in the cytoplasm, and when it is activated, it is transported to the membrane serosa  through a process called protein trafficking. PKCε has also been found to modulate the activity of α secretase by participating in M receptor-mediated sAPPα secretion . In the fibroblasts of AD patients, the function of PKC has been found to be lacking, leading to decreased responses of neurons to various growth factors and neurotransmitters; additionally, decreased PKC activity together with decreased trafficking ability of PKC subtypes is closely related to decreased memory and cognitive function [16, 17]. The downstream substrate of PKC, MARCKS, is expressed in the mouse brain and is associated with the spatial learning ability of animals. In senile plaques in AD, MARCKS coexists with PKC and Aβ, and the phosphorylation of MARCKS is considered to be a marker of PKC activation, which is associated with neuroprotection . PKCε is often associated with preconditioning neuroprotection and mediates ischaemic tolerance by activating ERK in brain slices [19, 20]. PKCα has also been implicated in both cell survival and cell death signalling pathways and is activated in the rat brain upon ischaemic preconditioning [21,22,23]. The activation of PKC is neuroprotective and can increase neuronal strength against apoptosis and maintain cellular homeostasis through the phosphorylation of HSP27 [24, 25]. PKCε induces neuroprotection against ischaemia by regulating many pathways, including the phosphorylation of the mitochondria K + ATP channel , increased synaptosomal mitochondrial respiration  and the activation of the extracellular signal-regulated kinase (ERK) pathway , via N-methyl-d-aspartate (NMDA) receptors , and by regulating gamma-aminobutyric acid (GABA) synapses . Our study demonstrated that donepezil is able to activate PKC and MARCKS by increasing their phosphorylation, which may be the molecular basis for its effects on cell viability and neuroprotection.
MARCKS phosphorylation is regarded as a marker of PKC activation. In their inactivated states, PKC isoforms are localized in the cytosolic fraction, but they translocate to the membrane fraction after activation. To determine whether donepezil can affect the subcellular distribution of PKC isoforms, PKCa and PKCε, the two main isoforms involved in the pathogenesis of AD, were evaluated after donepezil treatment. Since STED analysis technology is not available in our lab, we performed further experiments to support our point of view. To detect the subcellular distribution of the PKC isoforms, immunocytochemical staining was performed, and the subcellular distribution of the PKCα and PKCε proteins in the cytosolic and membrane fractions was also measured by Western blotting, as in our previous study . The results showed that donepezil decreased PKCα and PKCε expression in the cytosolic fraction but increased the expression of these two isoforms in the membrane fraction (Fig. 3). This implies that donepezil can induce the translocation of the PKCα and PKCε isoforms from the cytosolic fraction to the membrane fraction. Higher levels of the PKCα and PKCε isoforms were found in the membrane fraction in the donepezil-treated groups than in the vehicle-treated groups, supporting our conclusion that donepezil can activate the PKC signalling pathway. Activation of the PKC signalling pathway may thus become a key to the development of new drugs for the treatment of dementia in Alzheimer’s disease, which may indicate the clinical significance of our study.
Apoptosis and subsequent loss of neuronal function induced by various pathological and physiological factors may be the common pathological changes in neurodegenerative diseases, and protecting the function of the remaining neurons is one approach to treatment. As a specific and potent acetylcholinesterase inhibitor, donepezil is able to improve cognitive function in mild and moderate AD patients by increasing cholinergic nerve function and activating PKC to antagonize apoptosis. Donepezil can also increase the secretion of amyloid precursor protein α to support neurons and consequently exert its neuroprotective effect .
In our previous study, two cell lines were used in one experiment. When we investigated the effects of the protein kinase C activator TPPB on amyloid precursor protein (APP) processing, we used PC12 cells and SH-SY5YAPP695 cells, human neuroblastoma SH-SY5Y cells stably transfected with human wild-type APP695 cDNA . The same cells were used to study PMS777, a new cholinesterase inhibitor with anti-platelet activated factor activity . We used human neuroblastoma SK-N-SH cells and PC12 cells to study the effect of deprenyl on APP processing . These three experiments showed similar results, including the effects of these agents on APP expression, sAPPα secretion, and Aβ secretion, in the two cell lines [28,29,30]. In neuroprotective studies, we investigated the effects of erythropoietin on Aβ-induced neurotoxicity in SH-SY5Y cells and PC 12 cells, and the results showed the same trend [31, 32]. Although the two cell lines come from two different species, the experiments showed similar results. Thus, in the present study, we exposed PC12 cells to Aβ as an in vitro model for investigating the neuroprotective effects of donepezil.
Our study is clinically significant because extracellular deposits of Aβ protein is able to result in formation of senile plaques in the hippocampus, which is the pathological characteristics of AD associated with neurodegeneration. Among the fragments studied thus far, Aβ25-35 represents the shortest fragment of Aβ, which is processed in vivo by brain proteases. Aβ25-35 is the functional domain of Aβ required for its neurotoxic effect in dementia clinically, and it retains the toxicity of the full-length peptide . It is highly cytotoxic to neuronal cells and widely used both in vitro and in vivo in neuroscience research , including our own previous studies [2, 31, 32, 35]. Thus, Aβ25-35 was chosen for this experiment. The doses of Aβ25-35 in our study were used according to relevant studies [36, 37]. We firstly applied 5, 10, 20 or 50 μmol/L of Aβ25-35 to test the neurotoxicity of Aβ25-35 in the PC 12 cells. After this test, we chose the concentration of 20 μmol/L of Aβ25-35 for the protective experiment of donepezil. At the concentration of 20 μmol/L of Aβ25-35 to test the protective effect of donepezil, the cell viability was 56% (as determined by the MTT assay). If donepezil is added after Aβ administration, there may not be enough cells left to be rescued by donepezil, and the neuroprotective effect of donepezil would consequently be difficult to show. If a lower concentration of Aβ25-35 (5 or 10 μmol/L, for instance) is used, the neuroprotection afforded by donepezil may not be necessary, and thus its potential neuroprotective ability would also be difficult to show. Thus, we used Aβ25-35 at a concentration of 20 μmol/L (which damaged a portion of the cells) and added donepezil before Aβ administration to show its neuroprotective effect against Aβ.
Pretreating cells with donepezil before adding Aβ is a prophylactic approach that may not mimic the in vivo situation. Before treatment with Aβ25-35 in our study, the effect of donepezil on cell viability was tested with the concentration of 0.5, 1, 5, 10, 20 and 50 μmol/L, none of which had a negative effect on the PC 12 cells. Then, donepezil at 5, 10, 20 and 50 μmol/L was used to test its neuroprotective effect on the PC 12 cells after treatment with Aβ25-35. The protective effect of donepezil was dose dependent when used in the doses range of 5, 10, 20 and 50 μmol/L for the treatment of PC 12 cells. In the treatment of cells, we used the concentration μmol/L of the donepezil. When used in human for the treatment of AD, the dose of donepezil is mg/d and the highest dose of donepezil is 10 mg per day, which is probably to prevent possible side effects caused by larger-dose donepezil.
In conclusion, donepezil can antagonize Aβ25-35-induced neurotoxicity in PC 12 cells, and the activation of PKC may account for the neuroprotective effect of donepezil. Activation of PKC may become a key to develop novel approaches for the treatment of dementia in Alzheimer’s disease.