*3.6. Activation of ERK 1*/*2*

Figure 6 shows that Aβ (25–35) treatment significantly increased the expression of phosphorylated (p)-ERK 1/2 (*p* < 0.05). The pretreatments of Eq and E2 significantly prevented the Aβ (25–35)-induced activation of ERK 1/2 (*p* < 0.05). On the other hand, when the ER activity was inhibited by ICI-182,780, the effect of Eq and E2 on deactivation of ERK 1/2 was significantly reduced (*p* < 0.05).

**Figure 5.** The cyclin D1 expressions of the SH-SH5Y cells from different treatments. Protein expressions were assessed by Western blotting. Data were analyzed by a one-way ANOVA followed by the LSD post-hoc test and are representative of three independent experiments (*n* = 3). Values are presented as the mean + SD. Bars with different letters are significantly different at *p* < 0.05.

**Figure 6.** The phosphorylated (activated) ERK 1/2 expressions of SH-SH5Y cells from different treatments. Data were analyzed by a one-way ANOVA followed by the LSD post-hoc test and are representative of three independent experiments (*n* = 3). Values are presented as the mean + SD. Bars with different letters are significantly different at *p* < 0.05.

## **4. Discussion**

Evidence from previous clinical and experimental studies showed that estrogen replacement therapy may have beneficial effects on AD in postmenopausal women [25,26]. However, the use of estrogen as treatment is known to have side effects, such as the development of breast and endometrial cancers in women [23]. Phytoestrogens may be an alternative treatment for AD with fewer side effects. A previous study showed that the phytoestrogen, α-zearanol, elevated the cell survival of Aβ (25–35)-induced PC-12 cells by attenuating oxidative stress and apoptotic cell death in a manner similar to 17β-estradiol [27]. In the present study, both S-equol and 17β-estradiol were also found to increase cell survival followed by Aβ (25–35) treatment. These results would predict that phytoestrogen, S-equol, possessed putative neuroprotective effects against Aβ (25–35)-induced cytotoxicity on SH-SH5Y cells analogous to those of 17β-estradiol [28,29]. In addition, the result of the inhibition of ER with antagonist ICI-182,780 prior to the Eq and E2 treatments suggested that ER may have a role in the neuroprotection of S-equol and 17β-estradiol against Aβ (25–35) cytotoxicity. The critical roles of ERs have been implicated in the cognitive function [14]. The loss of ERα expression has been noted to more likely contribute to AD-related memory impairment and amyloidogenesis [30,31]. Our observations showed the downregulation of ERα protein expression in SH-SH5Y cells exposed to Aβ (25–35) alone, emphasizing the importance of the ERα functional role in response to Aβ (25–35)-induced cytotoxicity. Under normal conditions, the ERα function can be enhanced by its coactivators, such as SRC-1, for efficient transcriptional regulation [32]. The decreased SRC-1 protein expression in the Aβ (25–35)-treated group was seen in the present study showing the Aβ (25–35)-induced disruption of the SRC-1 coactivator. These results support the notion that Aβ (25–35)-induced perturbation of ERα was further evident from the corresponding decrease in the expression of SRC-1. Furthermore, S-equol or 17β-estradiol pretreatments efficiently attenuated the effects of Aβ (25–35) in the current study, demonstrating that to provide a neuroprotective effect, equol binds with ERα and recruits SRC-1 to enhance its effect. On the other hand, the actions of both compounds were blocked by anti-estrogen ICI-182,780 in the present study, observing that ERα is required for the neuroprotective response of S-equol or 17β-estradiol to Aβ (25–35) cytotoxicity.

17β-estradiol binding to ERα is able to trigger transcriptional regulation of target genes, such as cyclin D1 [33]. In this regard, a recent study has reported that 17β-estradiol bound ERα has a role in controlling cell cycles [34]. In the present data, we presume that downregulated ERα expression in the presence of Aβ (25–35) might partially contribute to aberrant cell cycles. In normal conditions, neuron cells are postmitotic and stay in the G0 phase, as indicated by the downregulation of proteins related to the cell cycle [35]. For instance, cyclin D1, a protein marker of the G0/G1 phase, is expressed at the beginning of the G1 phase and continually accumulates in the nucleus during the G1 phase in the presence of the cell cycle reactivation [36]. When the cells progress into the S phase, cyclin D1 can secrete into the cytoplasm and its overexpression can reduce cell sizes and shorten the G1 phase resulting in the accelerated entry into the S phase [37]. Likewise, our results showed that Aβ (25–35) caused cells to leave the postmitotic phase and reenter the cell cycle in parallel with the increasing level of cyclin D1. This finding is in line with previous studies which found that Aβ (25–35) toxicity induces cell-cycle reentry [9,38]. However, only a tendency toward a decrease in cell number of the G1 phase in the Aβ-treated group was observed in this study. Such observation might be ascribed to a more rapid cell cycle progression in response to a higher level of cyclin D1 followed by Aβ treatment as mentioned above. Alternatively, it is plausible that there is a high degree of variability in the G1-phase progression due to the differences in nature between cells, which indicates that the cell itself may enter into G1 or exit from G1 at different time points from its neighboring cells [39]. In presenilin (PS)-1 familial AD brains, the presence of cyclin D1 accumulation was observed to be linked to cell-cycle activation and subsequently led to cell death [40]. Our results are in accordance with previous findings showing that when exposed to 25 μM Aβ (25–35), SH-SY5Y cells accumulated in the S phase, indicating that they did not progress beyond the S phase accompanied by apoptosis [9,41]. Taken together, we speculate that changes in ERα and cyclin D1 expressions concomitantly occurring with aberrant cell cycle reentry appear likely to underlie the cytotoxic mechanisms of Aβ (25–35). Thus, apoptotic neuronal death is presumably the consequence of Aβ (25–35)-induced cytotoxicity [42]. However, it is noteworthy that more recent evidence indicates neuronal cell death triggered by a cell cycle reentry event could be independent of an apoptotic mechanism in AD [43]. More in-depth investigation is warranted to resolve this discrepancy. Nevertheless, S-equol prevented Aβ (25–35)-induced changes in the cell-cycle behavior, ERα, and cyclin D1 expressions, indicative of the neuroprotective potential of S-equol.

A common target for estrogen signaling and Aβ neurotoxicity is ERK 1/2 [9,10,44]. It was shown that ERK 1 and 2 are expressed in the pooled cerebrospinal fluid (CSF) of patients with AD, and elevated levels of ERK 1/2 in CSF are accompanied by increased levels of tau protein and the Aβ42 peptide [45]. Rapid activation of ERK 1/2 was reported in SH-SY5Y neuroblastoma cells exposed to Aβ (25–35) [9] and in mature hippocampal neurons [46]. Aberrant activation of ERK 1/2 was correlated with an elevated level of cyclin D1 that has been shown to be responsible for cell cycle reentry in neurons under Aβ-induced toxicity conditions, thereby potentiating the neuronal apoptosis responses [38]. The present data showed that Aβ (25–35) triggered ERK 1/2 activation, and pretreatments of S-equol and 17β-estradiol were able to prevent this response. In contrast, treatment with ICI-182,780 appeared to diminish the protective effects of S-equol and 17β-estradiol. These observations led us to propose that the neuroprotective mechanisms of the actions of S-equol and 17β-estradiol against Aβ (25–35) cytotoxicity might be mediated by the ERK1/2 pathways via ERα. Previous studies have shown that estrogen prevents cytotoxic effects of Aβ by activating MAPK which regulates ERK 1/2 expression and cyclin D1 to control cell cycle reentry [9,29]. Herein, we have shown that S-equol exhibited neuroprotective effects that mimicked the action of 17β-estradiol on Aβ (25–35)-treated SH-SY5Y cells through preventing cell cycle reentry downregulating cyclin D1 and ERα-mediated ERK 1/2 expressions, all of which might have involved suppression of Aβ (25–35)-induced cell cycle reentry by S-equol or 17β-estradiol pretreatments in the current study.
