Neuroprotective effects of E2 have been debated for decades [1–4]. Among these, the doses, duration, route and period of estrogen replacement have become the most important parameters. In this study, we demonstrated that in vitro physiological doses of E2 promoted cell proliferation and attenuated cell damage; where zas pharmacological doses of E2 inhibited cell proliferation and promoted cell damage. In vivo experiments further demonstrated that physiological levels of E2 replacement exhibited neuroprotective effects and supraphysiological levels of E2 replacement promoted neurodamage.
Previous reports have demonstrated that estrogen can regulate the development, maturation, survival, and function of multiple types of neurons in different brain regions . However, it is unclear whether the effects of physiological doses of E2 and pharmacological doses of E2 exert different effects on nerve cells. Interestingly, we observed that physiological doses of E2 (10 nM and 20 nM) increased the number of PC12 cells, however, pharmacological doses of E2 (10 μM and 20 μM) decreased the number of cells and induced the cells to lose their neurites. Mechanisms underlying this phenomenon is unclear. Previous research has reported that 0.5 nM E2 was ineffective in promoting proliferation of human neural progenitor cell (hNPC), as assessed by BrdU incorporation. Minimal effectiveness of E2 to promote proliferation was shown at 1 nM, maximal effectiveness at 100 nM and decrement at 250 nM, suggesting that the efficacy of E2 on proliferation was dose dependent . However, the effect of pharmacological doses of E2 on the proliferation of nerve cells has been rarely studied. To explore this question, PC12 cells received different concentrations E2 stimulus. The results of immunocytochemistry demonstrated that physiological doses of E2 could promote cell proliferation, which was consistent with previous findings . Additionally, Suzuki et al. has shown that physiological low doses of E2 protected neurons from brain ischemic injury by enhancing proliferation of newborn neurons. More importantly, our results have demonstrated, for the first time, that pharmacological doses of E2 inhibited the proliferation of nerve cells.
It is known that cell proliferation is determined by the DNA replication during the cell cycle, thus, flow cytometric analysis was used to detect the effects of E2 on the cell cycle. The results revealed that physiological doses of E2 significantly increased the percentage of cells in S + G2 phase of the cell cycle; however, pharmacological doses of E2 markedly decreased percentage of cell in S phase or S + G2 phase. In a previous study, 100 nM E2 has been shown to increase the expression of PCNA and CDK1/cdc2 in hNPCs , PCNA and CDK1/cdc2 are well defined and commonly used as markers of cell cycle. PCNA is a marker of the cells in early G1 phase and S phase of the cell cycle, and it acts as a homotrimer to increase the processing of leading strand synthesis during DNA replication . Whereas CDK1 exists as a component of CDK1/cyclin B complex, is required for transition from G2 to M phase . Taken together, we conclude that physiological doses of E2 promote DNA replication by acting on the PCNA and CDK1/cdc2 complexes. By contrast, it is possible that pharmacological doses of E2 attenuate DNA replication, however, the mechanism is still unclear.
To further investigate the effects of E2 on nerve cells after nerve injury, we applied an OGD-R experimental paradigm. OGD-R is composed of a hypoxic and a reoxygenation/reperfusion phase and it is recognized as an ideal in vitro model for the ischemic stroke. Our results demonstrated that physiological doses of E2 significantly increased cell apoptosis induced by OGD-R. Whereas pharmacological doses of E2 markedly attenuated cell apoptosis. Previous study have shown that 100 nM E2 was found to have a neuroprotective effect against CoCl2-induced apoptosis in PC12 cells by attenuating ROS production and modulating apoptotic signal pathway through caspase cascades, Bcl-2 family, cytochrome c, Fas/Fas-L as well as PI3K/Akt pathway . Studies have reported that chronic E2 treatment could rescue neurons destined to apoptosis or necrosis by interfering with apoptotic death cascades that activate caspase-3 . Our results demonstrated that physiological doses of E2 significantly reduced the expression of pro-apoptotic protein caspase-3, however, pharmacological doses of E2 increased the expression of caspase-3. These results verified that physiological doses of E2 exhibit the neuroprotective effects by modulating apoptotic signal pathway, whereas pharmacological doses of E2 may inhibit the pathway, but the mechanism need to be further explored.
We further examined the effects of different levels of E2 replacement in vivo on OVX rats following MCAO. We found that estrogens used in all previous studies of the neurodamaging effects of E2 were commercially manufactured slow-release pellets, which created an early, prolonged, supraphysiological peak plasma concentration . Therefore, we executed hormone replacement by daily subcutaneous injection of E2 on the back of the neck of the rats for 4 weeks, resulting in the tailored serum E2 level. The result showed that physiological levels of E2 replacement significantly improved neurological deficit scores, decreased infarct volume, reduced neuronal damage and inhibited the apoptosis in ischemic penumbra. However, superphysiological levels of E2 replacement worsen the ischemia–reperfusion injury by significantly lowering neurological deficit scores, increasing infarct volume, inducing neuronal damage and promoting the apoptosis in ischemic penumbra. Several studies [7, 8] have confirmed that low physiological levels of E2, which are strikingly similar to low-basal circulating levels found in cycling animals, exert profound neuroprotective actions by reducing apoptosis, enhancing proliferation of newborn neurons. Moreover, in the studies using commercial pellets, lower doses of estrogen tended to decrease ischemic damage, whereas higher doses tended to increase the damage . Furthermore, a recent study  showed that injection of estrogens into adult OVX female rats 30 mins before conditioning, the low physiological doses of 17β-estradiol and 17a-estradiol enhanced, whereas the superphysiological doses of 17β-estradiol and 17a-estradiol impaired, contextual fear conditioning, which relies on the integrity of the hippocampus and amygdala.
Nevertheless, some studies have shown neuroprotective effects of supraphysiological or pharmacological levels of E2. It has been shown that supraphysiological or pharmacological levels of E2 administered immediately before the onset , or as late as 6 h after the onset , of ischemic injury effectively protected the brain against ischemic injury. It appears that supraphysiological or pharmacological levels of E2 have both neuroprotective effects and neurodamaging effects depending on the timing of E2 administration. These studies [9, 10] reported that acutely administration of E2 but not long-term E2 replacement produced the neuroprotective effects of supraphysiological or pharmacological doses of E2. It is known that estradiol acts through different estrogen receptors (ERs) and activates distinct secondary messenger pathways at different time courses and involves various downstream mechanisms . In the classical chronic genomic mechanism, estradiol acts through soluble intracellular α or β receptors (ERα or ERβ), once these receptors were activated, they translocate to the nucleus where they function as ligand-dependent transcription factors . In contrast, fast non-genomic effects are mediated by classic receptors (ERα and ERβ) and specific G-protein–coupled receptors (GPR30 and ER-X) that regulate ligand-gated ion channels and neurotransmitter transporters . The GPR30 receptor is reported to be a novel estrogen receptor uniquely localized to the endoplasmic reticulum  and may act together with intracellular estrogen receptors to activate cell-signaling pathways to promote neuron survival after global ischemia [34, 35]. In a culture of cortical neurons, treatment with the GPR30 agonist G1 for 45 min attenuated the excitotoxicity induced by NMDA exposure. Additionally, acute neuroprotection mediated by GPR30 is dependent on rapid G-protein–coupled signals and ERK1/2 activation but independent of transcription or translation . Moreover, acute estradiol treatment protects CA1 neurons from ischemia-induced apoptotic cell death via the PI3K/Akt pathway . Therefore, the acute neuroprotective effects of estrogen maybe mediated via the fast non-genomic mechanism. Our results indicated that chronic supraphysiological doses of E2 replacement may exert neuroprodamaging effects, but during acute treatment for ischemic stroke, the supraphysiological or pharmacological doses of E2 may exert neuroprotective effects.
It is worthwhile to note several limitations in this study. First, we had not thoroughly investgate the molecular mechanisms underlying the effects of different doses of E2 on cell morphology, cell proliferation and cell apoptosis. Further studies are warranted to elucidate the mechanisms underlying the attenuation of cell proliferation and increase in cell apoptosis induced by pharmacological doses of E2. Second, the latest study has accurately demonstrated that E2 level in the hippocampus is approximately 8 nM in the male and 0.5–2 nM in the female, which is much higher than that in the circulation . This hippocampus-derived estrogen rapidly modulates dendritic spines , which has been extensively studied in relation to memory processes and synaptic plasticity. Therefore, the effects of estrogen replacement on the changes of estrogen level in the hippocampus should be further studied.