Ischemic stroke is the second leading cause of death worldwide, yet there are few therapeutic approaches available to treat this condition and those that exist, such as thrombolytic tissue plasminogen activator (tPA), are limited by a narrow therapeutic window . Two murine models are widely used to investigate the cellular and molecular mechanisms of stroke: transient and permanent MCAO. While the onset of ischemia can be precisely defined in animal models, this is rarely the case in human patients, and indeed, the onset of symptoms may not coincide with the onset of cerebral ischemia or they may not be noticed by the patient for some time . Thus, developing post-ischemia treatments represents a more realistic and clinically relevant therapeutic approach. As such, it is necessary to understand the pathophysiological events that occur in late stages of ischemia in order to develop therapeutic strategies that are effective over a broad time window following ischemia.
Previous studies demonstrated a protective effect of estradiol in pMCAO models when it was administered prior to the induction of ischemia, mimicking the circulating physiological levels of the hormone [18, 19, 49]. However, few studies to date have investigated the therapeutic effects of physiological estradiol treatment after the induction of ischemia. Our findings describe a potential mechanism of estrogen-induced neuroprotection in late stages of pMCAO in male rats, suggesting a significant clinical potential of estradiol to treat ischemic stroke.
Administration of estradiol (0.04 mg/kg) 6, 24, and 48 h post-pMCAO decreased the reactive gliosis 54 h post-pMCAO, as witnessed by the expression of GFAP and Iba1, an effect that was more pronounced in the ipsilateral cortex than the ipsilateral hippocampus. Differential down-regulation of the PI3K/Akt/GSK3/β-catenin pathway was observed in the cortex and hippocampus in the late stages of cerebral ischemia, while there were no changes in JNK phosphorylation following pMCAO in either region. Post-ischemic treatment with three doses of estradiol, beginning 6 h after the onset of pMCAO, partially recovered the activity of the PI3K/Akt/GSK3/β-catenin pathway, although this effect was more pronounced in the cerebral cortex (the region most affected by ischemia in this model) than in the hippocampus (secondary death cell area). Finally, the substantial decrease in pAkt (Ser-473) levels after pMCAO predominantly affected GFAP-negative cells and attending to their morphology and size, mostly neurons in the ischemic area.
Like many other neurodegenerative disorders, the reactive gliosis associated with ischemic stroke involves both astrocytes and microglia [44, 50, 51]. This response can vary depending on the severity and extent of brain damage, and it involves both positive elements, such as neurotrophins and anti-inflammatory components, and ‘negative elements’, including proteoglycans or components of myelin [39, 51–56]. It is therefore important to consider both these aspects of the reactive glial response when developing therapies for ischemic stroke. A strong correlation between the size of the infarct area and the accumulation of microglia has been described previously in the tMCAO model . However, have been describes that estrogens can either decreased [57, 58], or even increased [59, 60] the number of reactive astrocytes in some models of brain injury. The molecular causes for these differences are still unknown. Some authors postulate that this may represent the relative role of ERα and ERβ on the control of the neural inflammatory response in vivo, and it would depend on both type of injury and/or the CNS region . This is, to our knowledge, the first study to analyze the effect of estradiol on the accumulation of reactive glia (both astrocytes and microglia) during cerebral ischemia processes. Indeed, post-pMCAO treatment with estradiol significantly decreased in GFAP and Iba1 immunostaining in the ischemic area (cortex), reducing their levels to those seen in the control animals (SV). The reduction in reactive gliosis following estradiol treatment demonstrates an attenuation of ischemic damage, although the mechanisms underlying this effect are poorly understood. Estradiol treatment appears to up-regulate anti-inflammatory genes in the cortex, primarily via ER alpha, and it up-regulates the synthesis of ER alpha . Moreover, GSK3β inhibitors  attenuate the enhanced mRNA expression of proinflammatory mediators induced by ischemia in a pMCAO model, including iNOS, TNF-α and IL-1β typically released by activated glial cells.
The effect of estradiol may also be partially mediated by complementary effects in neurons. Indeed, multiple mechanisms may underlie the effects of neuroprotective agents, combining the inhibition of neuronal cell death and combating the detrimental effects of inflammation in the treatment of stroke . It is widely accepted that cell death in the ischemic core is mainly triggered by necrosis, due to restricted blood flow, while that occurring in the penumbra is predominantly mediated by apoptosis [4, 64, 65]. As apoptosis is a reversible process, research has focused on the apoptotic pathways involved in neuronal death after MCAO, with a view to identifying important therapeutic targets. Down-regulation of the PI3K/Akt pathway, a crucial regulator of neuronal cell survival , has been described up to 24 h after pMCAO induction [31, 67–69]. This pathway was down-regulated 54 h after the onset of pMCAO in the two tissues differentially affected by ischemia, the cerebral cortex (corresponding to the most affected region) and the ipsilateral hippocampus. Indeed, in both the ipsilateral cortex and hippocampus pMCAO induced a decrease in the levels of pAkt Ser-473 and to a lesser extent pAkt Thr-308, without significantly affecting total Akt levels, suggesting a loss of Akt activity. Phosphorylation of both Thr-308 and Ser-473 residues is required for maximum activation of Akt , and it is mediated by distinct kinases: phosphoinositide-dependent kinase 1 (PDK1) and mTORC2, respectively [71, 72]. The mechanism of activation of mTORC2 has not been fully characterized, although its direct role in Ser473 phosphorylation has been clearly demonstrated . Our findings are consistent with previous reports describing a decrease in p-PDK1(Ser241) and pAkt (Ser473) levels in the ipsilateral cortex when compared with the contralateral counterpart 24 h after of the induction of pMCAO . Furthermore, we found that post-pMCAO estradiol treatment led to partial recovery of pAkt Ser-473 levels in both ipsilateral regions (cortex and hippocampus), without affecting pAkt Thr-308 levels. These results suggest that estradiol regulates mTORC2 activity without substantially altering PDK-1 activity, although the underlying mechanism remains unclear. Alternatively, pAkt Ser-473 and pAkt Thr-308 may be differentially affected by the activation of specific phosphatases.
The decrease in pAkt Ser-473 levels in pMCAO animals (IV group) was not restricted to GFAP-positive cells but rather, it occurred predominantly in neurons. Moreover, post-pMCAO estradiol treatment reversed this effect, restoring the pAkt Ser-473 levels in these cells. These findings suggest that the effect of estradiol is not limited to the attenuation of activated gliosis in glial cells but also, that it affects the activity of the PI3K/Akt survival pathway in neurons. Indeed, we recently reported that estradiol may activate or cooperate with PI3K to phosphorylate and activate Akt [23, 74]. The decrease in Akt phosphorylation at both Ser473 and Thr308 residues suggested a reduction in Akt activity following pMCAO and thus, we analyzed the phosphorylation of GSK3α/β, a key downstream target of Akt. Phosphorylation of GSK3α/β by Akt at Ser-21/9 inactivates its kinase activity and may regulate cellular apoptosis . After 54 h of pMCAO, the decrease in pGSK3 Ser-21/9 detected in the cortex, and to a lesser extent in the hippocampus, is consistent with the observed reduction in Akt activity and it was correlated with an increase in GSK3 activity that may promote or mediate cell death [28, 76, 77]. Our data demonstrate that the pMCAO-induced decrease in cortical pGSK3 Ser-21/9 is rescued by estradiol treatment, virtually reverting to the levels seen in vehicle-treated animals (SV). Based on this finding, we propose two possible mechanisms of action by which estradiol may reduce reactive gliosis after pMCAO, through: (1) the direct inhibition of GSK3 in glial cells that subsequently alters the glial response; and (2) the direct modulation of Akt, and hence the GSK3 activity in neurons, resulting in a reduced pro-inflammatory response . Analysis of the ipsilateral hippocampi revealed no recovery of pGSK3 Ser-21/9 levels following estradiol treatment, suggesting that the activity of estradiol depends on the specific neurons and/or glial receptors differentially expressed in the cortex and hippocampus. Further studies will be aimed at investigating this hypothesis.
The activation of MAPK signaling pathways during ischemia plays an important role in apoptosis and inflammation. The inflammatory response increases the damage area, a process that is particularly pronounced during reperfusion . In animal models of stroke several inhibitors of the JNK pathway have protective effects [39, 56]. In our pMCAO model (without reperfusion), we detected no changes in total JNK or in the phosphorylation status of this enzyme at Thr183/Tyr185 residues after 54 h of pMCAO. Hence, the JNK signaling pathway does not appear to play a significant role in pMCAO, at least at 54 h after ischemia induction.
Further studies will be necessary to determine whether the neuroprotective effect of estradiol is receptor-specific, permitting the development of more selective treatments that enhance neuroprotection while avoiding some of the negative effects in non-neural cells.