- Top of page
- Materials and methods
Although estrogens are neuroprotective in a variety of neuroprotection models, the precise underlying mechanisms are currently not well understood. Here, we examined the role of protein kinase C (PKC) in mediating estrogen-induced neuroprotection in the HT-22 immortalized hippocampal cell line. The neuroprotection model utilized calcein fluorescence to quantitate cell viability following glutamate insults. 17β-Estradiol (βE2) protected HT-22 cells when treatment was initiated before or after the glutamate insult. The inhibition of PKC by bis-indolylmaleimide mimicked and enhanced βE2-induced neuroprotection. In contrast, the inhibition of specific PKC isozymes (α and β) by Go6976, inhibition of 1-phosphatidylinositol 3 kinase by wortmannin, or inhibition of protein kinase A by H-89, did not alter cell viability, suggesting a specific involvement of PKC in an isozyme-dependent manner. We further examined whether estrogen interacts with PKC in a PKC isozyme-specific manner. Protein levels and activity of PKC isozymes (α, δ, ε, and ζ) were assessed by western blot analysis and radiolabeled phosphorylation assays respectively. Among the isozymes tested, βE2 altered only PKCε; it reduced the activity and membrane translocation of PKCε in a manner that correlated with its protection against glutamate toxicity. Furthermore, βE2 reversed the increased activity of membrane PKCε induced by glutamate. These data suggest that the neuroprotective effects of estrogens are mediated in part by inhibition of PKCε activity and membrane translocation.
We and others have demonstrated that estrogens exert potent neuroprotective effects in tissue culture, as well as in animal models (Green and Simpkins 2000; Brinton 2001; Wise 2003). In vitro studies have shown that both 17β-estradiol (βE2), the naturally occurring potent feminizing estrogen, and 17α-estradiol, the biologically inactive isomer of βE2, reduce toxicity caused by serum deprivation, β-amyloid and exposure to glutamate receptor agonists (see review by Green and Simpkins 2000). In addition, βE2 reduces ischemic lesions in animals subjected to middle cerebral artery occlusion (Yang et al. 2000; Shi et al. 2001) and reduces ethanol withdrawal injury in a rat model (Jung et al. 2003, 2004).
Although accumulating evidence supports a role for PKC in certain forms of neuroprotection, direct evidence linking the neuroprotective effects of estrogens with its alterations in PKC signaling has been lacking. Previous studies have demonstrated that exposure to estrogens can regulate PKC activity and expression in both in vitro and in vivo models (Cordey et al. 2003; Jung et al. 2003; Hayashi et al. 2005). However, the observed effects of estrogen on PKC are quite varied as a result of the variety of cell types, exposure conditions and methodologies used to assess changes in the various PKC isozymes and their activity.
Using the immortalized hippocampal cell line (HT-22) and glutamate insults, the present study examined whether PKC is involved in glutamate toxicity and estrogen protection. We further focused on the involvement of specific PKC isozymes, in particular PKCε, in the neuroprotective effects of βE2. Our previous study reported that estrogen inhibits the protein levels and activity of PKCε associated with ethanol withdrawal insults (Jung et al. 2003; Jung et al. 2005). PKCε has also been reported to be involved in UV-induced cell death (Chen et al. 1999) and apoptotic neuronal insults (Knauf et al. 1999). To this end, we hypothesized that estrogen counteracts the PKC signaling pathway in an isozyme-specific manner as a part of its neuroprotective mechanisms.
- Top of page
- Materials and methods
The present study demonstrated that both PKC inhibitors and estrogen protect against glutamate-induced neurotoxicity in HT-22 cells and that estrogen inhibits the activity and the membrane translocation of PKC in an isozyme (PKCε)-specific manner. The involvement of PKCε in this neuroprotection is evident from the finding that glutamate activates membrane-bound PKCε whereas estrogen prevents the activation. Two other second messengers, PKA and PI3K, are known to be influenced by estrogen treatment and may be involved in signaling estrogen neuroprotection (Lindford et al. 2000; Yu et al. 2004). However, neither PKA nor PI3K appear to be involved in estrogen protection against glutamate toxicity in HT-22 cells because the specific inhibitors of PKA (H-89) and PI3K (wortmannin) failed to alter cell viability or affect estrogen protection.
Our data provide evidence that inhibition of PKC is neuroprotective. Previous studies, utilizing various models of neuroprotection, also support the concept that PKC inhibition is neuroprotective. The activation of PKC contributes to glutamate-induced neurotoxicity in primary neurons (Pizzi et al. 1996), whereas the inhibition of PKC protects cells from glutamate-induced toxicity (Felipo et al. 1993). The inhibition of PKC before an insult has also been shown to be cytoprotective against anoxia, glucose deprivation and nitric oxide toxicity (Boniece and Wagner 1993; Maiese and Boccone 1995).
Several mechanisms have been proposed for protection afforded by PKC inhibition. PKC-mediated signaling events (i.e. phosphorylation) may activate nucleases and/or proteases required for cell death pathways (Bertolotto et al. 2000). Alternatively, the inhibition of PKC has been shown to attenuate oxidative neuronal injuries induced by buthionine sulfoximine, H2O2 or a pro-oxidant iron (Goodman and Mattson 1994; Higuchi and Matsukawa 1999; Noh et al. 2000). Directly relevant to this hypothesis, oxidants and antioxidants selectively react with the regulatory and catalytic domains of PKC, resulting in activation and inhibition of PKC respectively (Gopalakrishna and Jaken 2000). Finally, PKC-mediated neuroprotection may also be a consequence of PKC's role in cellular apoptosis. Phorbol esters, which initially activate and subsequently down-regulate PKCs, induce apoptosis and cell protection respectively (McConkey et al. 1989; Araki et al. 1990; Deacon et al. 1997; Li et al. 1999). Taken together, these studies suggest that PKC activation under certain conditions may be a common critical step in a variety of neuronal insults.
In the present study, the neuroprotection induced by the combination of a PKC inhibitor with βE2 produced synergistic protection (greater than the additive effect of either agent alone). This suggests that the pathways of PKC and βE2 interact with each other. To test this possibility, we examined the effects of βE2 on individual PKC isozymes (α, δ, ε and ζ) in HT-22 cells. Among these enzymes, only PKCε was altered by βE2, suggesting an isozyme-specific link between the pathways of PKC and βE2. The isozyme specificity is further supported by our finding that Go6976, which selectively inhibits PKCα and PKCβ but not PKCε (Martiny-Baron et al. 1993), failed to alter cell viability. Moreover, our finding that βE2 reversed the glutamate-induced increase in the activity of membrane PKCε suggests a counteraction between PKCε and βE2 in this neuroprotection. The effects of βE2 on PKCε were more prominent in the membrane than in the cytosol: βE2 suppressed both activity and protein levels of PKCε in the membrane, whereas it increased the protein levels of cytosolic PKCε but did not significantly alter the activity in this cell compartment. The reasons for this are not clear at present and require further investigation.
The effect of estradiol on PKC activity could only be demonstrated in the PKCε-specific assay, as no significant effects on total PKC activity were measured. PKCε contributes a significant portion of the total PKC activity, ranging from 30 to 40% of the total (Table 3). The small reduction in total PKC activity by βE2 is therefore probably due to effects of estrogens on PKCε activity.
When stimulated, inactive PKCs in the cytosol translocate to the membrane or cytoskeletal component of cells, interact with anchoring proteins, and finally phosphorylate designated substrates (Csukai and Mochly Rosen 1999; Mochly-Rosen and Kauvar 2000). Our results indicate that βE2 alters the intracellular distribution of PKCε, but not that of PKCα or PKCδ, in a manner that is consistent with a reduced activity of PKCε. Exposure to βE2 increased the protein levels of PKCε in the cytosol and correspondingly decreased levels in the membrane, suggesting that estradiol suppresses the membrane translocation of PKCε.
The mechanisms by which estrogens alter PKC activity are unclear. It is possible that estrogen deactivates a cell surface receptor for PKC. Binding of PKC to cell surface receptors triggers phosphoinositide breakdown and releases calcium from intracellular stores, both of which mediate the downstream events of PKC (Morley et al. 1992). Alternatively, estrogens may alter PKC activity through antioxidant properties. The HT-22 cell line lacks ionotropic glutamate receptors and so glutamate kills the HT-22 cells by a mechanism not involving excitotoxic glutamate receptors (Zaulyanov et al. 1999). Instead, HT-22 cells contain the glutamate/cystine antiporter, which is required for the delivery of cystine into neuronal cells. Inhibition of cystine uptake by excess extracellular glutamate ultimately leads to a reduction in endogenous antioxidant glutathione and cell death (Murphy et al. 1989; Tan et al. 1998). In agreement with this, glutamate at millimolar concentrations induced oxidative stress in the same cell line (Davis and Maher 1994). At concentrations that suppress the membrane translocation and activity of PKC in HT-22 cells, βE2 demonstrates potent antioxidant activity (Behl et al. 1995; Goodman et al. 1996; Green et al. 1997b). Oxidants selectively react with the regulatory domain to activate PKC, whereas antioxidants appear to interact with the catalytic domain to inhibit cellular PKC activity (Gopalakrishna and Jaken 2000). These findings support the hypothesis that estrogen inactivates PKC, prevents membrane translocation, and thereby protects cells from oxidative insults (Jung et al. 2004, 2005).
Growing evidence suggests that the novel PKC isozymes such as PKCε and PKCδ are generally pro-apoptotic in nature (Cross et al. 2000) and that their inhibition is protective (Carpenter et al. 2002; Petrovics et al. 2002). The activation of PKCε was shown to be required for UV-induced apoptosis (Chen et al. 1999). In contrast, PKCε-negative mutants block apoptosis triggered by a variety of neuronal insults (Knauf et al. 1999). βE2 exerts neuronal protection by increasing the expression of the anti-apoptotic protein Bcl-xL (Pike 1999), by decreasing levels of nip-2 mRNA, which encodes a pro-apoptotic protein (Meda et al. 2000), and by blocking the DNA degradation that accompanies glutamate insults (Behl et al. 1997). Relevant data come from our previous study, in which estrogen was shown to suppress PKCε in ethanol withdrawn rats and attenuate apoptotic cell death in the same rat group (Jung et al. 2003).
On the other hand, in certain models of neuroprotection PKCε has exhibited somewhat controversial effects, such as protection against mild ischemic brain insults (Raval et al. 2003). Cordey et al. (2003) reported that estrogen activates PKC in the cortical neurons to exert protection against β-amyloid toxicity. Although the reasons for these discrepancies are not clear, there are a few explanations. First, the PKC assay used by Cordey et al. (2003) neither discriminated between PKC isozymes nor specifically measured PKCε activity, whereas an isozyme-specific assay was used in our study. Second, they applied low concentrations of estrogen (nanomolar range) to cerebrocortical neurons whereas pharmacological concentrations (nanomolar to micromolar) were applied to hippocampal neurons in our study. In fact, when we examined the effects of βE2 on PKC in the hippocampus and cortex of rats withdrawn from ethanol, we saw a similar phenomenon; E2 increased the activity of PKCε in the cortex but decreased it in the hippocampus (M. E. Jung, unpublished observation). Given this, one might speculate that PKC, in particular PKCε, is a molecular sensor that modulates signal homeostasis depending upon concentration, brain region or the nature of insults.
In summary, we assessed PKCs to determine their roles in estrogen protection. Inhibition of two other protein kinases (PKA and PI3k) failed to alter cell viability regardless of estrogen presence. In contrast, we found that PKC inhibition was protective. Three lines of evidence suggest a counteracting link between βE2 and PKC signaling pathways in this neuroprotection model. First, the combined treatment of βE2 and a PKC inhibitor BIM produced a synergistic protection against glutamate toxicity. Second, βE2 inhibited the membrane translocation of PKCε in a manner that correlated with its protection. Finally, βE2 reversed the glutamate-induced increase in the activity of PKCε. These findings support the hypothesis that PKC may be an operative factor in glutamate-induced neuronal cell death (Favaron et al. 1990) and that estrogen suppresses the PKC signaling pathway in an isozyme (PKCε)-specific manner to exert neuroprotection.