Hippocampi were rapidly dissected from 2 to 3 month-old male Sprague-Dawley rats, transferred to oxygenated, ice-cold cutting medium containing (in millimolar): sucrose (220), NaCl (20), NaHCO₃ (26), glucose (10), KCl (2.5), NaH₂PO₄ (1.25), MgSO₄ (2), and cut into 400 μm thick transverse slices using a McIlwain tissue chopper. Hippocampal slices were maintained in a recovery chamber with artificial cerebrospinal fluid (aCSF) medium, containing (in millimolar): NaCl (124), KCl (2.5), CaCl₂ (2.5), MgSO₄ (1.5), NaH₂PO₄ (1.25), NaHCO₃ (24), d-glucose (10), and saturated with 95%O₂/5%CO₂, for 1 h at 37°C. Hippocampal slices were then transferred into screw-cap microfuge tubes containing 3 ml of freshly oxygenated aCSF medium in the presence of various drugs.

Synaptoneurosomes were prepared from 2- to 3-month-old male Sprague-Dawley rats, as previously described (25), with minor modifications. Briefly, cortices were rapidly dissected and placed into chilled modified Krebs solution (mKrebs) containing (in mM): NaCl (118.5), KCl (4.7), KH₂PO₄ (1.18), NaHCO₃ (24.9), MgSO₄ (1.18), CaCl₂ (2.5), d-glucose (10), HEPES (1), pH 7.4, and equilibrated with 95%O₂/5%CO₂. Cortices were homogenized in a 7-ml Kontes tissue Dounce homogenizer with 10 strokes. Homogenized tissue was filtered through a 100-μm nylon mesh filter (BD Falcon) and the resulting suspension was filtered again through a 5-μm pore size Acrodisc syringe filter with a Supor membrane (Pall Life Sciences). The filtrate was then centrifuged at 1000 × g for 15 min at 4°C, washed once, and centrifuged again. The pellet was resuspended in mKrebs and various drugs were directly added to the pre-warmed (5 min, 37°C) synaptoneurosome suspension for the indicated periods of time.

Slices were incubated with 10 nM 17β-estradiol (Calbiochem) at 37°C for the indicated periods of times. Alternatively, slices were incubated with 100 ng/ml BDNF (Millipore), 100 nM 4,4′,4′′-(4-propyl-[1H]-pyrazole-1,3,5-triyl)trisphenol (PPT), 10 nM diarylpropionitrile (DPN), and 100 nM (±)-1-[(3aR*,4S*, 9bS*)-4-(6-bromo-1,3-benzodioxol-5-yl)-3a,4,5, 9b-tetrahydro- 3H-cyclopenta[c]quinolin-8-yl]-ethanone (G1) (all from Tocris Bioscience) for 30 or 60 min. We selected these concentrations for estrogen receptors (ERs) specific agonists based on previous reports showing a relatively low selectivity of DPN (about 30-fold) for ERβ over ERα, as compared to the much higher selectivity of PPT (around 1000-fold) for ERα over ERβ (26) and of G1 for G protein-coupled receptor 30 (GPR30) over ERα and ERβ (27). In some experiments, slices were pre-treated for 30 min with different inhibitors or antagonists, including U0126 (5 μM), K252a (1 μM), ICI 182780 (1 μM), 4-[2-phenyl-5,7-bis(trifluoromethyl)pyrazolo[1,5-a]pyrimidin-3-yl]phenol (PHTPP, 1 μM), cycloheximide (25 μM) (all from Tocris Bioscience), calpain-inhibitor (10 μM, Calbiochem), (±)-(3aR*,4S*,9bS*)-4-(6-bromo-1,3-benzodioxol-5-yl)-3a,4,5,9b- tetrahydro-8-(1-methylethyl)-3H-cyclopenta[c]quinoline (G36, 1 μM, Azano Pharmaceuticals), H1152 (200 nM, Cayman chemicals), PF-3758309 (50–500 nM, ChemieTek), rapamycin (1 μM, Cell Signaling), and latruncullin A (0.5–5 μM, Sigma).

Actin polymerization was quantified by measurement of “rhodamine-phalloidin fluorescent enhancement,” as previously described (28), with little modifications. In brief, hippocampal slices (three to six pooled slices) were washed twice with fresh aCSF after treatment, and subsequently fixed in PBS containing 4% paraformaldehyde and 1% octyl-β-d-glucopyranoside for 15 min at room temperature. After two rinses with PBS, slices were homogenized and centrifuged at 1000 × g for 1 min. Lysates were incubated with 15–30 nM phalloidin-tetramethylrhodamine B isothiocyanate (TRITC, Invitrogen) for 30–45 min at room temperature. After three washes, lysates were collected in 200 μl/slice of PBS and fluorescent intensity (excitation and emission wavelength were 546 and 590 nM, respectively) was determined using a POLARstar Omega fluorescence polarization microplate reader (BMG Labtech).

RhoA activity was determined by pull-down of RhoA-GTP with Rhotekin binding domain (RBD)-linked agarose beads (Millipore), as described previously (6). Rhotekin RBD binds strongly to both RhoA and RhoC, but weakly to RhoB (29). RhoC is not present in brain, thus RhoA was specifically detected in blots probed with RhoA antibody. Briefly, samples (6–10 pooled slices) were homogenized in Mg²⁺ lysis buffer (25 mM HEPES, pH 7.5, 150 mM NaCl, 1% Igepal CA-630, 10 mM MgCl₂, 0.5 mM EDTA and 10% glycerol) containing a protease inhibitor Cocktail (Thermo Scientific). Protein levels were measured and equalized. Samples were incubated with Rhotekin RBD-agarose beads and gently rocked for 1 h at 4°C. Agarose beads were collected by centrifugation (30 s, 16000 × g, 4°C) and washed three times with Mg²⁺ lysis buffer, resuspended and boiled in Laemmli buffer for 5 min, and then separated by SDS-PAGE. Western blot analysis was performed using mouse anti-RhoA antibody (sc-418, Santa Cruz Biotechnology).

Protein lysates were prepared at a final concentration of 4–8 μg protein/μl by homogenization of rat hippocampal slices in ice-cold lysis buffer (50 mM Tris–HCl, pH 7.4, 1% Triton, 1 mM EDTA, 1 mM EGTA, 1 mM PMSF) containing a protease and phosphatase inhibitor cocktail (Thermo Scientific). After sample processing, 25–40 μg of denatured proteins were subjected to 6–15% SDS-PAGE and transferred onto polyvinylidene difluoride membranes. Membranes were blocked with 5% non-fat milk dissolved in Tris-buffered saline (TBS) for 30 min at room temperature and probed overnight with different primary antibodies at 4°C. Membranes were then washed three times with TBS containing 0.1% Tween 20 (TBST) for 10 min followed by incubation with Odyssey infrared-conjugated secondary antibodies diluted 1:10000 in Odyssey blocking buffer (LI-COR Biosciences) containing 0.1% Tween 20 and 0.05% SDS for 1 h at room temperature. After 3× 10 min washes with TBST, membranes were scanned using an Odyssey infrared imager (LI-COR Biosciences).

All primary antibody solutions were prepared in TBST containing 5% bovine serum albumin at a 1:1000 dilution, except otherwise stated. The following primary antibodies were used: anti-phospho-mTOR (Ser2448) and anti-mTOR (both from Cell Signaling), anti-phospho-ERK1/2 (Thr202/Tyr204) and anti-ERK1/2 (both at 1:2000; Cell Signaling), anti-phospho-Akt (Ser473 and Thr308) and anti-Akt (all from Cell Signaling), anti-phospho-p70S6K (Thr389 from Millipore and Thr421/Ser424 from Cell Signaling), anti-p70S6K (Cell Signaling), anti-phospho-cofilin (Ser3, Abcam) and anti-cofilin (Cell Signaling), anti-phospho-PAK1/2/3 (Invitrogen) and anti-PAK3 (1:500, Millipore), anti-spectrin (1:2000; Millipore), anti-Arc (1:5000; Millipore), anti-actin (1:10000; Millipore), and anti-PTEN (Cell Signaling).

Statistical comparisons were made by using one-way analyses of variance (ANOVA) followed by Dunnett’s multiple comparison test or two-way ANOVA followed by Bonferroni posttest analysis. Results were generally calculated as means ± standard error of the mean (SEM) from the indicated number of independent experiments and expressed as fold of the indicated control. Statistical significance was always referred to as: *p < 0.05, **p < 0.01, and ***p < 0.001, as compared to control (DMSO or time = 0 min); and ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001, as compared to BDNF or E2 alone; absence of symbol indicates p values >0.05, which were considered as not significant.

Based on our previous study on the regulation of the mTOR pathway by BDNF (24), stimulation of the BDNF receptor, TrkB, causes ERK-mediated calpain-2 activation, which leads to PTEN cleavage and subsequent mTOR phosphorylation via Akt stimulation. We first determined the time course of estradiol-mediated regulation of mTOR and of some of the signaling proteins upstream of mTOR in acute hippocampal slices. Both application of 10 nM estradiol increased mTOR phosphorylation within 15 min, and the effect persisted for at least 60 min (Figure 1A). Similarly, PTEN levels were decreased as early as 15 min and remained decreased for up to 1 h after estradiol treatment, although the effect was statistically significant only for the 30-min time point (Figure 1B). In contrast, activation of both ERK and Akt by estradiol was transient; ERK phosphorylation was increased within 5–15 min but not at later time points, whereas Akt activation (measured as increase in phospho-Akt levels at Ser473) peaked 15 min after estradiol application and gradually returned to baseline (Figures 1C,D). A similar time course was obtained for estradiol-induced phosphorylation of Akt at Thr308 (data not shown). We also assessed calpain activity in hippocampal slices by determining the levels of the calpain-specific spectrin breakdown product (SBDP) after estradiol treatment. However, we could not detect any changes in SBDP levels across the different time points, as compared to the control group (data not shown). This is likely due to the high basal levels of SBDP in hippocampal slices from adult rats, which has been previously documented (30). For this reason, we monitored the effects of estradiol on calpain activity using a different preparation, cortical synaptoneurosomes. We first verified that estradiol produced a similar pattern of ERK activation in synaptoneurosomes as in hippocampal slices. ERK phosphorylation was enhanced 5 min after estradiol treatment, but the effect returned to baseline levels faster (at 15 min) than in hippocampal slices (Figure 1E). Similarly, estradiol transiently increased the levels of SBDP in synaptoneurosomes (Figure 1F).

To determine the link between estrogen-stimulated calpain activity and mTOR-dependent protein synthesis, we treated hippocampal slices with 10 μM calpain-inhibitor III (CI-III) before estradiol application and then examined mTOR activation. Calpain inhibition completely blocked estradiol-elicited increase in mTOR phosphorylation (Figure 2A). Pre-treatment with CI-III also prevented PTEN degradation and Akt activation caused by estradiol (Figures 2B,C). Similar results were obtained for estradiol-induced mTOR activation and PTEN degradation in cortical synaptoneurosomes (data not shown). One of the main targets of mTOR in neurons is the translation factor p70SK6, which is activated by phosphorylation at two different sites, Thr389 and Thr421/Ser424 (31). We then examined whether estradiol activates p70S6K in hippocampal slices. Treatment with estrogen enhanced the phosphorylation of p70SK6 at Thr389, and the effect was blocked by calpain inhibition (Figure 2D). In contrast, the increase in p70SK6 phosphorylation at Thr421/Ser424 caused by estradiol was not significantly inhibited by CI-III (Figure 2E), possibly because phosphorylation at these sites is not only mediated by mTOR but also by ERK (31). To further confirm that calpain activity was required for estradiol-mediated regulation of protein synthesis, we determined the effect of the hormone on the levels of Arc, a protein known to be rapidly up-regulated in response to synaptic activity (32). Estradiol significantly enhanced Arc levels in cortical synaptoneurosomes and the effect was blocked by CI-III pre-treatment (Figure 2F).

Rapid calpain activation by estradiol and BDNF in dendritic spines is triggered by ERK-dependent phosphorylation (8, 28). If calpain activity is required for mTOR-mediated protein translation, ERK should be upstream of mTOR as well. When hippocampal slices were pre-treated with the ERK kinase specific inhibitor U0126 (5 μM), both estrogen- or BDNF-induced mTOR activation was prevented (Figure 3A). Since estradiol and BDNF were able to activate the same signaling cascade, we then tested the hypothesis that BDNF could mediate the effects of estradiol on mTOR activity. Pre-incubation of hippocampal slices with the TrkB receptor antagonist K252a (1 μM) blocked the increase in mTOR phosphorylation caused by estradiol (Figure 3B).

To further examine the mechanism by which estradiol activates mTOR, we used specific agonists and antagonists of the two major ER isoforms, ERα and ERβ. Neither the non-specific ER antagonist ICI 182780 (ICI, 1 μM) nor the specific ERβ antagonist PHTPP (1 μM) affected estradiol-induced mTOR phosphorylation (Figure 4A). Consistent with this result, application of the ERα and ERβ specific agonists PPT (100 nM) or DPN (10 nM), respectively, did not stimulate mTOR activity (Figure 4B), suggesting that the effects of estradiol on the mTOR pathway do not involve activation of classical ERs. Some of the non-genomic effects of estradiol have been show to be mediated by the GPR30 (33). Thus, we tested the effects of a novel and very specific GRP30 antagonist, G36 (1 μM) (27), on estradiol-induced mTOR activation. This compound completely suppressed estradiol-elicited mTOR activation without affecting its basal levels (Figure 4A). In addition, treatment with the GPR30 agonist G1 (100 nM) enhanced mTOR phosphorylation to a similar extent as estradiol (Figure 4B). These results confirm that stimulation of the mTOR pathway by estradiol involves GPR30 activation. Interestingly, treatment with ICI alone significantly enhanced basal mTOR activity (Figure 4A). As will be discussed below, this effect might be related to its agonist activity at GPR30.

Estrogen has been shown to rapidly stimulate actin cytoskeleton dynamics via the RhoA/ROCK/cofilin pathway (6). It can also activate Rac/PAK signaling in neurons (14), but the role of this pathway in cofilin phosphorylation and actin polymerization has not yet been directly determined. Similarly, the time dependency for estradiol regulation of these pathways is uncertain. Thus, we first studied the temporal profile of the effects of estradiol on actin polymerization and on cofilin, RhoA, and PAK activation in hippocampal slices. Estradiol enhanced both actin polymerization and cofilin phosphorylation within 5 min, although these effects were not statistically significant. However, a significant activation of both actin polymerization and cofilin phosphorylation was observed 30 and 60 min after estradiol application (Figures 5A,B). Estradiol also increased RhoA and PAK activities although with different temporal patterns; the stimulatory effect of estradiol on RhoA activity was significant only at 60 min (Figure 5C), whereas PAK phosphorylation increased within 15–30 min and then returned to basal levels by 60 min (Figure 5D).

To examine the relative contribution of ROCK and PAK on estradiol-induced actin polymerization, we pre-treated hippocampal slices either with H1152 or PF-3758309, potent and specific inhibitors of ROCK and PAK, respectively. PF-3758309 has about a 10-fold higher selectivity for PAK1/4 (K_i values of 14/19 nM, respectively), as compared to PAK2/3 (K_i values of 190/99 nM, respectively) (34). Thus, we selected two different PF-3758309 concentrations (50 and 500 nM) in order to discriminate between PAK isoforms. We first verified the specificity of these inhibitors on BDNF- and estradiol-induced PAK phosphorylation by using an antibody that recognizes the activated forms of PAK1/2/3 isoforms. PF-3758309 blocked PAK activation caused by BDNF in a dose-dependent manner, although the effect was completely suppressed at the highest concentration only. In contrast, PF-3758309 totally inhibited estradiol-induced PAK activation at all the concentrations tested (Figure 6A) suggesting that BDNF and estradiol activate different PAK isoforms. As expected, treatment with H1152 did not affect the increase in PAK activity caused by either BDNF or estradiol (Figure 6A). Interestingly, both H1152 and PF-3758309 prevented the increase in cofilin phosphorylation and in actin polymerization induced by estradiol (Figures 6B,C), indicating that both pathways are involved in cofilin-dependent regulation of actin cytoskeleton by estrogen.

We also tested the effects of ER and TrkB receptor antagonists on estrogen-mediated actin reorganization. Treatment with either ICI or PHTPP blocked the effects of estradiol on PAK and cofilin phosphorylation. In contrast, K252a failed to inhibit estradiol-induced PAK activation, but it prevented the increase in cofilin phosphorylation caused by the hormone (Figures 7A,C). Consistent with the above data, ICI suppressed estradiol-induced actin polymerization. K252a reduced the stimulatory effect of estradiol on actin polymerization, although the effect was not significant (Figure 7E). To further confirm the involvement of ERs in the cytoskeletal changes elicited by estrogen, we tested the effects of the specific ER agonists on the same endpoints. Both PPT and DPN significantly increased PAK and cofilin phosphorylation (Figures 7B,D). Similarly, the two agonists stimulated actin polymerization to a similar extent as estradiol (Figure 7F). These results suggest that both ERα and ERβ are involved in estradiol regulation of the actin cytoskeleton.

We next evaluated the relationship between protein synthesis and cytoskeletal reorganization after exposure to estrogen in order to determine whether these two events are independent. To address this question, we determined the effects of protein synthesis inhibitors on estradiol-elicited actin polymerization in hippocampal slices. Neither 25 μM cycloheximide nor 1 μM rapamycin blocked estradiol-induced actin polymerization (Figure S1 in Supplementary Material). Estrogen-mediated stimulation of dendritic protein translation has been shown to be ERK-dependent (22). However, under our experimental conditions, treatment with the ERK kinase inhibitor U0126 (5 μM) failed to prevent estradiol-induced actin polymerization (Figure S1 in Supplementary Material). These results indicate that protein synthesis is not required for estradiol-mediated regulation of cytoskeletal dynamics.

We then determined whether actin polymerization was necessary for estradiol-mediated mTOR stimulation. Thus, we pre-treated hippocampal slices with Latrunculin A (Lat A), an agent that sequesters actin monomers, thereby preventing filamentous actin (F-actin) assembly. Two concentrations of Lat A were used (0.5 and 5 μM) based on the different sensitivities of F-actin to this drug depending on its subcellular location; low concentrations of Lat A specifically target actin filaments within the dendritic shaft, while higher concentrations are required to affect dendritic spine F-actin (31). Treatment with 0.5 μM Lat A alone had no effect on basal mTOR activity. In contrast, the depolymerizing agent slightly reduced mTOR phosphorylation at 5 μM, although the effect did not reach statistical significance. Interestingly, Lat A caused a concentration-dependent reduction of estradiol-mediated mTOR phosphorylation in hippocampal slices. Under the same conditions, BDNF-induced mTOR activation was also impaired, but the effect was only significant at the highest concentration of Lat A tested (Figure 8A). Similarly, BDNF- and estradiol-induced phosphorylation of p70S6K both at Thr389 (Figure S2A in Supplementary Material) and at Thr421/Ser424 (Figure 8B) was totally suppressed by Lat A at all the concentrations tested. In contrast, the increase in Akt phosphorylation caused by BDNF and estradiol was significantly reduced by Lat A at the highest concentration only (Figure S2B in Supplementary Material). These results suggest that actin polymerization is required for mTOR-dependent protein translation elicited by BDNF and estradiol.

If actin polymerization regulates estradiol-mediated mTOR activation/phosphorylation, it could be due to an activity-dependent event triggered by activation of classical pathways (i.e., RhoA/ROCK and/or Rac/PAK) or be due to a structural requirement linking activated Akt to mTOR. We addressed this question by determining the effects of specific ROCK and PAK inhibitors on estradiol- and BDNF-induced mTOR and Akt activation. Treatment with H1152 did not affect mTOR phosphorylation stimulated by BDNF or estradiol in hippocampal slices. In contrast, 500 nM (but not 50 nM) PF-3758309 completely blocked BDNF- and estradiol-induced mTOR activation (Figure 9A). Likewise, the increase in phospho-Akt caused by BDNF and estradiol was prevented by PF-3758309 only at the highest concentration tested. Surprisingly, treatment with H1152 alone enhanced Akt phosphorylation but the effect was not statistically significant (Figure 9B). This result is in agreement with previous studies linking ROCK inhibition to stimulation of Akt signaling pathway in neuronal cells (35, 36). Overall, these findings suggest that activation of PAK2/3 signaling by estradiol and BDNF regulates mTOR signaling, possibly through stimulation of actin polymerization.